Convergent evolution of chromosomal sex-determining regions in the animal and fungal kingdoms

Abstract

Sexual identity is governed by sex chromosomes in plants and animals, and by mating type (MAT) loci in fungi. Comparative analysis of the MAT locus from a species cluster of the human fungal pathogen Cryptococcus revealed sequential evolutionary events that fashioned this large, highly unusual region. We hypothesize that MAT evolved via four main steps, beginning with acquisition of genes into two unlinked sex-determining regions, forming independent gene clusters that then fused via chromosomal translocation. A transitional tripolar intermediate state then converted to a bipolar system via gene conversion or recombination between the linked and unlinked sex-determining regions. MAT was subsequently subjected to intra- and interallelic gene conversion and inversions that suppress recombination. These events resemble those that shaped mammalian sex chromosomes, illustrating convergent evolution in sex-determining structures in the animal and fungal kingdoms.

Figure 1. Fungal MAT Locus Paradigms

Interaction of mating partners during the fungal sexual cycle is directed by bipolar or tetrapolar mating systems. The budding yeast S. cerevisiae is an ascomycete with bipolar mating (graphic at upper left). The maize pathogen U. maydis is a tetrapolar basidiomycete with multiple mating types conferred by two mating type loci (graphic at upper right). One (a) is biallelic and encodes pheromones and pheromone receptors, while the second (b) is multiallelic and encodes homeodomain transcription factors. In contrast, the human pathogen C. neoformans (lower graphic) is a basidiomycete with a bipolar system with only two mating types (a and α). The C. neoformans MAT locus encodes homeodomain transcription factors, pheromones and pheromone receptors, other elements of the pheromone activated MAPK cascade, and many genes whose role in mating, if any, is at present unknown.

Figure 2. Genes of the MAT Locus

Comparison of a and α alleles of the MAT genes in the three Cryptococcus lineages based on percent nucleotide identity between the coding sequences of the a and α alleles within that species. K_s values were calculated from comparison of the a and α alleles in each species. Genes with unusual K_s values are shown in red, and genes from regions flanking MAT are shown in grey. Phylogenetic trees are based on maximum likelihood analysis (scale bar = 0.05 substitutions per site), and are labeled with the phylogenetic class they represent. Further details are presented in Figures S1 and S2. *The Average K _s for CAP1 was calculated after excluding the indicated unusual values.

Figure 3. The Structure of MAT Is Highly Rearranged, with Divergent Gene Alleles Embedded in Syntenic Genomic Regions

The nonrecombining α (blue) and a (yellow) MAT alleles from the divergent but related species are depicted, spanning more than 100–130 kb and including 10 kb of common flank regions on the left and right demarcated by sharp borders with MAT. The original locations of ancient tetrapolar loci proposed to have given rise to MAT are shown in red (ancestral homeodomain locus) and green (ancestral pheromone/receptor locus), with the most ancient genes (encoding homeodomain transcription factors, pheromones and pheromone receptors) bulleted. Genes that show mating type-specific phylogeny are shown in black, and genes with species-specific phylogeny are white. Synteny between the genes with species-specific phylogeny is indicated with grey boxes. Pseudogenes are labeled in blue, and grey bars represent repeated elements in Cn. var. neoformans. Red arrows represent pheromone amplicons.

Figure 4. MAT Is Highly Rearranged between Species and Mating Types

The genomic region spanning the nonrecombining α (blue) and a (yellow) MAT alleles from the divergent but related species is depicted, with pink and green colored bars representing regions of synteny, and black lines the relative positions of genes whose position is not conserved. Black arrows depict mating type-specific genes. White arrows represent genes with a species-specific phylogeny. Red arrows represent pheromone amplicons.

Figure 5. MAT Genes Have Different Phylogenetic Histories

(A) The genes of the MAT locus can be separated into four distinct groupings based on phylogenetic class, synonymous substitution rate, and nucleotide identity. The C. gattii alleles in each phylogram are encircled in red. (B) The LPD1 gene defines an ancient border of MAT. The 5′ end of the coding region is species-specific, while the 3′ region is mating type-specific. The ancient homeodomain locus is shown in red, and the ancient pheromone/pheromone receptor locus in green. Maximum likelihood trees are shown. Scale bar represents 0.05 substitutions per site. Further details are provided in Figures S2 and S3.

Figure 6. Reconstructing the Ancient MAT Alleles by Inversion-Mediated Rearrangement

Plotting the synonymous mutation rate (K_s) of each protein coding gene in MAT reveals that the different classes of genes in the two least rearranged loci (Cn. var. grubii MATα and C. gattii MATa; see Figure 4) can be clustered by a single inversion. This may represent an ancient linked tetrapolar system—one cluster contains the pheromone and pheromone receptor genes (green bars), and the other a homeodomain-encoding gene (red bar). Transposon remnants are present at the extrapolated inversion breakpoint regions in Cn. var. grubii, as indicated (Tn). K_s cannot be calculated between the SXI1α and SXI2a genes, because these are unrelated and not alleles, in contrast to other genes in the locus.

Figure 7. A Model for the Evolution of MAT

Our evidence indicates that the ancient loci of a canonical tetrapolar system expanded to incorporate additional genes, beginning with two rounds of expansion of the pheromone/receptor locus: first to acquire genes including components of the pheromone-signaling MAPK cascade (ancient), and second to acquire genes whose role in mating is unknown (intermediate I). Next, the ancestral homeodomain locus acquired genes hypothesized to function in the dikaryon or meiosis (intermediate II). The tetrapolar loci in one mating type fused by chromosomal translocation, entrapping the most recently acquired species-specific gene set (recent) and creating a tripolar intermediate. A second locus fusion event then occurred, to link the two regions from the opposite mating type and create the bipolar ancestors of MAT. Subsequent inversion-mediated rearrangements have erased the discrete evolutionary strata.