Mapping 12 histone modifications across 12 distant species, researchers in Barcelona found the chemical signals that activate genes nearly identical since the last common ancestor, while silencing systems have been reinvented repeatedly.

Inside every cell with a nucleus, DNA is wound around proteins called histones, and chemical tags attached to those histones mark which stretches of the genome should be read and which should be kept quiet. The system is called chromatin regulation, and it dates back roughly two billion years to the last common ancestor of all eukaryotes.

Researchers at the Centre for Genomic Regulation in Barcelona have now compared that machinery across the eukaryotic tree, and found a sharp asymmetry. The instructions for turning genes on have barely changed. The instructions for turning them off have been rewritten again and again.

“The cell’s instructions for activating genes are essentially the same in a human, a sea anemone and a soil amoeba,” said Arnau Sebé-Pedrós, an ICREA Research Professor at the CRG and the study’s senior author. “But the instructions for silencing genes and other genomic elements like transposons have been continuously evolving” since the last common eukaryotic ancestor.

The work appears in Nature Genetics.

Twelve species, twelve tags

The team profiled 12 histone modifications across 12 phylogenetically diverse species, spanning amoebae, fungi, plants, algae, single-celled predators and animals. The selection deliberately included lineages largely absent from previous work, among them discobans, rhizarians, ichthyosporeans and cryptomonads. Several of the species had never had their chromatin mapped at all.

That breadth is the point. Most of what is known about chromatin comes from a handful of model organisms: humans, mice, fruit flies, yeast, a nematode, a mustard plant. They are convenient, but they cluster on a few branches of a very large tree, and conclusions drawn from them may describe those branches rather than eukaryotes in general.

“We’ve established so many new rules from looking at such few species,” said Sean Montgomery, an author on the paper. “It’s the power of looking at non-model organisms to see how evolution has brought about many differing solutions” to the same problems.

The method that made it possible

Mapping chromatin across species this different required a new technique. The team’s method, iChIP2, labels chromatin from many species with unique molecular barcodes so that all of them can be read in a single experiment, and it works from very small amounts of material.

Barcoding is what makes the comparison tractable. Profiling a dozen modifications in a dozen species separately would mean well over a hundred experiments, each with its own technical quirks; pooling barcoded samples removes much of the variation that would otherwise be mistaken for biology.

The project began in 2017, with a predecessor method used to study comb jellies and placozoans.

“We wanted to map epigenetic states in scarce cell types in mice and humans,” said David Lara-Astiaso, now at the Arc Institute in California. “Eventually, we managed to transform that precursor into a general method for mapping genome regulation across the tree of life — more streamlined, more sensitive, and finally able to handle the particularities of very different species.”

Universality proved out of reach. “We initially hoped to build a completely universal protocol, but species differ too much for that,” said Cristina Navarrete, a co-first author. “Plants and algae have cell walls that require specialised preparation, for example. Once a lab has extracted chromatin from their favourite species, iChIP2 takes over robustly, and from very small amounts of material.”

Many ways to be silent

The diversity the team found in silencing is structural, not just a matter of degree. In some species, one modification silences transposable elements while a different one silences unused genes. In others, the same modifications pile up on the same regions, with no such division of labour. In the amoeba Acanthamoeba, a mark that activates genes in animals has been repurposed to switch them off.

That last case is the most striking. The same chemical tag, in the same position on the same histone, means opposite things in two lineages, which tells you that the tag itself carries no inherent instruction. What it means depends on the proteins that read it, and those have evolved.

Histone modifications are often described as a code, with particular tags standing for particular instructions. The findings here suggest the metaphor holds only within a lineage. The writing system is shared; the vocabulary for silence is not.

Two billion years of conflict

The authors’ explanation for the asymmetry is an arms race. Genomes are under continual attack from parasitic DNA: transposable elements, or jumping genes, which copy and paste themselves into new positions, sometimes harmlessly and sometimes destructively, and endogenized viruses, viral sequences that have become permanent residents. In the human genome, transposable elements make up roughly half of all DNA.

Silencing machinery exists in large part to keep these elements quiet. But because the elements themselves evolve, they can acquire new sequences, and sometimes fragments of the chromatin machinery, allowing them to evade detection. The host must then change its methods, and the cycle continues.

Activation faces no such adversary. A gene that needs to be read does not resist being read, so a system that works has no reason to change. Silencing is a contest, and contests produce variety.

The stakes, Sebé-Pedrós said, are absolute. “If a species loses its repressive mechanisms completely, it can’t tolerate parasitic elements like transposable elements or endogenized viruses,” he said. “The result is that it’s no longer there. It’s dead.”

“Different branches of life have developed different molecular toolkits to do the same thing,” he said.

Why it matters beyond evolution

Faulty gene regulation underlies many diseases, cancers prominently among them, where genes that should be silent become active or the reverse. Knowing which features of the silencing machinery are ancient and conserved and which are lineage-specific inventions helps clarify which findings from model organisms transfer to humans and which do not.

The work also feeds into large genome-sequencing efforts such as the Earth BioGenome Project and the Wellcome Sanger Institute’s Tree of Life programme, which are producing genome sequences for enormous numbers of species. A sequence alone says little about how a genome is regulated; a method that can map regulation in any organism, from small samples, turns those sequences into something more.

---

Notes

1. Navarrete, C., et al. “Diversity and evolution of chromatin regulatory states across eukaryotes.” Nature Genetics, 2026. DOI: 10.1038/s41588-026-02672-1.
2. Center for Genomic Regulation. “Life speaks one universal language to activate genes. It speaks many more to silence them.” EurekAlert!, 3 August 2026. https://www.eurekalert.org/news-releases/1137882.