Tissue origin of circulating microRNAs and their response to nutritional and environmental stress in rainbow trout (Oncorhynchus mykiss).

Stresses associated with changes in diet or environmental disturbances are common situations that fish encounter during their lifetime. The stability and ease of measuring microRNAs (miRNAs) present in biological fluids make these molecules particularly interesting biomarkers for non-lethal assessment of stress in animals. Rainbow trout were exposed for four weeks to abiotic stress (moderate hypoxia) and/or nutritional stress (a high-carbohydrate/low-protein diet). Blood plasma and epidermal mucus were sampled at the end of the experiment, and miRNAs were assessed using small RNA sequencing. We identified four miRNAs (miR-122-5p, miR-184-3p, miR-192-5p and miR-194a-5p) and three miRNAs (miR-210-3p, miR-153a-3p and miR-218c-5p) that accumulated in response to stress in blood plasma and epidermal mucus, respectively. In particular, the abundance of miR-210-3p, a hypoxamiR in mammals, increased strongly in the epidermal mucus of rainbow trout subjected to moderate hypoxia, and can thus be considered a relevant biomarker of hypoxic stress in trout. We explored the contribution of 22 tissues/organs to the abundance of circulating miRNAs (c-miRNAs) in blood plasma and epidermal mucus influenced by the treatments. Some miRNAs were tissue-specific, while others were distributed among several tissues. Some c-miRNAs (e.g., miR-210-3p, miR184-3p) showed similar variations in both tissues and fluids, while others showed an inverse trend (e.g., miR-122-5p) or no apparent relationship (e.g. miR-192-5p, miR-194a-5p. Overall, these results demonstrate that c-miRNAs can be used as non-lethal biomarkers to study stress in fish. In particular, the upregulation of miR-210-3p in epidermal mucus induced by hypoxia demonstrates the potential of using epidermal mucus as a matrix for identifying non-invasive biomarkers of stress. This study provides information about the tissue sources of c-miRNAs and highlights the potential difficulty in relating variations in miRNA abundance in biological fluids to that in tissues.