Human whole mitochondrial genome sequencing and analysis: optimization of the experimental workflow

AimTo evaluate critical steps in Illumina® Human mtDNA Genome assay: target enrichment, limited-cycle PCR, and library normalization, in order to optimize the protocol for analysis of whole mitochondrial genomes from human reference samples.MethodsThree long-range high-fidelity DNA polymerases (Platinum^TM PCR SuperMix High Fidelity, LA Taq® Hot Start, and PrimeSTAR® GXL) were tested for their performance in the amplification of mtDNA fragments. Sequencing results of ten samples, as well as negative controls, which underwent library preparation with 12 and 15 cycles in limited-cycle PCR were compared. Additionally, two library normalization methods were compared: bead-based normalization vs quantification and individual normalization.ResultsPrimeSTAR® GXL performed best for mitochondrial DNA enrichment. Increment of amplification cycles to 15 in limited-cycle PCR step did not affect either the sequencing process or variant calling. Library quantification combined with individual library-by-library dilution outperformed bead-based normalization.ConclusionOptimizations described herein provide beneficial insights for laboratories aiming at implementation and/or advancement of similar massively parallel sequencing workflows (eg, small genomes, PCR amplicons, and plasmids).

Library preparation in massively parallel sequencing (MPS) protocols is a sensitive process, usually consisting of multiple elaborate steps. To ensure high quality of sequencing results, it is important to optimize library preparation according to characteristics of a particular target molecule. There are several critical aspects of Illumina® Human mtDNA Genome (1) assay for analysis of whole mitochondrial DNA (mtDNA) on MiSeq® instrument: initial enrichment of the target molecule (achieved, in this case, by long-range PCR); PCR step wherein index-adapter oligonucleotides are added and libraries are amplified (termed “limited-cycle PCR” step); and normalization of libraries prior to their pooling for sequencing.In this study, we aimed to test three long-range high-fidelity DNA polymerases for their performance in amplification of mtDNA fragments, in order to determine the one best suited for Illumina® assay, in which mitochondrial genomes are amplified in two large fragments (sizes 9.1 kb and 11.2 kb) (Supplementary Figure 1(Supplementary Figure 1)).Various optimizations and evaluations have been previously published (2-6), but, to our knowledge, none of them assessed the impact of increasing the number of amplification cycles in limited-cycle PCR. Therefore, we also aimed to test and observe how sequencing results were affected by this step and whether there were any adverse effects that would impact variant calling.Lastly, we aimed to compare two library normalization methods: bead-based normalization vs quantification and individual normalization. Nextera® XT Library Prep Kit (Illumina, San Diego, CA, USA) has been known to produce uneven read depth profiles (2-4,6-8). Therefore, a great risk in this protocol is potential loss of sequence information in regions that achieve very low read depth (ie, too low for analysis and subsequent variant calling), or possibly receive no reads at all. The choice of library normalization method may affect this through distribution of reads among the sequenced libraries. So, a method that provides more uniform distribution of reads would be preferable, which we hypothesized would be the method of quantification and individual normalization rather than bead-based normalization.