Repeated Evolution Versus Common Ancestry: Sex Chromosome Evolution in the Haplochromine Cichlid Pseudocrenilabrus philander

Abstract

Why sex chromosomes turn over and remain undifferentiated in some taxa, whereas they degenerate in others, is still an area of ongoing research. The recurrent occurrence of homologous and homomorphic sex chromosomes in distantly related taxa suggests their independent evolution or continued recombination since their first emergence. Fishes display a great diversity of sex-determining systems. Here, we focus on sex chromosome evolution in haplochromines, the most species-rich lineage of cichlid fishes. We investigate sex-specific signatures in the Pseudocrenilabrus philander species complex, which belongs to a haplochromine genus found in many river systems and ichthyogeographic regions in northern, eastern, central, and southern Africa. Using whole-genome sequencing and population genetic, phylogenetic, and read-coverage analyses, we show that one population of P. philander has an XX-XY sex-determining system on LG7 with a large region of suppressed recombination. However, in a second bottlenecked population, we did not find any sign of a sex chromosome. Interestingly, LG7 also carries an XX-XY system in the phylogenetically more derived Lake Malawi haplochromine cichlids. Although the genomic regions determining sex are the same in Lake Malawi cichlids and P. philander, we did not find evidence for shared ancestry, suggesting that LG7 evolved as sex chromosome at least twice in haplochromine cichlids. Hence, our work provides further evidence for the labile nature of sex determination in fishes and supports the hypothesis that the same genomic regions can repeatedly and rapidly be recruited as sex chromosomes in more distantly related lineages.

Keywords: genome sequencing; population genetics; sex determination; speciation; species complex; teleost fish.

Fig. 1.

—Phylogenetic relationships and sex determination in East African cichlids. (A) Schematic phylogenetic relationships of East African cichlids. Information on sex determination systems based on Böhne et al. (2016), Cnaani et al. (2008), Feulner et al. (2018), Gammerdinger et al. (2018a, b), Kudo et al. (2015), Parnell and Streelman (2013), Peterson et al. (2017), Roberts et al. (2016), Ser et al. (2010), and Yoshida et al. (2011). Haplochromine lineages are depicted in green. (B) Map of East Africa and a zoom on the sampling locations: 1, Lake Chila and 2, Mbulu creek. (C) Male specimen of Pseudocrenilabrus philander. (D) PCA on genome-wide variant data of all P. philander individuals of this study. PC1 separates the lake individuals from the creek population. PC3 separates males from females. The outlier MJB7 and the potential sex-reversed individual MJA8 are highlighted: dark gray: Lake Chila, light gray: Mbulu creek, red: females, and blue: males.

Fig. 2.

—Genomic signatures of male–female differentiation in Pseudocrenilabrus philander. (A) Male–female F_ST for individuals from Lake Chila (upper panel) and Mbulu creek (lower panel) along the reference genome of Oreochromis niloticus. Each dot represents a single F_ST value per 10 kb window. (B) Male–female F_ST and difference in nucleotide diversity between sexes (π_diff = π_males − π_females) along LG7. Each gray dot represents a single value per 10 kb window. Black line: smoothed value (loess parameter = 0.01) and red line: no difference in nucleotide diversity between males and females.

Fig. 3.

—Phylogenetic analysis within the two Pseudocrenilabrus philander populations based on markers on LG7 and using genome-wide variants on all LGs but LG7. Maximum likelihood phylogeny of LG7 and all other LGs except LG7 for Lake Chila (upper panel) and Mbulu creek (lower panel); blue: males, red: females, and asterisks: 100% bootstrap support.

Fig. 4.

—XY-sites in Pseudocrenilabrus philander from Lake Chila. (A) Distribution of potential XY sex-patterned sites across all LGs in the Lake Chila population normalized by total number of sites per LG. (B) Distribution of XY-sites along LG7 in 10 kb bins. (C) Distribution of all variant sites called on LG7 in 10 kb bins.

Fig. 5.

—Sex chromosome coverage in Pseudocrenilabrus philander from Lake Chila. (A) Coverage of perfect alignments of males (blue) and females (red) along the de novo assembled Lake Chila female X-chromosome (left) and for comparison along the de novo assembled LG6 (right). (B) Coverage of all alignments of males and females along the de novo assembled Lake Chila female X-chromosome (left) and for comparison along the de novo assembled LG6 (right); red and blue lines: smoothing spline, black dotted lines: normalized coverage of 1, and gray dotted line: normalized coverage of 0.5.

Fig. 6.

—K-mer comparison in males and females. (A) Counts of 37 bp k-mers in male and female Lake Chila Pseudocrenilabrus philander. (B) Counts of 37 bp k-mers in human males and females. Humans have strongly differentiated sex chromosomes. K-mers derived from the Y chromosome are expected to have zero counts in females; k-mers derived from the X chromosome should have half the count in males than in females. Potential Y-k-mers are highlighted with a blue circle, X-mers with a red circle.

Fig. 7.

—Topology weighting analysis of LG7. Topology weighting analysis using 1-, 5-, and 10-kb windows between the four “populations” Lake Chila males, Lake Chila females, Lake Malawi males, and Lake Malawi females.