Transcriptional diversity in synaptic gene sets is sufficient to discriminate cortical neuronal identity

Abstract

Synapse diversity has been described from different perspectives, ranging from the specific neurotransmitters released, to their diverse biophysical properties and proteome profiles. However, synapse diversity at the transcriptional level has not been systematically identified across all synapse populations in the brain. To quantify postsynaptic and identify specific synaptic features of neuronal cell types we combined the SynGO (Synaptic Gene Ontology) database with single-cell RNA sequencing data of the mouse neocortex. We show that cell types can be discriminated by synaptic genes alone with the same power as all genes. The cell type discriminatory power is not equally distributed across synaptic genes as we could identify functional categories and synaptic compartments with greater cell type specific expression. Synaptic genes, and specific SynGO categories, belonged to three different types of gene modules: gradient expression over all cell types, gradient expression in selected cell types and cell class- or type-specific profiles. This data provides a deeper understanding of synapse diversity in the neocortex and identifies potential markers to selectively identify synapses from specific neuronal populations.

Synaptic proteome diversity is primarily driven by gene regulation of glutamate receptors and their regulatory proteins

“…Thus, we propose that the differential regulation of these genes is a main feature of gene expression programs between neuronal types. This finding aligns with the recent observation that, in general, synaptic genes contribute to the classification of cortical neurons 63 .…”

“…Suppl_6a) had very different topologies, with a much higher overlap between neurons from different classes. Indicating that synaptic genes importantly contribute to the classification of hippocampal excitatory neurons, as it has been recently shown for cortical neurons 63 . To further validate this observation, we asked how many of the genes that contributed most to the classification were synaptic.…”

Connecting genomic results for psychiatric disorders to human brain cell types and regions reveals convergence with functional connectivity

“…Results for non-neuronal cells suggested a diverse range of significant GO terms consistent with the supercluster labels: astrocyte with biological adhesion, choroid plexus with cilium, microglia with immune response, oligodendrocyte with neuron ensheathment, oligodendrocyte precursor with gliogenesis, and vascular with vasculature development. GO terms for most neuronal cell types were dominated by synaptic biology, consistent with findings that forebrain neuronal cell identity is to a large extent driven by specific expression of synaptic genes 36 (exceptions were lower/upper rhombic lip and cerebellar inhibitory neurons from non-cortical regions). Fourth, the Methods section describes additional features of TDEP genes: (a) visualization of supercluster TDEP genes yielded groups of non-neuronal cells, neocortical excitatory neurons plus medium spiny neurons, and inhibitory interneurons plus non-cortical excitatory interneurons ( Figure S4A ); (b) TDEP genes tend to co-occur in genomic regions ( Figure S5 ); (c) visualizations for gene expression specificity and genomic co-occurrence were similar suggesting that TDEP genes tend to be located near each other; and (d) all neuronal TDEP genes accounted for 61-65% of the SNP-heritability for the largest brain trait GWAS (scz2020, bip2021, mdd2019*, neuroticism, education, and IQ).…”