Transcriptome-Wide Discovery And Dynamics Of Methyl-6-Adenosine

RNA modifications have been known to exist since the late 1960s, but the methods to detect them globally did not exist until recently. The release of an antibody specific to the RNA modification methyl-6-adenosine (m6A) enabled its transcriptome-wide mapping sites using MeRIP-Seq, a variation of RIP-seq that enriches for RNA fragments using the antibody in a pulldown assay and identifies them using next-generation sequencing. The dynamic epitranscriptome has the potential to be involved in multiple layers of regulation, from translation repression to nuclear export and splicing. The purpose of this dissertation was to develop methods both experimentally and computationally to map and identify m6A sites. As a dynamic modification, m6A and other RNA modifications change in frequency in response to cell stress and stimuli. Developing methods to detect these changes further elucidate its functional role. First, the limitations of the MeRIP-Seq protocol were demonstrated, from its input requirements to the consequences of batch effects, ribosomal RNA contamination, and IP efficiency. Second, computational methods were developed to analyze MeRIP-seq data, implementing MeRIPPeR as an m6A peak finder that performs with higher sensitivity than other peak finders. The consequences of choice of aligner and annotation were also discussed, as well as methods to correct for technical variation and batch effects introduced during the MeRIP-seq protocol. Third, additional computational methods were developed to identify changes in methylation sites, identifying differentially methylated peak regions to unravel the dynamics of m6A. These three methods were then applied to two case studies. In the first study, changes in m6A sites in response to heat shock and ribavirin treatment in the context of nuclear export were examined. The results showed a correlation between differentially methylated peak regions in introns and changes in nuclear/cytosolic ratios. The second study explored m6A’s role in adipogenesis in the porcine model, serving as the first study of m6A in pigs. Differentially methylated peak regions found in both studies implicate dynamic m6A sites in genes involved in RNA regulation, splicing, and nuclear export pathways. The results demonstrate the biological importance of m6A and further implicate it in RNA processing and regulation