A new article (subscription required) in Ecology Letters, a prestigious research journal, outlines an ecological and evolutionary research program on the relatively new area of epigenetic inheritance. Epigenetic refers to traits that are heritable across generations but that are not encoded by DNA. The most common form of epigenetic regulation involves methylation (binding of methyl groups to DNA) of genes, thereby inactivating them. Gene methylation can be environmentally induced and, according to some recent research highlighted in the article, passed onto offspring over multiple generations (hence, epigenetic inheritance).
This probably seems pretty dry so far, so why I am I posting about it? If epigenetic inheritance is common and important, as the authors think, then it would represent a form of ’soft inheritance’. For more than five decades the vast majority of biologists have accepted that inheritance is ‘hard’, that is, encoded in a DNA molecule that does not change in response to environmental conditions (this does not mean that environment can’t effect phenotypes). Before that time, the possibility of soft inheritance was widely debated. In fact, soft inheritance predates the theory of natural selection, and is most often associated with Lamarckian evolution. While soft inheritance remained a popular idea for a long time (Darwin himself did not reject the idea, especially in later editions of The Origin), hard inheritance was codified during the ‘Modern Evolutionary Synthesis‘. Most evolutionary models since that time are predicated on hard inheritance, meaning that a plausible mechanism for soft inheritance through epigenetic variation, as the authors are keen to point out, could precipitate a substantial revision of some evolutionary processes, though it certainly wouldn’t negate the vast majority of previous research.
Admittedly, this isn’t an area of research that I keep up with, but I’ll relate my current thoughts in light of this article. The main problem with epigenetic inheritance is that there is no proven long term preservation of trait variance. Studies to date demonstrate that gene methylation can be inherited for a few generations, but no one knows beyond that. The problem is that, since (de)methylation can respond to the environment, an environmental change could completely alter the epigenetic variance of a population in a single generation, thereby negating all the epigenetic evolution of previous generations. This is not the case with hard inheritance via DNA. The authors no doubt appreciate this problem, but argue that new data will show that epigenetic inheritance can sometimes last long enough to alter long term evolutionary outcomes. One area I could imagine epigenetic inheritance being important is in invasion of new habitats. Numerous studies show that invasion success is often low because small colonizing populations are often maladapted to new territories and have little genetic variation with which to adapt to a novel environment. If adaptive epigenetic variation allows successful colonization to proceed long enough for a population to accrue genetic variation, and hence respond to new selection pressures, then it could substantially increase colonization success. The broader implication is that epigenetic inheritance might then permit species to experience and adapt to environments they otherwise would not have been able to colonize.
The scenario described above borders on wild speculation, but I write it because I think it would be an area where epigenetic inheritance could help explain a legitimate evolutionary problem. Otherwise, I am rather skeptical of its significance, given that most natural populations harbor sufficient genetic variation to explain the tempo of evolutionary change. If anything, the problem is explaining why evolution often proceeds so slowly given how rapidly populations can respond to natural selection. What is perhaps most interesting about this article is the way that new research can revive ideas which seemed thoroughly disproven in the recent past.