Journal: Nature methods

Article Title: Multiplexed Cre-dependent selection yields systemic AAVs for targeting distinct brain cell types

doi: 10.1038/s41592-020-0799-7

Figure Lengend Snippet: A multiplexed selection approach to identify capsids with specific and broad tropisms. Steps 1–6 describe the workflow in Round-1 (R1) selection, steps 7–9 describe Round-2 (R2) selection using synthetic pool method, steps 1a, 2a, and 6a-b show the incorporation of deep sequencing to recover capsids after R1 and R2 selection, and steps 10–11 describe positive and/or negative selection criteria followed by variant characterization. b , Structural model of the AAV9 capsid (PDB 3UX1) with the insertion site for the 7-mer-i library highlighted in red in the 60-meric (left), trimeric (middle), and monomeric (right) forms. c , Empirical Cumulative Distribution Frequency (ECDF) of R1 DNA and virus libraries that were recovered by deep sequencing post Gibson assembly and virus production, respectively. d , Distributions of variants recovered from three R1 brain tissue libraries, Tek, SNAP25, and GFAP (n = 2 per Cre line), are shown with capsid libraries sorted by decreasing order of the enrichment score. The enrichment score of AAV-PHP.V2 variant, described later, is mapped on this plot.

Article Snippet: The following vector plasmids are deposited on Addgene for distribution ( http://www.addgene.org ) AAV-PHP.V1: 127847, AAV-PHP.V2: 127848, AAV-PHP.B4: 127849, and AAV-PHP.N: 127851.

Techniques: Selection, Sequencing, Variant Assay

Journal: Nature methods

Article Title: Multiplexed Cre-dependent selection yields systemic AAVs for targeting distinct brain cell types

doi: 10.1038/s41592-020-0799-7

Figure Lengend Snippet: Clustering analysis of positively enriched variants from Tek (left), GFAP (middle) and SNAP/Syn (right) synthetic pool brain libraries with size of nodes representing their relative enrichment in brain, and the thickness of edges (connecting lines) representing degree of relatedness. Distinct families (yellow) with the corresponding AA frequency logos (AA size represents prevalence and color encodes AA properties) are shown. b , The 7-mer insertion peptide sequences of AAV-PHP variants between AA positions 588–589 of AAV9 capsid are shown. AAs are colored by shared identity to AAV-PHP.B and eB (green) or among new variants (unique color per position). c , AAV9 (left) and AAV-PHP.V1 (right) mediated expression using ssAAV:CAG-mNeongreen genome (green, n = 3, 3 weeks of expression in C57BL/6J adult mice with 3×10 11 vg IV dose/mouse) is matched in fluorescence intensity in sagittal sections of brain (above) with higher magnification image from cortex (below). Magenta is αGLUT1 antibody staining for vasculature. d , Percentage of vasculature stained with αGLUT1 that overlaps with mNeongreen (XFP) expression in cortex. One-way ANOVA non-parametric Kruskal-Wallis test (P-value 0.0036), and follow-up multiple comparisons using uncorrected Dunn’s test (P-value of 0.0070 for AAV9 vs PHP.V1) are reported. **P ≤ 0.01 is shown, P > 0.05 is not shown; data is mean ± S.E.M, n= 3 mice per AAV variant, cells quantified from 4–2 images per mouse per cell-type. e , Percentage of cells stained with each cell-type specific marker (αGLUT1, αS100 for astrocytes, αNeuN for neurons, αOlig2 for oligodendrocyte lineage cells) that overlaps with mNeongreen (XFP) expression in cortex. Kruskal-Wallis test (P-value of 0.0078), and uncorrected Dunn’s test (P-value of 0.0235 for neuron vs vascular cells, and 0.0174 for neuron vs astrocyte, respectively) are reported. *P ≤ 0.05 is shown, and P > 0.05 is not shown; data is mean ± S.E.M, n= 3 mice, cells quantified from 4–2 images per mouse per cell-type. f , Vascular transduction by ssAAV-PHP.V1:CAG-DIO-EYFP in Tek-Cre adult mice (left) ( n = 2, 4 weeks of expression, 1×10 12 vg IV dose/mouse), and by ssAAV-PHP.V1:Ple261-iCre in Ai14 reporter mice (right) ( n = 2, 3 weeks of expression, 3×10 11 vg IV dose/mouse). Tissues are stained with αGLUT1 (magenta (left) and cyan (right)). g , Efficiency of vascular transduction (as described in d ) in Tek-Cre mice (n= 2, mean from 3 images per mouse per brain region). h , Efficiency of vascular transduction in Ai14 mice (n= 2, a mean from 4 images per mouse per brain region).

