By continuing to browse this site you agree to us using cookies as described in About Cookies
Notice: Due to essential maintenance the subscribe/renew pages will be unavailable on Wednesday 26 October between 02:00- 08:00 BST/ 09:00 – 15:00 SGT/ 21:00- 03:00 EDT. Apologies for the inconvenience.
Fondazione IRCCS Cà Granda Ospedale Maggiore Policlinico, Centro Dino Ferrari, Milano, Italy
Y. Torrente, Stem Cell Laboratory, Dipartimento di Fisiopatologia medico-chirurgica e dei Trapianti, Università degli Studi di Milano, Fondazione IRCCS Ca' Granda Ospedale Maggiore Policlinico, Via F. Sforza 35, 20122 Milan, Italy
The protein dysferlin is abundantly expressed in skeletal and cardiac muscles, where its main function is membrane repair. Mutations in the dysferlin gene are involved in two autosomal recessive muscular dystrophies: Miyoshi myopathy and limb-girdle muscular dystrophy type 2B. Development of effective therapies remains a great challenge. Strategies to repair the dysferlin gene by skipping mutated exons, using antisense oligonucleotides (AONs), may be suitable only for a subset of mutations, while cell and gene therapy can be extended to all mutations. AON-treated blood-derived CD133+ stem cells isolated from patients with Miyoshi myopathy led to partial dysferlin reconstitution in vitro but failed to express dysferlin after intramuscular transplantation into scid/blAJ dysferlin null mice. We thus extended these experiments producing the full-length dysferlin mediated by a lentiviral vector in blood-derived CD133+ stem cells isolated from the same patients. Transplantation of engineered blood-derived CD133+ stem cells into scid/blAJ mice resulted in sufficient dysferlin expression to correct functional deficits in skeletal muscle membrane repair. Our data suggest for the first time that lentivirus-mediated delivery of full-length dysferlin in stem cells isolated from Miyoshi myopathy patients could represent an alternative therapeutic approach for treatment of dysferlinopathies.
Muscular dystrophies are a heterogeneous group of inherited disorders characterized by progressive muscle wasting and weakness; they present large clinical variability regarding age of onset, patterns of skeletal muscle involvement, heart damage, rate of progression and mode of inheritance [1, 2]. Molecular genetic studies have revealed different causative mutations in genes that encode proteins involved in all aspects of muscle cell biology. Two clinical forms of autosomal recessive muscular dystrophies – Miyoshi myopathy (MM) and limb-girdle muscular dystrophy type 2B (LGMD-2B) – arise from genetic defects in the dysferlin gene (DYSF; 2p13, GenBank NM_003494.2) [3, 4]. In addition to previously described symptoms, mutations in DYSF are associated with a wide spectrum of phenotypes, ranging from isolated hyperCPKemia to severe disability . Dysferlin is a 230-kDa protein that is abundantly expressed in skeletal and cardiac muscles and plays a central role in sarcolemmal repair . Based on protein sequence analysis, it was classified as a member of the ‘ferlins’ family, sharing a common multiple-motif C2 domain and one transmembrane domain in the C-terminal part of the protein [7-9]. To date, the reported DYSF gene mutations include point mutations, small deletions and insertions, distributed throughout the entire coding sequence. No hotspots have been identified, and missense, nonsense and frameshift mutations have been reported [10, 11].
Although no therapy is available for muscular dystrophies, different studies have demonstrated that transcript rescue is a feasible approach in dystrophinopathies. Since the work of Lu and colleagues that demonstrated how exon skipping was a good technique to bypass nonsense mutation into mdx mouse , the group of Aartsma-Rus confirmed the feasibility of exon skipping into human Duchenne muscular dystrophy cells  while Goyenvalle and co-workers showed that U7 small nuclear RNA mediated exon skipping rescues the dystrophic muscle . All this evidence lead to recent clinical trials . This technique uses specific antisense oligonucleotides (AONs) to skip one or several exons carrying disease-causing mutations in order to obtain an in-frame functional protein [14, 16, 17]. Unfortunately, it is only applicable for proteins in which part of the amino acid sequence can be deleted without important deleterious impact on their overall function. In the dysferlin gene, it may be possible to skip some exons in the functionally redundant C2 domains, such as exon 32 [18, 19], but not others, such as exons 53 and 54 that encode the unique terminal transmembrane domain . Other techniques complementary to the exon skipping approach include the use of adeno-associated viruses (AAVs) to carry wild-type cDNA of specific mutated genes and perform gene replacement [21, 22] and/or the use of gene miniaturizing and transplicing approaches to overcome the packaging limits of AAVs [23, 24]. Promising results have been obtained with mini-gene replacement [25, 26], and Grose et al. were able to deliver a cassette with an optimized dysferlin cDNA using AAV5. Following AAV5 transfer in an animal model, they produced the dysferlin full-length transcript and protein, with expression levels sufficient to correct functional deficits in diaphragm muscle and in skeletal muscle membrane repair .
