That’s negation in the “contradicting” sense of the term, rather than as in nullification or reversal. That being said, Jeffrey Tomkins’ headline today is “Plant Epigenome Research Negates Evolution,” which in theory could mean that epigenetics is acting to actively prevent the changes that evolution is creating. This is not, however, the case – at least not here.
Biological research involves a lot of “model organisms,” one of which is the thale cress, Arabidopsis thaliana. The paper that Tomkins is talking about today - Patterns of population epigenomic diversity (open access) – compares the patterns of DNA methylation of thale cress plants adapted to different environments, which they found to be much larger than they expected. DNA methylation involves the attachment of a methyl group to a base of DNA, which could be thought of as acting as a speed bump for transcription, slowing it down or stopping it entirely but in a way that can be undone if the group is removed. Eukaryotes like animals and plants seem to use every potential mechanism available to regulate gene expression, and this is no exception. And just as the sequence of As, Gs, Cs, and Ts can be determined and is called a genome, so too can the pattern of methylation within that genome be mapped – this is your “epigenome.”
Preventing genes from being expressed in this way – which can be specific to the organism and even to the cell type – can have an important effect on the phenotype of the organism as a whole, though this study does not investigate what they are. Tomkins talks at length about how all this “presents a number of very serious problems for Darwinian evolution.”
First, the methylation of DNA is not a random event—it involves a complex array of molecular machines (proteins and RNAs) that attach the methyl tags according to environmental signals that are placed at specific DNA addresses all over the genome.
Yes, it’s true that the methylation is not “random,” but this is what would be expected from the product of evolution also. For his second point Tomkins adds that “complex cellular machinery and infrastructure” are also needed for the methylation to be interpreted. I would say that this is only partially true: methylation is always going to have an effect of some kind, but further “cellular machinery” can add to this. As for Tomkins claim that there are “environmental signals … placed at specific DNA addresses,” I’m not sure that’s correct.
Third, for all of this to work as the plant grows and makes seed to reproduce, there needs to be another level in the system to ensure that the chemical DNA tags are accurately copied along with the DNA when it is replicated to produce new cells. This is especially important in the reproductive cells, so that the next generation of plants have the same adaptive system.
Importantly, the requirement that the methylation be heritable is necessary for evolution to be able to act upon them – if this condition were not the case we might actually be closer to something that truly does “negate evolution.”
Fourth, not only does this highly complex “all or nothing” system have a zero probability of evolving through gradualistic DNA mutations, but it also presents problems for the idea of natural selection acting upon it. If the plant is presenting a system of adaptation that is, to a large extent, insulated from direct selection on so-called positive DNA sequence mutations (an exceptionally rare occurrence), then how can evolution progress?
Tomkins seems to think that natural selection, and thus evolution, cannot act on methylation – this is patently false, and is contradicted by the very paper he is talking about which says:
DNA methylation is a covalent base modification of plant nuclear genomes that is accurately inherited through both mitotic and meiotic cell divisions. However, similarly to spontaneous mutations in DNA, errors in the maintenance of methylation states result in the accumulation of single methylation polymorphisms (SMPs) over an evolutionary timescale. The rates of SMP formation are orders of magnitude greater than those of spontaneous mutations, which are in part, probably due to the lower fidelity of maintenance DNA methyltransferases and accompanying silencing machinery. Epiallele formation in the absence of genetic variation can result in phenotypic variation, which is most evident in the plant kingdom, as exemplified by the peloric and colorless non-ripening variants from Linaria vulgaris and Solanum lycopersicum, respectively. Although rates of spontaneous variation in DNA methylation and mutation can be decoupled in the laboratory, in natural settings, these two features of genomes co-evolve to create phenotypic diversity on which natural selection can act.
So, there is heritable variation, in the form of “SMPs” (analogous to mutations, but accumulating much more rapidly), and this variation creates diversity in the organism as a whole which can be acted upon by natural selection. Evolution is therefore possible – all this from just the first paragraph of the introduction of the paper!
There are, however, two other aspects to Tomkins argument. The first is that he claims that this is “all or nothing” – perhaps one pattern can evolve into another, but you can’t get from no methylation to the kinds of patterns observed here in the thale cress? It’s true that if you remove all methylation from a mouse you’ll kill it, but that just means that it has come to rely on its existence and not that it cannot slowly build up over time. Properly answering this question would require information that this study does not provide – you would need to look instead at organisms that have no methylation, or just a little, and those that can safely have it removed. As it is we certainly don’t have the information required to say that methylation could not have evolved from nought.
The other aspect is just typical Tomkins:
Clearly, the precise and timely regulation of environment-responsive gene networks contains yet one more layer of bio complexity that Darwinian evolution cannot account for.
“Ooh, it’s complex, and evolution can’t explain complexity!” – that, capped off with a “clearly” for good measure. I call bullshit.
As it happens, this DpSU has bearing on an earlier one that Tomkins doesn’t spot. Back in October Mr Thomas pointed at a study comparing the amount of methylation in chimpanzees and humans. At the time he said:
If humans and chimps are close relatives, then they should have similar DNA methylation patterns in the areas of chromosomes that they have in common such as similar gene sequences. However, this team found major differences.
He argued that this difference means that humans and chimps are not related. However, this new study shows that there are “major differences” even in members of the same (plant) species! This is no barrier for evolution by any means.