Skip to main content
NCBI home page
Animals : an Open Access Journal from MDPI logoLink to Animals : an Open Access Journal from MDPI
. 2025 Dec 5;15(24):3518. doi: 10.3390/ani15243518

Integrative Genomics and Multi-Tissue Transcriptomics Identify Key Loci and Pathways for Hypoxia Tolerance in Grass Carp

Wenwen Wang 1,2,†, Mengyang Chang 1,†, Suxu Tan 1,2, Yiming Hu 1,2, Xinlu Ren 1, Hongtao Xue 3, Lizheng Gao 1, Xiao Cao 4, Ya Wang 4, Qiyu Li 4, Zhenxia Sha 1,2,5,*
Editor: Marco Casu
PMCID: PMC12729821  PMID: 41463803

Simple Summary

Hypoxia is a major threat to the survival and performance of cultured fish. In this study, we combined genome-wide association analysis (GWAS) with multi-tissue transcriptome profiling to investigate the genetic basis of hypoxia tolerance in grass carp. We identified 21 SNPs, 6 InDels, and 16 candidate genes, including usf1 and trpv4, associated with hypoxia response. These results provide useful genomic markers and resources to support selective breeding for improved hypoxia tolerance in grass carp.

Keywords: grass carp, hypoxia, GWAS, RNA-seq, SNPs

Abstract

Hypoxia is a critical environmental stressor in aquaculture, significantly affecting the survival and growth performance of cultured fish. To explore the genetic basis of hypoxia tolerance in grass carp (Ctenopharyngodon idella), we integrated genome-wide association analysis (GWAS) and multi-tissue transcriptome profiling. A total of 2000 grass carp were subjected to hypoxic stress, from which the 150 most hypoxia-intolerant (HI) and 150 most hypoxia-tolerant (HT) individuals were selected based on the time to loss of equilibrium (LOE). GWAS using 3,730,919 SNPs and 851,595 InDels identified 21 SNPs and 6 InDels associated with hypoxia tolerance. Two SNPs on chromosomes 10 and 13 reached genome-wide significance, accounting for 2.7% and 4.8% of the phenotypic variance explained (PVE), respectively. Validation of identified SNPs was performed using kompetitive allele-specific PCR (KASP) analysis. Candidate genes within ±50 kb of these variants were enriched in steroid biosynthesis, insulin signaling, and glycosphingolipid biosynthesis pathways. Transcriptomic analysis of six tissues (brain, gill, intestine, kidney, liver, and spleen) revealed 1620, 1221, 796, 246, 210, and 58 differentially expressed genes (DEGs) in the HT group compared to the HI group, respectively. DEGs in the brain were primarily enriched in steroid metabolic processes and angiogenesis regulation, while those in kidney and spleen DEGs were associated with oxygen transport and erythrocyte development. Integrated analysis of GWAS and transcriptome data identified 16 shared genes, including usf1 and trpv4. These findings reveal key genomic loci and molecular pathways underlying hypoxia tolerance in grass carp, providing valuable markers for future selective breeding programs.

1. Introduction

Hypoxia, a condition characterized by reduced dissolved oxygen (DO) levels in water, is a major environmental stressor affecting aquatic ecosystems worldwide [1]. In natural habitats, hypoxia can result from eutrophication, algal blooms, and thermal stratification [1]. In aquaculture systems, hypoxia mainly occurs under the following conditions: rainy/overcast days, high-temperature weather, the period from night-time to early morning, excessively high stocking density, or deterioration of water quality [2]. Hypoxia can cause metabolic disorders, weakened immunity, slow growth, decreased reproduction, and even death in fish [3]. Teleost fish are good models for hypoxia studies because they inhabit aquatic habitats with varying oxygen levels, and they have evolved a suite of physiological and biochemical adaptations, including metabolic suppression, enhanced anaerobic glycolysis, and increased oxygen uptake efficiency [4,5]. Different fish species, and even individuals within the same species, show considerable variation in hypoxia tolerance [6,7]. The genetic basis underlying this interspecific and intraspecific variation in hypoxia tolerance is complex and continues to be a key area of research.

Genome-wide association study (GWAS) has successfully identified genetic markers for disease resistance, growth, and stress tolerance in various fish species [8,9]. Multiple GWAS have identified genetic variants linked to fish hypoxia tolerance. For example, in a study on large yellow croaker (Larimichthys crocea), seven significant hypoxia-associated single nucleotide polymorphisms (SNPs) were identified by GWAS, and the shared genes from GWAS and differentially expressed genes (DEGs) from transcriptome analysis were related to glucose transport and metabolism, erythropoiesis, ion regulation, DNA replication, and repair [10]. In hybrid catfish, four quantitative trait loci (QTL) were found to be associated with low oxygen tolerance by GWAS using the catfish 250 K SNP array [11]. For channel catfish (Ictalurus punctatus) under hypoxia stress, one significant QTL across strains and six significant QTL within strains were identified and candidate genes were predicted to function in oxygen metabolism and to participate in MAPK or PI3K/AKT/mTOR signaling pathways [12]. A total of 4 significant SNPs and 16 potential candidate genes were identified as being associated with hypoxia tolerance in golden pompano (Trachinotus ovatus) [13]. However, research on fish hypoxia tolerance using GWAS remains relatively limited, with even fewer applications in molecular breeding.

Grass carp (Ctenopharyngodon idella), an economically important native Chinese freshwater fish, has been widely cultivated worldwide. Grass carp is the fish species with the highest aquaculture output in China and has contributed significantly to food security and rural economies in Asia [14,15]. Currently, research on grass carp breeding primarily focuses on traits such as growth and disease resistance, as well as GWAS on genetic loci associated with these traits [16,17]. With the development of intensive farming practices, stress-resistant traits, such as hypoxia tolerance and ammonia tolerance, have gained increasing attention [18]. In this study, we employed a combined GWAS and transcriptomic approach to identify SNPs, insertions and deletions (InDels), and potential candidate genes associated with hypoxia tolerance in grass carp, offering molecular markers for breeding programs and advancing our understanding of adaptive responses to environmental stress in aquatic species.

2. Materials and Methods

2.1. Animal Welfare Statement

All handling of the fish was performed in accordance with the Guidelines for the Care and Use of Laboratory Animals of China and approved by the Ethics Committee of the Medical College of Qingdao University, China (QDU-AEC-2024004).

2.2. Hypoxia Treatment and Sample Collection

The experiment was conducted at Weishan County Nansihu Fishery Co., Ltd., Jining, China. The grass carp cultivated by this company originates from fry procured from a hatchery in Guangdong Province (multiple breeding families and non-random mating). In this study, a total of 2000 six-month-old grass carp with an average body weight of 23.50 g and body length of 12.58 cm were collected and distributed into five 150 L tanks (each containing 120 L of water). The fish were acclimated for one week in tanks at a water temperature of 21.0 ± 1.0 °C and a DO level of 6.0 ± 0.5 mg/L by aeration. During the hypoxic stress experiment, the aeration was stopped and DO level was monitored by a water quality meter (YSI ProQuatro, Yellow Springs, OH, USA). According to the previous studies, DO levels were decreased by adding sodium sulfite [11,12,19,20]. The water temperature, pH, and ammonia nitrogen were monitored every 2 h to ensure stable experimental conditions.

The time interval from when DO dropped to 0.1 mg/L until each fish exhibited loss of equilibrium (LOE) was recorded as the hypoxia-tolerance phenotype. After reaching LOE, fish were euthanized with tricaine methanesulfonate (MS-222), and body weight and body length were measured. Fin clips were collected from the first 150 fish that reached LOE (hypoxia-intolerant fish, HI) and the last 150 fish (hypoxia-tolerant fish, HT) for genomic DNA extraction. In addition, tissues including brain, gill, intestine, kidney, liver, and spleen from the first three HI fish and the last three HT fish were collected for total RNA extraction, resulting in 36 tissue samples (3 replicates × 2 groups × 6 tissues). Each collected sample was placed in a cryogenic vial, immediately frozen in liquid nitrogen, and stored at −80 °C for subsequent analysis.

