Ex Vivo Gene Editing of the Dystrophin Gene in Muscle Stem Cells Mediated by Peptide Nucleic Acid Single Stranded Oligodeoxynucleotides Induces Stable Expression of Dystrophin in a Mouse Model for Duchenne Muscular Dystrophy
Department of Neurology, David Geffen School of Medicine, University of California Los Angeles, Los Angeles, California, USA
Duchenne muscular dystrophy (DMD) is a fatal disease caused by mutations in the dystrophin gene, which result in the complete absence of dystrophin protein throughout the body. Gene correction strategies hold promise to treating DMD. Our laboratory has previously demonstrated the ability of peptide nucleic acid single-stranded oligodeoxynucleotides (PNA-ssODNs) to permanently correct single-point mutations at the genomic level. In this study, we show that PNA-ssODNs can target and correct muscle satellite cells (SCs), a population of stem cells capable of self-renewing and differentiating into muscle fibers. When transplanted into skeletal muscles, SCs transfected with correcting PNA-ssODNs were able to engraft and to restore dystrophin expression. The number of dystrophin-positive fibers was shown to significantly increase over time. Expression was confirmed to be the result of the activation of a subpopulation of SCs that had undergone repair as demonstrated by immunofluorescence analyses of engrafted muscles using antibodies specific to full-length dystrophin transcripts and by genomic DNA analysis of dystrophin-positive fibers. Furthermore, the increase in dystrophin expression detected over time resulted in a significant improvement in muscle morphology. The ability of transplanted cells to return into quiescence and to activate upon demand was confirmed in all engrafted muscles following injury. These results demonstrate the feasibility of using gene editing strategies to target and correct SCs and further establish the therapeutic potential of this approach to permanently restore dystrophin expression into muscle of DMD patients. Stem Cells2014;32:1817–1830
Duchenne muscular dystrophy (DMD) is caused by mutations in the dystrophin gene, which result in a complete absence of dystrophin protein throughout the body [1-3]. Dystrophin deficiency greatly compromises the structural integrity of muscle tissue, and leads to rapid muscle necrosis, weakness, and ultimately, premature death between the second and third decades of life. The dystrophin protein contains four structural domains including the N-terminal, rod, cysteine-rich, and C-terminal domains, each playing key functional roles in the maintenance of membrane integrity.
An understanding of the various defects in the dystrophin gene that cause DMD has enabled the development and testing of different therapeutic approaches [4, 5]. The majority of those approaches are aimed at restoring shorter, although still functional dystrophin protein. Preclinical and clinical studies have confirmed the beneficial effects that some of those strategies can achieve in animal models and DMD patients. For instance, the use of antisense oligonucleotides to redirect mRNA splicing to restore protein expression has attained some success in restoring functional dystrophin protein in animal models for the disease as well as in clinical trials in DMD patients [6-9]. However, the level of expression is limited, as are the long-term beneficial effects. Strategies focused on restoring permanent levels of full-length dystrophin expression remain the best option to treat DMD [10-13]. Among those, readthrough of premature stop codons has shown some promise [14-16].
Gene correction mediated by single-stranded oligodeoxynucleotides (ssODNs) in muscle cells has shown to induce single base-pair alterations at the genomic level which are stably inherited through cell division [10-13]. Correction takes advantage of innate repair mechanisms present in the cells and capable of directing single base substitutions in the genomic DNA targeted for repair.
We have recently demonstrated that ssODNs composed of peptide nucleic acid bases (PNA-ssODNs) targeting the single base mutation in the mdx5cv mouse are capable of achieving a higher frequency of gene correction than their unmodified counterparts rendering this technology a realistic possibility for the treatment of DMD [11, 13]. However, our results have also evidenced the presence of factors, other than correction frequencies, that contribute to the long-term stability of dystrophin expression into muscles following correction [13, 17]. In fact, the progressive loss of dystrophin expression detected after treatment in muscles that received ssODNs compared to earlier time points clearly demonstrates that correction of mature myofibers alone is not sufficient to protect muscle from degeneration . These results prompted us to test the ability of PNA-ssODNs to target and correct satellite cells (SCs) and to determine the ability of cells that had undergone correction to stably restore dystrophin expression following transplantation into muscle of mdx/nude mice [18, 19].
SCs are a class of stem cells found between the basal lamina and sarcolemma of muscle fibers . They activate in response to injury or disease and are responsible for regenerating myofibers lost as a result of normal muscle turnover or in response to injury. Upon activation, SCs undergo an initial stage of cell division which results in the generation of two daughter cells. Of those, one will return to the quiescent stage while the other will continue to divide to give rise to progenies capable of fusing amongst each other to generate new myofibers or capable of fusing with preexisting fibers and repairing damaged ones.
