Chasing a moving target: Detection of mitochondrial heteroplasmy for clinical diagnostics

Abstract

Clinical interpretation of human mitochondrial DNA (mtDNA) variants has been challenging for technical and biological reasons but the involvement of dysfunctional mitochondria in many diseases makes it imperative to have a validated assay for detecting pathogenic variants. We have tested several methods to identify those best suited to detect and confirm mtDNA variants. The choice of methods is dependent on the amount of DNA available for testing and the sensitivity required for detecting low-level heteroplasmies. There is a tradeoff between a polymerases ability to amplify small amounts of DNA and its ability to generate accurate sequence. We report a simple method to measure heteroplasmy levels of large deletions from NGS data alone without need for qPCR or other methods. Use of HapMap samples for standardization needs to be done with caution as most have novel heteroplasmic sites that have arisen during immortalization/cell culture processes. Different batches of DNA can have variable sequence. In contrast, we observed no de novo heteroplasmies in healthy mother-child pairs studied using blood or saliva though the frequency of pre-existing heteroplasmies often changed dramatically across generations. Long-read nanopore sequencing of individuals with two heteroplasmies suggested a random distribution of variants on single molecules but technical artifacts prevent certainty on this finding. Urine provides an additional readily accessible source of mtDNA that can be used for bone marrow transplant recipients whose saliva/blood mtDNA may be contaminated by the BMT donors mtDNA. We have characterized cells suspended in urine via expression profiling and shown them to be primarily mucosal cells that are independent of blood. Understanding the pitfalls of the various mtDNA sequencing methods allows development of reliable and accurate tests suitable for clinical diagnostics.\n\nAuthor SummaryMitochondrial DNA is important for many diseases but it is present at many copies per cell so is harder to check for mutations compared to nuclear DNA. We have studied mitochondrial DNA in different ways to see how it changes across generations and in different locations in the body. The tests need to be much more sensitive than nuclear DNA tests so that we can detect mutations down to 1%. We have shown that mitochondrial DNA changes when cell lines are used but saliva, blood and cells in the urine can all be used for testing. Cells in the urine originate as mucosal cells and are independent of blood. We developed a new method for analyzing large deletions that means sequencing data alone can be used for measuring the frequency of deletions. We also followed a family with two variable sites to better understand how mitochondrial DNA changes from mother to child. In some children, the variants stayed the same while, in others, variants disappeared.