The story of Col-0 begins nearly a century ago in the European countryside. In the first half of the twentieth century, the German botanist Friedrich Laibach, who pioneered the use of Arabidopsis as a model organism, collected wild seeds near the city of Landsberg, a location that today corresponds to Gorzów in Poland. Among the material he gathered were the ancestors of what would eventually become one of the most important genetic resources in plant science. Laibach’s collections provided the raw stock from which later generations of researchers would select and stabilize the strains that now populate cold storage facilities and growth chambers across the globe.

The modern chapter of that story opened in the second half of the century, when George Rédei, a Hungarian-born geneticist who had moved to the United States, established his laboratory at the University of Missouri in Columbia. Working with material derived from the original collections, Rédei established the Arabidopsis line that took its name from its home city. He chose it deliberately: the Columbia line was stable and vigorous, qualities that made it ideal for the repetitive, demanding work of laboratory genetics. That decision, made on practical grounds, would echo through decades of plant biology, because Columbia-0 was ultimately selected as the very first plant genome to be sequenced, a landmark effort completed with its final publication in the year 2000.

From its humble beginnings as a wild plant gathered from the European countryside to its status as the laboratory workhorse of modern plant science, the history of Col-0 thus spans an entire century. During that century, the strain has been propagated in countless laboratories, passed between collaborators, deposited in stock centres, and redistributed to new generations of researchers. Each transfer and each round of seed multiplication represents an opportunity for change: a DNA replication error here, a spontaneous mutation there, a shift in the pattern of chemical tags that decorate the genome and silence or activate genes without altering the underlying sequence.

To understand just how much change has accumulated, the new study took an unusually comprehensive approach. The researchers assembled 78 supposedly identical Col-0 strains obtained from various laboratories and stock centres around the world, each with a documented propagation history. For every strain, they performed whole-genome sequencing and methylome sequencing, the latter revealing the distribution of methyl groups across the DNA, the primary form of epigenetic modification in plants. This dual strategy allowed them to locate both genetic mutations, which alter the sequence of DNA letters, and epimutations, which alter the pattern of methylation while leaving the sequence intact. For a selected subset of lines, the team added a transcriptome analysis, examining how the accumulated differences translated into changes in gene activity.

The results were striking in their detail. Reminiscent of earlier studies on the accumulation of genetic changes in natural populations, the analysis identified many single-nucleotide polymorphisms across the collection of supposedly identical strains. These were not limited to silent changes in non-coding regions: the researchers found non-synonymous substitutions that alter the amino acid sequences of proteins, and in some cases high-impact mutations affecting gene start or stop codons, the signals that define where a gene begins and ends. Such mutations have the potential to truncate or extend proteins dramatically, and their presence in laboratory strains demonstrates that even the most carefully maintained genetic resources are not frozen in time.

Intriguingly, however, these high-impact variants appeared at a lower rate than would be expected by chance. The most likely explanation, the researchers suggest, is a form of purifying selection imposed not by nature but by researchers themselves over the years. Lines of Col-0 that show obvious phenotypic divergence, such as altered growth, development, or morphology, are generally discarded by observant scientists who recognize that something has gone wrong with their stock. This ongoing, largely unintentional culling acts as a filter, removing the most damaging mutations before they can spread through the community. What remains, and what quietly accumulates instead, are the small-effect mutations and epimutations that do not produce conspicuous symptoms.

That is precisely where the danger lies for experimental reproducibility. Small-effect changes can slowly build up in a lineage and, over many generations, end up subtly modifying phenotypes in ways that are difficult to detect by eye but large enough to influence experimental outcomes. A gene whose expression differs slightly between two laboratories’ Col-0 stocks could alter a plant’s response to stress, its flowering time, or its immune reactions, confounding comparisons between results obtained in different places. The study’s methylome data show that epigenetic variation, which can arise and be inherited far more readily than DNA sequence changes, adds a further layer of divergence that standard genotyping would entirely miss.

Perhaps the most remarkable result of the analysis is its reconstruction of the strain’s genealogy. Using molecular dating and a calibrated genetic lineage reconstruction based on both genetic and epigenetic changes, the researchers traced the relationships among the 78 strains back to a most recent common ancestor living roughly 55 to 60 generations ago. That timescale corresponds almost exactly to about one generation per year since the establishment of the Columbia dynasty, a rate that matches the practical tempo of laboratory propagation, in which seeds are typically multiplied and passaged on an annual cycle. In effect, the family tree of Col-0 written in its genomes recapitulates the written history of its distribution, a striking convergence of molecular and documentary records.

The broader lesson extends well beyond Arabidopsis. Col-0 was chosen as the first plant genome to be sequenced precisely because it was assumed to be a stable, uniform reference, and it has served that role for more than two decades. But the new findings demonstrate that the notion of a single canonical Col-0 is an idealization. Every vial of seeds in every stock centre represents a snapshot of a lineage that has been quietly evolving under laboratory conditions, shaped by drift, by methylation changes, and by the well-meaning but incomplete selection of the researchers who maintain it. For the plant science community, the practical message is clear: documenting strain provenance, periodically re-sequencing key stocks, and being transparent about which Col-0 derivative was used in any given study are no longer optional refinements but essential safeguards for reproducibility. For biology at large, the century-long story of Columbia-0 offers a vivid reminder that even our most cherished laboratory standards are living things, and that living things, given time, always change.

Subject of Research: Genetic and epigenetic divergence among laboratory strains of the Arabidopsis thaliana reference line Columbia-0 (Col-0)

Article Title: Col-0 genetic divergence

Article References: Col-0 genetic divergence. (n.d.). https://doi.org/10.1038/s41477-026-02417-2

Image Credits: AI Generated

DOI: 10.1038/s41477-026-02417-2

Keywords: Arabidopsis thaliana, Col-0, genome sequencing, methylome, epimutations, single-nucleotide polymorphisms, purifying selection, plant genetics, laboratory strains, reproducibility, molecular dating, transcriptome