Here, we demonstrate that SCs isolated from a mouse model of DMD efficiently take up PNA-ssODNs and are amenable to gene repair. When transplanted into skeletal muscle of dystrophin-deficient mice, cells were able to restore low, although detectable, levels of dystrophin protein. Importantly, we demonstrate that the number of dystrophin-positive fibers significantly increases over time suggesting that correction of a portion of the transplanted cells is sufficient to actively regenerate muscle and to induce beneficial effects. These results clearly establish the importance of targeting SCs for the treatment of muscle disorders and pave the way for future studies aimed at determining the long-term therapeutic potential of cell-mediated regenerative medicine for muscle disorders using gene editing strategies.
Materials and Methods
Mice of the mdx5cv strain (B6Ros.Cg-Dmdmdx−5cv/J) and control C57 strain (C57BL/6J) were used as donor mice to obtain mdx5cv SCs and SCs isolated from wild-type mice, respectively. The mdx/nude strain was generated by backcrossing the mdx mouse with the nude mouse (CBy.Cg-Foxn1nu/J). All mice were purchased from The Jackson Laboratory (Bar Harbor, ME, http://jaxmice.jax.org). All procedures were carried out in accordance with the guidelines of the Administrative Panel on Laboratory Animal Care of the University of California, Los Angeles.
The 18 bp control (PNA-CTLC) and correcting (PNA-CORC) oligonucleotides were designed to be complementary to the transcribed strand (Fig. 1A) and were synthesized by Panagene, Inc. (Panagene, North Korea, http://www.panagene.com) . All oligonucleotides were HPLC purified and exhibited a single peak of the expected molecular weight as determined by MALDI TOF mass spectroscopy analysis.
SC Isolation and Transfection
SCs were isolated from hind limb muscles of mdx5cv mice (6–8 weeks of age) as described previously [10, 21, 22]. Briefly, muscles were incubated in Dulbecco's modified Eagle's medium (DMEM) with 0.2% (wt/vol) collagenase II (Life Technologies, Carlsbad, CA, http://www.lifetechnologies.com). SCs were released from bulk fibers through an additional digestion in 20 ml of Hams F-10, 10% horse serum (HS), 0.5 U/ml dispase (Life Technologies), and 38 U/ml collagenase type II (U.S. Biological, http://www.usbio.net). Cells were sedimented by centrifugation and the supernatant containing the SCs was filtered and centrifuged. The pellet was then resuspended in growth medium consisting of Ham's F-10 nutrient mixture (Mediatech, Herndon, VA, http://www.cellgro.com) supplemented with 20% fetal bovine serum (FBS), penicillin, and streptomycin and plated in six-well plates (2 × 104 cells/well) coated with extracellular matrix (Sigma, St. Louis, MO, http://www.sigmaaldrich.com, 1:500). At this stage, the preparation contained primarily SCs as determined by flow cytometry and immunostaining analyses (Fig. 1 and Supporting Information Fig. S1). Cells were transfected 2 hours later with PNA-ssODNs (150 pmol/µl) as previously described . SCs isolated from age-matched C57 mice were used as a positive control, and received a sham transfection to control for the procedure. Cell differentiation was induced by maintaining the cells in low serum medium (differentiation medium) consisting of DMEM supplemented with 2% HS, penicillin, and streptomycin.
Single fiber cultures were prepared as described previously . Tibialis-anterior (TA) muscles of transplanted mice were digested in collagenase II and single fibers were cultured in proliferation medium containing 20% FBS (Mediatech) and βFGF for 24 hours prior to fixation and immunostaining analysis.
Analyses were performed using a Beckton Dickinson FACScalibur flow cytometry (Becton Dickinson, Franklin Lakes, NJ). Uptake of oligonucleotides was measured in cells trypsinized and harvested 24 hours after transfection. Analyses of intracellular epitopes were performed in cells fixed with 4% paraformaldehyde and permeabilized with staining buffer containing 0.1% Triton X-100. Cells were stained with an anti-rat CD34 antibody conjugated to FITC (Beckton-Dickinson, San Jose, CA, http://www.bd.com/us; 1:100). Data were acquired at 1 × 104 events per sample. Cells were distinguished from noncellular debris using forward and side scatter gating.
