How Embryos Police the Evolution of Duplicated Genes in Salmon

When an entire genome gets duplicated, evolution is handed an enormous gift: a spare copy of every gene, free to mutate, specialize or disappear without catastrophic consequences. For more than a century, biologists have suspected that these wholesale duplication events, which struck repeatedly during vertebrate history, helped fuel the rise of new organs, new body plans and new species. Yet the ancient origins of the earliest duplications have made it fiendishly difficult to watch the process in action. A landmark study published in Nature Ecology & Evolution now offers the clearest view yet, using salmon and trout as a living time capsule of genome evolution in progress.

An international consortium of researchers generated an extraordinary trove of data: 771 multiomics datasets spanning embryonic development and adult tissues in Atlantic salmon and rainbow trout, totalling 77.7 billion read pairs. The work was carried out under the European AQUA-FAANG project, part of the Functional Annotation of Animal Genomes initiative, and combined RNA sequencing to measure gene expression with ATAC-seq and ChIP-seq to map open chromatin and epigenetic marks. From these data, the team identified 319,718 robust open chromatin regions in salmon and 194,002 in trout, and used a computational approach called ChromHMM to classify the genome into functional states, pinpointing tens of thousands of active promoters and enhancers in each species.

The reason salmonids are so valuable for this kind of work lies in their peculiar genomic history. Roughly 80 to 100 million years ago, the ancestor of all salmon and trout experienced a whole genome duplication, an autotetraploidization event that doubled the entire genome within a single species. Unlike the ancient duplications that shaped all vertebrates more than half a billion years ago, this salmonid event is recent enough that the genome is still in the midst of rediploidization, the slow return to normal diploid inheritance. Around half of salmonid genes are still retained as functional duplicate pairs, known as ohnologues, sitting in large collinear blocks on duplicated chromosomes.

Crucially, rediploidization did not happen uniformly across the genome. Some duplicated regions resolved back into diploid inheritance before salmon and trout lineages split roughly 25 million years ago, giving their genes a long head start on divergence. Other regions resolved independently in each species, meaning their duplicate pairs are evolutionarily much younger. This stratification, called asynchronous rediploidization, allowed the researchers to compare duplicate genes at different stages of post-duplication evolution within the same genomes, effectively providing a series of snapshots of the same evolutionary process at different ages.

The central finding is striking: the pace of regulatory evolution after genome duplication is not random but is governed by the same developmental constraints that shape gene evolution across all vertebrates. Measuring expression divergence between duplicate pairs using Jensen-Shannon distance, the team found that divergence was highest in early embryos, before the maternal-to-zygotic transition, and fell steadily through embryogenesis, reaching its minimum at the pharyngula stage, which corresponds to the vertebrate phylotypic period. This is the stage, long recognized in evolutionary developmental biology, when embryos of different species look most alike and when pleiotropic genes are recruited across many developing tissues, creating a densely interconnected regulatory environment that resists change.

The correlation was remarkably tight. The pattern of decreasing expression divergence across embryogenesis was nearly identical between early and late rediploidization pairs, with a Pearson correlation of 0.97, and it also matched the divergence of orthologous genes between salmon and trout. In other words, the constraints acting on duplicate genes within a single genome mirror those acting on the same genes across species. The authors argue that these pleiotropic constraints do not merely limit divergence after duplication; they actively drive which duplicates are retained, because genes expressed during mid-to-late segmentation tend to be dosage-sensitive, and keeping both copies of such genes buffers against harmful imbalances in gene dose.

Adult tissues told a different and equally revealing story. Expression divergence between duplicates varied enormously across the body, with the brain showing the lowest divergence, consistent with its status as the tissue under the strongest purifying selection, while liver, ovary and testis showed the highest. Sexual maturation had little effect in most tissues, but mature testis displayed elevated divergence compared with immature testis. When the researchers compared orthologous genes between salmon and trout, they found that duplicates retained from the genome duplication had evolved more slowly than single-copy genes in gill, head kidney and distal intestine, though in the brain even single-copy genes evolved slowly, and in the ovary, maturation stage mattered more than duplication status.

Turning from genes to their regulatory elements, the team classified every active promoter and enhancer according to its fate after duplication: shared, meaning both duplicated copies remained active; alignable only, meaning the sequence was conserved but activity was lost from one copy; or exclusive, meaning the active element existed on only one chromosome. Promoters proved conservative, with 53.5 percent shared, closely matching the estimated 55 percent retention rate of duplicate gene pairs. Enhancers were far more dynamic, with only 33.3 percent shared and 43.8 percent falling into the alignable-only category, confirming that enhancers turn over faster than promoters after genome duplication, a pattern also seen when comparing regulatory elements across species.

The link between enhancer fate and gene divergence was direct. Duplicate gene pairs whose expression remained conserved had significantly more shared enhancers, while pairs with diverged expression had more exclusive enhancers, and their enhancer-usage profiles across tissues were measurably farther apart. In adult tissues, enhancer-usage distance tracked expression distance closely, whereas during embryogenesis it did not, suggesting that promoter changes and a specific subset of shared enhancers dominate regulatory divergence in embryos, while wholesale remodelling of the enhancer repertoire drives divergence in adult organs. Transposable elements added another layer: exclusive enhancers overlapped transposon sequences more than twice as often as shared enhancers, and DNA transposons of the TcMar class, the most active group in the salmon genome, were specifically enriched in enhancers of diverged duplicate pairs, implicating these jumping genes both in disabling old enhancers and in creating new ones.

Perhaps the most evocative result concerns conserved non-coding elements, ancient DNA sequences under purifying selection across hundreds of millions of years of vertebrate evolution. Among 24,007 enhancer-associated conserved elements in salmon, 61.4 percent were retained as shared duplicate pairs, and these were strongly enriched for the oldest categories, dating back to the ancestor of all bony vertebrates. The shared elements active during late somitogenesis, the phylotypic period, were packed with motifs for master developmental transcription factors such as POU5F1, Sox and Pax family proteins, precisely the regulators that orchestrate embryonic patterning. This suggests that duplicated enhancers carrying deeply conserved developmental instructions are functionally retained in duplicate, reinforcing the developmental lock that limits evolutionary change at the hourglass waist. The study, whose data are openly available through the FAANG portal and SalmoBase, not only illuminates how genomes recover from duplication but also provides a resource for breeding improved aquaculture strains and for conserving wild salmonid populations whose genomes carry this extraordinary evolutionary archive.

Subject of Research: Gene-regulatory evolution following whole genome duplication in salmonid fishes

Article Title: Developmental contexts of gene-regulatory evolution following whole genome duplication in salmonids

Article References: Baudement, M.-O., Morata, D. P., Gillard, G. B., Dewari, P. S., Gundappa, M. K., Podgorniak, T., Grønvold, L., Baranasic, D., Laurent, A., Giudicelli, F., Zunar, B., Carrera-García, E., Perquis, A., Brionne, A., Thi Nguyen, T., Ruiz Daniels, R., Merino, G. A., Thybert, D., Ilsley, G. R., … Macqueen, D. J. (2026). Developmental contexts of gene-regulatory evolution following whole genome duplication in salmonids. Nature Ecology & Evolution, 10(10), 1959-1981. https://doi.org/10.1038/s41559-026-03171-6

Image Credits: AI Generated

DOI: 10.1038/s41559-026-03171-6

Keywords: whole genome duplication, salmonids, Atlantic salmon, rainbow trout, ohnologues, rediploidization, enhancers, gene regulation, phylotypic period, transposable elements, conserved non-coding elements, comparative genomics