Molecular biomarkers of muscle growth and quality traits in Atlantic salmon:gene-nutrient interactions in fast and slow growing strains

The overall aims of the project were to contribute to sustainable aquaculture by investigating the mechanisms underlying complex growth and quality traits and by identifying biomarkers that can be used to test new raw materials for salmon diets.

Over the last 20 years fish farming has developed into a major industry in rural parts of Scotland, with a shop value of more than £1 billion in 2008. Scottish salmon is exported to more than 50 different countries. Since muscle is the end product of fish farming it is important to understand how its growth is regulated. Atlantic salmon have only just started to be domesticated as farm animals. An important step for developing improved strains of fish is be to identify the most important genes for regulating growth. Most of our knowledge about the regulation of muscle mass from studies on mammals including humans. However, we know that fish are different in that they are cold blooded with slower metabolism and they are also adapted to natural periods of fasting in the wild during the winter and during spawning migrations. One further complication is that salmon have undergone two rounds of whole genome duplication relative to the common ancestor to mammals. Many of these additional copies (called paralogues) have been retained and have potentially acquired new functions. This project set out to dramatically increase knowledge about how muscle growth is controlled in Atlantic salmon. Proteins controlling growth are made using information provided by messenger RNA from a large number of different genes. In this project we have determined the sequences
of more than 30 kinds of mRNA from Atlantic salmon that are known to be involved in controlling muscle mass in mammals. Furthermore, in order to discover novel fish genes controlling growth we used a technique called suppression subtractive hybridisation. This method allowed us to identify candidate genes that were uniquely expressed in fasted individuals and in fish fed for varying periods of time to stimulate growth. We screened candidates in a cell culture system and confirmed that 13 of the 40 candidates selected were likely to be involved in muscle formation. Since growth can be regulated both at the mRNA and protein level we developed a panel of antibodies to study protein expression. We were now able to compare the time course of gene and protein expression as salmon switched from a fasted state where there was net protein breakdown to a fed state in which there was net protein accretion. We found that IGF-I and several of its binding proteins as well of a number of the candidates from our gene discovery experiments showed a much greater response to feeding in fast than slow growing farniles. The regulatory regions of these genes and the proteins that control their transcription are good places to search for natural variations in gene sequence between the fast and slow growing salmon families in the future in order to develop selective breeding programs for superior domesticated strains. The cell culture system was also used to manipulate the levels of hormones and amino acids (the building blocks of proteins) and to determine the effects on gene expression and muscle formation. We found that amino acids alone stimulated the formation of muscle involving the local production of a growth factor called IGF-I. Some of the retained paralogues of muscle genes were also shown to have distinct expression patterns suggesting new functions relative to the ancestral gene.
The development of diets and feeding regimens that can improve growth rate and food conversion efficiency, whilst reducing production costs and improving environmental sustainability is important for the future development of aquaculture. With increasing global demand for fish meal and oil there is an increasing demand to move towards alternative raw material sources such as plant proteins. It is therefore necessary to test new raw materials from sustainable sources, often of plant origin, as well as identifying any bioactive components in raw materials that might provide specific health benefits for the fish or consumers. The production of new feeds and testing their efficiency using feeding trials is a major cost to the aquaculture industry. Traditionally, experiments investigating growth in fish have relied on measuring changes in mass and length which requires feeding trials that last for several weeks. In this project we fed replicated groups of salmon diets varying in the amount of fish meal or amino acids and investigated which genes or proteins might be able to predict growth. We found that the ratio of the messenger RNA for two genes (MLC2:MURF1) and the phosphorylation of a protein called Akt were good predictors of growth, explaining up to 90 of the individual variation in growth rate at the level of the tank. The big advantage of this "molecular method" of predicting growth is the reduced time and expense of conducting growth trials.

Full-length cDNAs for insulin-like growth factor I (IGF-I), insulin-like growth factor II (IGF-II), insulin-like growth factor binding protein (lGFBP) 1, 2, 4, 5, 6 and IGFBP-related protein 1 (IGFBP-rP1) and two gene paralogues of IGFBP-2 and IGFBP-5 were obtained. The transition from zero growth (restricted feeding) to fast growth (satiation feeding) was examined and characterised by constitutive upregulation of IGF-I and IGFBP-4, a transient increase in IGFBP-5.2, and downregulation of IGFBP-2.1, IGF-II, IGF2R, and IGFR1 a. IGFBP-4, was the only IGFBP constitutively upregulated at all time points following feeding, and was positively correlated with IGF-1. Refeeding fasted fish increased expression of genes involved in lipid metabolism and triacylglycerol synthesis (ALDOB, DGAT1 and LPL) 3-14 dafter refeeding, followed by ADIPOQ, MLC2, IGF-I and TALD01, at 14-32d after refeeding. The ubiquitin ligases MuRF1 and MAFbx were strongly downregulated. We concluded that in fasted salmon fat synthesis, decreased protein degradation and increased myofibrillar protein synthesis precedes any increase in genes involved in new muscle growth. To identify novel genes involved in muscle growth we produced 7420 high quality DNA sequences (ESTs) from SSH experiments that compared rapidly growing fish and fish that had zero growth rate. 40 putative candidate genes selected, 23 were confirmed as differentially expressed during refeeding of fasted fish To identify biomarkers of fast growth, the expression of >40 genes, protein expression (including IGFs, IGFBPs, AktP-Akt, mTor/P-mTor, 6S-RP/P-6S-RP), muscle immunocytochemistry and ultrastructure were examined in fast and slow growing strains over a time course following fasting and feeding. FOXD1, Mid11P, ASB2, DRG1, IGF-I, IGFBP-4, IGFBP-5.1, IGFBP-5.2 were expressed at higher levels in fast than slow growing families correlating with a higher abundance of myogenic progenitor cells. We also analysed the expression of genes and proteins in fish fed diets containing different fish meal or amino acid levels to achieve a range of growth rates. The average MLC2/MuRF1 values for each tank compared to the average TGC for each tank, was highly correlated (r2=0.88, p=0.0001, r=0.94, p=0.001). Combining the replicate tanks increased the correlation (r2=0.96, p=0.0001, r=0.98, p=0.0001). The ratio of MLC2 to MuRF1 and to a lesser extent the phosphorylation state of Akt and mTor were identified as promising molecular markers of growth.
The expression of myoD1 gene paralogues, and their regulation by amino acids and insulin-like growth factors, was investigated in primary cell cultures. MyoD1 b, myoD1 c and myoD1 a were differentially expressed during myotube maturation. Depriving cells of amino acids and serum led to a rapid increase in pax7, and decreased myoD1c and PCNA as cells became quiescent. Amino acid replacement resulted in a rapid increase in myoD1 c. slower increase in myoD1b and PCNA and decreased pax7 expression as cells entered the cell cycle. Our results show that myod1c and myoD1 b expression correspond to G 1 and S/G2 phases of the cell cycle, and that myoD1 band myoD1 care expressed in proliferating cells whilst myoD1 a is expressed in differentiating cells. In cells starved of amino acids and serum, IGF-I mRNA decreased 10-fold which was reversed by amino acid replacement. IGFBP-5.1 and IGFBP-5.2 expression increased with amino acid replacement and synergistic increases in expression of IGF-I, IGFBP5.2 and IGFBP-4, were observed with hormone and amino acids addition. IGF-I and IGF-II directly stimulated IGFBP-6 expression, but not when amino acids were present. It was concluded that amino acids are sufficient to stimulate myogenesis and IGF-I production is controlled by paracrine pathways.