| Phylogenetic analysis of mitochondrial DNA: Detection of mutations in patients with occipital stroke | ||
|---|---|---|
| Prev | Next | |
We found CSGE, as used here to study the mtDNA genotypes of 43 population samples belonging to haplogroup U, to be an ideal method for screening mutations and polymorphisms in mtDNA. It turned out to be a rapid method with a high capacity and it dramatically alleviated the laborious task of sequencing. A sensitivity of 0.96 was achieved in detecting sequence differences in mtDNA, with a specificity of 1.0. The phylogenetic network for the Finnish mtDNA haplogroup U constructed on the basis of these findings proved to be an unambiguous tree with few homoplasies and pointed to several previously unidentified common polymorphisms.
Ten percent of young patients with an occipital brain infarct are thought to have a mitochondrial disorder and 6% to have the common MELAS mutation 3243A>G. considering the mitochondrial genetics of patients who suffer from an occipital stroke and do not harbor 3243A>G, we found that all those with migraine as a probable aetiology for stroke belonged to mtDNA haplogroup U, suggesting that this genotype confers a risk of occipital stroke. More specific investigations revealed an association of migrainous stroke with haplogroup U5.
In addition to the five patients with migrainous stroke, we analyzed the complete mtDNA coding sequence by CSGE and the HVS I in nine other patients with occipital stroke. The network analysis showed all five patients with migrainous stroke and five others to belong to the cluster U5, one to haplogroup K and the remaining three to the haplotypes U2, U4, and U* intervening between U5 and haplogroup K, suggesting a ratio of 10:1:3, whereas the corresponding ratio in the general population is 10:1:0.75.
Sequence analysis of the entire coding region revealed that the haplotypes in the five patients with migrainous stroke differed from those observed in the controls in only one case (#285). Therefore, if mtDNA poses a risk of migrainous stroke, this risk may be defined by clusters of polymorphisms at nt 11467-12308-12372 or 3197-9477-13617. None of these substitutions may be considered pathogenic as such, but mild pathogenicity cannot be ruled out.
Four patients harboured potentially pathogenic mutations. Two of these were in transfer RNA genes and one in the 16S ribosomal RNA gene. Three mutations led to amino acid replacements. The 4295A>G in tRNAIle has previously been considered part of the aetiology of cardiomyopathy in a child. In our patient it was found to be heteroplasmic with a proportion of 97% of mutant DNA in blood. The 8296A>G in tRNALys proved to be homoplasmic in blood, but it has previously been associated with diabetes mellitus and hearing impairment. The mutation 1850T>C in the gene coding for 16S rRNA is located in a dinucleotide pair that is highly conserved between species.
Similar phylogenetic networks will be required for the purposes of medical genetics as well as population genetics. Such networks would help in distinguishing between a rare polymorphism and a pathogenic mutation in clinically affected persons. Likewise, they would enable more detailed comparisons to be made between and within populations and allow more accurate phylogenetic relationships to be determined.