Background: The heads of annelids (earthworms, polychaetes, and others) and arthropods (insects, myriapods, spiders, and others) and the arthropod-related onychophorans (velvet worms) show similar brain architecture and for this reason have long been considered homologous. However, this view is challenged by the 'new phylogeny' placing arthropods and annelids into distinct superphyla, Ecdysozoa and Lophotrochozoa, together with many other phyla lacking elaborate heads or brains. To compare the organisation of annelid and arthropod heads and brains at the molecular level, we investigated head regionalisation genes in various groups. Regionalisation genes subdivide developing animals into molecular regions and can be used to align head regions between remote animal phyla. Results: We find that in the marine annelid Platynereis dumerilii, expression of the homeobox gene six3 defines the apical region of the larval body, peripherally overlapping the equatorial otx+ expression. The six3+ and otx+ regions thus define the developing head in anterior-to-posterior sequence. In another annelid, the earthworm Pristina, as well as in the onychophoran Euperipatoides, the centipede Strigamia and the insects Tribolium and Drosophila, a six3/optix+ region likewise demarcates the tip of the developing animal, followed by a more posterior otx/otd+ region. Identification of six3+ head neuroectoderm in Drosophila reveals that this region gives rise to median neurosecretory brain parts, as is also the case in annelids. In insects, onychophorans and Platynereis, the otx+ region instead harbours the eye anlagen, which thus occupy a more posterior position. Conclusions: These observations indicate that the annelid, onychophoran and arthropod head develops from a conserved anterior-posterior sequence of six3+ and otx+ regions. The six3+ anterior pole of the arthropod head and brain accordingly lies in an anterior-median embryonic region and, in consequence, the optic lobes do not represent the tip of the neuraxis. These results support the hypothesis that the last common ancestor of annelids and arthropods already possessed neurosecretory centres in the most anterior region of the brain. In light of its broad evolutionary conservation in protostomes and, as previously shown, in deuterostomes, the six3-otx head patterning system may be universal to bilaterian animals.

The question of why animals vary in their ability to regenerate remains one of the most intriguing questions in biology. Annelids are a large and diverse phylum, many members of which are capable of extensive regeneration such as regrowth of a complete head or tail and whole-body regeneration, even from few segments. On the other hand, some representatives of both of the two major annelid clades show very limited tissue regeneration and are completely incapable of segmental regeneration. Here we review experimental and descriptive data on annelid regeneration, obtained at different levels of organization, from data on organs and tissues to intracellular and transcriptomic data. Understanding the variety of the cellular and molecular basis of regeneration in annelids can help one to address important questions about the role of stem/dedifferentiated cells and “molecular morphallaxis” in annelid regeneration as well as the evolution of regeneration in general.