RNA- Programmed DNA Methylation

Appropriate epigenetic modification of DNA via cytosine methylation has been shown to be vital to the proper functioning of cellular processes in eukaryotes. High levels of methylation in promoter regions are associated with transcriptional silencing. The ability to precisely target methylation to specific sites in promoters could be used as a tool for gene silencing and for understanding the relationship between DNA methylation and cell phenotype.   In addition, since abnormal hypomethylation is associated with diseases such as leukemia, atherosclerosis and, Alzheimer’s disease, precise targeting of methylation to a single promoter is a potential therapeutic strategy. The typical strategy for targeting DNA methylation involves end-to-end fusion of a DNA binding domain (DBD), such as zinc fingers and TALE, to a cytosine DNA methyltransferases (MTase), typically bacterial enzymes or human DNMT3a. The two main drawbacks of this approach are (1) methylation occurs at non-target sites since the MTase domain remains active to methylate these sites when the DBD is not interacting with DNA and (2) new sequence recognition requires DBD redesign. Our solution to the first problem is to use the target site for methylation as a template for assembling the active form of an otherwise ineffective split methyltransferase (sMTase).  This strategy has been shown to result in very specific targeting in bacteria (~80% methylation at the target site and ≤0.8% methylation at other CpG sites) using zinc fingers as the DBD. The second problem can be addressed by the CRISPR-Cas9 genome engineering tool, which has not yet been applied to target DNA methylation. A key feature of nuclease- null Cas9 (dCas9) as a DBD is its ability to bind specific sites defined by the single-stranded guide RNA (sgRNA) sequence.  A carefully selected sgRNA allows for highly specific targeting without the requirement for DBD redesign. An ideal dCas9-sMTase would have many exemplary features: the precision of true single-site targeting of methylation, ease of targeting via simple design of sgRNA, and multiplexed targeting of several sites via multiple sgRNAs. The later advantage addresses the concern that single site methylation might not be sufficient to have a desired phenotypic effect. To this end, we designed fusion proteins of dCas9 and a sMTase.  The dCas9-sMTase provides high precision and control over CpG methylation in bacteria and human cells.