Viral vectors are attractive tools to express genes in neurons. Transduction of neurons with a recombinant, replication-deficient Sindbis viral vector is a method of choice for studying the effects of short-term protein overexpression on neuronal function. However, to which extent Sindbis by itself may affect neurons is not fully understood. We assessed effects of neuronal transduction with a Sindbis viral vector on the transcriptome and proteome in organotypic hippocampal slice cultures, and analyzed the electrophysiological properties of individual CA1 neurons, at 24 h and 72 h after viral vector injection. Whereas Sindbis caused substantial gene expression alterations, changes at the protein level were less pronounced. Alterations in transcriptome and proteome were predominantly limited to proteins involved in mediating anti-viral innate immune responses. Sindbis transduction did not affect the intrinsic electrophysiological properties of individual neurons: the membrane potential and neuronal excitability were similar between transduced and non-transduced CA1 neurons up to 72 h after Sindbis injection. Synaptic currents also remained unchanged upon Sindbis transduction, unless slices were massively infected for 72 h. We conclude that Sindbis viral vectors at low transduction rates are suitable for studying short-term effects of a protein of interest on electrophysiological properties of neurons, but not for studies on the regulation of gene expression.

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.

Soluble oligomeric amyloid-β (Aβ) is a prime suspect to cause cognitive deficits in Alzheimer's disease and weakens synapses by removing AMPA-type glutamate receptors (AMPARs). We show that synapses of CA1 pyramidal neurons become vulnerable to Aβ when they express AMPAR subunit GluA3. We found that Aβ-oligomers reduce the levels of GluA3 immobilized at spines, indicating they deplete GluA3-containing AMPARs from synapses. These Aβ-driven effects critically depended on the PDZ-binding motif of GluA3. When GluA3 was expressed with a single amino acid mutation in its PDZ-binding motif that prevents GRIP binding, it did not end up at spines and Aβ failed to trigger synaptic depression. GluA3 with a different point mutation in the PDZ-motif that leaves GRIP-binding intact but prevents its endocytosis, was present at spines in normal amounts but was fully resistant to effects of Aβ. Our data indicate that Aβ-mediated synaptic depression requires the removal of GluA3 from synapses. We propose that GRIP-detachment from GluA3 is a critical early step in the cascade of events through which Aβ accumulation causes a loss of synapse.