Additionally, stem cells alone or, better, combined with gene correction methods have received much attention for their potential use in therapies for human degenerative diseases [28-30]. In 2004, our group isolated from human blood a subpopulation of CD133+ stem cells that, when transplanted into a DMD animal model (the scid/mdx mouse), restored dystrophin expression, participated in skeletal muscle regeneration and regenerated the satellite cell pool . Recently, dystrophic CD133+ stem cells were transduced with a lentivirus carrying a construct designed to skip exon 51 of dystrophin; following transplantation into scid/mdx mice, they fused in vivo with regenerative fibers, restructuring the dystrophin-associated protein complex .
In the present work, we isolated blood-derived CD133+ stem cells from two patients carrying different mutations in the dysferlin gene. The first patient had two mutant alleles (one deletion of exon 22 and one large deletion between exons 25 and 29), while the second patient had a homozygous deletion of exon 55. We verified the feasibility of rescuing dysferlin of blood-derived CD133+ stem cells isolated from the first patient by means of exon skipping using AONs. AON-treated blood-derived CD133+ stem cells failed to express dysferlin in vivo after transplantation into immune/dysferlin-deficient scid/blAJ mice . These data together with the ineligibility for exon skipping of the second patient (lacking exon 55) prompted us to design a lentivector for expression of the full-length dysferlin transcript. We showed that such a vector allowed dysferlin expression of blood-derived CD133+ stem cells isolated from MM patients to correct functional deficits in skeletal muscle membrane repair of transplanted scid/blAJ mice.
Patient phenotype and genotype description
This study included patients exhibiting a typical MM phenotype. Patient 1 was a 34-year-old man diagnosed with MM at the age of 17 years. At the time of diagnosis, his right quadriceps muscles showed evidence of a myopathic pattern with fiber size variability, increased connective tissue, necrosis and interstitial cellularity. Immunostaining of peripheral blood mononuclear cells (PBMCs) (Fig. 1C) and western blot analysis (Fig. 1E) showed an absence of DYSF protein. We isolated CD14+ mononuclear cells from the patient's PBMCs to perform detailed genetic analyses of the DYSF gene. Exploration of the entire DYSF coding sequence and intronic boundaries resulted in the identification of one deletion in exon 22, leading to a premature codon stop (c.2077delC, p. His693Thrfs*). A second mutation was identified using genomic quantitative real-time PCR; it consisted of a large genomic deletion encompassing exons 25–29 and was predicted to cause an in-frame deletion (p.Tyr838_Arg1058del) . The absence of protein observed in western blot and immunofluorescence analyses suggested a destabilization and/or degradation of this predicted in-frame truncated RNA or protein. RT-PCR was performed in order to detect a shorter in-frame mRNA. Surprisingly, we failed to amplify the deleted mRNA using primers located around the deletion (namely, in exons 20 and 31), and we amplified only the allele carrying the exon 22 deletion (Fig. 1A). We designed primers overlapping the junction 24–30, specific for the allele carrying the large deletion, but still observed no amplification (Fig. 1B).
Patient 2 was a 45-year-old man who was diagnosed with MM at 22 years of age. At the time of diagnosis, his right quadriceps muscles showed evidence of a myopathic pattern with necrosis and an increased interstitial cellularity. RT-PCR analysis confirmed the presence of dysferlin mRNA in patient 2 (Fig. 1D), whereas immunostaining of PBMCs (Fig. 1C) and western blot (Fig. 1E) showed an absence of dysferlin protein. Gene sequencing showed a homozygous deletion of exon 55 (c.6233_6240delCCTTCAGC, p.Pro2078Leufs*92) of the DYSF gene, causing a stop codon.
Feasibility of exon skipping of exons 22–23, 25–29, 22–29 of the DYSF gene
To verify the feasibility of an exon skipping approach in patient 1, we constructed several deleted dysferlin cDNAs. We obtained a dysferlin protein fused to the C-terminal part of enhanced green fluorescent protein (EGFP) in the N-terminal of the protein, allowing its expression in mammalian cells. The exons of interest were confirmed to be deleted, including 22–23, mimicking the skipping of the mutation of one allele of patient 1; 25–29, mimicking the skipping of the mutation in the second allele of patient 1; and 22–29, to investigate the possibility of skipping a high number of exons. We transfected full-length and deleted Δ22–23, Δ25–29, Δ22–29 dysferlin plasmids into HEK cells, to evaluate production of the deleted dysferlin proteins. Then 48 h after transfection we measured the percentage of dead cells (respectively, full-length dysferlin 5.1%; non-transfected cells 1.9%; Δ22–23 18.8%; Δ25–29 14.4%; Δ22–29 18.7%) and the percentage of the EGFP signal by fluorescence activated cell sorting (FACS), confirming the expression of truncated EGFP-dysferlin plasmid and the efficiency of transduction (Fig. 2A). RT-PCR analysis from transfected HEK cells showed the expression of dysferlin mRNA for all of the truncated isoforms (Fig. 2B). These data were validated by western blot analysis: we identified the full-length dysferlin protein corresponding to a molecular weight of ~ 268 kDa and other truncated dysferlins characterized by lower molecular weights (Fig. 2C). Furthermore, the expression of the deleted form Δ22–29 demonstrated that all exons between 22 and 29 could be removed to produce a truncated protein (Fig. 2C). Unfortunately, the same approach could not be used to create a Δ51–55 EGFP dysferlin, mimicking the skipping of the mutation of patient 2, because this block of exons encodes two crucial elements, the 3′UTR sequence and the transmembrane domain of the protein .