Cell Grafting and Muscle Harvesting
Hind limbs of host mice (6–8 weeks of age) were irradiated with 18 Gy and mice were allowed to recover for 4 days [24-26]. To promote regeneration, TA muscles of recipient mice were injured 4 days prior to engraftment using a single intramuscular injection (50 µl) of cardiotoxin (Calbiochem, Gibbstown, NJ, http://www.emdmillipore.com) resuspended at a concentration of 100 ng/µl [27, 28]. Three days after cardiotoxin injection, cells were engrafted into hind limbs of host mice. All transplantation procedures were performed using cells maintained in growth media for 12 hours following explant. Plates were washed three times with phosphate buffered saline (PBS) to eliminate debris and nonadherent cells. Cells were then trypsinized, centrifuged, and resuspended in a final volume of 50 µl of injectable grade saline solution. Following the initial dose optimization study (Fig. 3C), all subsequent transplantation procedures were performed using a single batch of cells prepared from mdx5cv or wild-type mice and at a dose of 5,000 cells per engrafted muscle. Some mice received a second injection (50 µl) of cardiotoxin (100 ng/µl), 3 weeks after grafting, in ipsilateral TA muscles. TA muscles were isolated, embedded in Tissue-Tek O.C.T compound (Sakura Finetek U.S., Inc., Torrance, CA, http://www.sakuraus.com) and muscles were transversely cut along the longitudinal axis into sections of 10 µm thickness at intervals of 300 µm and mounted on slides.
Immunofluorescence and Western Blot Analyses
Immunostaining analyses of cells in culture were performed using mouse antibodies to Pax7 (Developmental Studies Hybridoma Bank, Iowa City, IA, http://dshb.biology.uiowa.edu; 1:100), MyoD (BD Pharmingen, San Jose, CA; 1:250), desmin (Sigma; 1:500), and myogenin (BD Pharmingen, San Jose, CA, http://www.bdbiosciences.com; 1:250), incubated overnight, and counterstained with an Alexa 488-conjugated goat-anti-mouse immunoglobulin (Ig) (Life Technologies).
Single fibers isolated from TAs of mdx/nude mice engrafted with SCs were permeabilized with 0.5% Triton X-100 (Sigma), blocked with 20% goat serum, and incubated overnight with a mouse-anti-Pax7 antibody. Myofibers were counterstained with an Alexa 488-conjugated goat-anti-mouse Ig (Life Technologies) secondary antibody. Dystrophin immunostaining of cultured cells was performed as previously described [10, 11, 13] using a monoclonal antibody (Mandys-8, Sigma; 1:200) followed by an Alexa 546-coupled goat-anti-mouse (H+L) (Life Technologies; 1:250).
For immunohistology, sections were incubated using a polyclonal antibody against dystrophin (Thermo Scientific Neomarkers, Fremont, CA, http://www.thermoscientific.com; 1:100) and detected with the Alexa 546-coupled goat-anti-rabbit secondary antibody (Life Technologies; 1:500) . Hoechst staining (1:15,000) was used to visualize the nuclei within myofibers. Consecutive sections isolated from injected muscles were immunoassayed using Mandys-1011 (1:50) or Mandys-18 (1:100) to confirm the expression of full-length dystrophin [11, 13, 30]. Specific antibody binding was detected with the Alexa 546-coupled goat-anti-mouse secondary antibody (1:1,000). Reduction of nonspecific binding of the secondary antibody was achieved using a papain digested whole goat-anti-mouse antibody at a concentration of 25 µg/ml as previously described [13, 30, 31].
For the detection of quiescent and activated SCs, cryosections of TA muscles (10 µm) were incubated with a rabbit polyclonal antibody against MyoD (Santa Cruz Biotechnology, Inc., Santa Cruz, CA, http://www.scbt.com; 1:200) and mouse monoclonal antibody against Pax7 (Hybridoma Bank; 1:100) labeled with a Zenon labeling kit (Life Technologies) according to manufacturers' instructions. Rabbit polyclonal antibodies were detected with the Alexa 546-coupled goat-anti-rabbit secondary antibody as described above.
Western blot analysis was performed using an antibody directed toward the rod domain (Mandys-8; 1:400) of the dystrophin protein [10, 11, 32]. α-Actin was used as a sample loading control and was detected using a rabbit-anti-actin antibody (Sigma).
DNA and RNA Analyses
Gene correction was assessed in all cultures used for the transplantation procedures using cells maintained in culture for up to 2 weeks following transfection of the control and correcting PNA-ssODNs. Analyses at the genomic level were performed using DNA (200 ng) digested with HphI and subjected to amplification using the Forw-ex10 primer and Rev-int10 primers or the Forw-ex22 primer and Rev-int23 primers (Fig. 2C and Supporting Information Table S1) [11, 13]. Analysis of mRNA transcripts was performed using previously described protocols [11, 13]. Polymerase chain reactions (PCRs) on reverse transcribed cDNA (PCR) reactions were carried out using the forward primer (Forw-ex9) and the reverse primer (Rev-ex10in) (Fig. 2D and Supporting Information Table S1).
