Genome Editing of Expanded CTG Repeats within the Human DMPK Gene Reduces Nuclear RNA Foci in the Muscle of DM1 Mice

Abstract

Myotonic dystrophy type 1 (DM1) is caused by a CTG repeat expansion located in the 3' UTR of the DMPK gene. Expanded DMPK transcripts aggregate into nuclear foci and alter the function of RNA-binding proteins, leading to defects in the alternative splicing of numerous pre-mRNAs. To date, there is no curative treatment for DM1. Here we investigated a gene-editing strategy using the CRISPR-Cas9 system from Staphylococcus aureus (Sa) to delete the CTG repeats in the human DMPK locus. Co-expression of SaCas9 and selected pairs of single-guide RNAs (sgRNAs) in cultured DM1 patient-derived muscle line cells carrying 2,600 CTG repeats resulted in targeted DNA deletion, ribonucleoprotein foci disappearance, and correction of splicing abnormalities in various transcripts. Furthermore, a single intramuscular injection of recombinant AAV vectors expressing CRISPR-SaCas9 components in the tibialis anterior muscle of DMSXL (myotonic dystrophy mouse line carrying the human DMPK gene with >1,000 CTG repeats) mice decreased the number of pathological RNA foci in myonuclei. These results establish the proof of concept that genome editing of a large trinucleotide expansion is feasible in muscle and may represent a useful strategy to be further developed for the treatment of myotonic dystrophy.

Keywords: CRISPR-Cas9; DMPK; gene therapy; myotonic dystrophy; nucleotide repeat disorders.

Figure 1

CRISPR-SaCas9 Genome Editing to Target DMPK CTG Repeats (A) Scheme of selected Sa sgRNA target sites located in the 3′ UTR of the DMPK gene between the stop codon (stop) and the polyadenylation signal (pA), flanking the CTG repeats [(CTG)n]. The sgRNAs in black resulted in higher percentages of indels, as assessed by TIDE (Table S1). (B) Genomic PCR analysis showing the deletion of the region flanking the CTG repeats in HeLa cells. Cells were transfected with plasmids expressing SaCas9 and the indicated sgRNA couples (combinations of sgRNAs downstream of the CTG repeat region 12A, 12B, 13A, and 23 and sgRNAs upstream 1, 4, 7, and 8). NT, non-transfected cells; −, SaCas9-expressing plasmid without sgRNA. White and black arrows indicate undeleted and deleted PCR amplicons, respectively.

Figure 2

CRISPR-SaCas9 Lentiviral Vectors Delete CTG Repeats in DM1 Patient-Derived Muscle Line Cells (A) Scheme of lentiviral vector constructs containing SaCas9 and sgRNA sequences. LTR, long terminal repeat; CMV, cytomegalovirus promoter; NLS, nuclear localization signal; HA, human influenza hemagglutinin epitope; pA, polyadenylation signal; U6, human U6 small nuclear RNA (snRNA) gene promoter; sgRNAup, sgRNA sequence targeting regions upstream of the CTG repeats; sgRNAdw, sgRNA sequence targeting regions downstream of the CTG repeats; hPGK, human phosphoglycerate kinase gene promoter. (B) PCR amplicons of the genomic region containing the CTG repeats in DM1 myoblasts transduced with increasing MOIs of lentiviral vectors expressing SaCas9 and the indicated sgRNA couples. Triangle in gradient colors reflects the MOI, from 5 in white to 100 in black. A total MOI of 5, 10, 20, 50, and 100 was used for the two vectors at 1:1 ratio. White and black arrows indicate undeleted and deleted PCR amplicons, respectively. Genomic DNA from non-transduced cells (NT) and cells transduced with only one lentiviral vector (SaCas9 or sgRNA, MOI 50) were used as controls. (C) Percentage of DMPK CTG repeat deletion (% DEL) quantified from agarose gel images shown in (B). (D) Percentage of DM1 myoblasts without nuclear foci visualized by FISH images after treatment with the indicated MOIs of lentiviral vectors SaCas9 and sgRNA couple 4-23. Histograms show average values from three independent biological replicates ± SD. Statistical analysis by two-tailed Student’s t test. *p < 0.05; ns, not significant. Error bars represent SD.