Restoration of dysferlin expression in vitro by AONs and lentivirus full-length DYSF
Both exons 22 and 23 had to be removed to restore an ORF and to allow the production of a ‘truncated’ but functional protein that could ameliorate the clinical phenotype (Fig. 3A). According to previously reported data [34, 35], several regions of both exons were analyzed with regard to recognizing exon splicing enhancer in addition to the donor/acceptor splice sites and allowing the skipping of targeted exons. After AON treatment of normal human myoblasts, RT-PCR analysis encompassing exons 22 and 23 was performed using a combination of primers located from exon 20 to exon 26 to ensure amplification of shorter fragments carrying an exon 22–23 deletion or larger ones, as described in previous studies [36, 37]. We observed several skipped products corresponding to deletions of exon 23 alone, exons 22 and 23, exons 22–24, or exon 24 alone (Fig. 3A). In both treated and untreated samples, we observed deletion of exon 24 alone (data not shown). This unexpected skipped product suggested an alternative transcript expressed at very low level in myoblasts in culture. The AON treatment of blood-derived CD133+ stem cells isolated from patient 1 led to the expression of a skipped dysferlin (Δ22–24) with a low efficiency (Fig. 3A).
We next developed a strategy based on complete dysferlin delivery by lentiviral vector. Importantly, this approach represented a valid alternative to exon skipping technology for the treatment of patient 2, for whom we could not remove the exons located in the C-terminal part of the protein (such as exons 51–55) because their removal would impair dysferlin protein expression and function . Moreover, other regions required skipping of three or more exons to restore a correct ORF, a result hardly obtainable with a classical exon skipping strategy, as we showed in patient 1. To overcome these limitations, we subcloned the complete dysferlin transcript into a lentiviral pRRL backbone and tested its efficiency in blood-derived CD133+ cells isolated from patients 1 and 2. We constructed a lentiviral vector (pRRLSINPPTPGK H dysferlin WPRD, or LV-FL DYSF) in which expression of human full-length dysferlin cDNA was driven by human PGK1 promoter. We used RT-PCR analysis to test the LV-FL DYSF transduction efficiency on CD133+ stem cells isolated from the patients. Total mRNAs from engineered CD133+ stem cells were analyzed by nested RT-PCR to confirm the presence of a full-length DYSF transcript (Fig. 3C). To further analyze the LV-FL DYSF transduction efficiency we synthesized a second lentiviral vector using the same pRRL backbone subcloned with GFP transcript. Our lentivirus construction recapitulates the transduction efficiency of previously tested and published lentivirus vectors [38, 39]. We performed FACS analysis to determine the expression of GFP in infected CD133+ normal blood derived cells and we showed a percentage of 60.9% of GFP positive cells demonstrating a good efficiency of transduction as previously described [40, 41] (Fig. 3B).
Failure of dysferlin expression after engraftment of AON-treated blood-derived CD133+ stem cells isolated from MM patients
We injected 2 × 105 AON-treated blood-derived CD133+ stem cells isolated from patient 1 into the tibialis anterior (TA) of 5-month-old scid/blAJ mice (two groups of n =5 for patient cell batch) after verifying the expression of a truncated dysferlin by RT-PCR (Fig. 3A). At 72 h after the injection, human α-centromeric positive fibers were detected in all injected muscles, but dysferlin was not detected at the surface of these human α-centromeric positive fibers (data not shown). One month after the intramuscular transplantation of AON-treated blood-derived CD133+ stem cells isolated from patient 1 (n =5), we did not observe the presence of dysferlin positive myofibers.