The ratios of gene correction obtained by targeting PNA-ssODNs to targeting ssODNs were calculated using a standard ΔΔCt method (2−ΔΔCt) . All data were normalized to GAPDH which was used as control [17, 34]. All reactions were performed on a MyiQ single-color detection system (Bio-Rad Laboratories, Hercules, CA, http://www.bio-rad.com) as described .
Amplicons were separated on 1.5% agarose gels and PCR products were purified using the Qiagen gel extraction kit (Qiagen). DNA sequencing was carried out using an Applied Biosystems ABI377 automated sequencer.
Quantitative Measurements of Muscle Fibers
The number of central nuclei in muscles was determined in sections immunostained for dystrophin and counterstained with Dapi (Life Technologies). A minimum of 500 fibers per muscle were analyzed for each treatment group. Fibers were considered to have centrally-located nuclei if there was at least one nucleus within the central portion of the dystrophin-positive fiber .
The relative amount of dystrophin expression restored within individual myofibers was measured in muscle sections immunoassayed for dystrophin using an Applied Imaging Ariol SL-50 . For each region, the minimum intensity value recorded (representative of the cytoplasm or background intensity) was subtracted from the maximum intensity value (corresponding to the sarcolemma) to correct each measurement for background intensity. Values are reported as percentage of the control.
Data are presented as means and standard deviations (SDs). Comparisons between groups were done using Student's t test assuming two-tailed distribution and equal variances.
Oligonucleotide Design and Transfection of SCs
SCs were isolated from purified single fibers obtained by enzymatic digestion of mdx5cv muscles followed by gentle trituration and dissociation [36, 37]. Flow cytometry was used to assess the purity of the preparation and demonstrated that the majority of the cells expressed CD34, a marker of SCs (Fig. 1B, 1C). Immunostaining analyses confirmed the purity of the preparation and also revealed the presence of distinct populations of myogenic cells in preparations obtained from muscles of mdx5cv mice compared to those obtained from wild-type mice (Fig. 1D and Supporting Information Fig. S1). The significantly higher percentage of cells expressing MyoD, desmin, and myogenin detected in muscles isolated from mdx5cv mice are likely to be due to the degenerative process that characterizes muscles lacking dystrophin and that results in the activation of SCs that are responsible for replacing myofibers that have degenerated as also observed by others .
Flow cytometry analysis clearly demonstrated that fluorescently labeled PNA-ssODNs are efficiently delivered into cells maintained in culture for 2 hours following isolation and subsequently transfected for 10 hours (Fig. 1E and Supporting Information Fig. S2). Time course analyses demonstrated that approximately 95% of the transfected cells had taken-up the PNA-ssODNs when analyzed 24 hours following transfection. Fluorescence remained high for at least 48 hours, but rapidly declined between 72 and 96 hours following oligonucleotide delivery (Supporting Information Fig. S2B).
Gene Correction Mediated by PNA-ssODNs In Vitro
The frequencies of gene repair achieved in SC cultures were assessed 2 weeks following isolation and transfection. Immunostaining analyses revealed the expression of dystrophin in some of the cells propagated in vitro and then induced to differentiate for 48 hours (Fig. 2A). Results were confirmed by Western blot analyses (Fig. 2B). Dystrophin was absent in untransfected cells or cells treated with PNA-CTLC. Cells transfected with PNA-CORC showed a clear band at the expected molecular weight although expression was much lower than that obtained from SCs isolated from wild-type muscles and maintained in culture for the same period of time (Fig. 2B). Quantitative analysis of protein expression obtained from multiple independent experiments demonstrated that the level of full-length dystrophin protein ranged between 1% and 2% of that of wild-type cells (mean 1.7% ± 0.49%).
Real-time PCR was used to validate the results obtained by Western blot analysis (Fig. 2C). A band of the exact molecular weight as that obtained in cells isolated from wild-type mice was clearly detected only in mdx5cv cells that received PNA-CORC, but not in cells treated with the control PNA-ssODN, further confirming the specificity of the correction process mediated by targeting oligonucleotides (Fig. 2C). The level of full-length dystrophin expression detected was similar to that obtained by Western blot analysis (mean 1.5% ± 0.64%).
To further confirm that the results obtained were due to correction of the dystrophin gene, genomic DNA was isolated from cells treated with the targeting or control PNA-ssODNs and subjected to restriction DNA digestion to abrogate all copies of the dystrophin mdx5cv gene refractory to correction (Fig. 2D). Amplification of a specific product was detected in mdx5cv cells transfected with the targeting oligonucleotide, but not in cells treated with the control PNA-ssODN. Direct sequencing of the amplicon confirmed correction at the genomic DNA level (not shown). The frequencies of gene repair ranged between 1.5% and 2.1% as determined by quantitative PCR thus confirming the results obtained at the mRNA and protein levels. Altogether these results demonstrate that PNA-CORC can target and induce the desired single base-pair alteration at the DNA level and that correction results in the expression of full-length dystrophin.
