Examination of alternative splice code of neurexins for synaptic specification

The brain is composed of a large number of cells types that assemble together into highly specific circuits. Precise connectivity is crucial to ensure proper brain functions and requires mechanisms that generate molecular diversity to encode certain aspects of neuronal wiring. One possible mechanism to generate molecular diversity is alternative splicing. For example through alternative splicing Neurexin (Nrxn1-2-3) genes have the potential to give rise to more than 12’000 protein isoforms (Tabuchi and Sudhof, 2002). Neurexin has been demonstrated to be involved in synapse formation and functions (Reissner et al., 2013). Importantly, studies have reported that alternative splicing plays a pivotal role in the function of Neurexin as it regulates their interaction with a large variety of ligands (Reissner et al., 2013). This in turn can promote the differentiation of distinct postsynaptic structures (Chih et al., 2006). Therefore, Neurexins constitute ideal candidates to encode certain parameters of synaptic connectivity. However, a central question has remained unraveled. Indeed, the spatial logic of Neurexin isoforms expression in the brain is not well understood.
Here, I report that by using bichromatic reporters, alternative splicing is differentially regulated between neuronal and non-neuronal cell populations and that the alternative splicing activity within a cell population can exhibit different levels of cell-to-cell variations. By profiling Nrxn mRNA repertoires in genetically-defined neuronal cell populations, I have identified highly divergent splice insert incorporation choices in two fundamentally different neurons populations in the hippocampus. Indeed, exon 21 which encodes for splice insert at alternative splice segment 4 (AS4) in Nrxn is predominantly incorporated in mRNA in Parvalbumin interneurons compared to excitatory Camk2 pyramidal neurons. Finally I investigated the function of Neurexin isoforms containing the exon 21 in vivo by conditionally deleting them in Parvalbumin interneurons population. Anatomical analyses indicated that synaptic density and vesicle docking were unaltered. However, mice in which isoforms containing splice insert at AS4 in Nrxn 1 and 3 were deleted, displayed an impaired short-term memory formation.
Thus, my study has provided evidences that alternative splicing regulation of Nrxn genes is genetically encoded and that deletion of cell-type specific isoforms impairs neuronal functions. This highlights the relevance of cell-type specific regulation of alternative splicing.