Repurposing CRISPR-Cas13 systems for robust mRNA trans-splicing

Abstract

Type VI CRISPR enzymes have been developed as programmable RNA-guided Cas proteins for eukaryotic RNA editing. Notably, Cas13 has been utilized for site-targeted single base edits, demethylation, RNA cleavage or knockdown and alternative splicing. However, the ability to edit large stretches of mRNA transcripts remains a significant challenge. Here, we demonstrate that CRISPR-Cas13 systems can be repurposed to assist trans-splicing of exogenous RNA fragments into an endogenous pre-mRNA transcript, a method termed CRISPR Assisted mRNA Fragment Trans-splicing (CRAFT). Using split reporter-based assays, we evaluate orthogonal Cas13 systems, optimize guide RNA length and screen for optimal trans-splicing site(s) across a range of intronic targets. We achieve markedly improved editing of large 5' and 3' segments in different endogenous mRNAs across various mammalian cell types compared to other spliceosome-mediated trans-splicing methods. CRAFT can serve as a versatile platform for attachment of protein tags, studying the impact of multiple mutations/single nucleotide polymorphisms, modification of untranslated regions (UTRs) or replacing large segments of mRNA transcripts.

Fig. 1. CRAFT efficiently rescues EGFP expression via precise RNA trans-splicing.

a Schematic of 5′ CRAFT RNA editing. b bright field (top) and fluorescent images (bottom) of HEK293 cells transfected with Psp-dCas13b, 5′ rcRNA, Psp-dCas13b and non-targeting 5′rcRNA, or Psp-dCas13b and a on-target 5′rcRNA. c Quantitation of flow cytometry for 5′CRAFT targeting LMNA intron showing percent GFP positive cells (left) and mean fluorescence intensity (MFI) (right) (Data are mean ± s.d. from n = 3 individual samples). d Schematic of 3′ CRAFT RNA editing. e bright field (top) and fluorescent images (bottom) of cells transfected with Rfx-dCas1d, 3′ rcRNA, Rfx-dCas13d and non-targeting 3′rcRNA, or Rfx-dCas13d and on-target 3′rcRNA. f Quantitation of flow cytometry for 3′CRAFT targeting LMNA intron showing percent GFP positive cells (left) and MFI (right) (Data are mean ± s.d. from n = 3 individual samples). g Schematic diagraming modified 5′rcRNAs (left) that lack the direct repeat (-DR), separate the spacer from the direct repeat (+stuffer), and original rcRNA. Quantitation by flow cytometry for of EGFP rescue using 5′rcRNA (Data are mean ± s.d. from n = 3 individual samples). h Quantification of EGFP rescue using orthogonal protein/RNA partners in 5′CRAFT; rcRNAs that contain the direct repeat from either PguCas13b (blue bar) or PspCas13b (green bar) along with either dPguCas13b enzyme (blue triangle) or dPspCas13b enzyme (green circle) (Data are mean ± s.d. from n = 3 individual samples). i Schematic diagraming modified 3′rcRNAs. Quantitation by flow cytometry for of EGFP rescue using 3′rcRNA (Data are mean ± s.d. from n = 3 individual samples). j Quantification of EGFP rescue using orthogonal protein/RNA partners in 5′CRAFT; rcRNAs that contain the direct repeat from either PguCas13b (blue bar) or PspCas13b (green bar) along with either LwaCas13d (blue bar) or RfxCas13d (green bar) along with either dLwaCas13d enzyme (blue triangle) or dRfxCas13d enzyme (green circle) (Data are mean ± s.d. from n = 3 individual samples). Statistical significance was determined by unpaired two-tailed Student t-test. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001; ns not significant. Source data are provided as a Source Data file.

