Estimating Epigenetic Mutation Rates

Generation-spanning maps of Arabidopsis thaliana DNA methylation allow researchers to compute how quickly epigenetic marks appear and disappear in the plant’s genome.

May 11, 2015
Jenny Rood


Epigenetic mutations, such as changes in the patterns of DNA methylation, occur much often more in the thale cress (Arabidopsis thaliana) genome than genetic mutations that alter the DNA bases, but at a low enough rate to be subject to natural selection, according to a study published today (May 11) in PNAS.

Methylation on the DNA base cytosine can regulate the expression of genes and transposable elements in organisms including the model plant A. thaliana, in which 14 percent of the cytosines are methylated. While some methylation marks are maintained from generation to generation in A. thaliana, others are rapidly acquired or lost over time.

To understand if these methylation changes can be subject to natural selection, researchers from the University of Groningen in the Netherlands, the University of Georgia, and the University of Minnesota mapped cytosine methylation at single-base resolution over as many as 32 generations of three different lines of the plant. The scientists then designed a model to estimate how quickly methylation at a particular location was gained or lost between generations and fit the model to their experimental data on the known locations of the methylated cytosines. They found that methylation changes are common throughout the genome. Some regions were more likely to lose methylation than to gain it, while others, such as transposable elements, were as much as 30 times more likely to gain methylation than lose it.

“Epigenetic mutations are about 100,000 times more likely than DNA sequence mutations,” study coauthor Frank Johannes of Groningen said in a statement. Importantly, however, the epigenetic mutation rate is still low enough to be subject to natural selection, the authors wrote in their paper. With the first estimates of cytosine methylation gain and loss in A. thaliana, “we are now in a position to quantify these dynamics precisely on a genome-wide scale,” Johannes added.