Mitochondria are complex and dynamic organelles found in most eukaryotic cells, and involved in many cellular functions, most notably the production of cellular energy through the oxidative phosphorylation (OXPHOS) machinery. In mammals, mitochondria contain their own genome that encodes thirteen essential polypeptides for the OXPHOS system, alongside mitochondrial rRNAs and tRNAs required for their expression. Gene expression in mitochondria is distinct from the bacterial and eukaryotic cytosolic counterparts, and many aspects of the processes of transcription and translation are not fully understood. Given the recent advancements in genome engineering in mitochondria, in particular the development of base editing, it is now possible to investigate these processes in more detail than ever before. Using these reverse genetics approaches, this work aims to utilise this new toolkit to probe different aspects of mitochondrial gene expression and broaden the understanding of these processes. The work covers mitochondrial transcription, ribosome biogenesis, translation initiation and protein synthesis chronologically.

In the first section of this work, following the discovery of a second light strand promoter (LSP2) in mitochondria in vitro by collaborators, cytosine base editing was used to probe its activity in living cells. By the precise mutation of two individual sites in LSP2, one which increased promoter activity and another which decreased it in vitro, it was possible to demonstrate the same effect in living cells on transcription in living cells and confirm its activity as a bona fide mitochondrial promoter. This promoter likely allows mitochondrial genome replication and gene expression to occur from two sites, and its discovery changes the understanding of these processes.

In the next chapter, the role of m4C methyltransferase METTL15 in ribosome biogenesis and rRNA stability was studied. The METTl15 gene was first inactivated in cells using CRISPR/Cas9, and the lack of m4C modification on the mitochondrial 12S rRNA shown. METTL15 knock-out cells exhibited a decrease in mitochondrial translation rates, reduced OXPHOS steady-state levels and perturbed mitoribosomal biogenesis. The levels of the mitoribosomal small subunit were severely depleted in these cells, with proteins close to the modification site showing the most significant reduction. Through the complementation of catalytically inactive METTL15 into knockout cells, the role of METTL15 as a chaperone in ribosome biogenesis, rather than its enzymatic activity, was shown. Base editing of the 12S rRNA was then performed to demonstrate the importance of the integrity of the rRNA decoding centre in ribosome activity.

The following chapter explores the roles of enzymes NSUN3 and ALKBH1 in modifying tRNAMet, allowing it to decode non-conventional AUA methionine codons. Ribosome profiling data showed that, in the absence of NSUN3 and ALKBH1, the mitochondrial ribosome stalls at AUA codons, confirming a long-held hypothesis on the role of these enzymes in aiding AUA recognition by tRNAMet. The mutation of the conventional AUG start codons of CYTB and ATP6 was performed to corroborate these findings, yielding unexpected results on mRNA processing.

In the final chapter, a library of cytosine base editors was designed and optimised for the precise ablation of each mitochondrially-encoded protein. The library was used to generate near-homoplasmic knockout cells of each protein, with minimal off-targets, allowing for the systemic investigation of the role of each mitochondrially-encoded protein. The library was further used to test the emerging adenine base editors.

Taken together, this work provides a comprehensive use of genetic engineering to study different aspects of mitochondrial gene regulation, making important discoveries and contributions to the field.