Single-step genome-wide association study for resistance to Piscirickettsia salmonis in rainbow trout (Oncorhynchus mykiss)

Abstract

One of the main pathogens affecting rainbow trout (Oncorhynchus mykiss) farming is the facultative intracellular bacteria Piscirickettsia salmonis. Current treatments, such as antibiotics and vaccines, have not had the expected effectiveness in field conditions. Genetic improvement by means of selection for resistance is proposed as a viable alternative for control. Genomic information can be used to identify the genomic regions associated with resistance and enhance the genetic evaluation methods to speed up the genetic improvement for the trait. The objectives of this study were to i) identify the genomic regions associated with resistance to P. salmonis; and ii) identify candidate genes associated with the trait. We experimentally challenged 2,130 rainbow trout with P. salmonis and genotyped them with a 57 K SNP array. Resistance to P. salmonis was defined as time to death (TD) and as binary survival (BS). Significant heritabilities were estimated for TD and BS (0.48 {+/-} 0.04 and 0.34 {+/-} 0.04, respectively). A total of 2,047 fish and 26,068 SNPs passed quality control for samples and genotypes. Using a single-step genome wide association analysis (ssGWAS) we identified four genomic regions explaining over 1% of the genetic variance for TD and three for BS. Interestingly, the same genomic region located on Omy27 was found to explain the highest proportion of genetic variance for both traits (2.4 and 1.5% for TD and BS, respectively). The identified SNP in this region is located within an exon of a gene related with actin cytoskeletal organization, a protein exploited by P. salmonis during infection. Other important candidate genes identified are related with innate immune response and oxidative stress. The moderate heritability values estimated in the present study show it is possible to improve resistance to P. salmonis through artificial selection in the current rainbow trout population. Furthermore, our results suggest a polygenic genetic architecture and provide novel insights into the candidate genes underpinning resistance to P. salmonis in O. mykiss.