Development of GEARBOCS and its Applications:

A) Experimental scheme for GEARBOCS-mediated in vivo genome editing in cortical astrocytes. GEARBOCS AAVs are produced with AAV-PHP.eB capsid and retro-orbitally injected into the loxP-STOP-loxP Cas9 or Cas9-EGFP mice at P21. Brains are prepared for immunohistochemistry and subsequent confocal fluorescent imaging at P42. B) Schematic of GEARBOCS vector showing its four essential elements for CRISPR/Cas9-based genome editing: 1) A human U6 promoter that can drive the expression of a unique guide RNA (gRNA) which is cloned into the gRNA cloning site (Sap1 restriction enzyme); 2) Donor Insertion Site (DIS) wherein the donor DNA (e.g. encoding epitope tags, EGFP, mCherry etc.) can be cloned between the BamH1 and Sal1 restriction sites, 3) Cre expression cassette driven by astrocyte-specific gfaABC1D promoter with microRNA targeting cassette; 4) AAV2 Inverted terminal repeats (ITR) for AAV packaging and expression. C). Guide RNA target site present in the genome is recognized by the unique gRNA/Cas9 complex through its recognition sequence and PAM sequence (green box) followed by the double stranded break at cut site (between red and yellow box). D). Donor DNA designed to clone into the GEARBOCS has the donor sequence (Light blue Box) flanked by the guide RNA target sites at both ends. GEARBOCS can accommodate up to 2kb size donors for AAV mediated cargo delivery and in vivo genome editing. E-F Schematic of GEARBOCS mediated genome editing mechanism in astrocytes. In GEARBOCS without donor model, AAV mediated delivery of GEARBOCS into the mouse cortex in loxP-STOP-loxP Cas9 mice leads to the Cre-mediated expression of spCas9 in astrocytes through the gfaABC1D promoter in GEARBOCS. E) In the absence of a donor sequence, U6 driven gRNA, and spCas9 cause double strand breaks in the gene of interest, which is followed by the imprecise non-homologous end joining (NHEJ) repair process, leading to indels and subsequent gene knockout (KO). F) In GEARBOCS with donor, guided by the gRNA, spCas9 makes double stranded breaks both within the gene of interest and the GEARBOCS vector around the donor sites. The excised donor fragment from the GEARBOCS can integrate into the genome by homology-independent targeted integration (HITI). The donors are designed to be in frame with the gene of interest. If the endogenous tagging of a protein of interest in astrocytes (TagIn) is needed, the tag is knocked in frame with both the START and STOP codons of the gene of interest. If the donor has its own STOP codon and polyA tail, then this will lead to GeneTrap.

GEARBOCS efficiently targets mouse cortical astrocytes:

A). Schematic of the experimental scheme. GEARBOCS AAV is generated with AAV-PHP.eB capsid and retro-orbitally injected into the loxP-STOP-loxP Cas9-EGFP mice at P21. Three weeks post viral injection (P42), brains are collected for analysis. B) Confocal immunofluorescent image of a sagittal section from AAV-GEARBOCS-injected loxP-STOP-loxP Cas9-EGFP mouse brain showing the prevalence of Cas9-EGFP expression (green). Scale bar=50μm. C). Confocal immunofluorescent image from V1 cortex (indicated as dotted box in B) showing the Cas9-EGFP positive cells in different cortical layers. Scale bar=30μm. D-F) Confocal immunofluorescent images showing EGFP positive cells (green) co-stained with different cell type-specific markers (magenta) such as (D) NeuN (neurons), (E) Olig2 (oligodendrocytes), and (F) Sox9 (astrocytes). Scale bar=20μm. G) Cell type-specificity of GEARBOCS-Cre expression quantified by as the percentage of each cell types co-localized with EGFP positive cells in the V1 cortex. H) Cell type efficiency of GEARBOCS-Cre expression quantified by taking the percentage of EGFP positive cells co-localized with cell type marker in the V1 cortex. (G-H: n=6 animals, Error bars show SEM).