2.3. Genomic DNA Extraction and Sequencing

Genomic DNA was extracted from 300 fin samples using the PureLink Fast DNA Tissue Kit (Invitrogen, Carlsbad, CA, USA), following the manufacturer’s protocol. The DNA concentration and quality were assessed using a nucleic acid analyzer (OSTC, Beijing, China). High-quality genomic DNA was stored at −20 °C for subsequent library construction. The average DNA concentration was 80–100 ng/μL, and the OD260/280 ratios ranged from 1.8 to 2.0. Libraries were constructed with the NEBNext® UltraTM DNA Library Prep Kit (NEB, Ipswich, MA, USA). Following quality control, paired-end libraries with 350 bp insert size were constructed. The DNA libraries were sequenced on the DNBSEQ-T7 platform (MGI Tech, Shenzhen, China), generating paired-end 150 bp reads with an average depth of ~10× per sample.

2.4. Genotyping and Filtering

The raw DNA sequencing data were assessed for quality and filtered using fastp software (version 0.23.4) [21] to remove adapter contamination and low-quality reads. After filtering, high-quality clean reads were generated and subsequently mapped to the genome of the C. idella (GCA_019924925.1) [22] using BWA software (version 0.7.17) [23]. Duplicate reads were marked and removed using Picard Tools (version 2.25.0) (https://github.com/broadinstitute/picard, 4 April 2025). Variant calling was performed to identify SNPs and InDels variants using GATK (version 4.5.0.0) [24]. SNPs were subjected to hard filtering with the following parameters: QualByDepth (QD) < 2.0, FisherStrand (FS) > 60.0, RMSMappingQuality (MQ) < 40.0, ReadPosRankSumTest (ReadPosRankSum) < −8.0, MappingQualityRankSumTest (MQRankSum) < −12.5, and SOR ≥ 3.0. For InDel filtering, the parameters were: QD < 2.0, MQRankSum < −12.5, FS > 200.0, ReadPosRankSum < −8.0, and SOR > 10.0. After variant filtering, PLINK software (version 1.9) [25] was used for additional quality control with the following thresholds: (1) missing genotype rate per SNP (-geno) < 0.01, (2) individual missing rate (-mind) > 0.02, (3) minor allele frequency (-maf) < 0.05, and (4) Hardy–Weinberg equilibrium (-hwe) < 0.001. To categorize the functional effects of the variants, SnpEff software (version 5.0) was utilized for annotation based on the annotated genome of C. idella.

2.5. Linkage Disequilibrium (LD) and Population Structure Analyses

Only autosomal SNPs with MAF > 0.05 were retained for LD and population structure analyses. LD analysis was conducted using PopLDdecay software (version 3.42), which calculates the LD coefficient (r2) between pairs of variants (SNPs and InDels) [26]. The genome-wide LD decay pattern as a function of physical distance was visualized using the Plot_OnePop Perl script provided with the PopLDdecay package. To investigate potential genetic relatedness within the population, population structure analysis was performed. Principal component analysis (PCA) was conducted using PLINK (version 1.9) [25], and the first two principal components (PC1 and PC2) were plotted to visualize sample clustering. Additionally, Admixture software (version 1.3.0) [27] was used to estimate population structure. The number of ancestral populations (K) was set from 1 to 15, and the optimal K value was determined based on the minimum cross-validation (CV) error. Visualization of the population structure results was performed using the R package ggplot2 (version 4.2.1).

2.6. Genome-Wide Association Study (GWAS)

GWAS was conducted using GEMMA software (version 0.98.5) based on a linear mixed model (LMM) [28]. First, the genetic kinship between individuals was estimated using the parameter: -gk 2. The resulting genetic relatedness matrix was then applied in the GWAS analysis, with model adjustments made based on the results from PCA. The GWAS model used is expressed as follows:

y = Wα + Xβ + Zμ + ϵ

where y is the phenotype vector, W is the covariate matrix (including intercept and PC covariates), α is the vector of fixed effects, X is the genotype matrix, β is the marker effect, Z is the random effect design matrix, μ is the random effect vector, and ϵ is the residual.

To account for multiple testing, we applied the Bonferroni correction method, setting the significance threshold at 0.05/N for significant associations and 1/N for suggestive associations, where N represents the total number of variants analyzed. Manhattan and quantile-quantile (Q-Q) plots were generated using the CMplot package (version 4.5.10) to visualize the GWAS results. The phenotypic variance explained (PVE) value of the variants was calculated based on the published formula. The SnpEff software (version 5.0) [29] was used to construct the genome annotation database of C. idella using genome annotation files and reference genome sequences. Subsequently, the identified SNPs from GWAS were annotated. Considering the genome-level LD in C. idella, genomic regions flanking the detected SNPs (50 kb upstream and downstream) were identified for candidate genes. Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analyses of candidate genes were performed using the clusterProfiler package (version 4.6.2) [30] in R. Protein–protein interaction (PPI) network of candidate genes was constructed using String (https://cn.string-db.org/, 4 April 2025).

2.7. Total RNA Extraction, Transcriptome Sequencing and Data Processing

Total RNA was extracted from 36 samples, including brain, gill, intestine, kidney, liver, and spleen from three HI fish and three HT fish using TRIzol reagent (Qiagen, Hilden, Germany). RNA integrity was assessed using Agilent 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA, USA), and samples with RNA Integrity Number (RIN) ≥ 7.0 were used for library preparation. The sequencing libraries were constructed using Stranded RNA LibraryPrep Kit (lllumina, San Diego, CA, USA). The libraries were sequenced on the DNBSEQ-T7 platform (MGI Tech, Shenzhen, China), generating paired-end 150 bp reads.

The raw reads in FASTQ format were initially processed using fastp [21], which removed low-quality reads to generate clean reads. These clean reads were subsequently aligned to the reference genome (GCA_019924925.1) using HISAT2 (version 2.0.5) [31]. Gene expression levels were quantified as Fragments Per Kilobase of exon model per Million (FPKM), and read counts were obtained using featureCount (version 2.0.4) [32]. PCA was conducted in R (version 4.2) to assess the biological duplication of samples. Differential expression analysis was performed using DESeq2 (version 1.40.2) [33], with a threshold of adjust p-value < 0.05 and fold change > 2 to identify DEGs. Hierarchical clustering analysis of DEGs was performed to illustrate expression patterns across different groups and samples using R (version 4.2). A radar map to visualize the expression of upregulated or downregulated DEGs was generated using the ggplot2 package in R (version 4.2). Furthermore, Gene Ontology (GO) and KEGG pathway analyses of DEGs were carried out using the clusterProfiler (version 4.6.2) [30] package in R to elucidate the functional roles of the identified DEGs.

2.8. Validation of Significant SNPs

In this study, two SNPs [(the SNP at position 33,445,030 bp on chromosome 5 (SNP7) and the SNP at position 36,269,337 bp on chromosome 10 (SNP11)] that showed strong statistical associations with hypoxia tolerance based on GWAS and subsequent filtering criteria were selected for KASP genotyping. A total of 300 individuals were used to validate the genotype–phenotype relationship of these loci. Each SNP genotype was called using fluorescence signal clustering in the Bio-Rad CFX96 system, and the accuracy was verified by randomly re-genotyping 10% of the samples. The primers were designed on https://tools.goodbtk.com/SNPPrimer (20 April 2025) and listed in Supplementary Table S1. PCR reactions were conducted with a final volume of 2 μL in each well, containing 1 μL Flu-Arms 2× PCR mix V4 (GoodBTK, Guangzhou, China), 0.02 μL primer F1 (10 μM), 0.02 μL primer F2 (10 μM), 0.06 μL primer R (10 μM), DNA template (5–50 ng), and water to a total volume of 2 μL. Genotype-phenotype association was re-evaluated using Chi-square test between allelic groups.

