PROJECT SUMMARY/ABSTRACT
Highly disruptive truncating mutations to protein-coding genes in mitochondrial DNA (mtDNA) affect nearly 10%
of all cancers and predominantly arise heteroplasmically, affecting a fraction of the total mtDNA pool. Although
decades of investigation into pathogenic mtDNA variants in the germline have established that they profoundly
disrupt normal mitochondrial oxidative phosphorylation, the effects of such mutations in cancer cells are largely
unknown. The fundamental barrier to rigorous interrogation of mtDNA mutations in cancer cells has been a lack
of tools for genetically engineering mtDNA. Recently, a new mtDNA-editing technology pairing TALE binding
domains to a DddAtox cytosine base editor (DdCBE) has been successfully used to introduce point mutations
into mtDNA, revolutionizing the ability to genetically manipulate mtDNA with high precision. In parallel, our team
recently discovered that truncating mtDNA mutations are under strong positive selection in specific genetic
contexts (subunits of NADH dehydrogenase/Complex I, “CI”) and cancer lineages (colorectal, kidney, and thyroid
cancers), and that the heteroplasmic dosage and transcriptional phenotype of these mutations are readily
detectable in single cell sequencing data. These convergent discoveries motivated our team to engineer DdCBEs
to introduce truncating mutations to several CI and non-CI mtDNA genes in cell lines, enabling for the first time
a functional interrogation of truncating mtDNA mutations in cancer cells. Using these tools, we propose
integrative computational/experimental studies to test the overarching hypothesis that CI-truncating mutations
produce physiologically significant and therapeutically actionable metabolic changes in tumors. In Aim 1, we will
computationally investigate mtDNA mutation patterns across ~100,000 tumor samples, identifying recurrent
mutant alleles and co-incident driver mutations in nuclear DNA. In parallel, we will express DdCBEs to model CI-
and non-CI truncating mutations in colorectal cancer cell lines, and define the molecular phenotypes conferred
by CI truncating mutations using transcriptomic, metabolomic, and isotope tracing experiments. Our Preliminary
Data indicates that the phenotype of CI-truncating mutations depends on their heteroplasmic dosage. Thus, in
Aim 2 we will use transient expression of DdCBEs to produce isogenic panels of colorectal cancer cell lines at
characteristically distinct mutation dosages. Using a combination of single cell and bulk molecular profiling, we
will test the hypothesis that CI-truncating mutations rewire tumor cell metabolism towards a pro-proliferative
configuration in a dosage-sensitive manner. Finally, hypothesizing that CI-truncating mutations induce genetic
dependencies absent in mtDNA-wild-type cells, Aim 3 will use DdCBE-engineered cell line models to identify,
validate, and mechanistically study novel synthetic lethalities associated with CI-truncating mutations. The
results of these studies will deliver a new, detailed understanding of the function, dosage sensitivity, and
therapeutic vulnerability of one of the most common genetic insults in the cancer genome.