Article Snippet: The following vector plasmids are deposited on Addgene for distribution ( http://www.addgene.org ) AAV-PHP.V1: 127847, AAV-PHP.V2: 127848, AAV-PHP.B4: 127849, and AAV-PHP.N: 127851.

Techniques: Expressing, Fluorescence, Staining, Variant Assay, Marker, Transduction

Journal: Nature methods

Article Title: Multiplexed Cre-dependent selection yields systemic AAVs for targeting distinct brain cell types

doi: 10.1038/s41592-020-0799-7

Figure Lengend Snippet: Transduction by AAV-PHP.B4–B6 and C1 variants, as well as B, eB, and AAV9 controls in sagittal brain and liver sections. Fluorescence intensity is matched with AAV-PHP.eB across each set of images (column-wise). The white box on the sagittal brain images marks the thalamus and not the precise region of the figures to the right. Vectors are packaged with ssAAV:CAG-2xNLS-EGFP genome ( n = 3 per group, 1×10 11 vg IV dose/adult C57BL/6J mouse, 3 weeks of expression). Tissues are stained with cell-type specific markers (magenta): αNeuN for neurons, αS100 for astrocytes and αOlig2 for oligodendrocyte lineage cells. Liver tissues are stained with a DNA stain, DAPI (blue). b , The percentage of αNeuN + , αS100 + and αOlig2 + cells with detectable nuclear-localized EGFP in the indicated brain regions are shown (n=3 per group, 1×10 11 vg dose). A two-way ANOVA with correction for multiple comparisons using Tukey’s test is reported with adjusted P-values (****P ≤ 0.0001, ***P ≤ 0.001, **P ≤ 0.01, *P ≤ 0.05, is shown, and P > 0.05 is not shown on the plot; 95% CI, data is mean ± S.E.M. The dataset comprises a mean of 2 images per region per cell-type marker per mouse).

Article Snippet: The following vector plasmids are deposited on Addgene for distribution ( http://www.addgene.org ) AAV-PHP.V1: 127847, AAV-PHP.V2: 127848, AAV-PHP.B4: 127849, and AAV-PHP.N: 127851.

Techniques: Transduction, Fluorescence, Expressing, Staining, Marker

Journal: Nature methods

Article Title: Multiplexed Cre-dependent selection yields systemic AAVs for targeting distinct brain cell types