3. Results

3.1. Phenotype Statistics

The first grass carp to lose equilibrium was observed at 0.68 h after the addition of sodium sulfite. As the duration of hypoxia stress increased, individual fish successively lost equilibrium. The average time for the first 150 fish to lose balance was 3.49 h, while that for the latter 150 fish was 23.62 h (Figure 1). Notably, a significant difference (p =7.65 × 10−192) was detected in the time to LOE between the HI and HT groups (Figure 1), highlighting marked inter-individual variation in the hypoxia tolerance trait of grass carp.

Figure 1.

Figure 1

Open in a new tab

Comparison of lost equilibrium time in the hypoxia-intolerant (HI) and hypoxia-tolerant groups (HT). Asterisk (***) indicates statistical significance between HI and HT groups using Welch’s t-test, *** p < 0.001.

3.2. Genotyping and Population Structure

After genotyping and quality filtering, a total of 3,730,919 SNPs and 851,595 InDels were identified across 300 fish. These SNPs and InDels were widely distributed across the 24 chromosomes (Figure 2A). Based on the 893.2-Mb reference genome, the combined variants correspond to an average density of approximately 5.13 markers per kb, indicating a generally uniform genome-wide distribution without strong clustering signals. The MAF distributions of SNPs and InDels are shown in Figure 2B. The genome-wide average density rate was calculated as 243 bp/SNP and 963 bp/InDel. Of these variants, 1,214,563 (32.55%) SNPs and 279,885 (32.87%) InDels were located in introns, 1,319,775 (35.37%) SNPs and 294,830 (34.62%) InDels were identified in intergenic regions, and 69,245 (1.86%) SNPs and 5007 (0.59%) InDels were located in exons (Figure 2C).

Figure 2.

Figure 2

Open in a new tab

The characteristics of SNPs and InDels in grass carp. (A) Density and distribution of SNPs and InDels among 24 chromosomes. From outermost to innermost, the tracks represent: chromosomes, SNP density, InDel density, deletion density, and insertion density. (B) The minor allele frequency (MAF) distribution of SNPs and InDels. (C) Frequency of genomic region positions for SNPs and InDels.

The PCA using high-quality SNPs revealed that the grass carp used in this study belonged to several discrete sub-populations (Figure 3A). When K = 12, the CV value reached its minimum (Figure 3B), indicating this was the optimal model for describing the population genetic structure. This suggests that the population consisted of 12 genetically distinct sub-populations. In addition, the genetic relatedness matrix showed weak genetic relatedness among the grass carp used in this study (Figure 3C), reflecting a high level of genetic diversity within the population. The LD analysis revealed a maximum r2 value of 0.44, followed by a sharp decrease to 0.1 at a distance of 400 bp (Figure 3D). Population structure analysis, including PCA (Supplementary Figure S1A) and genetic relatedness matrix (Supplementary Figure S1B) using InDels, was also conducted, supporting the findings from the SNP analysis.

Figure 3.

Figure 3

Open in a new tab

Population structure analyses using SNPs of grass carp. (A) Principal component analysis (PCA) analysis. (B) Variation in cross-validation (CV) error at different K values. (C) Heatmap of genomic relatedness. (D) Linkage disequilibrium (LD) decay.

3.3. GWAS

GWAS for the HI and HT groups was conducted using 150 genotyped samples each. Association analyses of HI and HT traits were performed using LMM with 3,730,919 SNPs and 851,595 InDels. The threshold p value for genome-wide statistical significance was 0.05/3,730,919 = 1.34 × 10−8 [−log10 (p value) = 7.87] for SNPs. The threshold p value for the suggestive association was 1/3,730,919 = 2.68 × 10−7 [−log10 (p value) = 6.57] for SNPs. For the InDels, the significant and suggestive thresholds were −log10 (0.05/851,595) = 7.23 and −log10 (1/851,595) = 5.93, respectively.

The Manhattan plots of the GWAS results are shown in Figure 4. A total of 21 SNPs were identified in association with hypoxia tolerance in grass carp, located on chromosome 2, 3, 4, 5, 6, 7, 10, 13, 14, 16, 19 and 20 (Table 1, Figure 4A). Two SNPs on chromosome 10 and 13 were significantly associated with hypoxia tolerance with PVE values of 4.8% and 2.7%, respectively (Table 1). A total of 6 suggestive InDels were detected on chromosome 1, 2, 3, 10, 11 and 15 (Figure 4B). The QTL shared between the GWAS results of SNPs and InDels were located on chromosome 2 and 10. The Quantile-quantile (Q-Q) plots confirmed the reliability and validity of our GWAS analysis (Supplementary Figure S2), demonstrating that the statistical model was appropriately specified for this study.

Figure 4.

Figure 4

Open in a new tab

Manhattan plots of genome-wide association analysis for hypoxia tolerance in grass carp. (A) Manhattan plot for SNPs. (B) Manhattan plot for InDels. The solid line indicates the threshold for significance, and the dotted line indicates the threshold for suggestive association. The red boxes represent the QTL shared between the GWAS results of SNPs and InDels.

Table 1.

The significant and suggestive SNPs associated with hypoxia tolerance in grass carp.

SNPs Chromosome Position Allele PVE (%) p Value Association
1 13 31,907,229 A/C 4.8 4.2568 × 10−9 significant
2 10 11,648,443 G/A 2.7 9.5401 × 10−9 significant
3 20 32,115,202 G/A 6.1 4.1755 × 10−8 suggestive
4 20 32,115,197 T/A 6.0 4.1763 × 10−8 suggestive
5 7 22,787,017 A/C 3.5 6.1038 × 10−8 suggestive
6 2 4,792,764 G/A 2.2 1.1057 × 10−7 suggestive
7 5 33,445,030 A/G 4.0 1.1164 × 10−7 suggestive
8 4 9,957,664 C/T 1.9 1.1423 × 10−7 suggestive
9 2 4,101,177 C/T 2.0 1.203 × 10−7 suggestive
10 7 22,787,042 T/C 3.3 1.2183 × 10−7 suggestive
11 10 36,269,337 A/G 5.2 1.268 × 10−7 suggestive
12 20 29,628,336 C/A 4.1 1.4681 × 10−7 suggestive
13 3 31,672,079 T/C 2.8 1.5794 × 10−7 suggestive
14 19 7,384,912 T/C 2.6 1.7027 × 10−7 suggestive
15 14 24,079,263 A/T 3.9 1.8668 × 10−7 suggestive
16 5 31,480,201 A/T 1.8 1.8728 × 10−7 suggestive
17 19 16,977,642 G/T 2.5 2.0578 × 10−7 suggestive
18 19 2,891,834 C/T 1.7 2.0833 × 10−7 suggestive
19 6 13,385 G/A 1.2 2.1928 × 10−7 suggestive
20 16 23,173,806 A/G 3.0 2.5773 × 10−7 suggestive
21 20 1,808,957 G/T 5.5 2.6181 × 10−7 suggestive
Open in a new tab

3.4. Genes Within the QTL Regions

To provide insights into the potential candidate genes associated with hypoxia tolerance, we examined the ±50 kb genomic regions surrounding the significant and suggestive SNPs and InDels. A total of 89 genes were identified in the SNP-associated regions (Supplementary Table S2) and 34 genes in the InDel-associated regions (Supplementary Table S3). To further investigate their functional relevance, KEGG pathway enrichment analysis was performed (Figure 5A). These candidate genes were significantly enriched in pathways such as steroid biosynthesis, insulin signaling, glycosphingolipid biosynthesis, IgA production, and glucagon signaling.

