Gene-editing in patient and humanized-mice primary muscle stem cells rescues dysferlin expression in dysferlin-deficient muscular dystrophy

Abstract

Dystrophy-associated fer-1-like protein (dysferlin) conducts plasma membrane repair. Mutations in the DYSF gene cause a panoply of genetic muscular dystrophies. We targeted a frequent loss-of-function, DYSF exon 44, founder frameshift mutation with mRNA-mediated delivery of SpCas9 in combination with a mutation-specific sgRNA to primary muscle stem cells from two homozygous patients. We observed a consistent >60% exon 44 re-framing, rescuing a full-length and functional dysferlin protein. A new mouse model harboring a humanized Dysf exon 44 with the founder mutation, hEx44mut, recapitulates the patients' phenotype and an identical re-framing outcome in primary muscle stem cells. Finally, gene-edited murine primary muscle stem-cells are able to regenerate muscle and rescue dysferlin when transplanted back into hEx44mut hosts. These findings are the first to show that a CRISPR-mediated therapy can ameliorate dysferlin deficiency. We suggest that gene-edited primary muscle stem cells could exhibit utility, not only in treating dysferlin deficiency syndromes, but also perhaps other forms of muscular dystrophy.

Fig. 1. Isolation of pure MuSC populations from two patients carrying a homozygous DYSF c.4872_4876delinsCCCC founder mutation.

a MuSC derived from patient 1 (upper panel) and 2 (lower panel) stained for the myogenic markers DES, PAX7, MYOD, MYF5, and the proliferation marker KI-67. Nuclei are stained with Hoechst. The percentage of cells expressing each marker is shown in the corresponding colors. Scale bars: 50 µm. Several MuSC populations (Patient 1: n = 2; Patient 2: n = 8) were prepared from a single muscle biopsy obtained from each patient. b PCR genotyping of DYSF exon 44 in genomic DNA from a control and the two patients carrying the homozygous c.4872_4876delinsCCCC mutation.

Fig. 2. Screening for mutation-specific sgRNAs to rescue the DYSF exon 44 reading frame in patient hiPSC.

a Strategy to rescue the DYSF reading frame by Cas9-induced NHEJ. b hiPSC from patient 1 and a control (not carrying the DYSF exon 44 mutation) were transfected with a plasmid encoding eSpCas9(1:1) in combination with three different mutation-specific sgRNAs. Venus positive cells were selected by FACS-sorting and processed for DNA analysis via NGS. c Sanger-sequencing chromatograms of patient 1 hiPSC after FACS-sorting. The protospacer sequences are underlined in black. The PAM sequences are underlined in red. The dotted vertical lines indicate the expected DSB sites. d Allele frequencies of Venus+ patient 1 and control hiPSC transfected with eSpCas9(1.1) and sgRNA#1-3 (as on the left) were determined by NGS. The most frequent indels are indicated. e Single cell-derived clones were expanded from Venus+ patient 1 hiPSC transfected with eSpCas9(1.1) and sgRNA#3 (n = 22 clones). Upper pie chart: Distribution of unedited, heterozygously edited and homozygously edited clones. Lower pie chart: Indel distribution among all edited clones. f Sanger-sequencing chromatogram from an hiPSC clone carrying a homozygous +1A insertion at the DSB site (purple arrow). g Western blot analysis of dysferlin protein expression in two edited hiPSC clones from patient 1, homozygous for the +1A insertion, compared to unedited hiPSC from patient 1 and two controls, as well as MuSC and myotubes (MT) from two controls. α-tubulin was used as loading control (n = 1 independent experiment). Source data are provided as a Source Data file.

Fig. 3. SpCas9 mRNA reframes DYSF exon 44 with >60% efficiency in MuSC from both patients with high precision and safety.

