Applying Transcriptomic Profiling to the Formulation of Sustainable Feeds for Atlantic Salmon (Salmo salar)

Dietary replacement of fish meal (FM) and oil (FO) by animal and plant alternatives can compromise growth performance of farmed Atlantic salmon. After a 14-week feeding trial, Atlantic salmon fed a diet based on animal by-product meal and vegetable oil showed lower growth rates compared to others fed an FM/FO-based diet. Interestingly, feeding a vegetable meal/oil-based diet resulted in growth rates similar to the FM/FO-based diet. To investigate the transcriptomic background of these differences in growth, a 44K microarray experiment was conducted on liver samples from these three dietary groups. The microarray experiment identified forty-one differentially expressed features (Rank Products, FDR<10%) that were annotated as metabolic and growth-relevant genes. Fifteen of these genes were selected for validation through real-time quantitative polymerase chain reaction (qPCR). Only two genes failed the qPCR validation, while eleven of them were confirmed as significantly affected by diet. To evaluate the potential of these genes as growth biomarkers, the qPCR data was subjected to a stepwise discriminant analysis. The multivariate analysis generated two powerful discriminant functions able to differentiate individuals among dietary groups with a 77% accuracy (cross-validated). Only four genes were included as predictive variables: fabp3a (lipid transport), adssl1c (nucleotide synthesis), sgk2a and htra1a (cell growth and proliferation). These genes are currently being utilized as molecular biomarkers for the development of a liver multiplex qPCR. This new platform will constitute a powerful tool for both academia and industry to characterize diet performance in reared Atlantic salmon.