Transplantation of SCs in mdx/Nude Mice
A dose response study was used to assess the ability of SCs that had undergone gene repair to restore dystrophin expression in dystrophin-deficient mice. Experiments were conducted in the mdx/nude strain, a known mouse model for DMD. This strain has been used extensively in the field and has shown to provide efficient engraftment following transplantation. Furthermore, the impaired immune system in immunocompromised mdx/nude mice has been associated with reduced fibrosis rendering comparative analyses between individual experiments more reliable especially in older mice [18, 19].
To determine the minimum number of SCs needed to be transplanted in order to detect an effect, we isolated SCs from mdx5cv mice, allowed them to seed for 2 hours and then transfected them with the targeting PNA-ssODN as described above (Fig. 3A). As positive control, we used SCs obtained from wild-type mice that had undergone the same isolation procedure, but that were treated with transfection reagent alone [27, 28]. Dystrophin-positive fibers were clearly detected in all muscles that received SCs (Fig. 3B, 3C). No statistically significant differences were detected in the number of dystrophin-positive fibers in muscles that received mdx5cv SCs treated with PNA-CORC at any of the dosages analyzed (Fig. 3C). These results suggested that, under our experimental procedures, high dosages of SCs are required to achieve limited, although detectable, dystrophin expression in muscle. Therefore, all subsequent analyses were performed using a dose of 5,000 cells per transplanted TA.
Retention of Stem Cell Properties Following In Vitro Transfection with PNA-ssODNs
A hallmark characteristic of any stem cell population is the ability to emerge from a quiescent state in response to the needs of its environment and to divide to produce daughter cells which are ultimately responsible for repairing injured or diseased myofibers. To better characterize the effects of transplanting SCs which have been transfected with oligonucleotides, we injured TA muscles of engrafted mice and analyzed them for the expression of dystrophin 2 weeks after injury (Fig. 4A) [26, 39, 40]. An increase in the number of dystrophin-positive fibers was clearly detected in muscles that had undergone injury following transplantation compared to uninjured muscles (Fig. 4B). No significant differences in the number of dystrophin-positive fibers were observed in muscles of irradiated mice that were subjected to cardiotoxin injection 3 weeks following irradiation and analyzed 2 weeks later (Supporting Information Fig. S3).
Immunostaining analyses were used to assess for the presence of activated and differentiated SCs. Sections isolated from engrafted muscles were immunoassayed for the expression of Pax7 and MyoD, markers of quiescent and activated SCs, respectively. Pax7-positive cells were clearly detected in muscles of control mice or muscles of mdx/nude mice engrafted with SCs, but not in muscles of irradiated mice that did not receive SCs and that underwent cardiotoxin injury (Fig. 4C). Expression was confined to few isolated cells located adjacent to the muscle membrane. The majority of those cells coexpressed Pax7 and MyoD, suggesting that engrafted cells had retained the ability to divide and to give rise to muscle progenitor cells (Fig. 4D and Supporting Information Fig. S4).
To further clarify the effects of SC activation and repair in muscles, we analyzed consecutive sections of isolated TAs and followed the distribution of dystrophin-positive fibers along the length of the engrafted muscles (Fig. 5). On average, the longitudinal distribution of dystrophin in muscles that had undergone cardiotoxin injury following transplantation was greater than that detected in resting muscles and was confirmed in both, muscles that received mdx5cv SCs transfected with the targeting PNA-ssODN as well as in muscles engrafted with SCs isolated from wild-type mice. Moreover, the number of fibers comprising each individual cluster expressing dystrophin was one- to two-fold higher than that observed in muscles that received SCs but that had not undergone injury following transplantation (Fig. 5B, 5C). Altogether, these results provide strong evidence that at least a few of the engrafted SCs had maintained their ability to return to quiescence and that, when prompted, they were able to exit the quiescent state and actively contribute to muscle regeneration.
Persistence of Dystrophin Expression Over Prolonged Periods of Time
The long-term therapeutic application of gene correction in SCs was assessed 24 weeks after transplantation in mice that received mdx5cv SCs transfected with PNA-CORC and compared to that achieved after engraftment of SCs isolated from wild-type muscle. Large clusters of dystrophin-positive fibers were detected in all muscles that received SCs, but not in TAs of untreated mice (Fig. 6A). Interestingly, the number of dystrophin-positive fibers detected in muscles that received mdx5cv SCs treated with the targeting PNA-ssODN was threefold higher than that observed 5 weeks after transplantation suggesting that SCs that had undergone gene repair were able to actively contribute to muscle regeneration (Fig. 6B).