In vivo restoration of dysferlin expression in mouse after intramuscular delivery of LV-FL DYSF blood-derived CD133+ stem cells isolated from MM patients
As we obtained dysferlin expression in vitro, we tested the ability of blood-derived CD133+ stem cells isolated from patients 1 and 2 that were transduced with LV-FL DYSF to restore dysferlin expression in muscles of scid/blAJ mice. We performed a single intramuscular injection of 2 × 105 engineered cells in the untreated (n =4) or notexin (ntx) pretreated TA of 5-month-old scid/blAJ mice (n =9). Ntx injection was used to allow the regenerating fibers to increase the fusion of the human engineered blood-derived CD133+ stem cells. Moreover, ntx pretreatment exacerbated the pathology of scid/blAJ mice that show a less severe phenotype than the A/J dysferlin-null mice . One month after the transplantation, clusters of human α-centromeric positive fibers were detected in transverse muscle sections, presumably located near the injection sites, with a higher percentage in ntx-treated than untreated muscles (29.5 ± 3.77 versus 5.14 ± 2.15 of human positive nuclei per section) (Fig. 4A). The presence of human dysferlin was thus investigated in ntx-pretreated muscle sections of scid/blAJ dysferlin-null mice. Human dysferlin was detected in a patchy distribution of the muscle fibers of transplanted scid/blAJ mice (37.2 ± 5.60 of human positive dysferlin myofibers per section; < 10% of total myofibers) (Fig. 5). The majority of human dysferlin positive myofibers co-express human α-centromeric positive fibers (Fig. 5A). Human dystrophin expression was detected on serial sections using the NCL-Dys3 antibody (39.7 ± 3.58 of human positive dystrophin myofibers per section; < 10% of total myofibers) (Fig. 4B). RT-PCR analysis detected dysferlin expression in all treated muscles (Fig. 5B). Western blot analysis confirmed the findings previously described and showed expression of the dysferlin protein in all transplanted dystrophic muscles (Fig. 5C). To analyze whether the dysferlin restoration after intramuscular transplantation of engineered blood-derived CD133+ stem cells isolated from patients 1 and 2 was associated with improvement of muscle health, we investigated the tissue characteristics of treated muscles. Unfortunately, the injected muscles showed dystrophic features similar to untreated muscles given that there are no differences in the percentage of centronucleated fibers (7.466 ± 2.02 in injected scid/blAJ mice and 8.52 ± 2.9 in untreated scid/blAJ mice; n =9). In order to exclude less efficiency of in vivo engraftment of engineered stem cells we verified the success of intramuscular transplantation on scid/blAJ dysferlin-null mice of human blood-derived CD133+ stem cells isolated from healthy subjects. The number of human dysferlin positive muscle fibers in transplanted scid/blAJ mice was similar to the number obtained after transplantation of engineered blood-derived CD133+ stem cells (39.8 ± 7.13 of human positive dysferlin myofibers per section; < 10% of total myofibers). Clusters of human dystrophin positive myofibers were also found (41.2 ± 1.71 of human positive dystrophin myofibers per section; < 10% of total myofibers). The percentage of centronucleated fibers (8.104 ± 1.10 in injected scid/blAJ mice and 8.36 ± 2.4 in untreated scid/blAJ mice; n =9) and dystrophic features of human blood-derived CD133+ stem cell injected muscles were similar to untreated dystrophic muscles.
As dysferlin absence causes a defect in membrane repair, we tested the ability of single fibers from transplanted muscles to reseal the membrane after laser wounding, as previously described by Bansal et al. . Laser wounding experiments were performed in the presence of FM-143 dye on single fibers isolated from the TA of injected and non-injected scid/blAJ mice. FM-143 fluoresces more brightly in a lipid environment; thus, in membranes that are wounded but not healed, the fluorescence increases as the FM-143 entering the sarcolemma binds to internal membranes. When the membrane damage is resealed, further intracellular entry of the dye is blocked and the fluorescence stops increasing. Single muscle fibers isolated from transplanted scid/blAJ mice which received the AON-treated blood-derived CD133+ stem cells showed similar behavior after the laser wounding assay compared with the results with not-treated dystrophic scid/blAJ mice (data not shown). We thus administered a single injection of 2 × 105 engineered LV-FL DYSF blood-derived CD133+ stem cells isolated from patients 1 and 2 into the TA of 5-month-old scid/blAJ mice (two groups of n =5 for patient cell batch). At 1 month after the injection, human α-centromeric nuclei and human dysferlin double positive fibers were easily detected in all TA injected muscles but not in not-injected TA contralateral muscles (data not shown). Single fibers (n =100) from control c57Bl mice (n =5), untreated scid/blAJ mice (n =5) and transplanted scid/blAJ mice (n =5) were isolated and plated in dishes with Ca2+ NaCl/Pi. In all cases, a patch of fluorescence formed within 10 s after the damage at the lesion site (Fig. 6A). Full videos for treated and untreated mice can be downloaded from http://www.mediafire.com/download/9gao6v6no6t4gno/Scid-AJ_LV-DYSF.avi and http://www.mediafire.com/download/rf2bjcdjrffkkro/Scid-AJ_untreated.avi. It is known that the increase in signal intensity rapidly stops in wild-type dysferlin positive fibers, whereas it continues to augment in dysferlin negative fibers . Here we found that the increase of the signal was much lower in the engineered LV-FL DYSF blood-derived CD133+ stem cell transplanted fibers from the scid/blAJ mice than from the untreated mice (P <0.0001); notably, the trend of the curve for treated samples resembled that of control samples (Fig. 6A). For all experiments, we plotted the intensity of fluorescence coming into the fibers versus seconds. After the laser damage experiments single fibers were collected and total mRNA was extracted. RT-PCR analysis was performed to analyze dysferlin expression in treated samples. We confirmed that only the single fibers which expressed dysferlin were able to repair the membrane damage after laser wounding, while the absence of dysferlin expression correlated with the absence of differences in fluorescence intensity (Fig. 6B).
