Genomic characterization of the Atlantic cod sex-locus

Abstract

A variety of sex determination mechanisms can be observed in evolutionary divergent teleosts. Sex determination is genetic in Atlantic cod (Gadus morhua), however the genomic location or size of its sex-locus is unknown. Here, we characterize the sex-locus of Atlantic cod using whole genome sequence (WGS) data of 227 wild-caught specimens. Analyzing more than 55 million polymorphic loci, we identify 166 loci that are associated with sex. These loci are located in six distinct regions on five different linkage groups (LG) in the genome. The largest of these regions, an approximately 55 Kb region on LG11, contains the majority of genotypes that segregate closely according to a XX-XY system. Genotypes in this region can be used genetically determine sex, whereas those in the other regions are inconsistently sex-linked. The identified region on LG11 and its surrounding genes have no clear sequence homology with genes or regulatory elements associated with sex-determination or differentiation in other species. The functionality of this sex-locus therefore remains unknown. The WGS strategy used here proved adequate for detecting the small regions associated with sex in this species. Our results highlight the evolutionary flexibility in genomic architecture underlying teleost sex-determination and allow practical applications to genetically sex Atlantic cod.

Figure 1. Sex linked segregation of genotypes in Atlantic cod.

Over 55 million polymorphisms are compared in 110 males and 116 females of Atlantic cod using Fishers’ exact test. Six distinct regions are identified with a significant increase in p-values, i.e. above 6SD of the mean. Highest numbers of genotypes with most extreme p-values are found on LG11 (a), whereas reduced numbers with lower p-values are found on LG08, LG14, LG15 and LG17 (b). P-values were calculated using PLINK (v.1.90p) and −log transformed. Unplaced scaffolds have been concatenated for visualization.

Figure 2. PacBio read alignments of a female Atlantic cod specimen.

Long read alignments (grey) towards the male gadMor2 reference genome overlap those regions containing sex-associated genotypes (red box) on LG08 (a,b) and LG11 (c). Note the different genomic scale (in Kb) for each of the sub-panels. Small indels (black dots) within the alignments are a typical feature of PacBio read data. Read alignments are visualized using the Integrative Genomics Viewer.

Figure 3. Annotation surrounding the sex-associated region of Atlantic cod on linkage group 11.

Nine gene models (the arrow shows transcriptional direction) have been annotated within a 200 Kb window around the loci with sex-linked genotypic segregation. The histogram (orange) shows the number of sex associated loci (with p-values more than 6 SD from the mean) in windows of 1 Kb. Gene models with cDNA evidence from Atlantic cod (blue) have been annotated with gene names and are in conserved synteny (order and transcriptional direction) with those in three-spined stickleback and spotted gar. Three genes (ptpn2a, cep76 and id4) are in conserved synteny with medaka. Gene models without cDNA evidence (grey) do not have obvious homology with known genes. Average read depth (grey line) is calculated in 1 Kb windows and can be compared to the genome-wide average coverage (red dashed line).