Figure 5.

Figure 5

Open in a new tab

Analysis of candidate genes associated with SNP and InDels. (A) KEGG enrichment analysis of candidate genes. (B) Results of protein–protein interaction network (PPI) analysis of candidate genes.

Furthermore, we constructed a PPI network to identify potential hub genes (Figure 5B). Fourteen key candidate genes were highlighted, including protein phosphatase 2, regulatory subunit B, gamma b (ppp2r5cb), CDP-diacylglycerol--inositol 3-phosphatidyltransferase (cdipt), Zgc: 112,271 protein (bola2), translocase of outer mitochondrial membrane 7 homolog (tomm7), SLX1 homolog B, structure-specific endonuclease subunit (slx1b), glutaredoxin 2 (glrx2), ubiquitin carboxyl-terminal hydrolase L5 (uchl5), dystonin (dst), G protein-coupled receptor kinase interacting ArfGAP 2a (git2a), glycolipid transfer protein a (gltpa), ankyrin repeat domain 13A (ankrd13a), all-trans retinoic acid-induced differentiation factor (atraid), sorting nexin 17 (snx17), and eukaryotic translation initiation factor 2B, subunit 4 delta (eif2b4). These genes may play central roles in the regulation of hypoxia tolerance in grass carp.

3.5. Differential Expression Between HI and HT Groups

RNA-seq was conducted in brain, intestine, kidney, liver, gill, and spleen for both the HI and HT groups. A total of 1791 million raw reads were produced, with 1710 million clean reads retained after quality filtering (Supplementary Table S4). The raw data have been submitted to NCBI SRA with the BioProject ID PRJNA1306682. The average mapping rate was 91.98% across all samples (Supplementary Table S4). Comparative analysis of expression levels between HI and HT groups revealed tissue-specific differential gene expression: 1620, 1221, 796, 246, 210, and 58 DEGs were identified in the brain, intestine, kidney, liver, gill, and spleen, respectively (Supplementary Figure S3).

GO analysis revealed that DEGs in brain tissue were significantly enriched in biological processes (BP) such as steroid metabolic process, cholesterol biosynthetic process, cholesterol metabolic process, positive regulation of angiogenesis, and positive regulation of blood vessel development (Figure 6A). In the intestine, GO terms consistently enriched with those in the brain included steroid metabolic process, sterol biosynthetic process, cholesterol metabolic process, and related terms. Furthermore, DEGs were also enriched in tRNA metabolic process, tRNA aminoacylation for protein translation, tRNA aminoacylation, outer mitochondrial membrane organization, etc. (Figure 6B). DEGs in the kidney were enriched in innate immune response in mucosa gas transport, mucosal immune response, oxygen transport, erythrocyte development/differentiation and heme metabolic process (Figure 6C). The enriched GO terms included negative regulation of inclusion body assembly, regulation of angiogenesis, regulation of inclusion body assembly, regulation of vasculature development, positive regulation of angiogenesis, etc. in the liver (Figure 6D). GO BP terms, such as protein hydroxylation, peptidyl-proline hydroxylation to 4-hydroxy-L-proline, non-proteinogenic amino acid metabolic process, were enriched in the gill (Figure 6E). In addition, DEGs in the spleen were enriched in gas transport, oxygen transport, one-carbon compound transport, postsynaptic membrane assembly and erythrocyte development (Figure 6F).

Figure 6.

Figure 6

Open in a new tab

GO enrichment analysis of DEGs between HI and HT groups in different tissues: (A) brain; (B) intestine; (C) kidney; (D) liver; (E) gill; (F) spleen. Gene Ontology (GO) terms are categorized into three domains: BP (Biological Process), representing biological objectives or processes; CC (Cellular Component), referring to the cellular locations where gene products function; and MF (Molecular Function), describing the biochemical activities of gene products.

KEGG analysis of DEGs in different tissues was carried out (Figure 7). The enriched KEGG pathways were as follows: in the brain (Figure 7A), they included steroid biosynthesis, cytokine-cytokine receptor interaction, steroid hormone biosynthesis, amino sugar and nucleotide sugar metabolism, and the JAK-STAT signaling pathway; in the intestine (Figure 7B), they were proteasome, terpenoid backbone biosynthesis, aminoacyl-tRNA biosynthesis, steroid biosynthesis, and amino sugar and nucleotide sugar metabolism; in the kidney (Figure 7C), they were proteasome, hematopoietic cell lineage, protein processing in the endoplasmic reticulum, and porphyrin metabolism; in the liver (Figure 7D), they were Toll-like receptor signaling pathway, protein processing in the endoplasmic reticulum, IL-17 signaling pathway, Steroid biosynthesis, and Thyroid hormone signaling pathway; in the gill (Figure 7E), they were autoimmune thyroid disease, hematopoietic cell lineage, and viral myocarditis; in the spleen (Figure 7F), they were vasopressin-regulated water reabsorption, arachidonic acid metabolism, cell adhesion molecules, biosynthesis of unsaturated fatty acids, and alpha-Linolenic acid metabolism.

Figure 7.

Figure 7

Open in a new tab

KEGG analysis of DEGs between HI and HT groups in different tissues: (A) brain; (B) intestine; (C) kidney; (D) liver; (E) gill; (F) spleen.

3.6. Joint Analysis of GWAS and RNA-Seq

To identify candidate genes associated with hypoxia tolerance in grass carp, we determined 16 shared genes between the potential candidate genes identified by GWAS and DEGs obtained by RNA-seq analysis (Table 2). Eight, four, four, and two genes showed differential expression in the brain, kidney, intestine, and liver, respectively. Notably, among them, upstream transcription factor 1 (usf1) was annotated for its function in “response to hypoxia”, and transient receptor potential cation channel, subfamily V, member 4 (trpv4) for its “ATP binding” function.

Table 2.

Shared genes between the potential candidate genes identified by GWAS and DEGs obtained by RNA-seq analysis.

Gene Symbol Chromosome Position Gene Annotation Tissue Showing Differential Expression Gene Regulation
flo11 20 29,675,108–29,692,533 flocculation protein FLO11 Brain Down
ncs1a 5 31,515,563–31,537,983 neuronal calcium sensor 1a Brain Up
myh 5 31,409,911–31,463,184 myosin heavy chain, fast skeletal muscle-like Brain Down
ipo13b 20 29,625,934–29,674,126 importin 13b Brain; Kidney Up; Down
tubb2 2 4,049,561–4,054,417 tubulin, beta 2A class Iia Brain Down
snx17 20 1,844,003–1,958,074 sorting nexin-17 Brain Up
scara5 20 1,781,417–1,809,123 scavenger receptor class A, member 5 Brain Down
trpv4 5 33,489,408–33,514,568 transient receptor potential cation channel, subfamily V, member 4 Brain Down
usf1 15 26,021,801–26,026,293 upstream transcription factor 1 Intestine; Kidney Down; Down
LOC127501978 20 32,059,939–32,098,516 ncRNA Intestine Up
dio3b 20 32,118,645–32,121,485 iodothyronine deiodinase 3b Intestine Up
uchl5 10 8,839,438–8,845,469 ubiquitin carboxyl-terminal hydrolase L5 Intestine Down
tmod4 16 23,160,790–23,168,927 tropomodulin 4 Kidney Down
a2m 15 25,990,646–26,004,007 alpha-2-macroglobulin Kidney Up
cratb 19 7,391,651–7,439,445 carnitine O-acetyltransferase b Liver Up
rgs13b 10 8,824,407–8,827,219 regulator of G protein signaling 13b Liver Down
Open in a new tab

3.7. Validation of SNP Genotype Frequency by qPCR

Two statistically significant SNPs identified in the GWAS analysis—the SNP located at 33,445,030 bp on chromosome 5 (SNP7) and the SNP located at 36,269,337 bp on chromosome 10 (SNP11)—were selected for KASP genotyping. Both SNPs showed a strong association between genotype and hypoxia tolerance, with individuals carrying the GG genotype exhibiting significantly higher tolerance compared with those carrying AG or AA genotypes (Figure 8A,B). For both loci, GG/GG represented the superior genotype, whereas AG/AA corresponded to inferior hypoxia-tolerance performance (Figure 8C).

