Science1 publisher2 min readPublished
Duplicated salmon and trout genes changed expression least while embryo organs formed
Researchers led from Edinburgh's Roslin Institute found duplicated salmon and trout genes changed expression least in late embryos, across nearly 800 datasets. The study's lead, Dan Macqueen, ties that window to the stage when fish, frog and human embryos look most alike.
The Scientist · Science desk

What happened
- The salmonid ancestor's genome doubled around 100 million years ago, and many duplicated genes are still being retained, lost or repurposed, said Sigbjorn Lien of the Norwegian University of Life Sciences.
- Macqueen, the study's overall lead at the Roslin Institute, said expression change after the duplication depended strongly on life stage and varied across tissues.
- The sequencing came through AQUA-FAANG, a 6 million euro EU-funded collaboration of research organisations and industry partners across Europe.
- The data are freely available, and Macqueen presents them as a resource for future research and for breeding strategies in salmon and trout aquaculture.
Compiled by The ScientistSomething wrong?How this is made
Why it matters
- constraint Searches for duplicate copies that took on new roles are better aimed at other life stages and tissues, since regulatory change in late embryos appears to have been removed by selection.
- capability With the datasets free, a group studying one salmon or trout gene can check which copy of a pair is active in which tissue and stage without paying for new sequencing.
- precedent Macqueen's timing match sets up the same test in other lineages with duplicated genomes, to see whether the late-embryo constraint shows up outside salmonids.
A duplicated gene starts as two identical copies. Over time one of them can change how or when it is switched on, and that shift in expression is where a new role would begin [11]. The study tracked that kind of change, following when and where genes switch on and off across tissues and early development in Atlantic salmon and rainbow trout [4].
Age is the reason to use these fish. The duplication in the ancestor of all living vertebrates is more than 500 million years old, and at that age many details are hard to recover [2]. It is more than five times as old as the salmonid event [2]. Lien said the salmon and trout genomes "contain clearer traces of WGD compared to most other vertebrates, and thereby offer a unique window into how evolution unfolds after WGD" [3].
The central result is about where the copies were held in place. Duplicated genes had the least scope to change expression during late embryonic development, when organs are forming quickly [8]. The researchers' explanation is that a change to one gene at that stage can affect several parts of the embryo, so harmful changes are more likely to be removed by natural selection [9].
Macqueen said this period of "constraint" matches the period when vertebrate species from fish to frogs to humans "appear most similar in terms of morphology, suggesting fundamental rules are at play" [10]. That puts two observations side by side in time: low expression divergence in two species of one fish family, and peak body-plan similarity across vertebrates. I think the match is suggestive. On its own it is a correlation, and it does not show that a single constraint produces both.
The scale figures measure depth of sequencing. The 80 billion sequence pairs Lien cited [5] work out to about 100 million pairs per dataset on average [1]. They are spread across two related species. A paper listed with the release covers early embryogenesis in the duplicated Atlantic salmon genome [13].
On this evidence, salmonids hold up as a place to watch duplicates take on new jobs: the event is young, and the data resolve change by life stage and by tissue. The step from a change in expression to a new trait is still ahead. The release does not report how many duplicate pairs diverged, by how much, or any trait traced to a repurposed copy.
What to watch
- The full paper's counts of how many duplicate gene pairs diverged in expression, by how much, and in which tissues.
- Follow-up work that traces a specific salmon or trout trait to a repurposed duplicate copy.
- Use of the AQUA-FAANG maps in salmon or trout breeding programmes, the application Macqueen proposes.