Figure 3

Deletion of Expanded CTG Repeats and Focus Disappearance in DM1 Myoblasts Treated with CRISPR-SaCas9 DM1 myoblast clones were isolated from the bulk population after transduction with lentiviral vectors; isolated clones were analyzed for the presence of nuclear foci (A) and the presence of DMPK CTG repeats (B–F). (A) FISH-IF images of a representative DM1 myoblast clone without foci (DM1-Delta clone 22). DM1 clones non-transduced (DM1) or transduced with an MOI 50 of a lentiviral vector expressing SaCas9 (DM1-Cas9) or sgRNA_4-23 only (DM1-sgRNA) were used as controls. SaCas9 (α-HA) is shown in red, GFP is in green, RNA foci are in yellow [(CAG)7], and nuclei are in blue (DAPI). Scale bar, 10 μm. (B) PCR analysis of DMPK 3′ UTR in DM1-Delta clones 10, 3, 17, and 22 amplified with primers F1-DMPK-3UTR and R2-DMPK-3UTR, annealing regions surrounding the cutting sites of sgRNA 4 and 23. Ctrl, control myoblasts; DM1, non-transduced DM1 clone; DM1-sgRNA, DM1 clone transduced with only lentiviral vector expressing sgRNA_4-23. PCR amplicons with undeleted (a; white arrowhead) and deleted (b; black arrowhead) CTG repeats are shown. (C) Sequencing chromatograms of deleted PCR products of (B) from DM1-Delta clones 10, 3, 17, and 22, showing resection of the CTG repeats and resulting DNA end joining. (D) Sequence alignment of deleted and undeleted PCR products of (B) of ∼1 kb (a) and ∼0.4 kb (b), respectively. Indels and/or deletions of the CTG repeats are indicated by dashes; nucleotide substitutions are in red. Target sequences of sgRNA 4 and 23 are indicated in blue and pink, respectively; PAM sequences are underlined; and the CTG repeats region [(CTG)n] is highlighted in black. (E) Schematic representation of the DMPK gene and exon 15, indicating the relative positions of EcoRI cutting sites, the 1 kb Alu polymorphism (Alu), and the annealing region of probe B1.4. (F) Southern blot showing the genomic deletion of 2,600 CTG repeats in DM1-Delta clones. Genomic DNA was digested with EcoRI and hybridized with probe B1.4 shown in (E). Bands corresponding to the two alleles can be distinguished by size because the Alu insertion is 1 kb. (G) Number of CTG repeats [(CTG)n] in each allele of the DMPK gene, with (1) and without (2) the Alu insertion, and the expected size of EcoRI bands, in control (Ctrl), DM1, DM1-Delta, and DM1-sgRNA clones. ctg, unexpanded CTG repeats; CTG_exp, expanded CTG repeats; Δ, deletion of the CTG repeats; white and blue arrows, allele with and without the Alu insertion.

Figure 4

Reversion of Splicing Abnormalities in DM1 Patient-Derived Muscle Cells by CRISPR-SaCas9 Deletion of Expanded CTG Repeats Splicing profiles and quantification of LDB3 exon 11-, ATP2A1 exon 22-, MBNL1 exon 7-, DMD exon 78-, INSR exon 11-, and BIN1 exon 11-containing transcripts in differentiated myoblasts from DM1-Delta clones 10, 3, 17, and 22 compared to control (Ctrl), DM1, and DM1-sgRNA clones. Graphs show average values from independent biological replicates ± SD (n = 6 for Ctrl and DM1, n = 3 for the other samples). Statistical analysis by two-tailed Student’s t test. *p < 0.05, **p < 0.01, ***p < 0.001; ns, not significant; ˂DL, below detection limit. Error bars represent SD.