Figure 8.

Figure 8

Open in a new tab

The genotype frequency of the identified SNPs in the hypoxia intolerance (HI) and hypoxia tolerance (HT) groups. (A) The genotype frequencies of SNP7, the SNP at position 33,445,030 bp on chr 5. (B) The genotype frequencies of SNP11, the SNP at position 36,269,337 bp on chr 10. (C) The genotype combination frequencies for SNP7 and SNP11 (with the order being SNP7 followed by SNP11). Asterisk (*) indicates statistical significance between HI and HT groups using Chi-square test, * p < 0.05, *** p < 0.001.

4. Discussion

Given the widespread adoption of high-density aquaculture and the increasing prevalence of hypoxic waters globally [34,35,36], aquatic hypoxia has become a pervasive challenge for various fish populations, including marine fish, freshwater fish, aquaculture fish, and wild fish, and dramatically impacts fisheries and aquaculture worldwide. To date, there are no GWAS researches on hypoxia tolerance in grass carp. For the first time, we used GWAS to identify SNPs and InDels associated with hypoxia tolerance in grass carp. Considering that the population structure is characterized by non-random mating and multiple breeding families, we took this into account during the analysis to ensure the reliability of the associations. In combination with transcriptome analysis, potential candidate genes were further identified. This study provided novel insights into the adaptive strategies of grass carp under hypoxia stress and offered a theoretical basis and genetic markers for molecular breeding.

The time to LOE, a widely used indicator of hypoxia tolerance in fish [12,37,38], showed a striking difference between HT and HI groups of grass carp, with average values of 23.62 h and 3.49 h, respectively. This marked variation suggested the presence of heritable genetic components underlying hypoxia tolerance in grass carp, consistent with observations in other fish species such as channel catfish [12,20] and hybrid catfish [11]. Such phenotypic divergence provides a solid foundation for GWAS and genomic selecting breeding.

From the GWAS results, 21 SNPs and six InDels associated with hypoxia tolerance, distributed across 12 chromosomes were identified, indicating that hypoxia tolerance trait in grass carp are regulated by micro-efficient multiple genetic loci. The low individual phenotypic variance explained (PVE: 4.8% and 2.7%) by the identified two significant SNPs aligns with GWAS findings for the polygenic architecture of complex traits, where minor-effect loci collectively drive phenotypic variation. This reflects the trait’s inherent polygenicity, alongside potential limitations in variation coverage or sample size that limit detection of small-effect loci. Notably, these SNPs remain biologically relevant, as they were localized to hypoxia-responsive genes and they are potentially correlated with gene expression. Future studies with larger samples or multi-omics integration will help uncover additional loci and enhance cumulative PVE, advancing understanding of the trait’s genetic basis. Overlapping QTL regions on chromosomes 2 and 10 were detected for both SNPs and InDels, suggesting these regions may harbor conserved genetic elements regulating hypoxia responses. The candidate genes within ± 50 kb of these variants were enriched in pathways such as steroid biosynthesis, insulin signaling, and glycosphingolipid biosynthesis. Steroid biosynthesis, in particular, was consistently highlighted in both GWAS and transcriptome analyses. Previous researches have already demonstrated that hypoxia affects steroid biosynthesis [39,40]. Steroids, including corticosteroids, are known to modulate metabolic and stress responses in fish [41]. Hypoxia-inducible factor 1 (HIF1) is an oxygen-regulated transcriptional activator [42], and plays a significant role in the production of steroids [39].

Transcriptome profiling revealed distinct DEG patterns across six tissues, reflecting specialized roles in hypoxia adaptation. The most DEGs (1620) were identified in the brain, enriched in steroid metabolic processes and angiogenesis regulation. Steroids may modulate neural activity and energy metabolism in the brain, a tissue highly sensitive to oxygen deprivation. Enhanced angiogenesis (e.g., via upregulated genes in “positive regulation of blood vessel development”) could improve oxygen delivery to neurons, a strategy observed in hypoxia-tolerant fish [43]. DEGs in these six tissues were enriched in oxygen transport and erythrocyte development, indicating enhanced hematopoiesis and oxygen-carrying capacity in HT fish. This aligns with the well-known role of the HIF pathway in promoting erythropoiesis [44] and suggests that HT fish may optimize oxygen transport to tissues. Enrichment of steroid biosynthesis and proteasome pathways in the liver and intestine highlights their roles in metabolic adjustment and protein turnover. The liver, a central metabolic organ, may shift to anaerobic metabolism under hypoxia [45].

The PPI network analysis of the GWAS-identified genes revealed 14 key genes, such as cdipt and uchl5. Among these genes, cdipt, which encodes a key enzyme in phosphatidylinositol synthesis, was significantly reduced in mice after 12 h and 24 h of hypobaric hypoxia exposure [46]. This finding demonstrates the involvement of lipid metabolism in the hypoxic stress response and highlights the critical role of this gene. Furthermore, its down-regulation in the liver of common carp (Cyprinus carpio L.) under high-temperature stress suggested a conserved role for cdipt in responding to diverse environmental stresses [47]. This study found that HIF-1α transcriptionally activates uchl5 expression [48]. Given that HIF1 is an oxygen-regulated transcription activator, this result demonstrated that uchl5 has a specific function in hypoxic stress. Collectively, these results provide a foundation for further studies on the roles of these genes in environmental stress responses in fish.

The 16 genes shared between GWAS candidates and DEGs represent high-priority targets for functional validation in subsequent studies. The functional association between the validated SNPs and gene expression remains elusive. Demonstrating causality would require further verification, such as CRISPR/Cas9 editing, transgenic overexpression, or knockdown studies in grass carp. Among these, usf1 was downregulated in the intestine and kidney of HT fish. Previous studies have demonstrated that USF1 is a known regulator of metabolic genes [49], and plays important roles in response to hypoxia [50,51,52]. Jiang et al. found that increased protein levels and DNA binding of USF1 mediate the inhibition of CYP19 gene expression in the human placenta by hypoxia and Mash-2 [51,52]. In other studies of fish undergoing hypoxia stress, the expression or methylation changes in usf1 were also detected [53,54]. The downregulation of usf1 in this study may facilitate metabolic reprogramming under hypoxia. trpv4 was downregulated in the brain of HT fish, which mediates calcium signaling and has been linked to oxygen sensing in vertebrates [55,56,57]. The alpha-2-macroglobulin (a2m), a high-molecular weight homotetrameric glycoprotein [58], was upregulated in the kidney of HT fish. In studies of vertebrates, A2M-AS1 is downregulated in cardiomyocytes and exerts protective effects against hypoxia/reoxygenation-induced cardiomyocyte injury by suppressing apoptosis and oxidative stress [59,60]. Functional validation of these genes in fish models is necessary to elucidate their precise roles in hypoxia tolerance.

