A correspondent asked me about a recent paper in the journal Nature, “Mutation bias reflects natural selection in Arabidopsis thaliana,” aka the commonly studied flowering weed, thale cress. The abstract states, “Since the first half of the twentieth century, evolutionary theory has been dominated by the idea that mutations occur randomly with respect to their consequences. Here we test this assumption with large surveys of de novo mutations in the plant Arabidopsis thaliana.” They show that “epigenome-associated mutation bias reduces the occurrence of deleterious mutations in Arabidopsis, challenging the prevailing paradigm that mutation is a directionless force in evolution.”
That mutation is “directionless” or “random” is a traditional axiom of evolutionary biology. In fact, the paper states, “The random occurrence of mutations with respect to their consequences is an axiom upon which much of biology and evolutionary theory rests.” My correspondent wanted to know what it means to consider that some mutations may be “non-random” after all. She supposed that she was asking a “dumb question.”
Exactly the Question to Ask
Actually, it’s not in the least a dumb question — it’s exactly the right question to ask! In the context of this paper, what “non-random” means is that mutations are less likely to occur in gene-coding DNA — especially in what they call “essential genes.” This overturns two standard assumptions of the modern theory of evolution.
In evolutionary biology, it’s generally thought that mutations are “random” in two respects:
- Mutations occur with equal likelihood across the entire genome. So there’s no part of the genome that is MORE or LESS likely to experience mutations than any other part of the genome. This is supposed to mean mutations are not directed or concentrated, but in a sense are randomly distributed across the genome.
- Mutations occur without regard to the needs of the organisms, meaning they are random and not directed for or against what the organisms needs to survive.
The Nature study found evidence against both (1) and (2). In Arabidopsis, some parts of the genome are LESS likely to experience mutations, and those parts of the genome that experience fewer mutations tend to be the REALLY important parts of the genome that you wouldn’t want to be mutated because in those sections, mutations would most likely break genes that are very important to the organism.
A Look at the Specifics
Now let’s get into more specifics. In the genomes of most higher organisms, only a small percentage of the DNA represents genes that encode proteins. The Nature study found that sections of the Arabidopsis genome that encode genes are LESS likely to experience mutations than the “intergenic” regions — the sections of the genome between genes that don’t encode proteins. They found that “the frequency of mutation was 58% lower in gene bodies than in nearby intergenic space.”
They further found that “essential genes,” such as those basic genes responsible for translation (e.g., converting the information in DNA into proteins), had even LOWER mutation rates compared to other genes that had more specialized functions.
Please also note this important point: The study was able to directly measure mutations after they occurred in the plant but before mutations could have been affected by natural selection, which might “weed out” certain mutations that have deleterious effects. So the authors think they have provided a true and accurate measure of mutations as they occur in the DNA.
Or to put it another way, mutations don’t occur randomly in the sense that some parts of the genome are less likely to experience mutations than other parts of the genome. Instead, in a sense mutations DO occur with respect to the needs of the organism. I don’t mean that the non-randomness of mutations identified in this study could help organisms build new complex traits. That’s not indicated. Rather, the non-randomness of mutations seems to be designed to minimize mutations in the places where they would do the most damage to the organism’s basic functions.
Implications for Evolutionary Biology
The implications for evolutionary biology are profound. If mutations aren’t equally distributed across the genome, and aren’t random with respect to the needs of the organism, then two basic tenets of the standard neo-Darwinian model are false. This also could spell trouble for neo-Darwinism because it suggests that mutation rates are lowest in areas where mutations would presumably be needed to foster evolution — i.e., they are lowest in the genes.
If mutation rates are low in the gene-coding DNA, then it will take even longer for new complex traits to arise by mutating functional genes. This exacerbates what Darwin-skeptics call the “waiting time” problem, where it takes too long for necessary mutations to arise — far longer than the amount of time allowed by the fossil record.
Return of the Waiting-Time Problem
Intelligent design proponents have already identified the waiting time problem as a fundamental mathematical obstacle to neo-Darwinian evolution. Our colleagues published a paper in the Journal of Theoretical Biology last year, “On the waiting time until coordinated mutations get fixed in regulatory sequences,” which did mathematical modelling of the waiting time to generate traits requiring N mutations to provide an advantage. The paper found a serious challenge to neo-Darwinism:
[T]he fossil record is often interpreted as having long periods of stasis, interrupted by more abrupt changes and “explosive” origins. These changes include, for instance, the evolution of life, photo-synthesis, multicellularity and the “Avalon Explosion”, animal body plans and the “Cambrian Explosion”, complex eyes, vertebrate jaws and teeth, terrestrialization (e.g., in vascular plants, arthropods, and tetrapods), insect metamorphosis, animal flight and feathers, reproductive systems, including angiosperm flowers, amniote eggs, and the mammalian placenta, echolocation in whales and bats, and even cognitive skills of modern man. Based on radiometric dating of the available windows of time in the fossil record, these genetic changes are believed to have happened very quickly on a macroevolutionary timescale. In order to evaluate the chances for a neo-Darwinian process to bring about such major phenotypic changes, it is important to give rough but reasonable estimates of the time it would take for a population to evolve so that the required multiple genetic changes occur. [Internal citations omitted.]
Following the standards of the field, the study in Journal of Theoretical Biology adopted standard evolutionary assumptions that mutations are random — i.e., equally likely across the entire genome and occurring without respect to the needs of the organism. But the new study in Nature suggests that both these assumptions are false — and false in a way that probably makes it harder for neo-Darwinism to evolve new traits.