Figure 5

rAAV-Mediated CRISPR-SaCas9 Delivery in the Muscle of DMSXL Mice Results in the Deletion of Expanded DMPK CTG Repeats (A) Schematic representation of rAAV9 constructs: the expression of SaCas9 and sgRNAs 4-23 is under the control of the SPc5-12 and U6 promoters, respectively. The sgRNA construct contains an EGFP-Kash reporter under the Desmin promoter (Desm). ITR, inverted terminal repeat; SPc5-12, synthetic muscle-specific promoter; Int, intron; NLS, nuclear localization signal; HA, human influenza hemagglutinin epitope; pA, polyadenylation signal; EGFP-K, EGFP fused to Kash peptide; U6, human U6 small nuclear RNA (snRNA) gene promoter. (B) Representative immunofluorescence images of tibialis anterior muscle cross-sections from homozygous (HMZ) DMSXL mice 4 weeks after intramuscular injection of CRISPR-SaCas9 rAAV9 vectors, showing expression of SaCas9 (α-HA, in red) and sgRNAs (EGFP-K, in green) within muscle fibers. DAPI was used for nuclear staining. 21% and 76% of myonuclei were positive for SaCas9 and GFP, respectively, and 18% were positive for both. A higher magnification of myofibers from the upper panels (white boxes I., II., and III. in merged image) is shown in the panels below. Scale bars, 50 μm (upper panels) and 25 μm (I., II., and III.). (C) Genomic PCR of DMPK 3′ UTR from nine HMZ DMSXL mice (HMZ 1–9) 4 weeks after the injection of PBS (−) in the left TA muscle and rAAV9 vectors expressing SaCas9 and sgRNAs 4-23 (+) in the contralateral TA. The amplified band of ∼0.4 kb corresponds to the edited PCR amplicons with the deletion of 1,200 CTG repeats. (D) Sequence of deleted PCR products showing the end-joining site (black arrowhead) of sgRNA targets 4 (blue) and 23 (pink) after double-stranded breaks. (E) Alignment of unedited (DMSXL) and edited (Δ) DMPK 3′ UTR sequences from genomic DNA, showing the sharp cutting position at nucleotide N3 upstream of the PAM of sgRNAs 4 and 23. This representative sequence was obtained after Sanger sequencing of PCR products from TA muscles injected with rAAV-SaCas9 and sgRNA 4-23 (number of TA analyzed is equal to 9).

Figure 6

Indel Examination in DM1 Patient Cells and DMSXL Mice after Treatment with CRISPR-Cas9 Genomic deep sequencing of the DMPK 3′ UTR region with (DEL, primers F1-R1) or without CTG repeat deletion (sgRNA4, primers F1-R2; sgRNA23, primers F2-R1). Indel analysis was performed by alignment with the sequence resulting from a cut between nucleotides N3 and N4 at targets 4 and 23 for PCR amplicons with CTG repeat deletion or, alternatively, with the respective unmodified genomic sequence. (A) Percentage of reads with indels in bulk population of DM1 cells (DM1 bulk) and in TA muscle of DMSXL mice treated with CRISPR-Cas9 (+). Untreated DM1 cells and TA muscle injected with PBS were used as negative controls (−). (B) Indel distributions (deletions in black and insertions in red) upstream and downstream of the expected cutting sites (0) in PCR amplicons with (DEL) and without CTG repeat deletion (sgRNA4 and sgRNA23), generated from gDNA of treated DM1 cells and of representative TA muscles containing the lowest (73 DEL) and the highest (74 DEL) percentages of reads with indels. (C) Percentage of reads with indels (insertions + deletions) in the groups shown in (B).

Figure 7

CRISPR-SaCas9 Expression in DM1 Muscle Decreases Nuclear Foci (A) Representative confocal images of a TA muscle section stained with antibodies against laminin (α-LMN, red), FISH [(CAG)7, yellow], and DAPI (blue) (upper panels) from HMZ DMSXL mice. TA muscle section from WT animal shows the FISH background. A higher magnification of myofibers from the upper panels (white box in merged image) is shown in the panels below. Arrows indicate myonuclei with nuclear foci. Scale bars, 50 μm (upper panels) and 25 μm (lower panels). (B) Percentage of myonuclei containing foci in TA muscle fibers from DMSXL mice at 4 weeks after PBS (64.83% ± 9.25%) or rAAV9-SaCas9 + rAAV9-sgRNA_4-23 (49.25% ± 8.42%) intramuscular injection. Data are represented as means ± SD (N = 10 HMZ mice). Statistical analysis with two-tailed Student’s t test. ***p < 0.001. (C) Total number of myonuclei per fiber in the TA of wild-type (WT) and HMZ mice at 4 weeks after the injection of either PBS or rAAV9-SaCas9 + rAAV9-sgRNA_4-23 vectors. Data are represented as means ± SD (N = 3 for WT mice; N = 10 for HMZ mice). Statistical analysis with two-tailed Student’s t test. ns, not significant. Error bars represent SD.