Evolution Put No Extra Locks on Vital Genes, a Study in E. coli and Yeast Finds
A 1972 idea held that evolution has lowered the mutation rate of the genes cells cannot live without. A new analysis of E. coli and baker's yeast finds no sign of it.
Genes that cells cannot live without do not appear to mutate less often than other genes in the gut bacterium E. coli or in baker's yeast, according to a peer-reviewed study that challenges an idea about evolution first proposed in 1972.
The paper, by Sofia Radrizzani and Laurence D. Hurst of the University of Bath, appeared in PLOS Biology on November 12, 2026. [1] It tests the hypothesis that natural selection has favored lower mutation rates in a genome's most important genes, as if evolution had put extra locks on the vault. In these two organisms, the study finds no sign of the lock.
An idea with a long life
The proposal goes back to Cox in 1972 and has been revived since. The University of Bath said it most recently resurfaced in 2023. [2] The logic is appealing. A mutation in an essential gene can be fatal to a cell, so a lineage that copies those genes more carefully ought to leave more descendants.
The paper's introduction lists the evidence that kept the idea alive. One claim, from the plant Arabidopsis, held that essential genes mutate at up to half the rate of other gene classes. Another held that bacterial genes on one of the two strands of DNA, the lagging strand, mutate more often than those on the other. [1]
Two steps that are easy to confuse
The trouble is that mutation and selection are separate steps, and researchers usually see only the second. A mutation arises, and then natural selection decides whether it persists. Count the genetic changes that survive in a population and the harmful ones in essential genes may already be gone. Those genes then look as if they mutate less, even if the changes arose at the same rate. [1][2]
It is like counting cracks in the bridges still standing and concluding that the ones that fell never had any.
Hurst described the effect in his own term. Harmful mutations in important genes, he said, may vanish through natural selection before they can be sampled, leaving what he called "ghost mutations." [2]
Looking before the filter
To get around the problem, the Bath team turned to mutation-accumulation data. These experiments are designed to reduce selection's power to remove harmful mutations, so that changes can be counted closer to the moment they arise. [2]
No such design is perfect, so the authors ran three tests for signs that selection had still skewed the results. They found no evidence that it had done so substantially. [2]
The result was plain. The researchers found no lower mutation rate for essential genes in either E. coli or yeast. Essentiality was judged from experiments that remove genes one at a time and measure what happens to the cells. [1][2]
Radrizzani was blunt about what that means for the older claims.
"The claim that mutation is slower for important genes is a triumph of hype over substance," she said. [2]
Other proxies for importance fail too
The team also tried other ways of measuring how much a gene matters. Genes that make more abundant proteins did not show lower mutation rates in either species. Neither did genes with stronger signals of purifying selection over evolutionary time, meaning those whose sequences have changed least across species. [1][2]
There was one small exception. The paper reports a slight link between higher transcription, the copying of a gene into RNA, and lower mutation rates in E. coli. It did not appear in yeast. The abstract concludes that the results do not support a consistent effect of gene importance across the two species. [1]
The bacterial strand effect fared no better. In E. coli the researchers found no difference in mutation rate between genes on the two replication strands, and the paper says the same absence has been found in other prokaryotes, the group of single-celled organisms without a nucleus that includes bacteria. [1]
The paper also flags analytical problems in some earlier work, along with results from strains lacking mismatch repair, the cell's proofreading machinery for DNA copying errors, that do not line up with each other. [1]
Mutation is not uniform
The study does not say every stretch of DNA mutates equally. It points to other features that track with mutation rate. The clearest is replication timing: genes copied later in the cell's DNA-copying cycle tend to mutate more often, particularly in yeast. [1][2]
That shifts the question from whether evolution tunes mutation to protect important genes toward which physical features of the genome push mutation up or down. The researchers' evidence says the second is the better lead in these organisms.
What it does not settle
The work covers E. coli and baker's yeast, not people. It does not establish that essential genes in humans or in other organisms mutate at the same rate as their neighbors, and it did not test human mutation rates directly. [1]
That leaves the plant claim and the wider question open. Whether the pattern holds in more species, especially plants and animals, is the obvious next test. So is whether other experimental designs, free of the biases the paper identifies, reproduce its result. Researchers also still need to explain why transcription shows a faint link to mutation in E. coli but none in yeast. [1]
Every edition in brief, three times a day, on our Telegram channel, on Bluesky and on Threads.