5. Conclusions

This study integrated GWAS and multi-tissue transcriptome profiling to explore the genetic basis of hypoxia tolerance in grass carp. A total of 21 SNPs and 6 InDels associated with hypoxia tolerance were identified, with 2 SNPs on chromosome 10 and 13 reaching genome-wide significance. Candidate genes within these variants were enriched in pathways crucial for metabolic adaptation, including steroid biosynthesis and insulin signaling. Transcriptome analysis revealed distinct tissue-specific differential gene expression patterns under hypoxic stress, and DEGs were enriched in processes related to steroid biosynthesis, oxygen transport, and erythrocyte development. Integration of GWAS and transcriptome data highlighted 16 shared genes, notably usf1 and trpv4, as key players in hypoxia tolerance. These findings provide valuable genomic resources and molecular markers for selective breeding programs aimed at enhancing hypoxia tolerance in grass carp. We will further validate the function of candidate genes and explore their potential application in molecular breeding programs.

Supplementary Materials

The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/ani15243518/s1, Figure S1: Population structure analyses using InDels of grass carp. A. PCA analysis. B. Heatmap of genomic relatedness; Figure S2: Q-Q plots of grass carp subjected to hypoxia treatment. A. SNPs. B. InDels; Figure S3: The number of differentially expressed genes in the brain, intestine, kidney, liver, gill, and spleen between the HI and HT groups; Table S1: Primers used in the KASP analysis and qPCR experiments; Table S2: The candidate genes associated with significant and suggestive SNPs; Table S3: The candidate genes associated with suggestive InDels; Table S4: Summary of RNA-seq sequencing.

animals-15-03518-s001.zip (469.1KB, zip)

Author Contributions

Conceptualization, W.W., M.C. and Z.S.; Methodology, W.W. and M.C.; Software, M.C.; Investigation, W.W., Y.H., X.R., H.X., L.G., X.C., Y.W. and Q.L.; Resources, W.W. and M.C.; Data curation, M.C.; Writing—original draft, W.W., M.C. and S.T.; Writing—review and editing, W.W., M.C., S.T., Y.H., X.R., H.X., L.G., X.C., Y.W., Q.L. and Z.S.; Validation, W.W., X.R., H.X., L.G., X.C., Y.W. and Q.L.; Supervision, Z.S.; Funding acquisition, Z.S. All authors have read and agreed to the published version of the manuscript.

Institutional Review Board Statement

All animal protocols used in our experiments were approved by the Ethics Committee of the Medical College of Qingdao University, China (QDU-AEC-2024004), approved on 16 January 2024.

Informed Consent Statement

Not applicable.

Data Availability Statement

The RNA-seq raw data have been submitted to the NCBI SRA under BioProject ID PRJNA1306682, which is accessible via the provided link: https://www.ncbi.nlm.nih.gov/bioproject/?term=PRJNA1306682, 15 August 2025. The raw genome resequencing data and phenotypic data may be made available upon request by contacting the corresponding author.

Conflicts of Interest

W.W., M.C., S.T., Y.H., X.R., H.X., L.G., and Z.S. declared that there are no conflicts of interest. X.C., Y.W., and Q.L., who are employees of Weishan County Nansihu Fishery Company, confirmed that there are no conflicts of interest.

Funding Statement

This work was supported by the Biological Breeding-Major Projects (2023ZD0405501) and Taishan Scholar Youth Project of Shandong Province, China.

