Nanopore single-molecule sequencing to investigate mitochondrial DNA CpG methylation in Parkinson’s disease

By way of introduction, Theresa Lüth (Ph.D. student, Institute of Neurogenetics, University of Lübeck, Lübeck, Germany) gave an overview of Parkinson’s disease (PD) and the role played by mitochondrial dysfunction. PD, which is characterised by tremor, bradykinesia, and rigidity as a result of cell death of the dopaminergic neurons, is the fastest growing neurological disease, with more than 7 million affected individuals worldwide.

Two to five percent of PD cases are explained by monogenic causes, for example biallelic mutations in PRKN, a recessively inherited gene implicated in early-onset PD.

All samples in this research study were obtained from biallelic mutation carriers for PD. The sample size was noted to be small with blood samples taken from five individuals affected by PD and three healthy control subjects. Additionally, neuronal-derived DNA samples were obtained from four individuals affected by PD and three healthy control subjects.

Context of the role of mitochondria in PD pathogenesis was given, detailing the involvement of mitochondrial DNA (mtDNA) dysfunction and mtDNA maintenance impairment. The role of mtDNA epigenetics in PD remains unresolved, with conflicting reports on the significance and biological relevance of mtDNA methylation levels. Theresa discussed the methodological limitations of using short-read sequencing technology to evaluate methylation profiles and the challenges of determining the role played by mtDNA methylation in PD.

To investigate mtDNA methylation in PD samples, Theresa utilised the Ligation Sequencing Kit together with MinION and GridION devices to perform whole-genome sequencing of DNA obtained from whole blood and iPSC midbrain neurons. Methylation calling was performed using Nanopolish and Megalodon.

Overview of results

The aim was to obtain 10 Gb of data per basecalled sample. Coverage at 250x and 828x was observed in blood samples and neuronal-derived DNA samples, respectively. As anticipated, low levels of methylation levels were observed. Using a stepwise approach, it was determined that 100x coverage was sufficient for reliable detection of mtDNA CpG methylation.

To remove the potential impact of contamination with nuclear mitochondrial (NUMT) DNA, which are 500 bp or shorter in length, the team set the minimum alignment length to 1 kb during filtering. The data and end results were largely unchanged.

The false positive rate of mtDNA CpG methylation was determined by implementing a negative control with long-range PCR spanning the entire mtDNA genome. Methylation calling of amplicon sequencing data enabled a baseline to be determined which was then used to identify false positive calls and correct the data. Finally, the false positive rate was calculated for blood and neuronal samples using Nanopolish (0.029 and 0.028 respectively) and Megalodon (0.046 and 0.048 respectively) using the method outlined by Goldsmith, C. et al (2021).

Further downstream analysis was performed to determine variability in detection of CpG methylation in blood and neuronal derived mtDNA. Although low levels were observed globally, the level of methylation in blood-derived mtDNA was significantly lower than that of neuronal samples.

Lower mtDNA CpG methylation levels in PD samples compared to healthy controls were observed using both Nanopolish and Megalodon. It was evident that mtDNA methylation is associated with mtDNA gene expression and therefore could contribute to mitochondrial dysfunction and PD pathogenesis.

Theresa closed her presentation by sharing plans to expand the pipeline presented to include additional methylation profiles.

Authors: Theresa Lüth