We identified DNA markers linked to sex determining genes in six closely related species of tilapiine fishes. The mode of sex determination differed among species. In Oreochromis karongae and Tilapia mariae the sex-determining locus is on linkage group (LG) 3 and the female is heterogametic (WZ-ZZ system). In O. niloticus and T. zillii the sex-determining locus is on LG1 and the male is heterogametic (XX-XY system). A more complex pattern was observed in O. aureus and O. mossambicus, in which markers on both LG1 and LG3 were associated with sex. We found evidence for sex-linked lethal effects on LG1, as well as interactions between loci in the two linkage groups. Comparison of genetic and physical maps demonstrated a broad region of recombination suppression harboring the sex-determining locus on LG3. Sex-specific recombination suppression was found in the female heterogametic sex. Sequence analysis showed the accumulation of repetitive elements in this region. Phylogenetic analysis suggests that at least two transitions in the mode of sex determination have occurred in this clade. This variation in sex determination mechanisms among closely related species makes tilapias an excellent model system for studying the evolution of sex chromosomes in vertebrates.

BackgroundThe probable influence of genes and the environment on sex determination in Nile tilapia suggests that it should be regarded as a complex trait. Detection of sex determination genes in tilapia has both scientific and commercial importance. The main objective was to detect genes and microRNAs that were differentially expressed by gender in early embryonic development.ResultsArtificial fertilization of Oreochromis niloticus XX females with either sex-reversed ΔXX males or genetically-modified YY ‘supermales’ resulted in all-female and all-male embryos, respectively. RNA of pools of all-female and all-male embryos at 2, 5 and 9 dpf were used as template for a custom Agilent eArray hybridization and next generation sequencing. Fifty-nine genes differentially expressed between genders were identified by a false discovery rate of p < 0.05. The most overexpressed genes were amh and tspan8 in males, and cr/20β-hsd, gpa33, rtn4ipl and zp3 in females (p < 1 × 10−9). Validation of gene expression using qPCR in embryos and gonads indicated copy number variation in tspan8, gpa33, cr/20β-hsd and amh. Sequencing of amh identified a male-specific duplication of this gene, denoted amhy, differing from the sequence of amh by a 233 bp deletion on exonVII, hence lacking the capability to encode the protein motif that binds to the transforming growth factor beta receptor (TGF-β domain). amh and amhy segregated in the mapping family in full concordance with SD-linked marker on LG23 signifying the QTL for SD. We discovered 831 microRNAs in tilapia embryos of which nine had sexually dimorphic expression patterns by a false discovery rate of p < 0.05. An up-regulated microRNA in males, pma-mir-4585, was characterized with all six predicted target genes including cr/20β-hsd, down-regulated in males.ConclusionsThis study reports the first discovery of sexually differentially expressed genes and microRNAs at a very early stage of tilapia embryonic development, i.e. from 2 dpf. Genes with sexually differential expression patterns are enriched for copy number variation. A novel male-specific duplication of amh, denoted amhy, lacking the TGF-β domain was identified and mapped to the QTL region on LG23 for SD, thus indicating its potential role in SD.Electronic supplementary materialThe online version of this article (doi:10.1186/1471-2164-15-774) contains supplementary material, which is available to authorized users.

Evidence supports that sex determination (SD) in tilapia is controlled by major genetic factors that may interact with minor genetic as well as environmental factors, thus implying that SD should be analyzed as a quantitative trait. Quantitative trait loci (QTL) for SD in Oreochromis niloticus were previously detected on linkage groups (LG) 1 and 23. Twenty-one short single repeats (SSR) of >12 TGs and one single nucleotide polymorphism were identified using the unpublished tilapia genome sequence on LG23. All markers showed two segregating alleles in a mapping family that was obtained by a cross between O. niloticus male (XY) and sex-reversed female (ΔXY) yielding 29 females (XX) and 61 males (XY and YY). Interval mapping analysis mapped the QTL peak between SSR markers ARO172 and ARO177 with a maximum F value of 78.7 (P < 7.6 × 10−14). Twelve adjacent markers found in this region were homozygous in females and either homozygous for the alternative allele or heterozygous in males. This segment was defined as the sex region (SR). The SR encompasses 1.5 Mbp on a single tilapia scaffold (no. 101) harboring 51 annotated genes. Among 10 candidate genes for SD that were tested for gene expression, anti-Müllerian hormone (Amh), which is located in the center of the SR, showed the highest overexpression in male vs. female embryos at 3 to 7 days postfertilization.

Recent studies have revealed that the major genes of the mammalian sex determination pathway are also involved in sex determination of fish. Several studies have reported QTL in various species and strains of tilapia, regions contributing to sex determination have been identified on linkage groups 1, 3, and 23. Genes contributing to sex-specific mortality have been detected on linkage groups 2, 6, and 23. To test whether the same genes might control sex determination in mammals and fishes, we mapped 11 genes that are considered putative master key regulators of sex determination: Amh, Cyp19, Dax1, Dmrt2, Dmrta2, Fhl3l, Foxl2, Ixl, Lhx9, Sf1, and Sox8. We identified polymorphisms in noncoding regions of these genes and genotyped these sites for 90 individuals of an F 2 mapping family. Mapping of Dax1 joined LG16 and LG21 into a single linkage group. The Amh and Dmrta2 genes were mapped to two distinct regions of LG23. The Amh gene was mapped 5 cM from UNH879 within a QTL region for sex determination and 2 cM from UNH216 within a QTL region for sex-specific mortality. Dmrta2 was mapped 4 cM from UNH848 within another QTL region for sex determination. Cyp19 was mapped to LG1 far from a previously reported QTL region for sex determination on this chromosome. Seven other candidate genes mapped to LG4,

Genetic markers in tilapia species associated with loci affecting sex determination (SD), sex-specific mortality or both were mapped to linkage groups (LG) 1, 2, 3, 6 and 23. The objective of this study was to use these markers to fine-map the locus with the greatest effect on SD in Oreochromis niloticus. Our parental stock, full-sibs of Nile tilapia (Swansea origin), were divided into three groups: (i) untreated, (ii) feminized by diethylstilbestrol and (iii) masculinized by 17α-methyltestosterone. We analysed the first group for association of microsatellite markers representing these five LGs. The strongest association with gender was found on LG23 for marker UNH898 (χ(2) ; P=8.6×10(-5) ). Allele 276 was found almost exclusively in males, and we hypothesized that this allele is a male-associated allele (MAA). Sex-reversed individuals were used for mating experiments with and without the segregating MAA. Mating of individuals lacking the MAA resulted in all-female progeny. Mating of two heterozygotes for MAA gave rise to 81 males and 30 females. Analysis of association between gender and genotypes identified the MAA in 98.6% of males as opposed to 8.0% of females (χ(2) ; P=2.5×10(-18) ). Eight markers that flank UNH898 were genotyped to map the locus on LG23 within a confidence interval of 16-21 cM. Mating of homozygous individuals for MAA is underway for production of all-male populations.