MECHANISMS AND CONSEQUENCES OF DNA METHYLATION IN TWO MODEL SPECIES

Epigenetic modifications are known to regulate gene expression in a heritable manner, and can broadly be divided into three interacting classes: DNA methylation, histone modifications, and chromatin interactions. While the trans acting factors that establish, maintain, and remove DNA methylation are well-known, the cis acting mechanisms that direct DNA methylation to specific genomic locations remain elusive. Two gene classes offer insights into cis-acting mechanisms for DNA methylation: imprinted loci, and transposable elements. A locus spanning both is the murine Rasgrf1 locus. Rasgrf1 has a cis element, a series of tandem repeats, required for DNA methylation, but also harbors a long noncoding RNA, the pitRNA. The pitRNA is driven by the repeats and is targeted by piRNAs, small RNAs that mediate transposable element methylation in the mammalian male germline. However, the effects of the pitRNA versus the repeats have not yet been separated. My work, where I used CRISPR/Cas9 genome editing to generate a targeted mutant system permitting inducible control of the pitRNA, is the first to query the sufficiency of the pitRNA independently. Using quantitative qPCR and targeted bisulfite sequencing, I demonstrated that expression of the lncRNA at physiological levels in the male germline is insufficient to impart DNA methylation at Rasgrf1. These findings were complimented by additional in vitro studies, where I identified Sp1 as a transcription factor that binds the repeats and is required for pitRNA expression. Sp1 binds secondary DNA structure and has recently been identified as a regulating factor at another imprinted gene. Together, these findings support an alternative, critical role for the repeats beyond their known role in regulating pitRNA expression. Beyond mechanism, DNA methylation in the context of disease are an area of active study, though its utility in non-traditional model organisms is nascent. The second focus of my thesis speaks to this. I performed reduced representation bisulfite analysis on two dog breeds with highly diverse morphology and disease risks. While this work is largely preliminary, two differentially methylated regions have direct association with differential disease risk between these two breeds, suggesting that the canine methylome could be used as method of disease surveillance