Homoplasy or plesiomorphy? Reconstruction of the evolutionary history of mitochondrial gene order rearrangements in the subphylum Neodermata.

Recent mitogenomic studies have exposed a gene order (GO) shared by two classes, four orders and 31 species ('common GO') within the flatworm subphylum Neodermata. There are two possible hypotheses for this phenomenon: convergent evolution (homoplasy) or shared ancestry (plesiomorphy). To test those, we conducted a meta-analysis on all available mitogenomes to infer the evolutionary history of GO in Neodermata. To improve the resolution, we added a newly sequenced mitogenome that exhibited the common GO, Euryhaliotrema johni (Ancyrocephalinae), to the dataset. Phylogenetic analyses conducted on two datasets (nucleotides of all 36 genes and amino acid sequences of 12 protein coding genes) and four algorithms (MrBayes, RAxML, IQ-TREE and PhyloBayes) produced topology instability towards the tips, so ancestral GO reconstructions were conducted using TreeREx and MLGO programs using all eight obtained topologies, plus three unique topologies from previous studies. The results consistently supported the second hypothesis, resolving the common GO as a plesiomorphic ancestral GO for Neodermata, Cestoda, Monopisthocotylea, Cestoda + Trematoda and Cestoda + Trematoda + Monopisthocotylea. This allowed us to trace the evolutionary GO scenarios from each common ancestor to its descendants amongst the Monogenea and Cestoda classes, and propose that the common GO was most likely retained throughout all of the common ancestors, leading to the extant species possessing the common GO. Neodermatan phylogeny inferred from GOs was largely incongruent with all 11 topologies described above, but it did support the mitogenomic dataset in resolving Polyopisthocotylea as the earliest neodermatan branch. Although highly derived GOs might be of some use in resolving isolated taxonomic and phylogenetic uncertainties, we conclude that, due to the discontinuous nature of their evolution, they tend to produce artefactual phylogenetic relationships, which makes them unsuitable for phylogenetic reconstruction in Neodermata. Wider and denser sampling of neodermatan mitogenomic sequences will be needed to infer the evolutionary pathways leading to the observed diversity of GOs with confidence.