Abstract

Cells are the biological units that compose every animal. The molecular toolkit of each of these cells might well recapitulate its evolutionary history. In the lab we make use of a cell-type approach to study core evolutionary events in bilaterians by comparing the cell-complement of distantly related body plans. In order to compare cell-types within and across organisms, it is necessary to use the molecular information in a system-wide and unbiased manner. We used the annelid model Platynereis dumerilii, an animal for which we have generated gene expression atlases for the pre-metamorphosis life stages, achieving single-cell resolution using high-throughput imaging and image analysis routines. To complement this resource, we have sequenced the transcriptomes of thousands of dissociated single cells that we can spatially locate using our atlas. Additionally, we have generated an ultrastructure dataset for a full animal at six days post-fertilization, to which we can map gene expression in order to guide the reconstruction of a full connectome. The integration of these different technologies into a common atlas has the potential of rendering the location, morphology, gene expression and connectivity information for each cell within an entire organism. Using this resource we have begun to assess specific cell types functionally. We have identified a cell population expressing the gene dbx1. This transcription factor specifies a unique neuronal population in the vertebrate spinal cord. These neurons are commissural and are involved in the control of the left-right alternation during locomotion. Using Crispr-Cas technology, we have successfully knocked-out dbx1 in Platynereis and analyzed its phenotype. Preliminary results show that dbx1 neurons have a reduced number of commissures and an altered locomotion pattern. These results raise the possibility of an ancestral commissural dbx1 neuronal population being present in the Urbilaterian ancestor, with a possible role in locomotion control.