Footnotes

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

References

  • 1.Diaz R.J., Breitburg D.L. The hypoxic environment. Fish Physiol. 2009;27:1–23. [Google Scholar]
  • 2.Robertson C.E., Wright P.A., Köblitz L., Bernier N.J. Hypoxia-inducible factor-1 mediates adaptive developmental plasticity of hypoxia tolerance in zebrafish, Danio rerio. Proc. Biol. Sci. 2014;281:20140637. doi: 10.1098/rspb.2014.0637. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Galic N., Hawkins T., Forbes V.E. Adverse impacts of hypoxia on aquatic invertebrates: A meta-analysis. Sci. Total Environ. 2019;652:736–743. doi: 10.1016/j.scitotenv.2018.10.225. [DOI] [PubMed] [Google Scholar]
  • 4.Richards J.G. Physiological, behavioral and biochemical adaptations of intertidal fishes to hypoxia. J. Exp. Biol. 2011;214:191–199. doi: 10.1242/jeb.047951. [DOI] [PubMed] [Google Scholar]
  • 5.Nikinmaa M., Rees B.B. Oxygen-dependent gene expression in fishes. Am. J. Physiol. Regul. Integr. Comp. Physiol. 2005;288:R1079–R1090. doi: 10.1152/ajpregu.00626.2004. [DOI] [PubMed] [Google Scholar]
  • 6.Virani N.A., Rees B.B. Oxygen consumption, blood lactate and inter-individual variation in the gulf killifish, Fundulus grandis, during hypoxia and recovery. Comp. Biochem. Physiol. A. 2000;126:397–405. doi: 10.1016/S1095-6433(00)00219-1. [DOI] [PubMed] [Google Scholar]
  • 7.Wannamaker C.M., Rice J.A. Effects of hypoxia on movements and behavior of selected estuarine organisms from the southeastern United States. J. Exp. Mar. Biol. Ecol. 2000;249:145–163. doi: 10.1016/S0022-0981(00)00160-X. [DOI] [PubMed] [Google Scholar]
  • 8.Liu C., Wen H., Zhang C., Zhang Y., Wang L., Sun D., Liu M., Dong Y., Li P., Li H. Deciphering the genetic basis and genomic prediction of heat tolerance trait from whole-genome resequencing in spotted sea bass (Lateolabrax maculatus) Aquaculture. 2025;598:741951. doi: 10.1016/j.aquaculture.2024.741951. [DOI] [Google Scholar]
  • 9.Yanez J.M., Barria A., Lopez M.E., Moen T., Garcia B.F., Yoshida G.M., Xu P. Genome-wide association and genomic selection in aquaculture. Rev. Aquac. 2023;15:645–675. doi: 10.1111/raq.12750. [DOI] [Google Scholar]
  • 10.Ding J., Gao Z., Wang J., Zhang Y., Wang X., Wu X., Zhu J., Shen W. Genome-wide association and transcriptome analysis provide the SNPs and molecular insights into the hypoxia tolerance in large yellow croaker (Larimichthys crocea) Aquaculture. 2023;573:739547. doi: 10.1016/j.aquaculture.2023.739547. [DOI] [Google Scholar]
  • 11.Zhong X., Wang X., Zhou T., Jin Y., Tan S., Jiang C., Geng X., Li N., Shi H., Zeng Q., et al. Genome-wide association study reveals multiple novel QTL associated with low oxygen tolerance in hybrid catfish. Mar. Biotechnol. 2017;19:379–390. doi: 10.1007/s10126-017-9757-5. [DOI] [PubMed] [Google Scholar]
  • 12.Wang X., Liu S., Jiang C., Geng X., Zhou T., Li N., Bao L., Li Y., Yao J., Yang Y., et al. Multiple across-strain and within-strain QTLs suggest highly complex genetic architecture for hypoxia tolerance in channel catfish. Mol. Genet. Genom. 2017;292:63–76. doi: 10.1007/s00438-016-1256-2. [DOI] [PubMed] [Google Scholar]
  • 13.San L.Z., Liu B.S., Liu B., Zhu K.C., Guo L., Guo H.Y., Zhang N., Jiang S.G., Zhang D.C. Genome-wide association study reveals multiple novel SNPs and putative candidate genes associated with low oxygen tolerance in golden pompano Trachinotus ovatus (Linnaeus 1758) Aquaculture. 2021;544:737098. doi: 10.1016/j.aquaculture.2021.737098. [DOI] [Google Scholar]
  • 14.Lin S., Milardi M., Gao Y., Wong M.H. Sustainable management of non-native grass carp as a protein source, weed-control agent and sport fish. Aquac. Res. 2022;53:5809–5824. doi: 10.1111/are.16080. [DOI] [Google Scholar]
  • 15.Yang C., Li Q., Su J., Chen X., Wang Y., Peng L. Identification and functional characterizations of a novel TRIF gene from grass carp (Ctenopharyngodon idella) Dev. Comp. Immunol. 2013;41:222–229. doi: 10.1016/j.dci.2013.05.018. [DOI] [PubMed] [Google Scholar]
  • 16.Yu C., Jiang Y., Zhang C., Wu M., Gui L., Xu X., Li J., Shen Y. Whole-genome resequencing of grass carp (Ctenopharyngodon idella) for genome-wide association study on GCRV resistance. Aquaculture. 2024;592:741243. doi: 10.1016/j.aquaculture.2024.741243. [DOI] [Google Scholar]
  • 17.Hao Y., Jia X., Yuan L., Liu Y., Gui L., Shen Y., Li J., Xu X. Genome-wide association study reveals growth-related SNPs and candidate genes in grass carp (Ctenopharyngodon idella) Aquaculture. 2023;577:739979. doi: 10.1016/j.aquaculture.2023.739979. [DOI] [Google Scholar]
  • 18.Gao Y.J., Liu Y.J., Chen X.Q., Yang H.J., Li X.F., Tian L.X. Effects of graded levels of histidine on growth performance, digested enzymes activities, erythrocyte osmotic fragility and hypoxia-tolerance of juvenile grass carp Ctenopharyngodon idella. Aquaculture. 2016;452:388–394. doi: 10.1016/j.aquaculture.2015.11.019. [DOI] [Google Scholar]
  • 19.Melnychuk M.C., Chapman L.J. Hypoxia tolerance of two haplochromine cichlids: Swamp leakage and potential for interlacustrine dispersal. Environ. Biol. Fishes. 2002;65:99–110. doi: 10.1023/A:1019602403098. [DOI] [Google Scholar]
  • 20.Wang X., Liu S., Dunham R., Liu Z. Effects of strain and body weight on low-oxygen tolerance of channel catfish (Ictalurus punctatus) Aquac. Int. 2017;25:1645–1652. doi: 10.1007/s10499-017-0125-2. [DOI] [Google Scholar]
  • 21.Chen S., Zhou Y., Chen Y., Gu J. fastp: An ultra-fast all-in-one FASTQ preprocessor. Bioinformatics. 2018;34:i884–i890. doi: 10.1093/bioinformatics/bty560. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Wu C.S., Ma Z.Y., Zheng G.D., Zou S.M., Zhang X.J., Zhang Y.A. Chromosome-level genome assembly of grass carp (Ctenopharyngodon idella) provides insights into its genome evolution. BMC Genom. 2022;23:271. doi: 10.1186/s12864-022-08503-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Li H. Aligning sequence reads, clone sequences and assembly contigs with BWA-MEM. arXiv. 2013 doi: 10.48550/arXiv.1303.3997.1303.3997 [DOI] [Google Scholar]
  • 24.McKenna A., Hanna M., Banks E., Sivachenko A., Cibulskis K., Kernytsky A., Garimella K., Altshuler D., Gabriel S., Daly M., et al. The Genome Analysis Toolkit: A MapReduce framework for analyzing next-generation DNA sequencing data. Genome Res. 2010;20:1297–1303. doi: 10.1101/gr.107524.110. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Purcell S., Neale B., Todd-Brown K., Thomas L., Ferreira M.A., Bender D., Maller J., Sklar P., De Bakker P.I., Daly M., et al. PLINK: A tool set for whole-genome association and population-based linkage analyses. Am. J. Hum. Genet. 2007;81:559–575. doi: 10.1086/519795. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Zhang C., Dong S.S., Xu J.Y., He W.M., Yang T.L. PopLDdecay: A fast and effective tool for linkage disequilibrium decay analysis based on variant call format files. Bioinformatics. 2019;35:1786–1788. doi: 10.1093/bioinformatics/bty875. [DOI] [PubMed] [Google Scholar]
  • 27.Alexander D.H., Novembre J., Lange K. Fast model-based estimation of ancestry in unrelated individuals. Genome Res. 2009;19:1655–1664. doi: 10.1101/gr.094052.109. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Zhou X., Stephens M. Genome-wide efficient mixed-model analysis for association studies. Nat. Genet. 2012;44:821–824. doi: 10.1038/ng.2310. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Cingolani P., Platts A., Wang L.L., Coon M., Nguyen T., Wang L., Land S.J., Lu X., Ruden D.M. A program for annotating and predicting the effects of single nucleotide polymorphisms, SnpEff: SNPs in the genome of Drosophila melanogaster strain w1118; iso-2; iso-3. Fly. 2012;6:80–92. doi: 10.4161/fly.19695. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Wu T., Hu E., Xu S., Chen M., Guo P., Dai Z., Feng T., Zhou L., Tang W., Zhan L., et al. clusterProfiler 4.0: A universal enrichment tool for interpreting omics data. Innovation. 2021;2:100141. doi: 10.1016/j.xinn.2021.100141. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Kim D., Paggi J.M., Park C., Bennett C., Salzberg S.L. Graph-based genome alignment and genotyping with HISAT2 and HISAT-genotype. Nat. Biotechnol. 2019;37:907–915. doi: 10.1038/s41587-019-0201-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Liao Y., Smyth G.K., Shi W. featureCounts: An efficient general purpose program for assigning sequence reads to genomic features. Bioinformatics. 2014;30:923–930. doi: 10.1093/bioinformatics/btt656. [DOI] [PubMed] [Google Scholar]
