The fundamental question of whether nucleosomal or naked DNA is the preferred substrate of plant and animal methyltransferases has long remained unresolved. Our research showed that genetic inactivation of a single DDM1/Lsh family nucleosome remodeler biases methylation toward inter-nucleosomal linker DNA. We found that DDM1 enables methylation of DNA bound to the nucleosome, suggesting that nucleosome-free DNA is the preferred substrate of eukaryotic methyltransferases in vivo. Furthermore, we showed that simultaneous mutation of DDM1 and linker histone H1 reproduces the strong linker-specific methylation patterns of species that diverged from flowering plants and animals over a billion years ago. Our results indicate that in the absence of remodeling, nucleosomes are strong barriers to DNA methyltransferases. Linker-specific methylation can likely evolve simply by breaking the connection between nucleosome remodeling and DNA methylation.
DNA methylation and histone H1 are known to mediate transcriptional silencing of genes and transposable elements, but how they interact has been unclear. We showed that H1 is enriched in methylated sequences, yet this enrichment is independent of DNA methylation. We found that loss of H1 disperses heterochromatin, globally alters nucleosome organization, and activates H1-bound genes, but only weakly de-represses transposable elements. However, H1 loss strongly activates hypomethylated transposable elements. We also found that hypomethylation of genes activates antisense transcription, which is modestly enhanced by H1 loss. Our results demonstrate that H1 and DNA methylation jointly maintain transcriptional homeostasis. Such functionality plausibly explains why DNA methylation, a well-known mutagen, has been maintained within coding sequences of crucial plant and animal genes.
The RNA-directed DNA methylation (RdDM) pathway preferentially targets euchromatic transposable elements. However, RdDM is thought to be recruited by methylation of histone H3 at lysine 9 (H3K9me), a hallmark of heterochromatin. How RdDM is targeted to euchromatin despite an affinity for H3K9me is unclear. We showed that loss of histone H1 enhances heterochromatic RdDM, preferentially at nucleosome linker DNA. Surprisingly, this does not require SHH1, the RdDM component that binds H3K9me. Furthermore, H3K9me is dispensable for RdDM, as is CG DNA methylation. Instead, we found that non-CG methylation is specifically associated with small RNA biogenesis, and without H1 small RNA production quantitatively expands to non-CG methylated loci. Our results demonstrate that H1 enforces the separation of euchromatic and heterochromatic DNA methylation pathways by excluding the small RNA-generating branch of RdDM from non-CG methylated heterochromatin.
Methylation within CG dinucleotides (mCG) is thought to be semi-conservatively inherited during DNA replication. Our work with Arabidopsis plants deficient in both H1 and DDM1 revealed that intermediate levels of mCG can exist over a large fraction of the genome. We found that these intermediate states can be stably inherited over many generations, that they result in intermediate expression of transposable elements, and that individual CG sites in these plants cycle rapidly between methylated and unmethylated states. Our results indicate that the existing semi-conservative model is incomplete, and that stable mCG in transposable elements is enabled by a high rate of untemplated de novo methylation. We also found that the de novo activity is contributed by distinct methylation pathways (RdDM and the CMT enzymes). Our results therefore argue that stable epigenetic inheritance of DNA methylation in transposable elements is not a function of any individual system, but instead requires pathway integration.
Because semi-conservative mCG epigenetic inheritance is imperfect, it has been unclear over how many cell cycles mCG can encode additional information independent of the underlying genetic sequence and thus function as an epigenetic genotype. We found that gbM (mCG in gene bodies) establishment, maintenance, and even loss, constitute a unified process mediated by the methyltransferase MET1. MET1 activity in gene bodies is suppressed by the histone variant H2A.Z and the DNA demethylase ROS1, producing localized mCG patterns. Any level of gbM can be stably inherited over tens of generations, but gbM patterns eventually converge to a single DNA sequence-dependent steady state. This sequence dependence explains the prevalence of gbM in nucleosomes and exons, and more broadly we can accurately predict the overall steady-state gbM patterns of individual genes from sequence alone. Nevertheless, gbM undergoes large stochastic epigenetic fluctuations that explain much of the observed population-scale gbM variance. These fluctuations can last for thousands of years in the absence of genetic change, thereby establishing gbM as an epigenetic genotype able to mediate evolution on this timescale.
These results have been published in peer-reviewed articles, and I have disseminated them through presentations at international scientific conferences.