a Schematic overview of experimental workflow. Patient and control MuSC were transfected with SpCas9 mRNA (3 µg per 150,000 cells), with or without sgRNA#3 (2 µg per 150,000 cells). b Sanger-sequencing chromatograms around the target site from patient MuSC transfected with or without sgRNA#3. The protospacer and PAM sequences are underlined in the upper unedited chromatogram. The dotted vertical line indicates the expected DSB site. c Allele frequencies in control and patient MuSC transfected with SpCas9 mRNA, with or without sgRNA#3, at day 4 and 9 post transfection determined by NGS (n = 3 replicates per donor and time point, mean ± SD). Data from the two control MuSC populations are plotted together. Ctrl: Control. d Frequency distribution of all indels in MuSC from the two patients at day 4 post-transfection with SpCas9 mRNA and sgRNA#3 (n = 3 replicates per patient, mean ± SD). e OTS identified by GUIDE-seq. The OTS sequences are aligned to the on-target site (top). The total number of GUIDE-seq reads aligned to each OTS from all samples, using both the GSP+ and GSP- primers, is plotted on the right. f All OTS identified by GUIDE-seq/CrispRGold (OTS 1-7), plus the top 10 and all exonic OTS predicted by CRISPOR (OTS 1, 3, 4 and OTS 8-21) for sgRNA#3 (Supplementary Table 3) were analyzed by NGS in patient MuSC collected 4–9 days after transfection with SpCas9 mRNA, with (n = 3) or without (n = 3 for OTS 2, 5, 6 and 7; n = 2 for all other OTS) sgRNA#3 (mean ± SD). Source data are provided as a Source Data file.

Fig. 4. SpCas9 mRNA-mediated reframing efficiently rescues dysferlin protein expression and correct localization in patient MuSC while maintaining their myogenic profile.

a Western blot analysis of dysferlin protein expression in edited patient MuSC. Unedited patient and control MuSC were used as controls. Vinculin was used as loading control. b The intensity of the dysferlin bands from a was quantified using ImageJ and normalized to Vinculin (n = 3 replicates, mean ± SD). c Dysferlin immunostaining of terminally differentiated control and patient MuSC transfected with SpCas9 mRNA, with or without sgRNA#3. Dysferlin membrane localization is indicated by white arrows. Scale bars: 25 µm. d Edited patient MuSC stained for DES, PAX7 and MYOD. Scale bar: 50 μm. e Fold-change of the percentage of patient cells expressing myogenic and proliferation markers at day 5 after transfection of SpCas9 mRNA, with or without sgRNA (normalized to untransfected cells) (n = 1 independent experiment per patient, mean ± SD). Source data are provided as a Source Data file.

Fig. 5. Reframed dysferlin maintains functional properties.

a Scheme showing the mRNA and amino-acid sequence of wild-type, mutant, and reframed DYSF exon 44/45. The c.4872_4876delinsCCCC founder mutation introduces a frameshift in exon 44 that results in a premature stop codon in exon 45. The SpCas9-induced +1A insertion rescues the reading frame and results in an exchange of four amino acids (p.1622, p.1623, p.1624, and p.1626) compared to the wild-type protein. b Upper: Superposition of the experimentally determined Ca²⁺-bound C2B domain of rat synaptotagmin-1 (pdb 1TJX) with the AlphaFold2-predicted C2F domain of human dysferlin (Expasy accession number: O75923). The Ca²⁺ binding site is highly conserved in both domains (magnified top view). Lower: The magnification shows the site of the four mutated amino acids in a surface-exposed loop connecting the β-sandwich with the Ca²⁺-binding site of the C2F domain. c Dysferlin and annexin A1 immunostaining of unedited and edited patient myotubes, compared to control myotubes, performed up to 20 min after laser irradiation. Arrows indicate the wounding area (repair site). Nuclei are counterstained with Hoechst. Scale bars: 20 μm (lower magnification) and 5 μm (zoomed-in images). The experiment was conducted 3 independent times with similar results, and 13–26 myotubes were successfully injured and analyzed per condition (Control: n = 13; Patient, unedited: n = 17; Patient, edited: n = 26).

Fig. 6. Generation and characterization of a novel humanized LGMD2B mouse model carrying the c.4872_4876delinsCCCC founder mutation.