GEARBOCS-mediated editing of Sparlc1 gene in astrocytes:

A). Schematic showing the first gRNA target site in Sparcl1 locus used in this study. B-D). Confocal immunofluorescence microscopy images showing a Cas9-EGFP positive astrocyte (green) transduced with AAV-GEARBOCS vector without the gRNA (Control). This astrocyte still has Sparcl1 staining (magenta). E-G). Confocal immunofluorescence microscopy images showing the loss of Sparcl1 staining (magenta) in a Cas9-EGFP (green) positive AAV-GEARBOCS-Sparcl1-KO transduced astrocyte. H-J) Confocal immunofluorescence microscopy images of a Cas9-EGFP-positive astrocytes (green) transduced with the AAV-GEARBOCS-Sparcl1-KO in a large field of view. K) Quantification of the percentage of Cas9-EGFP positive astrocytes that are Hevin positive or negative by immunohistochemistry. N = 6 animals, p <0.001 by student’s t-test. L) Schematic showing the gRNA target site in the mouse Sparcl1 gene for AAV-GEARBOCS-Sparcl1-TagIn-C-mCherry. M-P). Confocal Immunofluorescence microscopy images of a Cas9-EGFP positive astrocyte (green) transduced with AAV-GEARBOCS-Sparcl1-TagIn-C-mCherry showing the co-localization of Sparcl1 (magenta) with mCherry (cyan). Q-T). Confocal immunofluorescence image of Cas9-EGFP positive astrocytes (green) transduced with the AAV-GEARBOCS-Sparcl1-TagIn-C-mCherry in a large field of view. U) Quantification of the percentage of Cas9-EGFP positive astrocytes that are mCherry positive or negative by immunohistochemistry. N = 6 animals, p <0.001 by student’s t-test. Scale bars=20μm. V) Confocal immunofluorescence image of Cas9-EGFP positive astrocyte transduced with the AAV-GEARBOCS-Sparcl1-GeneTrap. Scale bars=20μm.

Astrocytic Sparcl1 is required for formation and maintenance of thalamocortical synapses:

A). Schematic of thalamocortical inputs into the primary visual cortex (V1). Thalamocortical inputs (magenta) are labeled with VGlut2. B) Schematic of AAV-GEARBOCS-Sparcl1-KO experimental scheme. AAVs were retro-orbitally injected into the loxP-STOP-loxP Cas9 mice at P21 and the brains were collected at P42. C-D) Confocal immunofluorescence images of thalamocortical excitatory synapses marked as close apposition of VGluT2 (magenta) and PSD95 (green). E) Quantification of synaptic density of thalamocortical co-localized puncta. F) Schematic of AAV-GEARBOCS-Sparcl1-KO experimental scheme. AAVs were retro-orbitally injected into the loxP-STOP-loxP Cas9 mice at P42 and the brains were collected at P65. G-H) Confocal immunofluorescence images of thalamocortical excitatory synapses marked as close apposition of VGluT2 (magenta) and PSD95 (green). I) Quantification of synaptic density of thalamocortical co-localized puncta. J) Schematic of intracortical synapses in the primary visual cortex (V1). Intracortical inputs (magenta) are labeled with VGlut1. K) Schematic of AAV-GEARBOCS-Sparcl1-KO experimental scheme. AAVs were retro-orbitally injected into the loxP-STOP-loxP Cas9 mice at P21 and the brains were collected at P42. L-M) Confocal immunofluorescence images of thalamocortical excitatory synapses marked as close apposition of VGluT1 (magenta) and PSD95 (green). N) Quantification of synaptic density of thalamocortical co-localized puncta. O) Schematic of AAV-GEARBOCS-Sparcl1-KO experimental scheme. AAVs were retro-orbitally injected into the loxP-STOP-loxP Cas9 mice at P42 and the brains were collected at P65. P-Q) Confocal immunofluorescence images of thalamocortical excitatory synapses marked as close apposition of VGluT1 (magenta) and PSD95 (green). R) Quantification of synaptic density of thalamocortical co-localized puncta. n = 6 animals per condition. p values were calculated using a linear mixed effects model to account for the multiple images taken from each animal. Error bars represent 1 standard error of the mean. Scale bars=10μm.

