During the course of the project, we have identified and characterized the exons and microexons of interest: those that are only present in vertebrates but not in any invertebrate animal, and that are neuronal-specific across vertebrates (what we call Vertebrate- Neural- Alternatively Spliced exons, or VN-AS exons). We have also characterized their regulation across tissues and neuronal differentiation time courses, and made a public database with all this information (VastDB; vastdb.crg.eu). To achieve this goal, we have also developed the ExOrthist, a tool to infer exon homology relationships.
We have also implemented the necessary methodology to selectively delete microexons in zebrafish, and generated 21 lines with microexon deletions, 18 of which corresponded to clean deletions with no effect on overall gene expression. We have assessed phenotypes in these lines at different organizational levels using different tests. These range from study of neuritogenesis in vivo and in culture, transcriptomic changes using whole-embryo RNA-seq, basic locomotion and sensory tests using Daniovision, and social behaviour assays with a custom set-up. We have identified several microexons with defects on neuritogenesis and/or social behavior, associated with specific transcriptomic patterns. For a few cases, we could infer and test the molecular functions for these VN-AS microexons.
Moreover, these analyses led to unexpected evolutionary findings, which we have followed up in different publications (e.g. Torres-Mendez et al 2019 and 2021, Marletaz et al 2018). In particular, we found that the programs of neural microexons originated in bilaterian ancestors, much earlier than the origin of vertebrates. This means that, although vertebrate-specific neural microexons are a key component of the VN-AS program, the regulatory machinery originated before the origin of our group. Also, it means that other lineages (e.g. flies) have evolved their own programs of neural lineage-specific microexons, which we have also investigated.
