P17.51FUNCTIONAL PREDICTION OF MITOCHONDRIAL DNA MUTATIONS IN GLIOBLASTOMA MULTIFORME

Gliobastoma multiforme (GBM) is the most common adult primary brain tumour. It is highly malignant with fewer than half of patients surviving for over one year following diagnosis. One reason for this is its highly heterogeneous nature, which renders the current standard-of-care treatment of surgical resection, radiotherapy, and temozolomide (TMZ) chemotherapy inadequate. Altered metabolism is a hallmark of cancer compared to non-cancer cells and central to this are mitochondria, which contain their own multi-copy circular DNA (mtDNA). At the heart of every mitochondrion is a collection of five large protein machines known as the respiratory chain (RC). The RC consists of approximately 70 subunits. MtDNA-encodes thirteen of the core catalytic subunits that form the RC. Variations in mtDNA can therefore directly affect metabolism and contribute to the pathology of numerous human diseases, including cancer. The acquisition of TMZ-resistance and decreased overall survival in GBM patients has been linked to altered RC activity but the contribution of mtDNA variations to this remains unexplored. To address this we have employed a novel dual approach for (1) discovering and (2) predicting the function of mtDNA variations in GBM patients using a combination of long PCR and next generation sequencing and 3D structural analysis, respectively. We have generated a set of 10 high quality mtDNAs each from GBM-biopsy derived cells with a mean depth of coverage per nucleotide of over 22,000. This enabled a total of 193 mtDNA variations to be detected, including those present at less than 1% heteroplasmy. Twenty five of these variations caused non-synonymous amino acid substitutions (AAS) in mtDNA-encoded Complex III and IV proteins. Using 3D structural analysis, we predict that 8 of these are highly likely to be functional, occurring in RC active sites, substrate/drug binding pockets and protein-protein interaction regions. The latter class will affect the assembly and stability of the RC complexes. All functional candidates were absent from the non-neoplastic astrocytic control, and mining the human mtDNA database revealed the majority never, or very rarely, occur in healthy subjects. Long PCR and next generation sequencing provides a high throughput, cost effective and sensitive method for detecting mtDNA variations in GBM. While 3D structural analysis reveals considerable insight into the functional role of mtDNA variations, including how they could affect metabolism and contribute to the differential survival and chemosensitivity of individual patients. This study forms an important first step towards determining the functional and clinical significance of mtDNA variations in GBM. The next step will be to validate the functional and clinical significance of the candidate variations so that better predictors of outcomes and more personalised therapies can be developed for patients in the future.