In vivo mouse cortical astrocytes express Vamp2:

A). Schematic of Vamp2 (dark blue) as a member of the SNARE complex. B). Confocal immunofluorescence microscopy image from an mCherry-positive astrocyte (magenta) transduced with AAV-gfaABC1D-mCherry-CAAX co-stained with a Vamp2 antibody (cyan). Scale bar=10μm. C). Zoom in image of a single optical section showing the presence of Vamp2 (cyan) inside the mCherry-CAAX filled astrocytes (magenta). Scale bar=10μm. D). IMARIS reconstructed image showing the spatial distribution of Vamp2 (cyan) within the mCherry-CAAX filled astrocyte domain (magenta); Scale bar-10μm. E). Schematic of mouse Vamp2 gene showing the gRNA target site and the experimental scheme showing the GEARBOCS-Vamp2-GeneTrap. AAV-GEARBOCS-Vamp2-GeneTrap was retro-orbitally injected into the loxP-STOP-loxP Cas9 mice at P21 and immunohistochemical analyses of mCherry and Vamp2 were carried out at P42. F). Confocal immunofluorescence image of an mCherry positive astrocyte (magenta) from AAV-GEARBOCS-Vamp2-GeneTrap transduced brain showing that Vamp2 staining within the mCherry-positive astrocyte is diminished. Scale bar=10μm. G). Zoom in image of a single optical section showing the reduction of Vamp2 staining inside the mCherry-CAAX positive GEARBOCS-Vamp2- GeneTrap astrocyte (magenta). Scale bar-10μm. H). IMARIS reconstructed image showing the decreased expression of Vamp2 (cyan) in a GEARBOCS-Vamp2-GeneTrap astrocyte (magenta). Scale bar-10μm. I). Schematic of the experimental scheme showing the GEARBOCS-Vamp2-TagIn with an HA tag. AAV-GEARBOCS-Vamp2-TagIn-HA was retro-orbitally injected into the loxP-STOP-loxP Cas9 mice at P21 and immunohistochemical analyses of HA and Vamp2 were carried out at P42. J-L). Confocal immunofluorescence image of a single optical section from an astrocyte transduced with the AAV-GEARBOCS-Vamp2-TagIn-HA showing the co-localization Vamp2 (cyan) with HA (Red). Scale bar-20μm. M). Zoom in image of a single optical section showing the co-localization of Vamp2 with HA (white). Scale bar=10μm.

Astrocytic Vamp2 is required for regulating excitatory and inhibitory synapse numbers:

A). Schematic of both AAV-GEARBOCS-Vamp2-GeneTrap and AAV-gfaABC1D-mCherry-CAAX plasmid and the experimental scheme of AAV injection and synapse number quantification assays. AAVs were generated with the AAV-PHP.eB capsid and retro-orbitally injected into the loxP-STOP-loxP Cas9 mice at P21 and the brains were collected at P42. For synapse assays we used presynaptic VGluT1 and postsynaptic PSD95 markers. B). Schematic showing layer II/III VGluT1 (magenta) intracortical presynaptic inputs in V1 cortex. C-D) Confocal immunofluorescence images of intracortical excitatory synapses marked as close apposition of VGluT1 (magenta) and PSD95 (green). E) Quantification of synaptic density of intracortical co-localized puncta. 6 mice per condition. F). Schematic showing layer II/III VGAT (magenta) inhibitory presynaptic inputs in V1 cortex. G-H) Confocal immunofluorescence images of intracortical excitatory synapses marked as close apposition of VGAT (magenta) and Gephyrin (green). I) Quantification of synaptic density of intracortical co-localized puncta. 6 mice per condition. p values were calculated using a linear mixed effects model to account for the multiple images taken from each animal. Error bars represent 1 standard error of the mean. Scale bars= 10μm.