Fig. 2. Screen of parameters that affect CRAFT-mediated RNA editing.

a Schematic guide tiling for 5′CRAFT rcRNAs along the target intron (top). Position and color of guide correlates with the quantification by flow cytometry presented as percent GFP positive cells and MFI (below) (Data are mean ± s.d. from n = 3 individual samples). b Lead guide candidate from the (a) (guide 4) was modified by extending the original 30 bp guide to 150 bp in 20bp step increments and evaluated in the splitGFP reporter assay. Quantitation by flow cytometry for the 5′CRAFT guide length: percent GFP positive cells (left) and MFI (right) (Data are mean ± s.d. from n = 3 individual samples). c The direct repeat in the rcRNA of the lead guide candidate from (a) was swapped with the direct repeat from alternative type VI CRISPR species and co-transfected with the cognate catalytically dead cas13 protein. Quantitation by flow cytometry is presented as percent GFP positive cells (left) and MFI (right) (Data are mean ± s.d. from n = 3 individual samples). d Schematic guide tiling for 3′CRAFT rcRNAs along the target intron (top). Position and color of guide correlates with the quantification by flow cytometry presented as percent GFP positive cells and MFI (below) (Data are mean ± s.d. from n = 3 individual samples). e Lead guide candidate from (d) (guide 2) was modified by extending the original 30 bp guide to 150 bp in 20 bp step increments and evaluated in the splitGFP reporter assay. Quantitation by flow cytometry for the 3′CRAFT guide length: percent GFP positive cells (left) and MFI (right) (Data are mean ± s.d. from n = 3 individual samples). f The direct repeat in the 3′rcRNA of the lead guide candidate of LMNA (guide 2) was swapped with the direct repeat from alternative type VI CRISPR species and co-transfected with the cognate catalytically dead cas13 protein. Quantitation by flow cytometry is presented as percent GFP positive cells (left) and MFI (right) (Data are mean ± s.d. from n = 3 individual samples). Statistical significance was determined by One-Way ANOVA with Tukey’s post-test. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001; ns not significant. Source data are provided as a Source Data file.

Fig. 3. CRAFT enables efficient rewriting across a diverse set of target introns.

a Schematic guide tiling for 5′CRAFT rcRNAs along the target intron (top). Target position and color of guide target sequence correlates with the quantification by flow cytometry presented as percent GFP positive cells and MFI (below) across three different introns: human RYR2 intron 95/96, human DMPK intron 13/14, mouse dmd intron 23/24 (Data are mean ± s.d. from n = 3 individual samples). b Schematic guide tiling for 3′CRAFT rcRNAs along the target intron (top). Target position and color of guide target sequence correlates with the quantification by flow cytometry presented as percent GFP positive cells and MFI (below) across three different introns: human RYR2 intron 95/96, human DMPK intron 13/14, human FXN intron 1/2 (Data are mean ± s.d. from n = 3 individual samples).

Fig. 4. CRAFT is amenable to viral delivery and rewriting of endogenous mRNA.

a Schematic of stable HEK293 splitGFP reporter cell line generation accompanied by Sanger sequencing traces. b EGFP rescue in HEK293 cells that stably express the splitGFP reporter following transfection of 5′CRAFT plotted as percent GFP positive cells. c EGFP rescue in HEK293 cells that stably express the splitGFP reporter following transduction with 5′CRAFT (AAV genome schematics shown left) and plotted as percent GFP positive (right). d EGFP rescue in HEK293 cells that stably express the splitGFP reporter following transfection of 3′CRAFT plotted as percent GFP positive cells. e EGFP rescue in HEK293 cells that stably express the splitGFP reporter following transduction with 3′CRAFT (AAV genome schematics shown left) and plotted as percent GFP positive cells (right). f Schematic of cis- (left) trans- (right) endogenous LMNA transcripts highlighting a silent (C > G) snp in the trans-spliced. g Schematic of cis- (left) trans- (right) endogenous LMNA transcripts highlighting a silent (T > C) snp in the trans-spliced. h Sanger sequencing of the cis- (top) and trans- (bottom) spliced LMNA transcripts. i frequency of 5′CRAFT editing in endogenous LMNA transcripts. j Sanger sequencing of the cis- (top) and trans- (bottom) spliced LMNA transcripts. k frequency of 3′CRAFT editing in endogenous LMNA transcripts. l confocal microscopy images of N-terminal Flag tag introduced by 5′CRAFT Dapi (nuclear staining), AlexaFluor594 (Lamin A/C), and Flag (AlexaFluor488) (representative images from n = 2 independent experiments). m confocal microscopy images of C-terminal Flag tag introduced by 3′CRAFT (staining panel same as l). n volcano plot generated by DEseq2 analysis of differential gene expression from bulk RNA-seq in HEK293 cells transfected with dPspCas13b and either a targeting or non-targeting guide in the 5′ rcRNA. o volcano plot generated by DEseq2 analysis of differential gene expression from bulk RNA-seq in HEK293 cells transfected with dRfxCas13d and either a targeting or non-targeting guide in the 3′ rcRNA. All data are mean from n = 3 individual samples. Statistical significance was determined by unpaired two-tailed Student t-test. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001; ns not significant. Source data are provided as a Source Data file.

