Genetic modification of the mitochondrial genome

Our long-term goal is to develop novel methods for controlled genetic manipulation of mammalian mitochondria in living cells. Several parallel lines of investigation have been pursued in order to identify the most promising approaches.

A part of our work is based upon experiments aimed at developing mitochondrially-targeted engineered nucleases to eliminate human mitochondrial DNA (mtDNA) molecules with pathogenic point mutations and large-scale deletions. One class of designer nuclease technologies utilise the fusion of an engineered zinc finger DNA binding domain to a endonuclease domain, producing a zinc finger-nuclease (ZFN). By targeting pairs of engineered nucleases, adjacently, to a site of interest, a specific DNA double-strand break (DSB) can be produced.

Typically, there are between ~100-10,000 copies of mtDNA per cell. Sequence variation between mtDNA molecules within a cell occurs frequently, a state known as mtDNA heteroplasmy.

Mitochondrial disease originating from a pathogenic heteroplasmic mutation in mtDNA depends on the ratio of mutated : non-mutated mtDNA present in a cell (Figure, red and green circles, respectively). The heteroplasmic threshold required for disease presentation is variable between mutations, typically requiring ~60-90% mutant mtDNA load. This threshold is dependent on: mutation, tissue distribution and likely other, as yet unknown factors. A significant majority of mtDNA-mediated mitochondrial disease is heteroplasmic in nature.

The heteroplasmic nature of mtDNA in the disease state allows an unusual gene therapy approach entailing the selective targeting and degradation of mutant mtDNA molecules, based on DNA sequence discrimination. Reducing mutant mtDNA copy number leads to replication of the spared wild-type molecules, as a result of the cell recovering to the initial, gross mtDNA copy number (Figure).

Our laboratory has developed, and is constantly refining mitochondrially-targeted ZFNs (mtZFNs) to target mammalian mtDNA in situ. After the initial description of an efficient method to deliver ZFPs to mitochondria, we provided proof-of-principle that ZFPs can access and bind sites on human mtDNA with single nucleotide specificity [1][2][3]. More recently, we used mtZFNs to target and cleave mtDNA harbouring the m.8993T>G point mutation associated with neuropathy, ataxia, retinitis pigmentosa (NARP) and the ‘common deletion’ (CD), a 4977bp repeat-flanked deletion associated with Kearns-Sayre and Pearson’s marrow pancreas syndromes [4] [5]. Our current efforts focus on trialing the mtZFN approach in a mammalian in vivo model of mitochondrial disease to provide pre-clinical data, useful for future development of treatment for any sufferer of mitochondrial disease owing to heteroplasmic deletion, mutation or rearrangement of mtDNA.