To further confirm that dystrophin-positive fibers detected in transplanted muscles expressed full-length dystrophin, we immunoassayed consecutive sections isolated from engrafted muscles with an antibody specific to a region of the dystrophin protein that is spliced out as a result of the mdx5cv mutation (Mandys-1011) or with an antibody recognizing the region of the dystrophin protein close to the mdx/nude mutation responsible for the absence of dystrophin in this strain (Mandys-18) [11, 30, 41, 42]. TAs isolated from age-matched mdx/nude mice showed small clusters of dystrophin-positive fibers immunoreactive to Mandys-1011 and limited or no expression was observed following analysis using an antibody specific to exon 32 of the dystrophin protein (Supporting Information Fig. S6). Virtually all the fibers that were immunoreactive to the region of dystrophin encoded by exons 10 and 11 were also recognized by the antibody raised against the more distal regions of the dystrophin protein (Supporting Information Fig. S6).
Results were confirmed at the molecular level (Supporting Information Fig. S7). Sections obtained from muscles engrafted with SCs were immunostained for dystrophin and dystrophin-positive fibers were mechanically excised. Total genomic DNA was isolated from purified fibers using standard methods and was subjected to HphI digestion to eliminate all copies of dystrophin DNA that could have arisen, in those fibers, from the contribution of noncorrected mdx5cv SCs [11, 13]. Amplification was carried out using a set of primers encompassing exon/intron 10 and a separate set of primers positioned across intron 22 and exon 23 of the dystrophin gene. Amplification of a product was obtained in all muscles isolated from untreated mdx/nude mice or muscles engrafted with SCs, but not in TAs isolated from mdx5cv mice, confirming the specificity of the amplification products. Direct sequencing of the PCR amplicons confirmed the presence of corrected sequences containing the desired T-to-A transversion in exon 10 of the dystrophin gene and lacking the G-to-T mutation characteristic of the mdx/nude mouse (Supporting Information, Fig. S7).
To ensure that the differences in dystrophin-positive fibers detected in muscles that received mdx5cv SCs transfected with the targeting PNA-ssODNs compared to those observed in muscles that received wild-type SCs were not the result of differences in engraftment or survival of cells that could have occurred over time, we determined the number of quiescent and activated SCs in muscles isolated 24 weeks after SC transplantation (Fig. 6C–6E). No statically significant differences were detected in the number of Pax7- or MyoD-positive cells among muscles engrafted with SCs suggesting that the results obtained in muscles that received mdx5cv SCs treated with the PNA-ssODN were not due to increases in the ability of cells that had undergone correction to home into muscle following transplantation and over prolonged periods of time (Fig. 6C, 6D). To gather further evidence that the increase in the number and distribution of dystrophin-positive fibers was the result of the contribution of SCs that had undergone gene repair, single fibers were isolated from engrafted muscles and analyzed for the expression of Pax7. Virtually no positives were detected in irradiated muscles of mdx/nude mice confirming the efficacy of the irradiation procedure to ablate SCs. In net contrast, Pax7-positive donor-derived SCs were evident in myofibers maintained in culture for 48 hours after explant and isolated from muscles engrafted with mdx5cv SCs treated with PNA-CORC (Fig. 6F). Similar results were obtained from muscles isolated from control mice that received wild-type SCs (not shown).
Morphological Analyses of Engrafted Muscles Following Transplantation of SCs
To better characterize the functional activity of SCs that had undergone repair ex vivo and that were transplanted into recipient muscles, we analyzed the localization of myonuclei in dystrophin-positive fibers and used it as an index of functional recovery . As expected, dystrophin-positive fibers isolated from muscles that received mdx5cv SCs transfected with the targeting PNA-ssODN or SCs isolated from wild-type muscle that were analyzed 5 weeks after transplantation all contained predominantly central nuclei in dystrophin-positive fibers demonstrating that those fibers were newly regenerated and/or repaired (Fig. 7A, 7B). A significant decrease in the percentage of dystrophin-positive fibers containing central nuclei was evident 24 weeks after engraftment (Fig. 7B). No statistically significant differences were detected in the percentage of dystrophin-positive fibers containing central nuclei in muscles that received mdx5cv SCs transfected with PNA-CORC compared to muscles that received SCs isolated from wild-type muscle. To gather further insights into the stability of dystrophin expression obtained over time and to determine whether the improvement in muscle morphology observed in muscles that received mdx5cv SCs transfected with PNA-CORC could be correlated to the level of dystrophin expressed within individual fibers as the result of the contribution of transplanted cells, we analyzed the distribution of dystrophin expression along the length of the muscle (Fig. 7C). In muscles that received wild-type SCs, the majority of the dystrophin-positive fibers distributed for greater than two-thirds of the length of the muscle suggesting that, over time, engrafted SCs were able to actively contribute to muscle regeneration. A shift in the distribution of positive fibers expressing dystrophin was also observed in muscles that received mdx5cv SCs transfected with PNA-CORC analyzed 24 weeks after transplantation compared to that detected 5 weeks after engraftment. Results were correlated to the level of dystrophin expression achieved within individual fibers (Supporting Information Fig. S8). An increase in the percentage of dystrophin expression was clearly detected in muscles that received SCs and that were analyzed 24 weeks after transplantation compared to that detected 5 weeks after transplantation. Importantly, at the later time point analyzed, the majority of the fibers that expressed dystrophin showed levels of expression equal to or greater than 50% of the intensity of dystrophin detected in wild-type mice. These results greatly support the hypothesis that the increase in dystrophin expression detected in muscles analyzed 24 weeks after engraftment was the result of the contribution of SCs and provide clues on the levels of dystrophin expression that can be achieved into muscle using regenerative approaches aimed at targeting SCs.
