How the genome becomes DNA methylated

Epigenetic information provides a fundamental control of the genome. Of the epigenetic modifications, the first to be described in mammals was cytosine methylation of DNA, generally thought of as a repressive modification. Aberrant DNA methylation patterns are associated with disease, notably cancer, and may provide a mechanism underpinning long-term adverse health outcomes in response to factors such as poor diet and environmental exposures. Although DNA methylation has been known for several decades, the principles that govern where in the genome the de novo DNA methyltransferases are recruited to determine cell-type specific patterns of DNA methylation are still poorly understood. Biochemical studies show that the mammalian de novo methyltransferases (DNMT3A, DNMT3B, DNMT3L) are sensitive to epigenetic information in chromatin in the form of post-translational modifications of histones. Notably, the presence of H3K4me3, the classical histone modification associated with active promoters, antagonizes recruitment of the methyltransferases to chromatin. A number of other positive and negative correlations exist between the DNMT3s and histone modifications, but the causal relationships are unclear. We recently induced a point mutation in a predicted histone binding domain of DNMT3A. This mutation unexpectedly causes gain of DNA methylation in regions of the genome normally hypomethylated and occupied by the repressive histone modification H3K27me3. Many of these regions coincide with key developmental regulator transcription factors such as Hox genes. In general, DNA methylation and H3K27me3 are mutually exclusive repressive modifications, so the pattern of ectopic methylation is highly unusual. In mice, the mutation results in growth retardation, likely associated with deregulation of developmental regulators. Mutations in neighbouring amino acids in the same protein domain of DNMT3A have recently been identified in human congenital disorders.

The student will initially generate constructs by which to probe the interaction of wild-type and mutant DNMT3A with chromatin and with major epigenetic regulators in order to understand the cause of the aberrant methylation patterns. This will be done by knocking-in a protein tag into the Dnmt3a gene in mouse embryonic stem cells (ESCs) by CRISPR/Cas9 engineering. Tagged DNMT3A will initially be evaluated for genomic recruitment by ChIP-seq, and the results integrated with available datasets for a variety of histone modifications and epigenetic regulator binding. Novel interaction/binding modalities will be tested by antibody co-precipitations. Therefore, the project will include molecular techniques involved in generating a knock-in allele in ESCs, epigenomic assays and bioinformatic analysis. The project could a platform for a PhD project to investigate molecular interactions between DNMT3s, histone modifications and histone modifiers in greater detail.

Building on some of the reagents developed above, the student will investigate the recruitment principles of DNMT3A and mutant forms in more detail. He/she will test the over-arching hypothesis that DNMT3A senses a combination of histone modifications through different domains of the protein, some of which are differentially represented on the various isoforms of the protein expressed at different developmental stages. The model is that the interactions have different inherent strengths, such that modulation of one interaction, for example through mutation, will alter the genomic distribution of DNMT3A and the target loci it methylates. The student would test this hypothesis through the following studies:
- Use epitope-tagged DNMT3A to examine its genomic distribution in relation to known pattern of histone modifications;
- Examine changes in distribution patterns as DNMT3A isoforms switch during differentiation;
- Use epitope-tagged DNMT3A to identify protein interactions and how they are altered by mutation;
Induce further mutations in domains of DNMT3A based on structural predictions, using a cellular context to allow efficient screening.

Website: http://www.babraham.ac.uk/our-research/epigenetics/gavin-kelsey