Evolutionary divergence of basal and activity-dependent exon splicing in cortical neurons

Abstract

Alternative splicing of mRNA exons in mammalian neurons increases diversity of the proteome and is regulatable by signaling pathways. However, the degree of conservation of basal and signal-dependent exon usage between human neurons and those from experimental models such as mice is incompletely understood. We previously showed that cortical neuronal activity-dependent gene transcription exhibits human/mouse differences, driven by evolutionary divergence of cis-acting promoter elements (Qiu et al. 2016). Since alternative exon usage influences brain development and cognition, is controlled by neuronal activity, and is disturbed in brain disorders, we investigated human/mouse differences in exon usage in cortical neurons. Comparing orthologous exons, basal exon inclusion levels showed human-mouse conservation, but also significant differences determined by cis-acting sequences: human-mouse conservation and divergence in exon usage was recapitulated in neurons from Tc1 mice carrying human chromosome-21 (hCh21). Activity-dependent changes in exon usage also exhibited significant conservation: gene structure was more likely to be conserved in activity-regulated exons, and exons regulated in both human and mouse neurons were enriched in RBFOX and SAM68 targets, and genes were centred on cytoskeletal organisation, mRNA transcription/processing, and synaptic signaling. However, divergence was also evident, and human-specific activity-dependent exon usage was dominated by genes involved in lipid biosynthesis, signaling and trafficking. Notably, the pattern of activity-dependent usage of hCh21 exons in human neurons was not recapitulated in mouse Tc1 neurons. Thus, unlike species-specific differences in activity-dependent gene transcription, cis-acting DNA sequence divergence is insufficient to explain inter-species differences in activity-regulated exon usage. Trans-acting factors involved in activity-responsive splicing have likely also diverged.