Genome-wide association study for identifying loci that affect fillet yield, carcass, and body weight traits in rainbow trout (Oncorhynchus mykiss)

Authors: GONAZLEZ-PENA, DIANELYS - Orise Fellow; Gao, Guangtu; BARANSKI, MATTHEW - Nofima; MOEN, THOMAS - Aquagen; Cleveland, Beth; KENNEY, BRETT - West Virginia University; Vallejo, Roger; Palti, Yniv; Leeds, Timothy - Tim

Submitted to: Frontiers in Genetics
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 11/2/2016
Publication Date: 11/22/2016
Publication URL: https://handle.nal.usda.gov/10113/63324
Citation: Gonazlez-Pena, D., Gao, G., Baranski, M., Moen, T., Cleveland, B.M., Kenney, B., Vallejo, R.L., Palti, Y., Leeds, T.D. 2016. Genome-wide association study for identifying loci that affect fillet yield, carcass, and body weight traits in rainbow trout (Oncorhynchus mykiss). Frontiers in Genetics. 7(203):1-14. doi:10.3389/fgen.2016.00203.

Interpretive Summary: Increasing the proportion of a whole fish that is edible fillet, or fillet yield, is a means to improve production efficiency, and thus is of interest to fish farmers and consumers. However, fillet yield poses a challenge to animal breeders aiming to improve this trait through selective breeding because it cannot be directly measured in breeding candidates. The continued development of rainbow trout genomics resources now allow breeders and scientists to accurately predict the genetic merit of fish lacking phenotypic data based on analysis of their DNA, and also to identify the genes or chromosome regions that affect the trait of interest. In this study, we conducted a genome-wide association study using a population of rainbow trout that was characterized for fillet yield and genotyped for more than 38,000 chromosome regions. The analysis identified chromosome regions harboring genes that affected fillet yield and other carcass and body weight traits, but none of those chromosome regions or genes had a major effect on these traits. The vast majority of genetic variation for these traits appears to be affected by a large number of chromosome regions, each with a small effect. Thus, this study suggests that genomic selection based on high-density genotypic data will increase the accuracy of selection for improving fillet yield by allowing prediction of the specific genetic merit of the breeding candidates, and thus will increase the speed of genetic improvement compared to traditional selective breeding.

Technical Abstract: Fillet yield (FY, %) is an economically important trait in rainbow trout aquaculture that affects production efficiency. Despite that, FY has not received much attention in breeding programs because it is difficult to measure on a large number of fish and it cannot be directly measured on breeding candidates. The recent development of a high-density SNP array for rainbow trout has provided the needed tool for studying the underlying genetic architecture of this trait. Here, we conducted a genome-wide association study (GWAS) for FY, body weight at 10 (BW10) and 13 (BW13) months post-hatching, head-off carcass weight (CAR), and fillet weight (FW) in a population of rainbow trout selectively bred for improved growth performance. The GWAS analysis was performed using the weighted single-step GBLUP method (wssGWAS). Phenotypic records of 1,447 fish (1.5 kg at harvest) from 299 full-sib families in three successive generations, of which 875 fish from 196 full-sib families were genotyped using a high-density SNP array, were used in the GWAS analysis for FY, FW, and CAR. A total of 17,174 and 15,810 records from 789 families representing 7 generations were available for BW10 and BW13, respectively. Complete pedigree data were available dating back to initial development of the population (8 generations). A total of 38,107 polymorphic SNPs were analyzed in a univariate model with hatch year and harvest group as fixed effects, harvest weight as a continuous covariate, and animal and common environment as random effects. A new linkage map was developed to create windows of 20 adjacent SNPs for use in the GWAS. Two non-overlapping windows of 20 SNPs each located on chromosome Omy9 explained the largest proportion of the genetic variance for FY (1.5% and 1.0%) and FW (1.2% and 1.1%). One window on Omy5 explained 1.4% and 1.0% of the genetic variance for BW10 and BW13, respectively. Three windows located on Omy27, Omy17 and Omy9 (the same window detected for FY) explained 1.7%, 1.7%, and 1.0%, respectively, of the genetic variance for CAR. Among the detected 100 SNPs, 55% were located directly in genes (intron and exons). Nucleotide sequences of intragenic SNPs were blasted to the Mus musculus genome to create a putative gene network. The network suggests that differences in the ability to maintain a proliferative and renewable population of myogenic precursor cells may have a significant effect on the variation in growth and fillet yield in rainbow trout.