Study: Adaptive tracking with antagonistic pleiotropy results in seemingly neutral molecular evolution (DOI: 10.1038/s41559-025-02887-1)
For a long time, evolutionary biologists have thought that the genetic mutations that drive the evolution of genes and proteins are largely neutral: they're neither good nor bad, but just ordinary enough to slip through the notice of selection.
Now, a University of Michigan study has flipped that theory on its head.
In the process of evolution, mutations occur which can then become fixed, meaning that every individual in the population carries that mutation. A longstanding theory, called the Neutral Theory of Molecular Evolution, posits that most genetic mutations that are fixed are neutral. Bad mutations will be quickly discarded by selection, according to the theory, which also assumes that good mutations are so rare that most fixations will be neutral, says evolutionary biologist Jianzhi Zhang.

The U-M study, led by Zhang, aimed to examine whether this was true. The researchers found that so many good mutations occurred that the Neutral Theory cannot hold. At the same time, they found that the rate of fixations is too low for the large number of beneficial mutations that Zhang's team observed.
To resolve this, the researchers suggest that mutations that are beneficial in one environment may become harmful in another environment. These beneficial mutations may not become fixed because of frequent environmental changes. The study, supported by the U.S. National Institutes of Health, was published in Nature Ecology and Evolution.
"We're saying that the outcome was neutral, but the process was not neutral," said Zhang, U-M professor of ecology and evolutionary biology. "Our model suggests that natural populations are not truly adapted to their environments because environments change very quickly, and populations are always chasing the environment."
Zhang says their new theory, called Adaptive Tracking with Antagonistic Pleiotropy, tells us something about how well all living things are adapted to their environments.
"I think this has broad implications. For example, humans. Our environment has changed so much, and our genes may not be the best for today's environment because we went through a lot of other different environments. Some mutations may be beneficial in our old environments, but are mismatched to today," Zhang said.
"At any time when you observe a natural population, depending on when the last time the environment had a big change, the population may be very poorly adapted or it may be relatively well adapted. But we're probably never going to see any population that is fully adapted to its environment, because a full adaptation would take longer than almost any natural environment can remain constant."
The Neutral Theory of Molecular Evolution was first proposed in the 1960s. Previously, scientists studied evolution based on the morphology and physiology, or appearance, of organisms. But starting in the 1960s, scientists were able to start sequencing proteins, and later, genes. This prompted researchers to look at evolution at the molecular level.
To measure beneficial mutation rates, Zhang and colleagues investigated large deep mutational scanning datasets produced by his and other labs. In this kind of scanning, the scientists created many mutations on a specific gene or region of the genome in model organisms such as yeast and E. coli.
The researchers then followed the organism over many generations, comparing them against the wild type, or the most common version existing in nature, of the organisms. This allowed the researchers to measure their growth and compare their growth rate to the wild type, which is how they estimated the effect of the mutation.
They found that more than 1% of mutations are beneficial, orders of magnitude greater than what the Neutral Theory allows. This amount of beneficial mutations would lead to more than 99% of fixations being beneficial and a rate of gene evolution that is much higher than the rate that is observed in nature. The researchers realized they had made a mistake in assuming an organism's environment remained constant.
To investigate the impacts of a changing environment, Zhang's research team compared two groups of yeast. One group evolved in a constant environment for 800 generations (each generation lasted 3 hours), while the second group evolved in a changing environment, in this case composed of 10 different kinds of media, or solution, that the yeast grew in. The second yeast group grew in the first media for 80 generations, in the second media for another 80 generations, and so on, for a total of 800 generations as well.
The researchers found that there were far fewer beneficial mutations in the second group compared to the first. Although the beneficial mutations occurred, they didn't have a chance to become fixed before the environment shifted.
"This is where the inconsistency comes from. While we observe a lot of beneficial mutations in a given environment, those beneficial mutations do not have a chance to be fixed because as their frequency increases to a certain level, the environment changes," Zhang said. "Those beneficial mutations in the old environment might become deleterious in the new environment."
However, Zhang says there is a caveat: The data they used came from yeast and E. coli, two unicellular organisms in which it's relatively easy to measure the fitness effects of mutations. Deep mutational scanning data collected from multicellular organisms would tell whether their findings from unicellular organisms apply to multicellular organisms such as humans. Next, the researchers are planning a study to understand why it takes so long for organisms to fully adapt to a constant environment.
Other authors of the study include former U-M graduate students Siliang Song and Xukang Shen and former U-M postdoctoral researcher Piaopiao Chen.