Extreme-Phenotype Genome-Wide Association Analysis for Growth Traits in Spotted Sea Bass (Lateolabrax maculatus) Using Whole-Genome Resequencing

Abstract

Spotted sea bass (Lateolabrax maculatus) is an important marine economic fish in China, ranking third in annual production among marine fish. However, a declined growth rate caused by germplasm degradation has severely increased production costs and reduced economic benefits. There is an urgent need to develop the fast-growing varieties of L. maculatus and elucidate the genetic mechanisms underlying growth traits. Here, whole-genome resequencing technology combined with extreme phenotype genome-wide association analysis (XP-GWAS) was used to identify candidate markers and genes associated with growth traits in L. maculatus. Two groups of L. maculatus, consisting of 100 fast-growing and 100 slow-growing individuals with significant differences in body weight, body length, and carcass weight, underwent whole-genome resequencing. A total of 4,528,936 high-quality single nucleotide polymorphisms (SNPs) were used for XP-GWAS. These SNPs were evenly distributed across all chromosomes without large gaps, and the average distance between SNPs was only 175.8 bp. XP-GWAS based on the Bayesian-information and Linkage-disequilibrium Iteratively Nested Keyway (Blink) and Fixed and random model Circulating Probability Unification (FarmCPU) identified 50 growth-related markers, of which 17 were related to body length, 19 to body weight, and 23 to carcass weight. The highest phenotypic variance explained (PVE) reached 15.82%. Furthermore, significant differences were observed in body weight, body length, and carcass weight among individuals with different genotypes. For example, there were highly significant differences in body weight among individuals with different genotypes for four SNPs located on chromosome 16: chr16:13133726, chr16:13209537, chr16:14468078, and chr16:18537358. Additionally, 47 growth-associated genes were annotated. These genes are mainly related to the metabolism of energy, glucose, and lipids and the development of musculoskeletal and nervous systems, which may regulate the growth of L. maculatus. Our study identified growth-related markers and candidate genes, which will help to develop the fast-growing varieties of L. maculatus through marker-assisted breeding and elucidate the genetic mechanisms underlying the growth traits.

Keywords: Lateolabrax maculatus; genome-wide association analysis; growth traits; whole genome resequencing.

Figure 1

The comparison of growth traits between the fast-growing and slow-growing L. maculatus and the distribution of the SNPs used for GWAS on the chromosomes. (A–C) The comparison of body weight (A), body length (B) and carcass weight (C) between the 100 fast-growing L. maculatus and the 100 slow-growing ones. (D) The distribution of SNPs used for GWAS on the chromosomes. The redder the color, the more SNPs within a 1 Mb region; the bluer the color, the fewer SNPs. **** refers to p value < 0.0001.

Figure 2

Q–Q plot for body weight, body length, and carcass weight based on the Blink and FarmCPU model. (A) Quantile–quantile plot for body weight based on the Blink model. (B) Quantile–quantile plot for body length based on the Blink model. (C) Quantile–quantile plot for carcass weight based on the Blink model. (D) Quantile–quantile plot for body weight based on the FarmCPU model. (E) Quantile–quantile plot for body length based on the FarmCPU model. (F) Quantile–quantile plot for carcass weight based on the FarmCPU model.

Figure 3

Manhattan plot for the GWAS of body weight, body length, and carcass weight in L. maculatus. (A,B) Manhattan plot of GWAS for body weight based on the Blink (A) and FarmCPU (B) models. (C,D) Manhattan plot of GWAS for body length based on the Blink (C) and FarmCPU (D) models. (E,F) Manhattan plot of GWAS for carcass weight based on the Blink (E) and FarmCPU (F) models.

Figure 4

Boxplot of body weight for different genotypes of L. maculatus. The y-axis represents the body weight of the individuals with different genotypes, and different colors indicate different genotypes, with the “n” representing the number of individuals for each genotype. (A) Body weight of individuals with different genotypes for the SNP on chr16:13133726; (B) Body weight of individuals with different genotypes for the SNP on chr16:13209537; (C) Body weight of individuals with different genotypes for the SNP on chr16:14468078; (D) Body weight of individuals with different genotypes for the SNP on chr16:18537358.

Figure 5

Protein–protein interaction network and tissue expression profile analysis of the growth-related candidate genes. (A) The protein–protein interaction network of the growth-related candidate genes. (B–F) Tissue expression profile analysis of PTPRA (B), SLC7A8 (C), PARK2 (D), ZNF436 (E), and SORCS2 (F).