Autosomal recessive forms of muscular dystrophy include clinically divergent LGMD-2B and distal MM. Although distinct in terms of weakness onset pattern, both disorders arise from defects in the gene encoding dysferlin . Since the discovery of the dysferlin gene in 1998 , mutational analyses in dysferlinopathy patients have revealed mostly single nucleotide changes, without mutational hotspots or any obvious correlation between genotype and distribution of muscle weakness . This wide spectrum of identified DYSF mutations increases the difficulty of finding feasible treatments for these diseases. While the exon skipping technique has opened interesting new avenues for DMD treatment [14, 16, 17, 45], it has been clearly demonstrated that several dysferlin exons could not be skipped as their suppression would be undoubtedly deleterious [46, 47]. Surprisingly, in 2006, Sinnreich et al. reported a positive genotype–phenotype correlation in an LGMD-2B family with two severely affected sisters and a mildly affected mother. Furthermore, the mother carried a lariat branch point mutation in intron 31 that allowed an in-frame skipping of exon 32 . Taking these findings as a proof of principle for the feasibility of skipping exon 32, Wein et al. demonstrated in a dysferlinopathic patient that exon 32 was efficiently skipped using AONs without consequences on protein function .
In the present study, we investigated this approach in dysferlinopathies by first isolating blood-derived CD133+ stem cells from two patients: patient 1 with two mutated alleles (deletions in exon 22 and in exons 25–29) and patient 2 with a homozygous deletion of exon 55. To verify the feasibility of the exon skipping approach in patient 1, we constructed several dysferlin deletion cDNAs (Δ22–23, Δ25–29, Δ22–29). We found that all exons between 22 and 29 could be removed safely. In particular, it was necessary to remove exons 22–23 to restore an ORF. Unfortunately, the exon skipping efficiency was low in myoblasts and blood-derived CD133+ stem cells and did not lead to the in vivo expression of dysferlin after intramuscular transplantation of AON-treated blood-derived CD133+ stem cells. Moreover, the functional assays performed on dystrophic scid/blAJ muscles treated with AON-treated blood-derived CD133+ stem cells confirmed the failure of this approach. To enhance skipping efficiency, additional AONs encompassing larger and/or different target regions should be tested, individually or in combination with the AONs used here.
A gene transfer approach could be an alternative route for treating recessive diseases such as dysferlinopathy. Millay et al. showed that replacement of the dysferlin gene in AJ mice completely rescued muscle pathology and fully restored muscular force . The AAV vector is the most commonly used viral vector in muscle gene therapy; however, there are major technical issues associated with using AAV in therapeutic approaches, including the limited packaging size of AAV vectors, which is below the size of dysferlin mRNA. Interestingly, Grose et al. recently described AAV5 dysferlin delivery as a promising therapeutic approach that could restore functional deficits in dysferlinopathic patients . As is well known from previously published work, dysferlin exons 51–55 encode the C-terminal part of the protein, which has a fundamental role in anchoring dysferlin at the membrane and cannot be eliminated without impairing protein function . Here, we developed a strategy based on delivery of full-length dysferlin by lentiviral vector to allow dysferlin expression in both patients 1 and 2. We created a lentivirus carrying the full-length dysferlin gene (LV-FL DYSF) and we demonstrated the expression of dysferlin from transduced blood-derived CD133+ stem cells isolated from both patients. After transplantation of these cells into scid/blAJ mice, we detected the expression of dysferlin mRNA and protein in murine muscles; moreover, we noted normal resealing activity of isolated single fibers from transplanted scid/blAJ mice compared with the non-transplanted ones. Unfortunately, this effect was not sufficient to restore the dystrophic murine phenotype in vivo. A recent study by Lostal et al. showed that myoferlin and a mini-dysferlin could compensate for dysferlin deficiency in an in vitro assay of sarcolemmal repair; however, they could not rescue in vivo muscular defects associated with dysferlin absence . These findings suggested that muscular pathology in dysferlinopathies may be related not only to a membrane fusion defect but also to vesicle trafficking and inflammation. From this point of view, combining cell therapy (to replace dystrophic fibers and reduce the inflammatory environment) with gene therapy (to prevent further muscular damage) could represent a promising approach to treat dysferlinopathies. Moreover, the use of human stem cells in our in vivo animal studies obliges us to select as target animal model the scid/blAJ mice that show a less severe phenotype than the A/J dysferlin-null mice. The percentage of dystrophin positive fibers obtained after blood-derived CD133+ stem cells in scid/blAJ mice, however, was lower than the percentage of dystrophin positive fibers usually observed following the transplantation of those cells in scid/mdx mice [31, 32]. This result suggested that blood-derived CD133+ stem cell transplantation is less efficient in scid/blAJ mice than in scid/mdx. We argued that the cells needed the presence of a strong dystrophic environment to rescue the dystrophic features. However, the cells have an advantage in comparison with other sources of myogenic stem cells that possess muscle regenerating capacities, such as mesoangioblasts: they are easily isolated from the blood of dystrophic patients at different times representing an accessible tool to develop and test novel therapeutic approaches with major impact on research costs and number of treated patients. Our present study provides a proof of principle of the feasibility of using engineered CD133+ blood stem cells in dysferlinopathy treatment. Further studies will be needed to better elucidate the efficacy of our approach with increasing number of transplanted cells and different methods of delivery [50-52]. However, the present results contribute to improve translational research on the treatment of dysferlinopathies using stem cells engineered with a LV-FL DYSF, such as future testing of other cell types that possess robust myogenic potency.
