Investigations of the human mt genome are in the focus of biological and medical scientific disciplines. Compared to nuclear DNA (nDNA), mitochondrial DNA (mtDNA) is more vulnerable to oxidative damage and undergoes a higher rate of mutation [1
]. Because of these features the analysis of the mt genome has become a proven tool in population genetics. A multi-copy genome without recombination which accumulates mutations allows for the establishment of phylogenetic trees [2
]. It was the information from the highly variable mitochondrial control region (CR) that lifted the secret of human evolution starting in Africa about 150000 years ago and gave an insight in human migration all over the world within the past 60000 years [3
]. Sequences of full mt genomes are necessary to decipher yet not defined haplotypes and assign them to their phylogeographic environment.
Mitochondrial DNA mutations in the coding region (codR) have been associated with several pathologies [5
] including cancer [6
]. During oxidative phosphorylation (OXPHOS) mitochondria produce reactive oxidative species (ROS) that potentially induce DNA mutations. Such an initial mutation is heteroplasmic with the mutated variant constituting a minority [10
]. In the course of several replications the heteroplasmic mutation may become dominant leading to cancer [8
]. This theory is based on the results of several investigations on cancer tissues [11
]. Unfortunately, numerous articles addressing that issue are erroneous as reviewed in [19
]. On the one hand it is the lack of phylogenetic knowledge and the ambiguous mtDNA alignment that led to false conclusions of mtDNA mutations to be tumor-specific rather than evolutionary caused. On the other hand, laboratory-, sequencing-, and analysis errors led to wrong base-calls [21
]. Hence, flawed data hamper a precise interpretation of the conjunction between mtDNA mutations and the complex process of tumor development.
For forensic as well as for phylogenetic purposes we have already successfully established evaluated sequencing strategies that proved to be useful in a number of investigations where precise base-calling was necessary for the CR [22
], however such stringency is lacking for the whole mt genome. The published protocols vary concerning the number and sizes of PCR products, the chemistry employed, and the number of sequencing primers [27
]. One review reports the use of 58 sets of unique sequencing primers to completely cover the mt genome, while another protocol provides 77 sequencing primers for the codR and 7 additional primers for the CR [27
]. There, sequencing is performed on 12 amplicons that cover the whole mt genome in an overlapping manner [29
]. In a recent protocol [28
] the amplification of the entire mt genome was conducted with only two overlapping amplicons, followed by 48 upstream and downstream sequencing reactions. Whereas amplicon sizes must be kept short for forensic samples for reasons of limited DNA quality and quantity, a reduction of the necessary amplicons is desirable for other applications, where usually fresh DNA is obtained. This simplifies the laboratory work and minimizes potential amplicon mix-up [19
]. Independent of the amplification strategy high sequence quality is required to achieve reliable base-calling.
We addressed this issue by presenting a set of 96 carefully selected sequencing primers that are embedded in a reliable and fail-safe sequencing strategy. The following criteria were applied to guide the development. (1) Each nucleotide reported in the consensus sequence should derive from at least two independent sequencing reactions using different primers (double strand coverage) to avoid the reporting of phantom mutations and other ambiguous base-callings. (2) We envision a minimum number of PCR products to reduce the chance for amplicon mix-up during the (manual) set-up of sequencing reactions and (3) we selected primers that produce sequences with an optimal signal-to-noise ratio to enable unequivocal assignment of point and length heteroplasmy.