Characterization of RNA editing and gene therapy with a compact CRISPR-Cas13 in the retina

Abstract

CRISPR-Cas13 nucleases are programmable RNA-targeting effectors that can silence gene expression in a transient manner. Recent iterations of Cas13 nucleases are compact for adeno-associated virus (AAV) delivery to achieve strong and persistent expression of various organs in a safe manner. Here, we report significant transcriptomic signatures of Cas13bt3 expression in retinal cells and show all-in-one AAV gene therapy with Cas13bt3 can effectively silence VEGFA mRNA in human retinal organoids and humanized VEGF transgenic mouse (trVEGF029, Kimba) models. Specifically, human embryonic stem cells (hESC)-derived retinal pigment epithelium cells show high expression of Cas13bt3 from virus delivery corresponding to a significant reduction of VEGFA mRNA. We further show that intravitreal delivery of Cas13bt3 by AAV2.7m8 can efficiently transduce mouse retinal cells for specific knockdown of human VEGFA in the Kimba mouse. Our results reveal important considerations for assessing Cas13 activity, and establish the Cas13bt3 RNA editing system as a potential anti-VEGF agent that can achieve significant control of VEGFA for the treatment of retinal neovascularization.

Keywords: CRISPR-Cas13; RNA editing; VEGF; gene therapy; ocular neovascularization.

Fig. 1.

Comparison of RNA-targeting tools for in vitro testing of VEGFA mRNA silencing. (A) Schematic of CRISPR-Cas13 mechanism of action. (B) VEGFA sgRNA sequence and targeting region on VEGFA mRNA sequence. (C) Schematic of collateral and (D) off-target cleavage. (E) Schematic for domain structure of VEGFA targeting shRNA and Cas13 (CasRx and Cas13bt3) vectors. (F) VEGFA mRNA knockdown and (G) protein expression from shRNA, CasRx, and Cas13bt3 in HEK293FT cells (n = 4). (H) Volcano plots showing differentially expressed genes (DEGs) from Cas13bt3 only, Cas13bt3-NT sgRNA, and Cas13bt3-VEGFA sgRNA compared to transfection control in HEK293FT cells. VEGFA gene is denoted by a black arrow. (I) Venn diagram of a total number of DEGs from Cas13bt3 only, Cas13bt3-NT sgRNA or Cas13bt3-VEGFA sgRNA transfected cells. (J) Transcripts per kilobase million (TPM) plot for VEGFA and candidate off-target genes. (K) Top 5 enriched GO molecular function (MF) terms from Cas13bt3 only, Cas13bt3-NT sgRNA and Cas13bt3 VEGFA sgRNA. NES above zero represents significant activation of the pathway, while NES below zero represents significant suppression of the pathway. Orange text denotes key factors that are upregulated; blue text denotes key factors that are downregulated. Data presented as mean ± SEM. Statistical analysis was conducted by one-way ANOVA and Tukey’s multiple comparison test (F and G) and Welch’s t test (J). ns: not significant, *P ≤ 0.05, ***P ≤ 0.001, ****P ≤ 0.0001.

Fig. 2.

Cas13bt3 treatment and single-cell RNA sequencing of 3D retinal organoids. (A) Schematic of 3D retinal organoid scRNA-seq experimental procedure. (B) Uniform Manifold Approximation and Projection (UMAP) plots show retinal cell type clusters for the untreated control, Cas13bt3-NT sgRNA, and Cas13bt3-VEGFA sgRNA treatment groups. (C) Infection rate of retinal organoids by cell type and treatment group. (D and E) Multi-volcano plots showing DEGs between Cas13bt3 positive and negative cells from the Cas13bt3-NT sgRNA and Cas13bt3-VEGFA sgRNA treatment groups. (F) Number of unique DEGs across retinal cell types. (G) Schematic of variable Cas13bt3 expression in retinal organoid cells and the relationship between Cas13bt3 expression and targeting efficiency. (H) Histograms showing the level of Cas13bt3 expression by number of reads across retinal cell types in the VEGFA sgRNA group. (I) Violin plots of VEGFA expression against the level of Cas13bt3 expression for macroglial cells and (J) RPE cells. The Kruskal–Wallis test was performed to determine significant changes between samples with P-value below 0.05 considered to be significant. P-values are denoted in read where significant. The red box indicates cells of interest for high VEGFA targeting activity.

Fig. 3.

In vivo VEGFA mRNA knockdown in humanized Kimba mice. (A and B) Schematic AAV2.7m8 viral mixture and experimental procedure for intravitreal treatment of Kimba mice. (C and D) Immunohistochemistry of retinal flatmount and cryosection showing transduction of AAV2.7m8 and expression of Cas13bt3 (Magnification: 20X and 40X, respectively). (E) Expression of Cas13bt3 from Kimba mice retinas. (F and G) Expression of human (hVEGFA) and mouse (mVEGFA) VEGFA mRNA in Kimba retinas. (H) VEGFA mRNA targeting regions in human and mouse transcriptome. (I) Central fluorescein fundus angiography (FFA) images were obtained from Kimba mice at baseline (6 wk) and endpoint (14 wk). (J) Quantification of leaky vessel density and (K) number of branch points from mouse retina at baseline (6 wk) and endpoint (14 wk) using AngioTool. Data presented as mean ± SEM. Statistical analysis was conducted by one-way ANOVA with Tukey’s multiple comparison test (E–G) and Multiple paired t-tests (J–K). *P ≤ 0.05, **P ≤ 0.01, ****P ≤ 0.0001.