Abstract
DNA provides the fundamental framework for heritability, yet heritable trait variation need not be completely 'hard-wired' into the DNA sequence. In plants, the epigenetic machinery that controls transposable element (TE) activity, and which includes DNA methylation, underpins most known cases of inherited trait variants that are independent of DNA sequence changes. Here, we review our current knowledge of the extent, mechanisms and potential adaptive contribution of epiallelic variation at TE-containing alleles in this group of species. For the purpose of this review, we focus mainly on DNA methylation, as it provides an easily quantifiable readout of such variation. The picture that emerges is complex. On the one hand, pronounced differences in DNA methylation at TE sequences can either occur spontaneously or be induced experimentally en masse across the genome through genetic means. Many of these epivariants are stably inherited over multiple sexual generations, thus leading to transgenerational epigenetic inheritance. Functional consequences can be significant, yet they are typically of limited magnitude and although the same epivariants can be found in nature, the factors involved in their generation in this setting remain to be determined. On the other hand, moderate DNA methylation variation at TE-containing alleles can be reproducibly induced by the environment, again usually with mild effects, and most of this variation tends to be lost across generations. Based on these considerations, we argue that TE-containing alleles, rather than their inherited epiallelic variants, are the main targets of natural selection. Thus, we propose that the adaptive contribution of TE-associated epivariation, whether stable or not, lies predominantly in its capacity to modulate TE mobilization in response to the environment, hence providing hard-wired opportunities for the flexible exploration of the phenotypic space. This article is part of the theme issue 'How does epigenetics influence the course of evolution?'