Investigating the potential role of recombination regulator PRDM9 in mitochondria /

Abstract

At present, 805 mitochondrial DNA (mtDNA) deletions have been described. Short direct repeat regions of DNA flank many of these deletions, suggesting that specific regions of the mtDNA molecule have a susceptibility to deletion formation. Despite this, the exact underlying cellular mechanisms facilitating mtDNA deletions are unclear. PR domain 9 (PRDM9) is a meiotic-specific protein responsible for determining the site of recombination in the nuclear genome. Through its zinc finger repeat region, PRDM9 binds a specific DNA consensus sequence, and acts as a methyl transferase, opening chromatin for DNA crossover events to occur. This is of interest as mitochondrial DNA also contains PRDM9 binding motif sites. This thesis outlines the experimental steps taken to determine if PRDM9 has any involvement in mtDNA maintenance and viability. Firstly, an in silico approach was used to screen mtDNA sequences from 31,551 individuals for the presence of the PRDM9 binding motif, identifying multiple putative binding sites in and around known deletion forming flanking regions. In addition, population and phylogenetic stratification showed differential mtDNA binding motif patterns, potentially explaining the variable deletion frequencies between mtDNA haplogroups and populations. Secondly, to test the potential interaction between PRDM9 and mtDNA, complete genotyping of the PRDM9 zinc finger repeat region in a cohort of 48 mitochondrial single deletion patients and 50 healthy controls was performed. However, there was no association between PRDM9 haplotype and the formation of mtDNA deletions. Heterozygous individuals were significantly increased in the patient cohort compared to controls although no particular allele was associated with mtDNA deletion. Finally, PRDM9 protein levels were interrogated in cell lines and tissue samples. However, due to timing of expression it was not possible to reliably detect nascent protein using commercially available antibodies. To overcome this, stable cell lines overexpressing Flag-tagged PRDM9 were created. Low levels of PRDM9 expression were detected by immunoblotting indicating overexpression had worked but also indicating that PRDM9 turnover in cells is likely rapid. iv Given the data presented, and despite the presence of multiple putative PRDM9 binding sites in almost all mitochondrial genomes studied, we conclude that it is unlikely that PRDM9 has a significant effect on the maintenance of mtDNA. However, to the best of my knowledge this is the first stable PRDM9 overexpression model created and it has provided a unique insight into some of the functions of this protein.