  • 33.Love M.I., Huber W., Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 2014;15:550. doi: 10.1186/s13059-014-0550-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Dipta B., Sood S., Mangal V., Bhardwaj V., Thakur A.K., Kumar V., Singh B. KASP: A high-throughput genotyping system and its applications in major crop plants for biotic and abiotic stress tolerance. Mol. Bio. Rep. 2024;51:508. doi: 10.1007/s11033-024-09455-z. [DOI] [PubMed] [Google Scholar]
  • 35.Pezner A.K., Courtney T.A., Barkley H.C., Chou W.-C., Chu H.-C., Clements S.M., Cyronak T., DeGrandpre M.D., Kekuewa S.A.H., Kline D.I., et al. Increasing hypoxia on global coral reefs under ocean warming. Nat. Clim. Change. 2023;13:403–409. doi: 10.1038/s41558-023-01619-2. [DOI] [Google Scholar]
  • 36.Tellier J.M., Kalejs N.I., Leonhardt B.S., Cannon D., Höök T.O., Collingsworth P.D. Widespread prevalence of hypoxia and the classification of hypoxic conditions in the Laurentian Great Lakes. J. Great Lakes Res. 2022;48:13–23. doi: 10.1016/j.jglr.2021.11.004. [DOI] [Google Scholar]
  • 37.Speers-Roesch B., Mandic M., Groom D.J., Richards J.G. Critical oxygen tensions as predictors of hypoxia tolerance and tissue metabolic responses during hypoxia exposure in fishes. J. Exp. Mar. Biol. Ecol. 2013;449:239–249. doi: 10.1016/j.jembe.2013.10.006. [DOI] [Google Scholar]
  • 38.Barnes R., King H., Carter C.G. Hypoxia tolerance and oxygen regulation in Atlantic salmon, Salmo salar from a Tasmanian population. Aquaculture. 2011;318:397–401. doi: 10.1016/j.aquaculture.2011.06.003. [DOI] [Google Scholar]
  • 39.Ariyeloye S., Watts D., Jaykar M.T., Ermis C., Krüger A., Kaden D., Stepien B.K., Alexaki V.I., Peitzsch M., Bechmann N., et al. HIF1α controls steroidogenesis under acute hypoxic stress. Cell Commun. Signal. 2025;23:86. doi: 10.1186/s12964-025-02080-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Ariyeloye S., Kämmerer S., Klapproth E., Wielockx B., El-Armouche A. Intertwined regulators: Hypoxia pathway proteins, microRNAs, and phosphodiesterases in the control of steroidogenesis. Pflug. Arch. Eur. J. Physiol. 2024;476:1383–1398. doi: 10.1007/s00424-024-02921-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Barton B.A. Stress in fishes: A diversity of responses with particular reference to changes in circulating corticosteroids. Integr. Comp. Biol. 2002;42:517–525. doi: 10.1093/icb/42.3.517. [DOI] [PubMed] [Google Scholar]
  • 42.Semenza G.L. HIF-1 and mechanisms of hypoxia sensing. Curr. Opin. Cell Biol. 2001;13:167–171. doi: 10.1016/S0955-0674(00)00194-0. [DOI] [PubMed] [Google Scholar]
  • 43.Gerri C., Marín-Juez R., Marass M., Marks A., Maischein H.M., Stainier D.Y.R. Hif-1α regulates macrophage-endothelial interactions during blood vessel development in zebrafish. Nat. Commun. 2017;8:15492. doi: 10.1038/ncomms15492. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Haase V.H. Regulation of erythropoiesis by hypoxia-inducible factors. Blood Rev. 2013;27:41–53. doi: 10.1016/j.blre.2012.12.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Ojaghi M., Pamenter M.E. Severe acute hypoxia upregulates anaerobic metabolism in non-reproductive but not queen naked mole-rats. J. Exp. Biol. 2025;228:jeb250397. doi: 10.1242/jeb.250397. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Gao J., Guo Z., Zhao M., Cheng X., Jiang X., Liu Y., Zhang W., Yue X., Fei X., Jiang Y., et al. Lipidomics and mass spectrometry imaging unveil alterations in mice hippocampus lipid composition exposed to hypoxia. J. Lipid. Res. 2024;65:100575. doi: 10.1016/j.jlr.2024.100575. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.Sun J., Zhao L., Wu H., Lian W., Cui C., Du Z., Luo W., Li M., Yang S. Analysis of miRNA-seq in the liver of common carp (Cyprinus carpio L.) in response to different environmental temperatures. Funct. Integr. Genom. 2019;19:265–280. doi: 10.1007/s10142-018-0643-7. [DOI] [PubMed] [Google Scholar]
  • 48.Jiao S., Gao Q., Da F., Guo L., Wang J., Guo J., Li J., Liu J. Hypoxia-induced UCHL5 expression enhanced radioresistance of Hela cells in cervical cancer. Int. J. Radiat. Med. Nucl. Med. 2021;45:31–40. [Google Scholar]
  • 49.Laurila P.-P., Soronen J., Kooijman S., Forsström S., Boon M.R., Surakka I., Kaiharju E., Coomans C.P., Van den Berg S.A., Autio A., et al. USF1 deficiency activates brown adipose tissue and improves cardiometabolic health. Sci. Transl. Med. 2016;8:ra313–ra323. doi: 10.1126/scitranslmed.aad0015. [DOI] [PubMed] [Google Scholar]
  • 50.Hu J., Stiehl D.P., Setzer C., Wichmann D., Shinde D.A., Rehrauer H., Hradecky P., Gassmann M., Gorr T.A. Interaction of HIF and USF signaling pathways in human genes flanked by hypoxia-response elements and E-box palindromes. Mol. Cancer Res. 2011;9:1520–1536. doi: 10.1158/1541-7786.MCR-11-0090. [DOI] [PubMed] [Google Scholar]
  • 51.Jiang B., Mendelson C.R. O2 enhancement of human trophoblast differentiation and hCYP19 (aromatase) gene expression are mediated by proteasomal degradation of USF1 and USF2. Mol. Cel. Biol. 2005;25:8824–8833. doi: 10.1128/MCB.25.20.8824-8833.2005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 52.Jiang B., Mendelson C.R. USF1 and USF2 mediate inhibition of human trophoblast differentiation and CYP19 gene expression by Mash-2 and hypoxia. Mol. Cel. Biol. 2003;23:6117–6128. doi: 10.1128/MCB.23.17.6117-6128.2003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 53.Liu B., Wen H., Yang J., Li X., Li G., Zhang J., Wu S., Butts I.A., He F. Hypoxia affects HIF-1/LDH-A signaling pathway by methylation modification and transcriptional regulation in Japanese flounder (Paralichthys olivaceus) Biology. 2022;11:1233. doi: 10.3390/biology11081233. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 54.Beemelmanns A., Ribas L., Anastasiadi D., Moraleda-Prados J., Zanuzzo F.S., Rise M.L., Gamperl A.K. DNA methylation dynamics in Atlantic salmon (Salmo salar) challenged with high temperature and moderate hypoxia. Front. Mar. Sci. 2021;7:604878. doi: 10.3389/fmars.2020.604878. [DOI] [Google Scholar]
  • 55.Parpaite T., Cardouat G., Mauroux M., Gillibert-Duplantier J., Robillard P., Quignard J.-F., Marthan R., Savineau J.-P., Ducret T. Effect of hypoxia on TRPV1 and TRPV4 channels in rat pulmonary arterial smooth muscle cells. Pflug. Arch.-Eur. J. Physiol. 2016;468:111–130. doi: 10.1007/s00424-015-1704-6. [DOI] [PubMed] [Google Scholar]
  • 56.Yang X.-R., Lin A.H., Hughes J.M., Flavahan N.A., Cao Y.-N., Liedtke W., Sham J.S. Upregulation of osmo-mechanosensitive TRPV4 channel facilitates chronic hypoxia-induced myogenic tone and pulmonary hypertension. Am. J. Physiol.-Lung C. 2012;302:L555–L568. doi: 10.1152/ajplung.00005.2011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 57.Butenko O., Dzamba D., Benesova J., Honsa P., Benfenati V., Rusnakova V., Ferroni S., Anderova M. The increased activity of TRPV4 channel in the astrocytes of the adult rat hippocampus after cerebral hypoxia/ischemia. PLoS ONE. 2012;7:e39959. doi: 10.1371/journal.pone.0039959. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 58.Rehman A.A., Ahsan H., Khan F.H. alpha-2-Macroglobulin: A physiological guardian. J. Cell Physiol. 2013;228:1665–1675. doi: 10.1002/jcp.24266. [DOI] [PubMed] [Google Scholar]
  • 59.Yu H., Pan Y., Dai M., Wang X., Chen H. Mesenchymal stem cell–originated exosomal Lnc A2M-AS1 alleviates hypoxia/reperfusion-induced apoptosis and oxidative stress in cardiomyocytes. Cardiovasc. Drugs Ther. 2023;37:891–904. doi: 10.1007/s10557-022-07339-7. [DOI] [PubMed] [Google Scholar]
  • 60.Song X.L., Zhang F.F., Wang W.J., Li X.N., Dang Y., Li Y.X., Yang Q., Shi M.J., Qi X.Y. LncRNA A2M-AS1 lessens the injury of cardiomyocytes caused by hypoxia and reoxygenation via regulating IL1R2. Genes Genom. 2020;42:1431–1441. doi: 10.1007/s13258-020-01007-6. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

animals-15-03518-s001.zip (469.1KB, zip)

Data Availability Statement

The RNA-seq raw data have been submitted to the NCBI SRA under BioProject ID PRJNA1306682, which is accessible via the provided link: https://www.ncbi.nlm.nih.gov/bioproject/?term=PRJNA1306682, 15 August 2025. The raw genome resequencing data and phenotypic data may be made available upon request by contacting the corresponding author.

Articles from Animals : an Open Access Journal from MDPI are provided here courtesy of Multidisciplinary Digital Publishing Institute (MDPI)

RESOURCES