Influence of DNA methylation on transcription factor binding

Eukaryotic transcription factors (TFs) are key determinants of gene activity, yet they bind only a fraction of their corresponding DNA sequence motifs in any given cell type. Chromatin has the potential to restrict accessibility of binding sites; however, in which context chromatin states are instructive for TF binding remains mainly unknown. This thesis explores the contribution of DNA methylation to constrained TF binding by studying CTCF as a known methylation-sensitive TF and applying a genome-wide approach to identify further sensitive factors in mouse stem and differentiated cells.
CTCF is perhaps the most prominent example for a TF that can be prevented from binding by DNA methylation in vivo. However, it is restricted by methylation only at a subset of its genomic binding sites, such as the H19/Igf2 imprinting control region (ICR). In order to understand this context dependency of CTCF methylation sensitivity, we compared CTCF binding in isogenic mouse stem cells with and without DNA methylation. Two features distinguish the fraction of sites that are bound only in the absence of DNA methylation: CpG-containing variants of the canonical CTCF motif as well as higher CpG density in the flanking regions. The H19/Igf2 ICR indeed fulfils these criteria and we show that CTCF methylation sensitivity there is independent of the complete ICR sequence, the chromosomal context and H3K9me3 marks.
In order to go beyond CTCF and identify more methylation-sensitive TFs a priori, we mapped DNase I hypersensitive sites, as an indicator of TF binding, in mouse stem cells with and without DNA methylation. Methylation-restricted sites are enriched for TF motifs containing CpGs, especially for those of NRF1. In fact, NRF1 occupies several thousand additional sites in the unmethylated genome, resulting in increased genic and non-genic transcription. Restoring de novo methyltransferase activity initiates remethylation at these sites and outcompetes NRF1 binding. Even strong overexpression of NRF1 is unable to prompt binding at methylated regions. This suggests that binding of methylation-sensitive TFs relies on additional determinants to induce local hypomethylation. In support of this model, deletion of neighbouring motifs in cis or of a TF in trans causes local hypermethylation and subsequent loss of NRF1 binding. This competition between DNA methylation and TFs in vivo reveals a case of cooperativity between TFs that acts indirectly via DNA methylation.
Nevertheless, the vast majority of TF binding events do not change upon removal of DNA methylation in stem cells. To investigate whether more TFs are affected in differentiated cells, for which DNA methylation is essential, we generated methylation-deficient neuronal cells that survive for several days in culture. Changes in genic transcription and chromatin accessibility are surprisingly limited in the absence of DNA methylation, although again a subset of TF motifs are enriched in methylation-restricted sites, such as NRF1 and HNF6. While this closely resembles the situation in stem cells, we observe a striking activation of specific classes of endogenous retroviruses (ERV) only in the differentiated methylation mutant. Several lines of evidence indicate that methylation-sensitive TF binding at the cAMP-responsive element (CRE motif) is responsible for ERV activation in differentiated methylation mutants including mouse cortex, which might provide a link to the ensuing cell death.
Taken together, only a low percentage of TF binding events are restricted by DNA methylation in stem or differentiated cells. However, a subset of factors is methylation-sensitive at CpG-containing motifs. These factors rely on other TFs to keep their motif in an unmethylated state and their aberrant binding can have devastating consequences by repeat activation. Understanding the influence of DNA methylation on TF binding constitutes one step towards better interpretation of the rapidly growing number of epigenetic and TF binding maps. The success of the approach taken here suggests that it can be applied to other chromatin components and modifications, which should enable comprehensive prediction of TF binding and ultimately gene expression in development and disease.