Sex chromosomes have long fascinated evolutionary biologists because they embody a paradox: a pair of chromosomes that once exchanged genetic material freely through recombination gradually ceases to do so, one side accumulating mutations and degradation as the other, the X, continues to recombine in females. In humans and other mammals this process unfolded over hundreds of millions of years, leaving a Y chromosome that has lost most of its ancestral genes. But in fish, sex chromosomes are famously labile. Closely related species, and sometimes even populations within a single species, can carry entirely different sex-determining systems, and within the catfish order Siluriformes the picture is especially tangled. The family Bagridae contains both XY and ZW systems, and even congeneric species of the genus Mystus can differ in which system they use. This instability makes fish, and catfish in particular, ideal laboratories for studying sex chromosome origination in real evolutionary time.

The research team, led by Jianjun Liu, Liuwang Nie, and Huaxing Zhou, working across the Anhui Academy of Agricultural Sciences, Anhui Normal University, and Jiangxi Agricultural University, exploited precisely that recency. The darkbarbel catfish diverged from its close relative, the yellow catfish (Tachysurus fulvidraco), only about 19.4 million years ago, and the yellow catfish is already known to carry young XY sex chromosomes on its chromosome 2. If the two species inherited their sex chromosomes from a common ancestor, their sex-determining regions should sit on homologous chromosomes. Instead, the team found something far more surprising: the two species’ sex chromosomes have no structural relationship to each other at all.

The experimental design began in a fish farm, where a single mature female and a single mature male were bred in isolation. One year later they produced 31 offspring, 18 of which were females and 13 males. Using full siblings eliminates genetic noise from unrelated individuals and population differences, ensuring that any consistent genetic divergence between the sexes can be attributed to the sex-determining system itself. Genomic DNA was extracted from each offspring and sequenced on an MGI DNBSEQ-T7 platform in paired-end 150-base-pair mode, generating roughly 491.31 gigabases of raw data at average depths of 18 to 26 times coverage per individual. After quality filtering, approximately 479.54 gigabases of clean data were retained and mapped against a public reference genome of the darkbarbel catfish.

The first analytical pass, a comparison of sequencing read coverage and SNP density between the sexes, turned up nothing. That null result was itself informative. In species with old, highly differentiated sex chromosomes, the Y chromosome has degraded so much that reads from X-linked regions in females map poorly, producing measurable coverage differences between males and females. The absence of any such signal in the darkbarbel catfish suggested its sex chromosomes are homomorphic, meaning the X and Y remain nearly identical in structure and sequence, as expected for a very young pair.

To localize the sex-determining region, the researchers turned to the fixation index, or FST, a statistic that measures the proportion of genetic variance attributable to differences between populations, in this case males versus females. Using VCFtools and calculating FST in 10-kilobase sliding windows, they applied a threshold of 0.5, a cutoff justified for studies with more than ten samples per sex, to flag regions of strong sex-linked differentiation. Only one region in the entire genome cleared that bar: a span from roughly 12 to 30 megabases on chromosome 3. Principal component analysis of SNPs across chromosome 3 separated males and females cleanly along the first axis, confirming significant genetic differentiation, while linkage disequilibrium decay analysis showed that males carry far longer-range correlations between SNPs than females, as measured by elevated r-squared values. That pattern is exactly what an XY system predicts: in males, the non-recombining Y-linked segment stays locked into large haplotype blocks, whereas females, with two recombining X chromosomes, show rapid LD decay.

The really striking findings emerged when the team compared chromosome 3 across three chromosome-level assemblies: the darkbarbel catfish, the yellow catfish, and the more distantly related Asian red-tail catfish (Hemibagrus wyckioides), which split from the darkbarbel lineage around 94 million years ago. Synteny analysis, performed with MCScanX to detect collinear blocks of genes, showed that the darkbarbel catfish’s chromosome 3 is homologous to chromosomes 7 and 16 of the yellow catfish, and to chromosomes 12 and 25 of the Asian red-tail catfish. In other words, chromosome 3 in the darkbarbel catfish exists as two separate chromosomes in both of its relatives. The chromosome was stitched together from two ancestral autosomes by a fusion event, and microsynteny analysis pinned the fusion junction to a narrow window between 23.86 and 23.91 megabases on chromosome 3.

