Growth performance, tissue composition, and gene expression responses in Atlantic salmon (Salmo salar) fed varying levels of different lipid sources

Abstract

We investigated the influence of dietary fatty acid (FA) composition on growth performance, muscle tissue FA composition, and lipid-related hepatic gene expression in Atlantic salmon. Currently the commercial feed industry primarily uses terrestrial-based fats and oils as the main lipid component in the diet, with minimal provision of fish oil (FO). Therefore, we compared the growth and tissue FA response of salmon smolts (mean initial weight 256gfish−1) after 16weeks on three different terrestrial-based diets, with inclusion at three increasing levels of substitution with FO. The lipid sources used in this study were: FO, camelina oil (CO), canola oil (CA), and camelina oil with poultry fat (CO+PF). The terrestrial-based lipid sources had different FA signatures, on account of varied proportions of the C18 FA: 18:2n−6, 18:3n−3, or 18:1n−9. Inclusion of terrestrial FA and removal of marine FA (20:5n−3, 22:6n−3) were reflected in the FA composition of the muscle, and affected growth performance (based on lipid source and inclusion level) and hepatic gene expression of several lipid metabolism-relevant genes. Multiple regression analysis revealed that dietary lipid source affected growth performance, even when inclusion level (as a replacement of FO) was not a factor; suggesting that either inclusion or removal of certain FA in particular lipid sources, even at minimal levels, influenced growth performance. A breakpoint analysis of final weight indicated that dietary levels of <1% each of 20:5n−3 and 22:6n−3 resulted in lower final weight. A multidimensional scaling plot showed that the muscle tissue FA composition of salmon fed diets with CO were distinguishable from salmon fed CA and PF, mainly on account of 18:3n−3, while salmon fed FO were distinct from the terrestrial-based treatments. Salmon fed CO diets maintained 22:6n−3 levels in muscle tissue, and increased in n−3 and n−6 intermediary biosynthesis products in the muscle suggesting endogenous production of 22:6n−3 via dietary 18:3n−3. In addition, significant correlations were observed among dietary and/or muscle tissue FA with elovl2, fadsd5, and srebp1 transcript expression, further supporting FA biosynthesis. Significant relationships were also observed among lipid and cholesterol metabolism-related genes with dietary and tissue sterol and 20:4n−6 content, and they were most often related to cholesterol catabolism (or reduction in synthesis), such as hmgcr, dhcr7, and pparβ1. Correlations among gene expression with dietary and tissue lipid classes and FA content provided evidence to support the link between dietary FA input and phenotypic response.