Accèder directement au contenu

Extent and phenotypic impact of TE-mediated transgenerational epigenetic variation

TE sequences are now widely recognized as functionally important components of the genome with a pervasive impact on its activity. Notably, TEs can bring the expression of neighbouring genes under their own control and create epiallelic variation because of the epigenetic silencing mechanisms, including DNA methylation, that target TEs to limit their mobilization. As a matter of fact, known epiallelic variants with phenotypic consequences in plants and mammals all implicate changes in DNA methylation at TE sequences located adjacent to or within genes. However, the contribution of TE-associated epigenetic variants to the phenotypic diversity encountered in nature remains unknown because of the confounding effects of DNA sequence polymorphisms.

To overcome this problem, we have created a large population of epigenetic recombinant inbred lines (epiRILs), starting with a cross between two individuals with near identical genomes but contrasted so-called methylomes. Specifically, one parent exhibits a normal DNA methylation pattern whereas the other parent has reduced DNA methylation over TE sequences and their immediate flanks. Using this epiRIL population, we could demonstrate that approximately a third of the differentially methylated regions (TE- DMRs) are stably inherited across at least 8 generations, with phenotypic consequences. In contrast, the remaining two thirds of the parental TE-DMRs reverted to the normal methylated state, at variable rates. Thus, TE-DMRs differ widely in their epigenetic properties, and we showed that differences are in large part determined by the efficiency with which individual TE sequences are targeted by the RNA directed DNA methylation (RdDM) machinery.

In addition, many of our experimentally-induced TE-DMRs overlap with those present in nature. Moreover, while a natural counterpart could be found for 40% of the stable TE-DMRs, the percentage drops to 10% for the TE-DMRs that revert at a high rate in the epiRIL population. These last two findings provide strong evidence that many natural epialleles are truly epigenetic, as they can exist independently of the sequence polymorphisms they are associated with (in cis or trans) in the wild. In addition, these bona fide natural epialleles likely differ in their stability primarily as a result of intrinsic differences in their targeting by the RdDM machinery.

Leveraging the knowledge gained from this epiRIL population, we now wish to determine how epigenetic variants arise in nature and determine their contribution to heritable phenotypic variation.

(A) Construction and (B) high-troughput phenotyping of the epiRIL population, which was constructed by first crossing a wt individual and an isogenic ddm1 mutant plant. (C) QTL mapping profiles for two independent flowering time measurements (FT1 and FT2), as well as their average (combined).