Here is the kicker: the fusion point lies inside the sex-determining region. Linkage disequilibrium mapping of SNPs that are heterozygous in males but homozygous in females revealed 219 strongly linked SNPs spanning 12.62 to 29.20 megabases, distributed across 34 haplotype blocks clustered into three subregions. The largest block, spanning 17.93 to 19.22 megabases, contains seven genes, including STK35, a serine/threonine kinase known to be highly expressed in human testes and implicated in gametogenesis, and FOXO6, a forkhead transcription factor highly expressed in the testes of the fish Spinibarbus hollandi. Both are plausible candidates for a role in sex determination or male fertility. The fact that the sex-determining region straddles the ancient fusion junction suggests the fusion itself may have been a creative force, bringing together genetic elements that later coalesced into a sex-determining locus.

To probe the selective forces acting on these nascent sex chromosomes, the team computed synonymous substitution rates (dS) and the ratio of nonsynonymous to synonymous substitutions (dN/dS) for every protein-coding gene in three genomic compartments: the sex-determining region, the flanking pseudoautosomal region (PAR) that still recombines between X and Y, and the autosomes. Orthologous coding sequences between the darkbarbel catfish and the yellow catfish were clustered with OrthoFinder, aligned with PRANK, and analyzed with the yn00 program from the PAML package, with 1000-repetition bootstrapping to generate confidence intervals. The results were strikingly ordered. Synonymous substitution rates were lowest on autosomes (0.0625), intermediate in the SDR (0.0650), and highest in the PAR (0.0672). Selection pressure showed the inverse pattern: dN/dS was highest on autosomes (0.2519), slightly lower in the PAR (0.2479), and significantly depressed in the SDR (0.2084). One-way ANOVA confirmed that all differences among the three compartments were highly significant, with p values below 0.0001.

A dN/dS value well below 1 signals purifying selection, the preferential removal of deleterious protein-altering mutations. The SDR’s significantly reduced ratio indicates that genes trapped in the incipient non-recombining region are under intense functional constraint, likely because they include genes essential to male development and fertility that cannot tolerate disruption. The graded pattern, with the SDR more constrained than the PAR, which in turn is more constrained than the autosomes, captures an incipient differentiation continuum. It is a genome-wide portrait of a chromosome pair at the very beginning of the trajectory that, given tens of millions of years, could produce a degenerate Y chromosome like our own.

The comparative picture seals the interpretation. Although the darkbarbel catfish and the yellow catfish share the same chromosome number and are separated by less than 20 million years of evolution, three fusion and fission events distinguish their karyotypes, and two of those events independently gave rise to young XY sex chromosomes in the two species, with no sequence-level similarity between them. Prior work in fish had documented sex chromosome fusions in which an autosome attaches to an existing sex chromosome, creating so-called neo-sex chromosomes. What this study proposes is different and arguably more provocative: that the fusion of two ordinary autosome pairs can be the founding event itself, creating a new chromosomal context in which a sex-determining locus arises, linkage disequilibrium builds up around it, and recombination suppression follows. Chromosomal fusion, in this view, is not merely a consequence of sex chromosome evolution but a potential driver of it.

The authors are careful to note that more species and more data will be needed to establish fusion as a general mechanism of sex chromosome origination. But the darkbarbel catfish offers an unusually clean case study, made possible by sibling-family sequencing, a justified FST threshold, and chromosome-level assemblies of multiple related species. As chromosome-level genome assemblies accumulate across the tree of life, researchers expect many more examples of sex chromosomes at this embryonic stage, each one a natural experiment in how the most fundamental genetic division in animals, the division between male and female, first takes hold in the genome. For now, a modest catfish from Chinese rivers has provided biologists with something rare: a look at evolution’s paperwork while it is still being written.