Materials and methods
Human samples were collected after obtaining signed informed consent from all participants according to the guidelines of the Committee on the Use of Human Subjects in Research of the Fondazione IRCCS Ca' Granda Ospedale Maggiore Policlinico di Milano (Milan, Italy). This study was approved by the ethics committee of the University of Milan, Italy (CR937-G), which also authorized the use of human blood and muscle tissues. CD14+ cells were isolated from two patients who suffered from MM. Genomic DNA was extracted from peripheral blood lymphocytes by standard procedures . In the first patient we found a 4-bp deletion in exon 22 leading to a premature codon stop (c.2077delC, p.His693Thrfs*), and in the second patient we found a homozygous deletion of exon 55 (c.6233_6240delCCTTCAGC, p.Pro2078Leufs*92). Total RNA was extracted with TRIzol reagent (Invitrogen Life Technologies, Carlsbad, CA, USA) following the manufacturer's protocol. The cDNAs were prepared using Superscript III First Strand Reverse Transcriptase (Invitrogen Life Technologies, Carlsbad, CA, USA) following the recommendations of the manufacturer. To identify the large deletion in patient 1, we performed RT-PCR using a forward primer located in exon 20 and a reverse primer located in exon 31. Internal primers overlapping abnormal junction 24–30 were used to amplify the mRNA carrying a large deletion. All other primers used are available upon request. Sequencing reactions were performed with a dye terminator procedure, using a capillary automatic sequencer CEQ™ 8000 (Beckman Coulter, Brea, CA, USA) according to the manufacturer's recommendations. The large deletion was delimited using Genomic Quantitative Real-Time PCR with specifically designed TaqMan® or SYBR-GREEN® assays (Applied Biosystems, Foster City, CA, USA) as previously reported by Krahn et al. . Sequence variations are described according to den Dunnen's recommendations: http://www.hgvs.org/mutnomen.
Dysferlin plasmid construction
As proof of principle of exon skipping feasibility, we tested whether the skipped dysferlin product (Δ22–23) could be correctly expressed. From the GFP-DFL plasmid (a kind gift from K. Bushby) – which contains the complete dysferlin cDNA cloned into the pcDNA4/TO/MycHIS vector  – we subcloned an EcoRI-AfeI fragment into a smaller vector (p-Shuttle) to allow PCR amplification using the following primer combinations: HdysF21-R/Hdysf24-F and Hdysf24-R/Hdysf30-F (Table 1). After verification of deletion by enzyme restriction and direct sequencing, the purified fragment EcoRI-AfeI was cloned back into the initial vector by homologous recombination, resulting in a plasmid encoding a Δ22–23 dysferlin protein fused to EGFP at the N-terminal. Additional plasmids encoding other truncated forms of dysferlin (Δ25–29, Δ22–29) were produced using the same approach. Dysferlin plasmid GFP-DFL contained unique cutting sites for both EcoRI and AfeI, before the ATG initiation codon and in exon 35, respectively. PCR products were ligated with the Quick Ligase Kit (Promega, Madison, WI, USA) and transformed in XL10-Gold Ultracompetent Cells (Invitrogen Life Technologies, Carlsbad, CA, USA). Isolated clones were digested with HindIII and XbaI restriction enzymes, and the clones showing the expected digestion pattern were sequenced to verify correct deletion and absence of mutations.
Table 1. List of primers
By homologous recombination in BJ5183 bacteria, deleted fragments were cloned back into the original plasmid. Briefly, the original plasmid was cut with BstEII (two different cutting sites), and the deleted insert was cut with EcoRI and AfeI. Both linear fragments were purified on agarose gel (Promega). We mixed 26 ng plasmid and 90 ng insert with BJ5183 bacteria and performed transformation by thermal shock. Several clones were amplified by bacterial mini-culture, and plasmid was extracted using the plasmid DNA extraction kit (Macherey-Nalgen, Duren, Germany). To obtain a sufficient quantity of plasmid DNA, each clone was used to transform XL10-Gold bacteria; plasmid DNA was then verified by enzymatic digestion and sequenced. This unusual method was applied because strong plasmid recombinations occurred with classical ligation procedures.
HEK transfection with dysferlin plasmids
HEK cells (LGC Standards, Queens road Teddington, Middlesex, Oly, UK) were maintained in DMEM supplemented with 10% fetal bovine serum and penicillin-Streptomyces. HEK cells were seeded at 1.5 × 105 cells per well in a 24-well plate, grown for 20 h and transfected with pEGFP dysferlin plasmids. The transfection mixtures for each sample contained 2 μL of Lipofectamine (Invitrogen Life Technologies, Carlsbad, CA, USA) and 1.5 μg pEGFP dysferlin plasmid in a 100-μL total volume of DMEM (Invitrogen Life Technologies, Carlsbad, CA, USA) without antibiotics and fetal bovine serum. FACS and western blot analyses were performed 48 h after transfection.
A Cytomics FC500 flow cytometer and cxp 2.1 software (Beckman Coulter, Brea, CA, USA) were used to visualize 3 × 104 GFP and HEK cells. Each analysis included at least 10 000–20 000 events for each gate. A light-scatter gateway was set up to eliminate cell debris from the analysis. The percentage of positive cells was assessed after correction for the percentage reactive to an isotype control conjugated to fluorescein isothiocyanate. Transfection was highly efficient for all tested plasmids, as demonstrated by the expression of the GFP reporter gene.