We have demonstrated the feasibility of using PNA-ssODNs to direct single-point mutations at the genomic level and to correct SCs ex vivo. Gene correction frequencies were lower than those detected in myoblasts and previously reported  suggesting the presence of intrinsic differences between those two cell populations. Differences in fluorescence intensity and persistence of fluorescently labeled PNA-ssODNs detected in SCs transfected in vitro compared to that previously observed in myoblasts treated under the same experimental procedures may in part be responsible for the differences in gene correction frequencies achieved in this study (Supporting Information Fig. S2). Similarly, differences in chromatin structure and chromatin folding typical of nondividing cells such as quiescent SCs may have restricted the accessibility of the ssODN to the region targeted and could in part be responsible for the low levels of gene repair observed. Nonetheless, correction was clearly detected in cells maintained in culture for up to 2 weeks following SC explant and was stably inherited through cell division.
The significant increase in the number of cells that expressed MyoD, desmin, and myogenin obtained from mdx5cv muscles compared to those detected in cultures of wild-type muscles demonstrated the presence of distinct populations of cells in our preparations including differentiated and committed muscle progenitors (Fig. 1D). This increase is likely to be the result of the degenerative process that characterizes dystrophic muscles and that leads to continuous rounds of SC activation necessary to form new myofibers. Accordingly, correction mediated by PNA-CORC may have occurred not only into quiescent SCs but also into activated SCs and muscle progenitor cells. The possibility that the expression of dystrophin detected following engraftment of cells isolated from mdx5cv mice transfected with PNA-CORC may be the result of a small fraction of the muscle progenitor cells present in isolated muscles that had undergone repair rather than SCs is highly unlikely. First, extensive data in the literature have demonstrated that the number of cells required to achieve an effect are at least 1 order of magnitude higher than those used in this study [29, 40, 44]. Second, it has been clearly demonstrated that the majority of differentiated muscle progenitor cells die shortly after transplantation [45-47]. Furthermore, the increase in the number of dystrophin-positive fibers detected 24 weeks following transplantation in muscle that received mdx5cv SCs treated with the targeting PNA-ssODN suggests that a large number of the corrected cells were able to proliferate. Although muscle progenitors capable of self-renewing have been described they represent only a small minority of the cell population . Thus, it is safe to conclude that the results obtained are likely to be due to a fraction of the SCs that had undergone repair, homed into muscles, and actively contributed to myofiber regeneration. Moreover, the results demonstrate that explant of SCs ex vivo and transfection in vitro does not compromise the ability of cells to engraft into muscles and to activate in response to injury or during homeostasis establishing the importance of targeting SCs for the treatment of DMD. Whether activated SCs that were present in the preparation obtained from mdx5cv muscles are amenable to gene repair and were able to return into quiescence following transplantation into mdx/nude mice remains to be established and is the focus of active studies in the laboratory.