Fig. 5. Comparison of SMaRT vs. CRAFT across multiple targets.

a Schematic 5′ SMaRT and CRAFT designs (top). Schematic 3′ SMaRT and CRAFT designs (bottom). b Trans-splicing efficiency was quantified as percent GFP positive cells measured by flow cytometry (below) at 5 different introns: LMNA intron 10/11 (5′ and 3′ p < .0001), RYR2 intron 95/96 (5′ and 3′ p < .0001), DMPK intron 12/13 (3′ p < .0001), mouse dmd intron 23/24 (5′ p < .0001), and RHO intron 1/2 (3′ p < .0001) (Data are mean ± s.d. from n = 3 individual samples). c Schematic of cis-spliced endogenous DMD transcript (top) and trans-spliced transcripts (bottom). The position of the single nucleotide polymorphism (A > G) is highlighted. d Frequency of 3′CRAFT editing in endogenous DMD transcripts. The y-axis is the percent of reads from targeted amplicon sequencing containing the snp mutation. The x-axis refers to specific treatment: dCas13 alone, rcRNA alone, dCas13 with an rcRNA that does not target the DMD intron, and dCas13 with an rcRNA that targets the DMD intron (Data are mean ± s.d. from n = 3 individual samples). e Comparison of CRAFT and SmaRT at the DMD locus. The y-axis is the percent of reads from targeted amplicon sequencing containing the A > G mutation. Guides targeting the same intronic location for SMaRT (gray) and CRAFT (blue) are plotted on the x-axis (Data are mean ± s.d. from n = 3 individual samples) Guide A (p = .0005), Guide B (p = .064), Guide C (p = .0007). Statistical significance was determined by unpaired students two-tailed t-test. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001; ns not significant. Source data are provided as a Source Data file.

Fig. 6. High-throughput guide selection through guide coupled barcode.

a Schematic of barcode approach. Briefly, a library of unique barcodes corresponding a specific spacer sequence was delivered to HEK293 cells with Rfx-dCas13d. Functional RNPs target to intron 10/11 of LMNA and undergo trans-splicing. RNA is harvested and the abundance of each barcode at the start of exon 11 was measured by targeted amplicon sequencing. This barcode corresponds to the guide associated with its trans-splicing. b enrichment plot for each guide targeting LMNA intron 10/11 as a function of position. Y-axis is barcode enrichment calculated by the (%barcode in trans-spliced RNA / %barcode in input plasmid DNA), and x-axis is the position along LMNA intron 10/11 of each guide. Each unique barcode (3) for a single guide is plotted as a triangle and the average of these enrichment scores is plotted as a black circle. The best guide is plotted in purple. c enrichment plot for each guide targeting DMPK intron 13/14 as a function of position. d Schematic of endogenous DMPK transcript (left) and edited DMPK transcript (right). Notably, there is a single G > T transition mutation installed between the edited transcript and the endogenous transcript. e Frequency of 3′CRAFT editing in endogenous DMPK transcripts. The y-axis is the percent of reads from targeted amplicon sequencing containing the snp mutation. The x-axis refers to specific treatment: dCas13 alone, rcRNA alone, dCas13 with an rcRNA that does not target the DMPK intron, and dCas13 with an rcRNA that targets the given intron (Data are mean ± s.d. from n = 3 individual samples). Statistical significance was determined by unpaired students two-tailed t-test. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001; ns not significant. Source data are provided as a Source Data file.