RT-PCR and western blot analyses
Truncated dysferlin Δ22–23, Δ25–29 and Δ22–29 isoforms mimicking the skipped allele of our patient were detected by RT-PCR and western blot analyses. Total RNA was extracted from cells, muscles and single fibers using TRIzol reagent. First-strand cDNA was prepared as previously described . To detect dysferlin mRNA, nested RT-PCR was carried out with 1 μg cDNA. For the first amplification, the final mix (Invitrogen Life Technologies, Carlsbad, CA, USA) consisted of 1× Taq buffer, 1.5 mm MgCl2, 0.2 mm dNTP mix, 2.5 units Platinum Taq DNA polymerase and 0.2 mm Hex20-F and 0.2 mm Hex26-R for full-length dysferlin and Δ22–23, and not treated or Hex20-F and 0.2 mm of Hex31-R for Δ25–29 and Δ22–29 (Table 1). PCR conditions were as follows: 36 cycles of 94 °C for 1 min, 58 °C for 1 min and 72 °C for 1 min. PCR products were purified using the Jetquick PCR Product Purification Spin Kit (Genomed, Lohne, Germany). The second round of amplification used 5 μL of purified PCR product and was performed only for full-length dysferlin and not treated; the same conditions were used as in the first-round PCR except for the MgCl2 (1.5 mm) and the choice of the primers: Hex21-F and Hex24-R (Table 1). All PCR products were analyzed on 2% agarose gels. Obtained bands were acquired using the UVIsave Imaging System (UVItec Ltd, Cambridge, UK). HEK cells were lysed directly in 1× sample buffer (1% SDS) with addition of commercially available cocktails of protease and phosphatase inhibitors (Complete and PhosSTOP; both from Roche, Mannheim, Germany). Total protein concentration was determined according to Lowry's method. Samples were resolved on 12% polyacrylamide gel and transferred to supported nitrocellulose membranes (Bio-Rad Laboratories, Hercules, CA, USA), and the filters were saturated in blocking solution (10 mm Tris, pH 7.4, 154 mm NaCl, 1% BSA, 10% horse serum, 0.075% Tween-20). Filters were incubated overnight at 4 °C with anti-dysferlin IgG (1 : 200; Novocastra, Newcastle, UK). Detection was performed with horseradish peroxidase conjugated secondary antibodies (DakoCytomation, Carpinteria, CA, USA), followed by enhanced chemiluminescence development (Amersham Biosciences, Piscataway, NJ, USA). Pre-stained molecular weight markers (Bio-Rad Laboratories, Hercules, CA, USA) were run on each gel. Bands were visualized by autoradiography using Amersham Hyperfilm™ (Amersham Biosciences, Piscataway, NJ, USA).
AON design and efficiency tests
Acceptor/donor splice sites and exon splicing enhancer sequences located in exon 22 and 23 were used as targets for AONs. Normal human cells were transfected with Nucleofector® using 10 or 20 μg of AONs, individually or in combinations. After 24–48 h, cells were harvested and RNA was extracted. For each transfection condition, 1 μg RNA was retrotranscribed, and 80 ng cDNA was used as the matrix for the first PCR program. Primers were located in exons 20 (fwd) and 26 (rev) in order to amplify shorter fragments carrying an exon 22–23 deletion or larger ones, as sometimes observed in previous studies [16, 37]. A second semi-nested PCR, using an internal forward primer located in exon 21 and the same reverse primer (exon 26), was performed using 2 μL of PCR product from the first PCR. The skipping efficiency of the tested AONs was very low and observed only with a high number of cycles in the nested RT-PCR (70 cycles). For AON sequences, see Table 2.
Table 2. AON sequences
Targeted site, exon
3′ (gt), ex22
5′ (ag), ex22
3′ (gt), ex23
5′ (ag), ex23
5′ (ag), ex23
Blood-derived CD133+ isolation
Blood CD133+ cells were collected from peripheral blood of the two patients suffering from MM. CD133+ stem cells were isolated as previously described . After determining the purity of the CD133+ cells through cytofluorimetric analysis, they were plated in a proliferation medium  at a density of 15 × 104 cells·cm−2.
Lentivector carrying the full-length dysferlin
Considering the deep optimization needed to overcome exon skipping difficulties, we developed a parallel strategy based on complete dysferlin delivery by lentivirus vector. We subcloned the complete dysferlin transcript into a lentivirus pRRL backbone (LV-DYSF) and tested its efficiency in our two patients. Complete dysferlin cDNA was cut with EcoRI and subcloned into pRRL-cPPT-hPGK-eGFP-WPRE constructs . Complete dysferlin was amplified from plasmid pEGFP dysferlin, using the following primers: DYSF-atg-NheI-Fw, 5′-ATTCGCTAGCATGCTGAGGGTCTTCATCCTCT-3′, and DYSF-TGA-NheI-Rv, 5′-ATTCGCTAGCTCAGCTGAGGGCTTCACCAGC-3′. The PCR product woas digested with NheI and subcloned into pRRL-cPPT-hPGK-eGFP-WPRE vector. The transduction efficiency of lentiviral vectors containing eGFP and the dysferlin cassette was tested using the same multiplicity of infection as previously described [40, 41]. Both lentiviral vectors showed the same transduction efficiency. To enlarge the vector tropism, lentiviral vectors were pseudotyped with the VSV-G virus and generated by transfection of the plasmid pCMVΔR8.74 into 293T cells, as previously described . The infectious particle titer (ip·mL−1) was determined by quantitative real-time PCR using genomic DNA of transduced cells as described elsewhere . Constructs were sequenced on an ABI3130xl Genetic Analyzer (Applied Biosystems, Carlsbad, CA, USA) and analyzed with sequencer software (Gene Codes Corporation, Ann Arbor, MI, USA).
