Mapping the sex determination locus in the Atlantic halibut (Hippoglossus hippoglossus) using RAD sequencing

Abstract

Background: Atlantic halibut (Hippoglossus hippoglossus) is a high-value, niche market species for cold-water marine aquaculture. Production of monosex female stocks is desirable in commercial production since females grow faster and mature later than males. Understanding the sex determination mechanism and developing sex-associated markers will shorten the time for the development of monosex female production, thus decreasing the costs of farming.

Results: Halibut juveniles were masculinised with 17 α-methyldihydrotestosterone (MDHT) and grown to maturity. Progeny groups from four treated males were reared and sexed. Two of these groups (n = 26 and 70) consisted of only females, while the other two (n = 30 and 71) contained balanced sex ratios (50% and 48% females respectively). DNA from parents and offspring from the two mixed-sex families were used as a template for Restriction-site Associated DNA (RAD) sequencing. The 648 million raw reads produced 90,105 unique RAD-tags. A linkage map was constructed based on 5703 Single Nucleotide Polymorphism (SNP) markers and 7 microsatellites consisting of 24 linkage groups, which corresponds to the number of chromosome pairs in this species. A major sex determining locus was mapped to linkage group 13 in both families. Assays for 10 SNPs with significant association with phenotypic sex were tested in both population data and in 3 additional families. Using a variety of machine-learning algorithms 97% correct classification could be obtained with the 3% of errors being phenotypic males predicted to be females.

Conclusion: Altogether our findings support the hypothesis that the Atlantic halibut has an XX/XY sex determination system. Assays are described for sex-associated DNA markers developed from the RAD sequencing analysis to fast track progeny testing and implement monosex female halibut production for an immediate improvement in productivity. These should also help to speed up the inclusion of neomales derived from many families to maintain a larger effective population size and ensure long-term improvement through selective breeding.

Figure 1

Sequencing and RAD-tag summary. Detail of the number of reads before and after filters (orange disk) followed by the reconstructed number of RAD markers and polymorphic RAD markers (orange circles).

Figure 2

Genetic linkage map. Map with linkage group assignment determined using syntenic markers with previously published H. hippoglossus maps. The positions on the left side of the chromosomes are in cM. The rectangles on the right hand side represent the number of markers at this position. Detailed data is provided in the Additional file 2.

Figure 3

Results from QTL-Analysis. (A) Association results for genotyped SNPs. SNPs with p-values achieving genome-wide significance (P < 7.2 × 10-8) are shown in black. (B) Regional analysis of the QTL on LG 13. Plot of the LOD score (sex-association QTL search) along the linkage group 13 for family B and C.

Figure 4

KASP assay and fine gene mapping on LG 13. Details of the 10 markers tested by KASP assay. From bottom to top: Location of the 10 markers (in the genetic map in cM and syntenic loci on the G. aculeatus, three-spined stickleback, group XIV in bp); KASP assay results. The outer circle diameters for the KASP assay results are proportional to the number of alleles tested. The inner (solid) disks represent the marker association with the phenotypic sex. Detailed data is provided in Additional file 4. When no informative polymorphism was found, “n.a.” is specified.

Figure 5

Combined marker sex prediction. (A) Confusion matrix of the JRip rules. Blue cells are correct predictions; white cells are the erroneous predictions. Overall the predictions are 97% accurate. (B) JRip rules based on the alleles detected using the KASP assays.