On the surface, that doesn’t make any sense. Both versions of the gene encode exactly the same amino acid. How could one possibly be less fit than the other?
The secret to understanding this is remembering that the gene’s DNA isn’t used directly to make a protein. Instead, it’s transcribed into an RNA copy called a messenger RNA, and that is directly translated to make the protein. And alterations in the DNA can affect the RNA’s three-dimensional structure, its stability in the cell, and the rate it’s translated into protein.
The researchers found that the mutations expected to be neutral often influenced the amount of messenger RNA present in the cell. They also appeared to influence the RNA’s ability to fold into a three-dimensional shape. And they were also likely to affect the efficiency of the messenger RNA in translating into a protein. Combined, these could account for why this group of mutations had a collective impact on fitness.
So, does that mean it’s time to throw out our idea that mutations in a gene that don’t alter its protein sequence are neutral? And with it, all the tools we use to study protein evolution that are based on this assumption?
The researchers give one major reason why this would be premature: Yeasts are kind of weird. To start, unlike animals, which mostly get a copy of a gene from their moms and another from their dads, yeast carries only one copy of every gene, so it will likely be more sensitive to subtle effects. Yeasts also live a lifestyle similar to bacteria, carrying a simplified genome and focusing on rapid reproduction—a relatively minor metabolic hit is more likely to slow them down.
And these effects are still very subtle. Even if you completely hammered the function of these proteins by creating a mutation that truncated the protein early, the fitness cost was slight (fitness was 0.94 of the yeast strain without any mutations). It’s not even clear this behavior would occur in other genes in yeast, much less genes in other organisms.
The other thing that the researchers note is that, in actual populations that are evolving over the long term, evolutionary pressures are constantly shifting due to environmental changes. So, it could be that these mutations are effectively neutral in a realistic environment, so when we look at similar mutations in a natural population, they appear to be neutral.
All these results are an important caution and make it worth the time and effort needed to sort this out carefully.