Owing to the nature of the genetic code, many mutations within the exons of protein-coding genes do not alter the amino acid content of the protein. Up until a few years ago it was commonly assumed that such synonymous mutations in mammals were neutrally evolving, i.e., chance alone determined their fate. This fitted both with theory and also with some tests.

“ ...we not only provide the evidence that selection acts on synonymous mutations, we show what the likely mechanisms are and why prior tests for selection failed to see any evidence.”

What the paper shows is that this assumption is no longer valid. Importantly, we also show how the selection acts and how this has implications for understanding human disease. It is probably because the paper inverts the prior understanding that it has attracted some attention.

The paper is a synthesis of work that, when taken together, challenges the orthodoxy. It is a bit more than just synthesis, however, in that we not only provide the evidence that selection acts on synonymous mutations, but we also show what the likely mechanisms are and why prior tests for selection failed to see any evidence. In short, the mechanism of possible selection that had always been assumed and tested for, is not the underlying cause in mammals.

The paper has importance on a number of fronts. First, there has been an extensive debate concerning the roles of selection and chance in molecular evolution. For many years now the neutralist (chance dominates) school, while initially in the ascendancy, has been in retreat. One of the few classes of mutation that were thought to be exclusively neutrally evolving was synonymous mutations in mammals. We now show that these too can be subject to selection and why. This rolls back the front line of the debate but also has importance as the assumption of neutrality of synonymous mutations has been important in estimates of how often mutation occurs.

Second, the work has implications for understanding human genetic disease. It was often assumed that synonymous mutations could not be candidates for human genetic diseases. The paper shows that this isn’t true and many diseases are in fact owing to such changes. We also now understand why the prior assumption was wrong. Third, the new understanding should allow us to better design novel genes to be employed in genetic engineering.

I had been one of those who regularly assumed synonymous mutations must be neutral. I had for a while been worried about the assumption, so with a new graduate student (Jean-Vincent Chamary), invented a novel test to see if the assumption was sound. Much to my surprise, we discovered that evolution at synonymous sites had a very different profile to evolution in flanking introns, not in terms of the rates of evolution, but in terms of which bases were conserved.

The only major problem we encountered along the way was from the editor of the journal to whom we submitted the paper. He asserted that it is well known that synonymous mutations in mammals are neural and therefore initially rejected it, until we managed to convince him that this wasn’t exactly known.

This led us to wondering about what the mechanism might be. I think it was when we could show evidence for discrete mechanisms that people began to be convinced.