Most of the research aimed at determining the efficacy of SCs in restoring dystrophin expression for the treatment of DMD following engraftment has implemented the use of reporter systems to identify donor cells into transplanted muscles [49-52]. Although these systems have proven to be valuable tools, expression of proteins like β-galactosidase (β-gal) or green fluorescent protein as experimental controls has been associated with a strong proinflammatory response toward the reporter protein and increased muscle damage rendering the assessment of the long-term therapeutic efficacy of SC-mediated regenerative approaches to DMD difficult to perform . Our experimental settings overcome these limitations by ensuring that the expression of dystrophin detected following engraftment of mdx5cv SCs treated with the targeting PNA-ssODN is the result of correction that has occurred at the genomic level and not the result of expansion of revertant fibers that may have occurred over time. First, muscles subjected to transplantation were irradiated to ablate all endogenous SCs. Although it has been previously shown that a fraction of SCs survive the procedure, these cells are extremely rare in mdx mice [54, 55]. The fact that the number of dystrophin-positive fibers in TA muscles of control mdx/nude mice that had undergone irradiation remained practically unchanged over the course of the analyses compared to non-irradiated mdx/nude control muscles clearly demonstrates that the majority of the SCs are radiation sensitive and did not survive the procedure (Supporting Information Figs. S3, S5). Second, expression of full-length dystrophin was confirmed using antibodies specific to regions of the dystrophin protein that are not expressed in revertant fibers. Approximately 93% of the fibers that were analyzed 24 weeks following transplantation of mdx5cv SCs treated with PNA-CORC and that were identified using an antibody specific to the C-terminal region of the dystrophin protein were also positive for the region of the dystrophin protein encoded by exons 10 and 11 (Mandys-1011) and exon 32 (Mandys-18) (Supporting Information Fig. S6). Finally, correction of the genetic defect was confirmed in fibers that expressed dystrophin at the genomic level using direct sequencing of PCR products obtained from excised fibers that expressed dystrophin (Supporting Information Fig. S7).
Taken together, our results clearly demonstrate that the loss of dystrophin expression typically seen using systems targeting only mature myofibers [13, 17] can be efficiently counteracted by correcting SCs. Moreover, the results provide strong evidence that correction of as little as 1% of the isolated SCs is sufficient to achieve sustained beneficial effects.
Crescent evidence demonstrates that SCs are composed of a heterogeneous population of cells [49, 50, 52, 56]. As such, it is possible that the effects observed in muscle following transplantation may be the result of a specific subpopulation of SCs more amenable than others to gene repair following transfection of the correcting PNA-ssODNs. In this report, we have focused on studying the efficacy of PNA-ssODNs in targeting the whole SC population and on limiting cell manipulation following explant. This approach has previously been shown to be able to achieve a higher rate of engraftment and to minimize the loss of regenerative potential following transplantation compared to that obtained using fluorescence-activated cell sorting, required for the identification of specific SC subtypes [26, 28, 57]. Importantly, the reproducibility of the results between independent analyses in vivo and the limited variability observed among replicate experiments suggest that our experimental conditions were able to guarantee minimal or no loss of regenerative potential of transplanted cells following transfection with targeting PNA-ssODNs.
Further studies are required to better understand the applicability of cell-mediated regenerative approaches to muscle disorders using gene editing strategies. Safety concerns need to be addressed in detail to ensure the specificity of PNA-ssODNs in the correction process. Furthermore, a better characterization of SCs that have undergone gene repair may be able to determine the mechanisms that control the repair process in these cells and could lead to the development of new strategies capable of improving targeting efficiencies of ssODNs.
Despite the encouraging results obtained in this report, limitations still exist which could ultimately prevent the use of cell transplantation therapies for the treatment of muscle disorders. Among those are the need to target a large number of muscles simultaneously and the need to overcome the limitations in the ability of engrafted SCs to migrate within adjacent myofibers so as to maximize the effects achieved in DMD patients. Approaches are currently being optimized aimed at using the circulatory system as the method of delivering stem cells and clinical trials are already well on their way to determine the potential beneficial effects achieved using systemic administration of pericytes in DMD patients [58, 59]. An alternative to the use of transplantation procedures is the use of approaches aimed at targeting and correcting SCs in situ. Current studies in our laboratory are testing the feasibility of delivering PNA-ssODNs into muscles after systemic administration. Approaches that can facilitate the uptake of ssODNs into SCs following the delivery of PNA-ssODNs are likely to have an advantage and could lead to the development of an effective therapy for DMD.
Conclusion and Summary
The results presented in this report strongly support further testing of ssODN-mediated applications for the treatment of muscle disorders and provide clear evidence that stable correction of dystrophin gene defects can be achieved by targeting muscle stem cells. Furthermore, the results demonstrate that correction in a limited number of SCs is sufficient to produce significant levels of dystrophin expression and that the amount of dystrophin being restored into individual muscle fibers generated from SCs that have undergone gene repair is therapeutically relevant. As such, gene correction mediated by oligonucleotides remains at the forefront of the approaches for DMD and this study opens new options for the treatment of many neuromuscular diseases.
We would like to thank Dr. Glenn Morris (MRIC Biochemistry Group, North East Wales Institute, Wrexham, U.K.) for kindly providing the Mandys-1011 and the Mandys-18 antibodies. This work was supported by a grant from the Muscular Dystrophy Association (175142) to C.B.
F.N.: performed experiments in vivo, analyzed data, and wrote the manuscript; C.B.: designed and directed the study, performed the experiments in vitro, analyzed and integrated the data, and wrote the manuscript.
Disclosure of Potential Conflicts of Interest
The authors indicate no potential conflicts of interest.