Prospecting Genomic Regions Associated with Columnaris Disease in Two Rainbow Trout Breeding Populations

Flavobacterium Columnare, the causative agent of Columnaris disease (CD), is distributed around the world in fresh water sources, infecting freshwater finfish species. Recently it has been identified as an emerging problem for the rainbow trout aquaculture industry in the US. Two live-attenuated vaccines have been commercialized, but their efficacy in rainbow trout is still not clear. The purpose of this study was to prospect genomic regions that explain large portion of the additive genetic variance of CD resistance in rainbow trout. Two important aquaculture populations were investigated. The National Center for Cool and Cold Water Aquaculture (NCCCWA) odd-year line, with resistance to bacterial cold water disease (BCWD); and the Troutlodge, Inc., May odd-year (TLUM) nucleus breeding population, which provided 54,350 and 36,265 pedigree records; in which 8,453 and 3,986 fish had CD resistance phenotype records, respectively. Fish that survived to 21 days post immersion challenge were recorded as resistant (phenotype=2) and those that did not were rated as susceptible (phenotype=1). Genotypes for 57k SNP (Affymetrix Axiom^®) were available for 1,185 and 1,137 fish from NCCCWA and TLUM, respectively. The SNP effects and variances were estimated using the weighted single-step genomic BLUP approach for genome-wide association (WssGBLUP), which uses pedigree, genotypes, and phenotypes from genotyped and ungenotyped animals. The weighting strategy accounted for 1Mb moving SNP-windows. Genomic regions that explained more than 1% of the additive genetic variance were considered associated with CD resistance. A total of 13 windows located on six chromosomes were found to be associated with CD resistance in the NCCCWA population: two windows, located at 59-60 Mb and 61-62 Mb on chromosome Omy17, explained 12% and 11.33% of the genetic variance for CD resistance, respectively. In the TLUM population, a total of 16 windows located on nine chromosomes were detected. Only three similar windows (located on two chromosomes) were detected in both populations. The results suggest that CD resistance has an oligogenic architecture, and the SNP windows found to be associated with CD are not informative enough for selection decisions across populations. A moderate positive genetic correlation has been previously shown between CD and BCWD resistance in the NCCCWA population. One factor which might have contributed to detecting different QTL for CD resistance in the two populations is the five generations of selective breeding for BCWD resistance applied to the NCCCWA population in contrast to no selection pressure for disease resistance in the TLUM population.