a Schematic overview of the strategy to generate the transgenic mice. SpCas9 and a gRNA were used to induce a DSB at the 3′ end of the murine Dysf exon 44 in C57BL/6N mouse zygotes. A targeting vector containing either the human wild-type or mutant DYSF exon 44 flanked by 1.4 kb homology arms (HA) was provided to enable exchange of the murine exon 44 for the human exon 44 via homologous recombination. The resulting knock-in alleles were called hEx44wt (wild-type human exon 44) and hEx44mut (human exon 44 with the c.4872_4876delinsCCCC mutation). b Western blot analysis of dysferlin protein expression in quadriceps muscles of 12-week-old homozygous hEx44wt and hEx44mut male mice compared to age- and gender-matched wild-type C57BL/6N mice (n = 4 per genotype). c RT-qPCR analysis of dysferlin mRNA (relative to Gapdh) in quadriceps muscles from 12-week-old homozygous hEx44wt and hEx44mut male mice compared to age- and gender-matched wild-type C57BL/6 N mice (n = 4, mean ± SD). Kruskal-Wallis with post hoc Dunn’s test (p = 0.0065). d Dysferlin immunostaining of M. tib. ant. from homozygous hEx44wt and hEx44mut mice. Nuclei were stained with Hoechst. Scale bar: 50 μm. e Gomori’s Trichrome staining of M. tib. ant. from 40-week-old homozygous hEx44wt and hEx44mut mice. Scale bar: 100 μm. Source data are provided as a Source Data file.

Fig. 7. SpCas9 mRNA reframes DYSF exon 44 and rescues dysferlin protein expression in MuSC from hEx44mut mice.

a Schematic overview of experiment setup. MuSC isolated from hind limbs of homozygous hEx44mut mice were transfected with SpCas9 mRNA plus sgRNA#3 or GFP mRNA. b Allele frequencies in unedited (n = 3) and edited (n = 6) MuSC from homozygous hEx44mut mice determined by NGS (mean ± SD). c Sanger-sequencing chromatogram of MuSC from homozygous hEx44mut mice transfected with SpCas9 mRNA plus sgRNA#3 compared to unedited cells. The protospacer and PAM sequences are underlined. The dotted vertical line indicates the expected DSB site. d Relative Dysf mRNA expression in edited and unedited MuSC from homozygous hEx44mut mice normalized to MuSC from homozygous hEx44wt mice (hEx44wt: n = 3; hEx44mut, unedited: n = 1, hEx44mut, edited: n = 2, mean ± SD). e Western blot analysis of dysferlin protein expression in edited and unedited MuSC from homozygous hEx44mut mice. MuSC from homozygous hEx44wt mice were used as control. f Quantification of dysferlin signal relative to α-tubulin from e using ImageJ. g Dysferlin immunostaining of reframed hEx44mut MuSC after differentiation into myotubes. Nuclei were stained with Hoechst. Scale bar: 20 μm. Source data are provided as a Source Data file.

Fig. 8. Exon 44 re-framed murine MuSC regenerate muscle and rescue dysferlin in hEx44mut mice.

a MuSC were isolated from homozygous hEx44mut donors, transfected with SpCas9 mRNA and sgRNA#3, and transplanted into the TA muscles of late (75 weeks old) or early (11–14 weeks old) symptomatic homozygous hEx44mut recipients. Prior to grafting, late symptomatic recipients received no pre-treatment. Recipient muscles of early symptomatic mice were focally irradiated, and cell transplantation was performed in the presence or absence of cardiotoxin (CTX) injection. b Graph depicting the maximum number of dysferlin-expressing, donor-derived fibers per cross-section for each grafted TA muscle. Source data are provided as a Source Data file. c Dysferlin immunostaining was performed on transversal cryosections of TA muscles from uninjured, late symptomatic recipients. Dysferlin positive, donor-derived myofibers were found in 1 out of n = 5 grafted muscles. Nuclei are stained with Hoechst. Scale bar: 50 μm. d Up to 50 dysferlin positive myofibers per section were found in 4 out of n = 5 irradiated-only muscles of early symptomatic recipients. Scale bar: 50 μm. e Up to 150 dysferlin positive myofibers per section were found in 2 out of n = 5 irradiated plus CTX-injured muscles of early symptomatic recipients. Scale bar: 50 μm. f, g Pax7-positive cells/nuclei (white arrowheads) were found adjacent to donor-derived, dysferlin expressing myofibers in irradiated-only (f) and irradiated + CTX (g) recipient muscles. Scale bars: 20 μm. h Pax7/laminin immunostaining of the graft in (e). The zoomed-in areas contain Pax7-positive cells/nuclei in the MuSC niche (white arrowheads). Nuclei are stained with Hoechst. Scale bars: 50 μm.