doi: 10.1038/s41592-020-0799-7

Figure Lengend Snippet: The design of the 3-mer-s PHP.B library with combinations of three AA diversification between AA 587–597 of AAV-PHP.B (or corresponding AA 587–590 of AAV9). Shared AA identity with the parent AAV-PHP.B (green) is shown along with unique motifs for AAV-PHP.N (pink) and AAV-PHP.eB (blue). b , Distributions of R2 brain and liver libraries (at AA level) by enrichment score (normalized to R2 virus library, with variants sorted in decreasing order of enrichment score). The enrichment of AAV-PHP.eB and AAV-PHP.N across all libraries are mapped on the plot. c , Heatmap represents the magnitude (log2 fold change) of a given AA’s relative enrichment or depletion at each position across the diversified region, only if statistical significance is reached on fold change (boxed if p-value ≤ 0.0001, two-sided, two-proportion z-test, p-values corrected for multiple comparisons using Bonferroni correction). Plot includes variants that were highly enriched in brain (>0.5 mean enrichment score, where mean is drawn across Vglut2, Vgat and GFAP, n = 1 library per mouse line (sample pooled from 2 mice per line)) and negatively enriched in liver (<0.0) (32 AA sequences). d , Clustering analysis of positively enriched variants from Vgat brain library is shown with node size representing the degree of negative enrichment in liver and the thickness of edges (connecting lines) representing degree of relatedness between nodes. Two distinct families are highlighted in yellow and their corresponding AA frequency logos are shown below (AA size represents prevalence and color encodes AA properties). e , The percentage of neurons, astrocytes and oligodendrocyte lineage cells with ssAAV-PHP.N:CAG-2xNLS-EGFP in the indicated brain regions is shown ( n = 3, 1×10 11 vg IV dose per adult C57BL/6J mouse, 3 weeks of expression, data is mean±S.E.M, 6–8 images for cortex, thalamus and striatum, and 2 images for ventral midbrain, per mouse per cell-type marker using 20x objective covering the entire regions). A two-way ANOVA with correction for multiple comparisons using Tukey’s test gave adjusted P-values reported as ****P ≤ 0.0001, ns for P > 0.05, 95% CI. f , Transduction by ssAAV-PHP.N:CAG-NLS-EGFP ( n = 2, 2×10 11 vg IV dose per adult C57BL/6J mouse, 3 weeks of expression) is shown with NeuN staining (magenta) across three brain areas (cortex, SNc (substantia nigra pars compacta) and thalamus).

Article Snippet: The following vector plasmids are deposited on Addgene for distribution ( http://www.addgene.org ) AAV-PHP.V1: 127847, AAV-PHP.V2: 127848, AAV-PHP.B4: 127849, and AAV-PHP.N: 127851.

Techniques: IF-P, Expressing, Marker, Transduction, Staining

Journal: Nature methods

Article Title: Multiplexed Cre-dependent selection yields systemic AAVs for targeting distinct brain cell types

doi: 10.1038/s41592-020-0799-7

Figure Lengend Snippet: Clustering analysis showing the brain-enriched sequence families of all variants described herein, either identified in prior studies (PHP.B-B3, PHP.eB) or in the current study (PHP.B4–B8, PHP.V1–2, PHP.C1–3). The thickness of edges (connecting lines) representing degree of relatedness between nodes. The AA sequences inserted between 588–589 (of AAV9 capsid) for all the variants discussed are shown below. b , Transduction of AAV9, AAV-PHP.V1 and AAV-PHP.N across three different mouse strains: C57BL/6J, BALB/cJ and FVB/NJ are shown in sagittal brain sections (right), along with a higher magnification image of the thalamus brain region (left). c , Transduction by AAV-PHP.B, AAV-PHP.C1–C3 in C57BL/6J and BALB/cJ mice are shown in sagittal brain sections (right), along with a higher magnification image of the thalamus brain region (left). b , c , The white box on the sagittal brain images represents the location of thalamus and not the precise area that is zoomed-in on the figure to the left. The fluorescence intensity is matched across all sagittal sections and across all thalamus regions acquired. The insets in AAV-PHP.V1 are zoom-ins with enhanced brightness. The indicated capsids were used to package ssAAV:CAG-mNeongreen ( n = 2–3 per group, 1×10 11 vg IV dose per 6–8 weeks old adult mouse, 3 weeks of expression. The data reported in b , c are from one independent trial where all viruses were freshly prepared and titered in the same assay for dosage consistency. AAV-PHP.C2 and AAV-PHP.C3 were further validated in an independent trial for BALB/cJ, n = 2 per group).

Article Snippet: The following vector plasmids are deposited on Addgene for distribution ( http://www.addgene.org ) AAV-PHP.V1: 127847, AAV-PHP.V2: 127848, AAV-PHP.B4: 127849, and AAV-PHP.N: 127851.

Techniques: Sequencing, Transduction, Fluorescence, Expressing