Signal conflicts in the phylogeny of the primary photosynthetic eukaryotes.

It is widely accepted that the first photosynthetic eukaryotes arose from a single primary endosymbiosis of a cyanobacterium in a phagotrophic eukaryotic host, which led to the emergence of three major lineages: Chloroplastida (green algae and land plants), Rhodophyta, and Glaucophyta. For a long time, Glaucophyta have been thought to represent the earliest branch among them. However, recent massive phylogenomic analyses of nuclear genes have challenged this view, because most of them suggested a basal position of Rhodophyta, though with moderate statistical support. We have addressed this question by phylogenomic analysis of a large data set of 124 proteins transferred from the chloroplast to the nuclear genome of the three Archaeplastida lineages. In contrast to previous analyses, we found strong support for the basal emergence of the Chloroplastida and the sister-group relationship of Glaucophyta and Rhodophyta. Moreover, the reanalysis of chloroplast gene sequences using methods more robust against compositional and evolutionary rate biases sustained the same result. Finally, we observed that the basal position of Rhodophyta found in the phylogenies based on nuclear genes depended on the sampling of sequences used as outgroup. When eukaryotes supposed to have never had plastids (animals and fungi) were used, the analysis strongly supported the early emergence of Glaucophyta instead of Rhodophyta. Therefore, there is a conflicting signal between genes of different evolutionary origins supporting either the basal branching of Glaucophyta or of Chloroplastida within the Archaeplastida. This second possibility would agree with the existence of the subkingdom Biliphyta, joining Glaucophyta and Rhodophyta.