Hardwiring of fine synaptic layers in the zebrafish visual pathway

Background Neuronal connections are often arranged in layers, which are divided into sublaminae harboring synapses with similar response properties. It is still debated how fine-grained synaptic layering is established during development. Here we investigated two stratified areas of the zebrafish visual pathway, the inner plexiform layer (IPL) of the retina and the neuropil of the optic tectum, and determined if activity is required for their organization. Results The IPL of 5-day-old zebrafish larvae is composed of at least nine sublaminae, comprising the connections between different types of amacrine, bipolar, and ganglion cells (ACs, BCs, GCs). These sublaminae were distinguished by their expression of cell type-specific transgenic fluorescent reporters and immunohistochemical markers, including protein kinase Cβ (PKC), parvalbumin (Parv), zrf3, and choline acetyltransferase (ChAT). In the tectum, four retinal input layers abut a laminated array of neurites of tectal cells, which differentially express PKC and Parv. We investigated whether these patterns were affected by experimental disruptions of retinal activity in developing fish. Neither elimination of light inputs by dark rearing, nor a D, L-amino-phosphono-butyrate-induced reduction in the retinal response to light onset (but not offset) altered IPL or tectal lamination. Moreover, thorough elimination of chemical synaptic transmission with Botulinum toxin B left laminar synaptic arrays intact. Conclusion Our results call into question a role for activity-dependent mechanisms – instructive light signals, balanced on and off BC activity, Hebbian plasticity, or a permissive role for synaptic transmission – in the synaptic stratification we examined. We propose that genetically encoded cues are sufficient to target groups of neurites to synaptic layers in this vertebrate visual system.

Abstract licence: CC BY

40 more references, whose wording the paper’s licence keeps private.

A person counts once a day per paper. We keep no cookie and no address: a reader is a hash that changes every day. Any month under five reads shows no number.