INTEGRATION OF GENOMIC, RNASEQ AND QTL RESOURCES IDENTIFIES VARIATION IN CANDIDATE GENES FOR GROWTH IN TURBOT (Scophthalmus maximus L.)

ROBLEDO D.; FERNANDEZ, C.; HERMIDA, M.; SCIARA, A.A.; ARRANZ, S.E.; ALVAREZ DIOS, J.A.; CAAMAÑO, J.A.; ALVAREZ-CASTRO J.M.; NETTELBLAD C. ; MARTINEZ, P.; BOUZA, C.

Santiago de Compostela

Simposio; The International Symposium on Genetics in Aquaculture XII; 2015

Classical familiar breeding programs have improved growth rate in turbot. However, marker assisted selection may further increase growth rate, because it is possible to identify specific genomic regions explaining a significant proportion of the phenotypic variance of such trait. Moreover, marker assisted selection could provide a more efficient selection of genetic variants that, otherwise, could be lost in classical breeding programs. The implementation  of marker assisted selection requires markers tightly linked to the causal mutation(s). The  increase of turbot genomic resources during the last years has enabled the construction  of linkage maps of appropriate density to detect growth-related QTL. The objective of this  study was to validate previously detected molecular markers associated to growth related  traits in turbot. Eighteen full-sib families derived from breeding programs were genotyped  for 39 markers at 11 linkage groups to look for association to weight, length and Fulton?s condition factor. The results indicate that 25 markers out of the 39 analysed showed significant  association in at least one family and for at least one trait. In addition, all the considered  linkage groups comprised at least one marker with significant association. Twenty markers out of the 39 analysed jointly explained 47.5% of the phenotypic variance of weight; 18 markers explained up to 57.7% of the phenotypic variance for length; and only one marker  was associated with Fulton?s condition factor, explaining 1.2% of the phenotypic variance.  The explained phenotypic variance when considering allelic variants, instead of markers,  was higher (87.3% for weight, 89.1% for length, and 11.4% for Fulton?s condition factor). The  conclusion is that a set of markers and alleles highly associated with growth in turbot is now  available. In practical terms and due to the sparse distribution of growth-related QTL in turbot across the genome, several markers from different linkage groups should be used to  increase genetic gain in selection programs. Keywords: growth, explained phenotypic variance, MAS, breeding program, molecular markerTurbot is a marine flatfish of commercial importance in Europe and China, the main goal of genetic breeding programs in this species is improving growth traits. Previous genomic studies have detected several growth-related quantitative trait loci (QTL), of interest for marker  assisted selection. In this study, integrative genomics was applied to dissect the genetic basis of growth traits in turbot. RNAseq was carried out to analyze muscle and liver transcriptomes under nutritional stress to look for genes involved in growth regulation. Muscle and  liver cDNA samples from fastened and control turbot were sequenced using Illumina HiSeq. A  total of 36,793,041 paired-end reads were generated, accounting for 20,447 genes after alignment against the draft turbot genome. Muscle and liver transcriptomes were used to look for  SNP variation at growth-related genes, which were anchored to structural QTL-mapping by  means of the recently assembled turbot genome as a way to tackle the genetic basis behind  phenotytes. Sixty-four genes were found to be close to growth QTL associated markers.  Forty-five genes-SNPs were selected based on the availability of feasible SNPs, the proximity  of QTL growth markers and functional relevance for growth, including the insulin growth  factors 1 and 2, leptin receptor, myostatin1, parvalbumin1 or growth hormone receptor2, which  were associated with growth traits in other vertebrates and teleosts. 60% of these SNPs  were located within untranslated regions, 31% in coding sequences and the rest in putative  splicing regions derived from RNAseq assembly, and thus putative candidates for functional  changes underlying growth phenotypes. Technically feasible SNPs (96%) were validated in a  wild Atlantic population, 91% of them were polymorphic (mean unbiased gene diversity=0.30,  range:0.06-0.51; mean minimum allele frequency=0.22, range:0.03-0.50) and all loci were in  Hardy-Weinberg equilibrium. Linkage mapping of informative SNP and association with phenotypic variation for growth traits, including a search for epistatic effects, were performed  using a large set of farmed families. These markers were also analyzed to look for adaptive  diversity in wild and domestic populations of turbot. Integrative genomics has successfully  identified SNP markers in growth-related genes suitable for population and family associa- tion studies, as a basis for marker assisted selection strategies.