Dual-targeting CRISPR-CasRx reduces C9orf72 ALS/FTD sense and antisense repeat RNAs in vitro and in vivo

Abstract

The most common genetic cause of frontotemporal dementia (FTD) and amyotrophic lateral sclerosis (ALS) is an intronic G₄C₂ repeat expansion in C9orf72. The repeats undergo bidirectional transcription to produce sense and antisense repeat RNA species, which are translated into dipeptide repeat proteins (DPRs). As toxicity has been associated with both sense and antisense repeat-derived RNA and DPRs, targeting both strands may provide the most effective therapeutic strategy. CRISPR-Cas13 systems mature their own guide arrays, allowing targeting of multiple RNA species from a single construct. We show CRISPR-Cas13d variant CasRx effectively reduces overexpressed C9orf72 sense and antisense repeat transcripts and DPRs in HEK cells. In C9orf72 patient-derived iPSC-neuron lines, CRISPR-CasRx reduces endogenous sense and antisense repeat RNAs and DPRs and protects against glutamate-induced excitotoxicity. AAV delivery of CRISPR-CasRx to two distinct C9orf72 repeat mouse models significantly reduced both sense and antisense repeat-containing transcripts. This highlights the potential of RNA-targeting CRISPR systems as therapeutics for C9orf72 ALS/FTD.

Fig. 1. CRISPR-CasRx can effectively lower sense and antisense repeat-containing transcripts and DPRs and prevent polyGR and polyPR formation in HEK293T cells.

a Strategy for targeting sense and antisense C9orf72 transcripts with gRNAs. Sense gRNAs target repeat-containing variants 1 and 3 but not variant 2. b Sense and antisense NLuc reporter assay designs. c Sense NLuc assays testing single plasmids expressing both CasRx and sense guides. d Antisense NLuc assays testing single plasmids expressing both CasRx and antisense guides. Data in (c, d) are shown as mean ± S.D. n = 2 biological repeats (all technical repeats shown on graph), one-way ANOVA with Dunnett’s test for post-hoc analysis. e, f Single vectors expressing a single guide (guide 10 for sense targeting or guide 17 for antisense targeting) or both guides 10 and 17 were used in our sense and antisense NLuc reporter assays. CasRx with sense and antisense targeting array can effectively reduce both e sense and f antisense C9orf72 DPR levels to a similar degree to single guide expressing plasmids indicating effective guide array maturation and multi-target engagement. All NLuc data normalised to FLuc and non-targeting (NT+NT) guide. Data in (e, f) shown as mean ± S.D. n = 3 biological repeats (all technical repeats shown on graph), one-way ANOVA with Dunnett’s test for post-hoc analysis. Source data are provided as a Source Data file.

Fig. 2. CRISPR-CasRx reduces endogenous sense and antisense repeat-containing C9orf72 RNA and DPRs in patient-derived iPSC-neurons without lowering variant 2 or causing detectable off-target effects.

a Experimental setup to determine target engagement in patient-derived iPSC-neurons. Created in BioRender. Cammack, A. (2024) https://BioRender.com/z76w550. b Schematic of RT-qPCR primers used to quantify C9orf72 transcripts. c–e RT-qPCRs for C9orf72 transcripts in neurons treated with CRISPR-CasRx + gRNA lentiviruses. Data presented as fold change compared to non-targeting (NT) gRNA. f, g Levels of polyGP and polyGA in iPSC-neurons treated with sense targeting gRNAs 8 and 10, shown as fold-change compared to average of non-targeting (NT) gRNA treatments across C9 lines. Data in (b–g) shown as mean ± S.D, n = 3 biological replicates (i.e. individual C9 lines), each shown as different colour data points (orange, green, and yellow). Experiments were performed on three separate inductions of C9 line 1 and in one induction each of C9 lines 2 and 3. p-values calculated with two-way ANOVA and Dunnett’s test for post-hoc analysis. h Volcano plot of DESeq2 analysis showing DEGs between neurons treated with sense-targeting guide 10 compared to non-targeting (NT) gRNA with C9orf72 transcripts grouped by those that contain intron1 and the repeats, and those that do not. Dotted lines indicate thresholds for fold change on x-axis (|log₂FoldChange|>0.5) and p value on y-axis (adjusted p < 0.05, two-sided Wald test, p-values are adjusted for multiple testing using Benjamini-Hochberg method). n = 3 independent inductions of C9 line 1. Source data are provided as a Source Data file.

