Thiouracil (TU)-tagging is an intersectional method for covalently labeling newly transcribed RNAs within specific cell types. Cell type specificity is generated through targeted transgenic expression of the enzyme uracil phosphoribosyl transferase (UPRT); temporal specificity is generated through a pulse of the modified uracil analog 4TU. This technique has been applied in mouse using a Cre-dependent UPRT transgene, CA>GFPstop>HA-UPRT, to profile RNAs in endothelial cells, but it remained untested whether 4TU can cross the blood-brain barrier (BBB) or whether this transgene can be used to purify neuronal RNAs. Here, we crossed the CA>GFPstop>HA-UPRT transgenic mouse to a Sepw1-cre line to express UPRT in layer 2/3 of visual cortex or to an Nr5a1-cre line to express UPRT in layer 4 of visual cortex. We purified thiol-tagged mRNA from both genotypes at postnatal day (P)12, as well as from wild-type (WT) mice not expressing UPRT (background control). We found that a comparison of Sepw1-purified RNA to WT or Nr5a1-purified RNA allowed us to identify genes enriched in layer 2/3 of visual cortex. Here, we show that Cre-dependent UPRT expression can be used to purify cell type-specific mRNA from the intact mouse brain and provide the first evidence that 4TU can cross the BBB to label RNA in vivo.

Background Developing cortical neurons express a tightly choreographed sequence of cytoskeletal and transmembrane proteins to form and strengthen specific synaptic connections during circuit formation. Nectin-3 is a cell-adhesion molecule with previously described roles in synapse formation and maintenance. This protein and its binding partner, nectin-1, are selectively expressed in upper-layer neurons of mouse visual cortex, but their role in the development of cortical circuits is unknown. Methods Here we block nectin-3 expression (via shRNA) or overexpress nectin-3 in developing layer 2/3 visual cortical neurons using in utero electroporation. We then assay dendritic spine densities at three developmental time points: eye opening (postnatal day (P)14), one week following eye opening after a period of heightened synaptogenesis (P21), and at the close of the critical period for ocular dominance plasticity (P35). Results Knockdown of nectin-3 beginning at E15.5 or ~ P19 increased dendritic spine densities at P21 or P35, respectively. Conversely, overexpressing full length nectin-3 at E15.5 decreased dendritic spine densities when all ages were considered together. The effects of nectin-3 knockdown and overexpression on dendritic spine densities were most significant on proximal secondary apical dendrites. Interestingly, an even greater decrease in dendritic spine densities, particularly on basal dendrites at P21, was observed when we overexpressed nectin-3 lacking its afadin binding domain. Conclusion These data collectively suggest that the proper levels and functioning of nectin-3 facilitate normal synapse formation after eye opening on apical and basal dendrites in layer 2/3 of visual cortex.

Developing cortical neurons must express numerous molecules to form and strengthen specific synaptic connections during circuit formation. Nectin-3 is a cell-adhesion molecule with known roles in the development and maintenance of hippocampal circuits. This protein and its binding partner, nectin-1, are selectively expressed in upper-layer neurons of mouse visual cortex, but their role in the development of cortical circuits is unknown. Here we block nectin-3 expression (via shRNA) or overexpress either full length (OE) or dominant-negative (DN, lacking a C-terminus) nectin-3 in developing layer 2/3 visual cortical neurons using in utero electroporation. We then assay dendritic spine densities at three developmental time points: eye opening (postnatal day (P)14), one week following eye opening after a period of heightened synaptogenesis (P21), and at the close of the critical period for ocular dominance plasticity (P35). shRNA knockdown of nectin-3 beginning at E15.5 or ~P19, increased dendritic spine densities at P21 or P35, respectively. Increasing nectin-3 binding/adhesion by expressing either DN nectin-3 (intact extracellular binding domain but lacking a Cterminus) or OE nectin-3 constructs beginning at E15.5, produced overall decreases in dendritic spine densities. However, DN nectin-3 expression alone prevented the increase in spine densities typically observed between P14 and P21. This suggests that nectin-3 intracellular signaling (reduced with DN expression) has a role in spine formation after eye opening. These data collectively suggest that both functional intracellular signaling and balanced overall expression of nectin-3 are required for normal synapse formation during the development of layer 2/3 visual cortical neurons.