4. Discussion
Given the global context of a scarcity of fish oil (New andWijkstroem, 2002) there is currently much focus on dietary fish
oil levels as well as on alternatives to fish oil. In this study, we initiated a nutrigenomic approach to obtain an overview of the consequences on rainbow trout liver of dietary changes using a
specific rainbow trout cDNA microarray spotted with 9023 annotated cDNAs. Irrespective of the diets, hybridisation of hepatic RNAs with microarrays showed highly expressed genes such as serum albumin, serotransferin and cytochrome c oxidase (data not shown). Only 41 genes were found to be statistically differentially expressed between livers of fish fed with or without fish oils. The relatively low number of variable genes may be explained by two non exclusive reasons. Firstly, we used a generic multi tissue microarray spotted with numerous genes which are not specific to the hepatic tissue alone. Secondly, the dietary treatments did not lead to major effects in fish liver: indeed, in our previous analysis using a candidate gene approach we observed only relatively few genes which were differentially expressed in liver between fish fed with or without added fish oil (Ducasse-Cabanot et al., 2007).