Multiple sex-associated regions and a putative sex chromosome in zebrafish revealed by RAD mapping and population genomics

Abstract

Within vertebrates, major sex determining genes can differ among taxa and even within species. In zebrafish (Danio rerio), neither heteromorphic sex chromosomes nor single sex determination genes of large effect, like Sry in mammals, have yet been identified. Furthermore, environmental factors can influence zebrafish sex determination. Although progress has been made in understanding zebrafish gonad differentiation (e.g. the influence of germ cells on gonad fate), the primary genetic basis of zebrafish sex determination remains poorly understood. To identify genetic loci associated with sex, we analyzed F(2) offspring of reciprocal crosses between Oregon *AB and Nadia (NA) wild-type zebrafish stocks. Genome-wide linkage analysis, using more than 5,000 sequence-based polymorphic restriction site associated (RAD-tag) markers and population genomic analysis of more than 30,000 single nucleotide polymorphisms in our *ABxNA crosses revealed a sex-associated locus on the end of the long arm of chr-4 for both cross families, and an additional locus in the middle of chr-3 in one cross family. Additional sequencing showed that two SNPs in dmrt1 previously suggested to be functional candidates for sex determination in a cross of ABxIndia wild-type zebrafish, are not associated with sex in our AB fish. Our data show that sex determination in zebrafish is polygenic and that different genes may influence sex determination in different strains or that different genes become more important under different environmental conditions. The association of the end of chr-4 with sex is remarkable because, unique in the karyotype, this chromosome arm shares features with known sex chromosomes: it is highly heterochromatic, repetitive, late replicating, and has reduced recombination. Our results reveal that chr-4 has functional and structural properties expected of a sex chromosome.

Figure 1. Mapping results for the F₂ of reciprocal crosses between *AB and NA wild type stocks.

(A) Family A, *AB grandsire. (B) Family B, NA grandsire. Sex is associated with a locus on chr-4 in both families, and a locus on chr-3 in Family B. LOD scores are plotted in order of marker position in cM. Significance thresholds were determined by permutation testing and are indicated by horizontal lines (red = 5%, gray = 10%).

Figure 2. Relationships between sex phenotype and haplotype on chr-3 and -4.

(A) In Family A, all individuals homozygous for the C allele at Marker ID29464 chr-4∶61,176,889 bp were male and 84% of individuals heterozygous at this marker were female. (B) In Family B, loci on chr-3 and -4 appear to interact to influence sex (marker ID32525 chr-4∶61,422,807bp; marker ID29552 chr-3∶20,676,406bp). Individuals homozygous for grandsire-derived alleles at the chr-4 locus developed as males independent of their chr-3 genotype and animals homozygous for granddam-derived alleles were all or nearly all female without regard to their chr-3 genotype. On the other hand, individuals that were homozygous for the grandsire-derived alleles at the chr-3 locus were male independent of their genotype at the chr-4 locus.

Figure 3. Genome wide differentiation between males and females.

(A) Family A. (B) Family B. Individual SNPs were tested for their non-random association with male or female phenotypes. Graphs plot the -log₁₀ p-value of a G-test across the genome and show peaks at the sar4 and sar3 locations. Red lines show the 0.1 percent false discovery rate within each family. Peaks in chr-14 are in genes that are members of a major multigene family in sar4, and may be mis-localized in the current genome assembly.

Figure 4. Sex-specific recombination rates across the genome.

(A) Female-specific recombination genome-wide. (B) Male-specific recombination genome-wide. (C) Female-specific recombination across chr-4. (D) Male-specific recombination across chr-4. Recombination rates were measured in a sliding window comparing the map distance between markers to the genomic location on the Zv9 assembly of the zebrafish genome. Recombination is generally elevated at telomeres but is reduced in males at sar4. For ease of visualization, recombination rates of excess of 35 cM/Mb at the top of the graph are indicated by *. In the female, these rates equal 88, 49, 67, 52, and 108 cM/Mb from right to left. In the male, these rates equal 94, 50, and 39 cM/Mb.

Figure 5. Repetitive nature of the hetetochromatic region of chr-4 compared to the whole zebrafish genome.

(A) Genome wide distribution (chr-1 to −25 across the top of the figure) of genetic elements enriched in the heterochromatic region of chr-4. Rows listed top to bottom: 5S RNA motif, micro RNAs (miRNAs), four different Ensembl multigene families (ENSFM), small nuclear RNAs (snRNAs), transfer RNAs (tRNAs), Kolobok-1 transposon, and pseudogenes. See Table 2 for specifics on these elements. (B) Within chr-4 (position in Mb listed across the top), enriched elements are found in the hetrochromatic region (green) delineated by the mir-430 cluster (light blue) at the centromeric side and sar-4 (red) near the telomere of the p arm. (C) Schematic of chr-4 with locations of the centromere, heterochromatic region and sar-4. (D) Replication banding of a cell cultured from a female zebrafish showing the late replication of chr-4q (arrows), unique among the 50 zebrafish chromosome arms. (E) Enlargement of the chr-4 pair; note heterochromatic chr-4q.

Figure 6. Genomic locations of candidate sex genes and the locations of mapped sex-associated regions.

The physical positions of genes with a potential influence on sex determination in zebrafish are marked in purple (see for names and locations the full list in Table S7). Red boxes indicate the 1.5 LOD drop intervals for sar3 and sar4 identified here in the *ABxNA cross and blue rectangles mark the 1.5 LOD droop intervals for sar5 and sar16 in the ABxIN cross . Dashed lines mark peaks of sex-associated loci identified herein (red) and by Bradley et al. (blue). Black stars indicate genes identified as candidates for functional sex determination genes in zebrafish by Bradley et al. . The red star indicates hsd17b1. Upper and lower bounds of the surrounding boxes represent the 1.5 LOD drop intervals for each peak.