Genome-Wide Association Analysis of Heat Tolerance in F2 Progeny from the Hybridization between Two Congeneric Oyster Species

Abstract

As the world's largest farmed marine animal, oysters have enormous economic and ecological value. However, mass summer mortality caused by high temperature poses a significant threat to the oyster industry. To investigate the molecular mechanisms underlying heat adaptation and improve the heat tolerance ability in the oyster, we conducted genome-wide association analysis (GWAS) analysis on the F₂ generation derived from the hybridization of relatively heat-tolerant Crassostrea angulata ♀ and heat-sensitive Crassostrea gigas ♂, which are the dominant cultured species in southern and northern China, respectively. Acute heat stress experiment (semi-lethal temperature 42 °C) demonstrated that the F₂ population showed differentiation in heat tolerance, leading to extremely differentiated individuals (approximately 20% of individuals die within the first four days with 10% survival after 14 days). Genome resequencing and GWAS of the two divergent groups had identified 18 significant SNPs associated with heat tolerance, with 26 candidate genes located near these SNPs. Eleven candidate genes that may associate with the thermal resistance were identified, which were classified into five categories: temperature sensor (Trpm2), transcriptional factor (Gata3), protein ubiquitination (Ube2h, Usp50, Uchl3), heat shock subfamily (Dnajc17, Dnaja1), and transporters (Slc16a9, Slc16a14, Slc16a9, Slc16a2). The expressional differentiation of the above genes between C. gigas and C. angulata under sublethal temperature (37 °C) further supports their crucial role in coping with high temperature. Our results will contribute to understanding the molecular mechanisms underlying heat tolerance, and provide genetic markers for heat-resistance breeding in the oyster industry.

Keywords: Crassostrea angulata; Crassostrea gigas; F2 progeny; GWAS; heat tolerance; oysters.

Figure 1

The plot of the daily number of deaths during heat stress and the Circos plot of an overview of the overall analysis results. (a) The plot of the daily number of deaths during heat stress. Each bar represents the number of deaths per day. The blue bar represents the “heat-sensitive (HS)” group, while red bar represents the “heat-tolerant (HT)” group. (b) The Circos plot of an overview of the overall analysis results. The genome information at different levels was represented using several circles, with each circle representing a specific aspect. Starting from the outermost circle, the length of each chromosome is depicted. The frequency histogram represents the gene density in different regions of the chromosome, and the scatter plot depicts the SNP density in different regions of the chromosome. The innermost circles depict different types of structural variations, with the orange line representing DUP (tandem duplication), the purple line representing DEL (large deletion), the green line representing INS (insertion), and the blue line representing INV (inversion). The connecting lines in the inner circle represent the BND (chromosome translocation) structural variation and its location on the chromosome.

Figure 2

Population structure and linkage disequilibrium (LD) decay analysis. (a) LD decay. The x-axis represents the physical distance between SNPs, while the y-axis shows the average r² values for each distance bin. (b) Principal components analysis (PCA). The blue and red dots represent individuals from the heat-sensitive (HS) and heat-tolerant (HT) groups, respectively. (c) Kinship among the resequencing samples. The color-coded matrix shows the pairwise kinship coefficients, with warmer colors indicating higher levels of relatedness.

Figure 3

The Manhattan and quantile–quantile (QQ) plots from GWAS analysis. (a) Manhattan plots. The red dashed line represents the genome-wide significant threshold (−log₁₀(1 × 10⁻⁶) = 6). (b) Quantile–quantile (QQ) plots. The red line represents the expected distribution of −log₁₀(1 × 10⁻⁶) values under the null hypothesis, while the deviations from this line reflect the p-value from which the regression of the phenotype on the SNP deviates from what is expected by chance.

Figure 4

The relative expression radio of candidate genes in heat stress experiment (37 °C) of C. gigas and C. angulata. The light blue and blue bars represent the gene expression levels of C. gigas under basal (0 °C) and heat stress (37 °C) conditions, respectively. The light red and red bars represent gene expression levels of C. angulata. The error bars represent the SD. Significant differences among groups were marked with * p < 0.05, ** p < 0.01, **** p < 0.0001.