Chemical and mechanical cues from the cerebrospinal fluid (CSF) can affect the development and function of the central nervous system (CNS). How such cues are detected and relayed to the CNS remains elusive. Cerebrospinal fluid-contacting neurons (CSF-cNs) situated at the interface between the CSF and the CNS are ideally located to convey such information to local networks. In the spinal cord, these GABAergic neurons expressing the PKD2L1 channel extend an apical extension into the CSF and an ascending axon in the spinal cord. In zebrafish and mouse spinal CSF-cNs originate from two distinct progenitor domains characterized by distinct cascades of transcription factors. Here we ask whether these neurons with different developmental origins differentiate into cells types with different functional properties. We show in zebrafish larva that the expression of specific markers, the morphology of the apical extension and axonal projections, as well as the neuronal targets contacted by CSF-cN axons, distinguish the two CSF-cN subtypes. Altogether our study demonstrates that the developmental origins of spinal CSF-cNs give rise to two distinct functional populations of sensory neurons. This work opens novel avenues to understand how these subtypes may carry distinct functions related to development of the spinal cord, locomotion and posture.

Multiple types of microvilliated sensory cells exhibit an apical extension thought to be instrumental in the detection of sensory cues. The investigation of the mechanisms underlying morphogenesis of sensory apparatus is critical to understand the biology of sensation. Most of what we currently know comes from the study of the hair bundle of the inner ear sensory cells, but morphogenesis and function of other sensory microvilliated apical extensions remain poorly understood. We focused on spinal sensory neurons that contact the cerebrospinal fluid (CSF) through the projection of a microvilliated apical process in the central canal, referred to as cerebrospinal fluid-contacting neurons (CSF-cNs). CSF-cNs respond to pH and osmolarity changes as well as mechanical stimuli associated with changes of flow and tail bending. In vivo time-lapse imaging in zebrafish embryos revealed that CSF-cNs are atypical neurons that do not lose their apical attachment and form a ring of actin at the apical junctional complexes (AJCs) that they retain during differentiation. We show that the actin-based protrusions constituting the microvilliated apical extension arise and elongate from this ring of actin, and we identify candidate molecular factors underlying every step of CSF-cN morphogenesis. We demonstrate that Crumbs 1 (Crb1), Myosin 3b (Myo3b), and Espin orchestrate the morphogenesis of CSF-cN apical extension. Using calcium imaging in crb1 and espin mutants, we further show that the size of the apical extension modulates the amplitude of CSF-cN sensory response to bending of the spinal cord. Based on our results, we propose that the apical actin ring could be a common site of initiation of actin-based protrusions in microvilliated sensory cells. Furthermore, our work provides a set of actors underlying actin-based protrusion elongation shared by different sensory cell types and highlights the critical role of the apical extension shape in sensory detection.