News: Scientists Uncover Two Million Ancient DNA Switches Controlling Plant Genes.

Published 10 AM EST, Fri Mar 20, 2026

An international project has uncovered millions of ancient DNA ‘switches’ that have been regulating plant genes for up to 300 million years – a discovery that could pave the way for more precise engineering of crop traits.

---

---

A landmark study published in Science has revealed that plant genomes contain millions of ancient regulatory DNA sequences, called conserved non-coding sequences (CNSs), that have been hiding in plain sight for decades. These sequences are not genes themselves, but the molecular switches that control when, where, and how strongly genes are expressed. Using a new computational platform that analyzed data from 284 plant species across 73 families, The Conservatory Project identified over two million of these sequences, some predating the emergence of flowering plants by more than 300 million years. The reason they went undetected for so long: plants routinely duplicate and reshuffle their entire genomes, burying the relationship between genes and their regulatory controls under layers of evolutionary noise.

The study reframes where the real power in a plant genome lives. Much like the genetic difference between humans and chimpanzees lies not in the genes themselves but in when and where those genes are activated, plant evolution is similarly governed by regulatory logic. These CNS sequences act as dials rather than switches — nudging gene expression in subtle, directional ways rather than triggering wholesale changes. This distinction matters enormously in applied breeding: editing a gene outright often produces phenotypic results too extreme to be useful, while targeting its regulatory sequences produces the kind of moderate, controllable trait shifts that agriculture actually needs.

The practical result is a publicly available dataset of thousands of regulatory elements that breeders and biotechnologists can now target with precision. The era of blunt genetic "knockouts" is giving way to fine-tuned cis-regulatory editing — modifications that shift traits like drought tolerance, pest resistance, and yield in agriculturally meaningful directions without the developmental dysfunction that comes from removing a gene entirely. As the researchers put it, these sequences were once dismissed as junk DNA. They are now understood to be the instruction manual for plant development.

For cannabis genetics, this research opens a meaningful new frontier. Traits central to commercial performance; terpene consistency, cannabinoid accumulation, stress-triggered hermaphroditism, auto-flowering timing, resin density — are not controlled by simple on/off genes. They are the product of finely regulated expression across development, which is precisely where CNSs operate. Identifying the conserved regulatory elements governing these traits would allow breeders to move from phenotype-first selection toward genotype-informed breeding: understanding why a plant expresses a trait, not just that it does, and reproducing it reliably at scale. For operations already built around systematic genetic quality control, regulatory genomics represents the next logical layer — the science that begins to explain the patterns breeders have been documenting for years.

Source: University of Cambridge