BACKGROUND: The availability of mitochondrial genomes has allowed for the resolution of numerous questions
regarding the evolutionary history of fungi and other eukaryotes. In the Gibberella fujikuroi species complex, the exact
relationships among the so-called “African”, “Asian” and “American” Clades remain largely unresolved, irrespective of the
markers employed. In this study, we considered the feasibility of using mitochondrial genes to infer the phylogenetic
relationships among Fusarium species in this complex. The mitochondrial genomes of representatives of the three
Clades (Fusarium circinatum, F. verticillioides and F. fujikuroi) were characterized and we determined whether or not the
mitochondrial genomes of these fungi have value in resolving the higher level evolutionary relationships in the complex.
RESULTS: Overall, the mitochondrial genomes of the three species displayed a high degree of synteny, with all the
genes (protein coding genes, unique ORFs, ribosomal RNA and tRNA genes) in identical order and orientation, as
well as introns that share similar positions within genes. The intergenic regions and introns generally contributed
significantly to the size differences and diversity observed among these genomes. Phylogenetic analysis of the
concatenated protein-coding dataset separated members of the Gibberella fujikuroi complex from other Fusarium
species and suggested that F. fujikuroi (“Asian” Clade) is basal in the complex. However, individual mitochondrial
gene trees were largely incongruent with one another and with the concatenated gene tree, because six distinct
phylogenetic trees were recovered from the various single gene datasets.
CONCLUSION: The mitochondrial genomes of Fusarium species in the Gibberella fujikuroi complex are remarkably
similar to those of the previously characterized Fusarium species and Sordariomycetes. Despite apparently
representing a single replicative unit, all of the genes encoded on the mitochondrial genomes of these fungi do not
share the same evolutionary history. This incongruence could be due to biased selection on some genes or
recombination among mitochondrial genomes. The results thus suggest that the use of individual mitochondrial
genes for phylogenetic inference could mask the true relationships between species in this complex.