Abstract
ABSTRACT
Cytosine methylation is perhaps the most dynamic and best-studied form of epigenetic modification. Occurring predominantly in the CG dinucleotide context within mammalian genomes, it is essential for normal embryonic development, X chromosome inactivation, genomic imprinting and transposon silencing. The uniqueness of CG methylation lies in its ability to be maintained following cell division in the absence of the signal which created it. Together, the global distribution of CG methylation forms an “epigenetic memory” that contributes to the differentiation and maintenance of distinctive somatic cell fates. Artificial manipulation of cytosine methylation (known as “synthetic epigenetics”) could potentially improve the creation and differentiation of developmentally potent cells, which will be instrumental in the advancement of regenerative medicine.
I have undertaken two experimental projects investigating the potential of synthetic epigenetics, whereby epigenetic modifiers of non-mammalian provenance were overexpressed in mouse embryonic stem cells (ESCs). In plant genomes, methylation occurs at CHG nucleotides (where H is any base other than G) in an analogous manner to mammalian CG methylation, due to the action of the CMT3 methyltransferase. Thus, in the first project, I created transgenic cell-lines with inducible overexpression of Arabidopsis CMT3, with the aim of creating methylation at CHG nucleotides. Overexpression of CMT3 was performed in three mouse embryonic stem cells (mESC) lines; two wild-type lines (E14 and V6.5) and a DNA methylation depleted line known as (DNMT-TKO). My experimental results to date suggest that CMT3 alone is inadequate for CHG methylation establishment in mammalian genomes, however, increased levels of methylation in the CG context were detected, suggesting it may have previously unappreciated maintenance capacity in this context.
The second experimental project explored the nature of DNA methylation removal. Cytosine methylation can be actively removed from DNA by the Ten eleven translocation (TET) enzymes, which oxidise 5-methylcytosine (5mC) to 5 hydroxymethylcytosine (5hmC), and further oxidised derivatives 5-formylcytosine (5fC) and 5-carboxylcytosine (5caC). Exactly how these analogues are returned to unmodified cytosine is unknown, partly because there are currently few molecular tools to control the transition between 5hmC, 5fC and 5caC. Unpublished in vitro observations suggest that Ten eleven translocation (TET) enzymes from the amœba Naegleria gruberi are far more effective at iteratively producing 5fC and 5caC, compared to mammalian homologues which produces 5hmC predominantly (T.P.
Jurkowski, personal communication). I created mouse embryonic stem cells with inducible expression of Naegleria TET and mutated Naegleria TET variants, and used bisulphite sequencing to assess their relative ability to demethylate DNA. I found that the amœba TET variants were ineffective at demethylating DNA compared to mammalian TET controls. Unexpectedly, some constructs even showed significant increased methylation. I hypothesise that the Naegleria TET variants I created act as non-functional competitors of endogenous TET when overexpressed in mammalian cells. While this likely precludes their widespread use in the field of synthetic epigenetics, they still may provide a useful experimental tool for further study of TET enzyme biology.