Fig. 1: Re-expression of IRSp53 in the adult IRSp53-cKO brain by BBB-penetrant PHP.eB.

a Diagrams showing the domain structure of the IRSp53 variant used in the present study and the structure of the pAAV variant carrying HA-tagged IRSp53 cDNA (pAAV-hSyn-DIO-mEGFP-GSG-P2A-HA-IRSp53) for the production of BBB-penetrant PHP.eB-IRSp53. Domains of IRSp53: IMD, IRSp53-MIM homology domain; CRIB-PR, CDC42/Rac interactive binding-proline rich domain; SH3, Src homology 3 domain; WW, WW domain; and PDZ-B, PDZ domain-binding motif. pAAV parts; ITR, inverted terminal repeat; hSynI, human synapsin I promotor; mEGFP, monomeric EGFP; HA-IRSp53, N-terminally HA-tagged IRSp53; P2A, a cleavable peptide linker. b Diagram showing injection of PHP.eB into the retro-orbital sinus for BBB penetration. c Diagram showing a conditional deletion of IRSp53 in Emx1-positive glutamate neurons in Emx1-Cre;Baiap2/Irsp53^fl/fl (IRSp53-cKO) mice and re-expression of IRSp53 in Emx1-positive glutamate neurons in these mice by the infection of PHP.eB-IRSp53 and Cre-mediated normalization of the orientation of the HA-IRSp53-EGFP cDNA construct. d Re-expression of IRSp53 in various brain regions of adult IRSp53-cKO mice, as shown by the presence of EGFP signals, representing IRSp53 expression, in brain regions including the medial prefrontal cortex (mPFC), anterior cingulate cortex (ACC), and hippocampus. Note that the hippocampal CA2 region shows stronger EGFP signals, as previously reported⁶⁶. PHP.eB-IRSp53 was injected into the retro-orbital sinus of IRSp53-cKO mice at postnatal week 8 and allowed to express IRSp53 for 4 weeks. S1, primary somatosensory cortex; M1/2, primary/secondary motor cortex; Cg1, anterior cingulate cortex; PrL, prelimbic region of the mPFC; IL, infralimbic region of the mPFC; CA2, CA2 region of the hippocampus; BLA, basolateral amygdala; LH, lateral hypothalamus. Scale bar, 1 mm. e Levels of IRSp53 proteins driven by PHP.eB-IRSp53 are ~60% of WT levels, as shown by immunoblot analysis of IRSp53 proteins in the cortex + hippocampus of WT, IRSp53-cKO, and PHP.eB-IRSp53-infected IRSp53-cKO mice (cKO-IRSp53) (3 months). IRSp53 signals were normalized to GAPDH signals (control) (n = 3 mice [WT], 3 [IRSp53-cKO], and 3 [IRSp53-cKO and PHP.eB-IRSp53]). f An example showing IRSp53 re-expression in the prelimbic region of the mPFC in IRSp53-cKO mice (3 months), as shown by immunofluorescence staining for EGFP and HA. Note that EGFP signals are mainly found in cell-body areas whereas HA signals are mainly found in neuropil areas, suggesting that HA-IRSp53 proteins separated from the fused EGFP by the proteolytic cleavage of the P2A linker were targeted to dendrite/synaptic areas. DAPI staining was used for nuclear staining. Scale bar, 50 μm. g Representative image of the colocalizations between EGFP and Vglut1/2 (glutamatergic neuron marker) in the prelimbic and infralimbic regions of the mPFC and the primary motor area in IRSp53-cKO mice infected with PHP.eB-IRSp53 (8–12 weeks), as shown by double fluorescence in situ hybridization. Scale bar, 20 μm. h Quantification of the colocalization between EGFP and Vglut1/2, calculated by EGFP-positive neurons among Vglut1/2-positive neurons (n = 8 slices from 2 mice [IRSp53-cKO and PHP.eB-IRSp53]). i Quantification of the colocalization between EGFP and Vglut1/2, calculated by Vglut1/2-positive neurons among EGFP-positive neurons (n = 8 slices from 2 mice [IRSp53-cKO and PHP.eB-IRSp53]). Error bars represent the standard errors of means.