Fig. 3. CRISPR-CasRx treatment rescues stressor-induced neurotoxicity in C9orf72 patient iPSC-derived spinal neurons (iPSC-SNs).

a Glutamate-induced toxicity was assessed 20 days after nucleofection with CRISPR-CasRx plasmids. b, c Cell death in control and C9orf72 iPSC-SNs expressing CRISPR-CasRx plasmids and single gRNAs, quantified as the ratio of propidium iodide (PI) positive spots to DAPI positive nuclei and Alamar Blue cell viability assay following 4-h exposure to 10 µM glutamate. n = 6 lines per condition. d Schematic of glutamate-induced excitotoxicity assay in control and C9 lines expressing CRISPR-CasRx dual guide plasmids. e, f Quantification of PI incorporation and Alamar Blue cell viability following 4-h exposure to 10 µM glutamate, demonstrating rescue by the dual guide CRISPR-CasRx. Data points for PI incorporation represent average percent cell death across 10 images per well. Data points for Alamar Blue assay represent average percent viability from 3 replicate wells for each condition. Two-way ANOVA with Tukey’s multiple comparison test was used to calculate statistical significance. Data presented as mean ± S.D. Source data are provided as a Source Data file.

Fig. 4. CRISPR-CasRx reduces G₄C₂ sense RNA in vivo in 149 R mice.

a Schematic of experimental paradigm. CasRx PHP.eB AAV and the 149 R C9orf72 AAV9 were co-injected via intracerebroventricular (ICV) injection into P0 C57BL6/J mice. CAG-CMV enhancer/chicken β-actin promoter. CBA chicken β-actin promoter, NLS nuclear localisation signal. Created in BioRender. Cammack, A. (2024) https://BioRender.com/u85i634. b RT-qPCR of the sense repeat-containing transcripts from the 149 R AAV performed on RNA isolated from the hippocampus 3 weeks post-injection. Each data point represents one animal. RT-qPCR data presented as fold change compared to non-targeting (NT+NT) CasRx PHP.eB AAV-treated mice. n = 8 NT+NT, n = 14 10+17. c MSD immunoassay of polyGP in bulk hippocampal tissue 3 weeks post-injection. n = 7 NT+NT, n = 14 10+17. Data in (b, c) are shown as mean ± S.D, two-sided unpaired t-test. Source data are provided as a Source Data file.

Fig. 5. CRISPR-CasRx simultaneously reduces sense and antisense repeat RNAs in C9orf72 BAC transgenic mice.

a Diagram of experimental paradigm. CasRx PHP.eB AAV (8E+9 vg per animal) was injected via intracerebroventricular (ICV) injection into postnatal day 0 (P0) C9orf72 BAC mice. Tissue was harvested 8.5 months post-injection. CAG-CMV enhancer/chicken β-actin promoter, NLS nuclear localisation signal. Created in BioRender. Cammack, A. (2024) https://BioRender.com/u85i634. b C9orf72 promoter methylation assay of prefrontal cortical DNA extracted from C9orf72 BAC mice, demonstrating no overt transgene silencing within the cohort. c–e RT-qPCRs for C9orf72 transcripts in C9orf72 BAC mice treated with CRISPR-CasRx + gRNA AAVs in bulk prefrontal cortex 8.5 months post-injection. RT-qPCR for c sense repeat-containing C9orf72 transcripts (variants 1 and 3). d sense variant 2 C9orf72 transcripts, or e antisense repeat-containing C9orf72 transcripts. RT-qPCR data presented as fold change compared to non-targeting (NT+NT) gRNA CasRx PHP.eB AAV-treated mice. For (c–e), n = 13 Guide 10+17; n = 11 Guide NT+NT. e, f Levels of f polyGA and g polyGP DPRs in bulk cortical tissue from 8.5 month old animals treated with CasRx PHP.eB AAV. For (e, f), n = 14 Guide 10+17; n = 12 Guide NT+NT. Data in (c–g) are shown as mean ± SD, two-sided unpaired t-test. Source data are provided as a Source Data file.