Lentiviral vector transduction
Dystrophic blood-derived CD133+ cells were transduced using 107–108 ip·mL−1. In 96-well tissue culture dishes, 2–4 × 104 cells were plated per well; then we added 100 μL of DMEM supplemented with 10% fetal bovine serum. Four hours post-transduction, 200 μL medium was also added to each well. The dishes were incubated for 24 h at 37 °C and 5% CO2, followed by washing and in vitro studies or in vivo transplantation. The expression of full-length dysferlin mRNA was demonstrated using RT-PCR analysis as previously described. Primers (Hdysf41-F/Hdysf56-R) are listed in Table 1.
Transplantation of the blood-derived CD133+ cells into immunodeficient scid/blAJ mice
A breeding colony of homozygous scid/blAJ mice was previously established . Their maintenance was authorized by the National Institute of Health and Local Committee, protocol number 10/10-2009/2010. All mice were fed ad libitum and allowed continuous access to tap water. Five-month-old scid/blAJ mice were deeply anesthetized with 2% avertin (0.015 mL·kg−1) prior to sacrifice by cervical dislocation, and all efforts were made to minimize suffering. Human engineered CD133+ cells isolated from dystrophic blood (2 × 105 cells in 7 μL NaCl/Pi) were injected into the right TA muscle of seven mice (six pre-treated with ntx) as previously described . One month after injection, muscle tissues were removed, frozen in liquid nitrogen-cooled isopentane and cryostat-sectioned.
Fluorescent in situ hybridization (FISH) and immunofluorescence analyses on transplanted muscles
Transplanted human cells were detected using a Texas Red-labeled human α-centromeric probe. FISH analysis was performed on frozen muscle sections. The slides were first treated for 30 min with Histochoice Tissue Fixative (Sigma-Aldrich, St. Louis, MO, USA) and sections were dehydrated in 70%, 80% and 95% alcohol. The denaturation was performed with 70% deionized formamide in 2× NaCl/Cit, and the slides were dehydrated again at −20 °C. The hybridization step was performed overnight at 37 °C. A Texas Red-labeled human α-centromeric (Exiqon A/S, Vedbaek, Denmark) probe was used to identify human cells. Nuclei were counterstained with 4′,6-diamidino-2-phenylindole (DAPI). Non-transplanted human and mouse muscle sections were used as positive and negative controls. Slides were observed using a Leica TCS SP2 confocal microscope (Leica, Hessen Wetzlar, Germany). A series of 10-μm transverse cryosections were cut over the injected scid/blAJ muscle length and were examined by immunohistochemical and hematoxylin and eosin staining. Dysferlin positive myofibers were detected with monoclonal anti-dysferlin IgG (NCL-DYSF; Novocastra) at a final dilution of 1 : 50, as previously described .
Nested RT-PCR and western blot for evaluation of full-length dysferlin expression
To verify the in vivo expression of full-length dysferlin, a series of 60 10-μm transverse cryosections from throughout the injected muscle length were collected in 1.5-mL Eppendorf tubes and mixed with 800 μL TRIzol reagent. cDNA was obtained as described above. To detect human dysferlin mRNA, nested RT-PCR was carried out with 1 μg cDNA using a final mix (Invitrogen Life Technologies, Carlsbad, CA, USA) of 1× Taq buffer, 1.5 mm MgCl2, 0.2 mm dNTP mix, 2.5 units Platinum Taq DNA polymerase, 0.2 mm Hex20-F and 0.2 mm Hex26-R. The program included 38 amplification cycles of 94 °C for 2 min, 92 °C for 1 min, 52.8 °C for 2 min and 72 °C for 2 min. PCR products were purified using the Jetquick PCR Product Purification Spin Kit (Genomed, Lohne, Germany) and a second round of amplification was performed using 5 μL of purified PCR product and the same conditions except for the MgCl2 (1.5 mm) and the primers (Hex21-F and Hex24-R). PCR conditions were as follows: 94 °C for 1 min, 65 °C for 1 min and 72 °C for 1 min (36 cycles). The PCR products were analyzed on 2% agarose gel. For western blot analysis, transplanted and non-transplanted muscles were homogenized with an electric homogenizer using a 1% Nonidet P-40 detergent buffer containing 20 mm Tris, pH 8, 137 mm NaCl, 2 mm EDTA, 10% glycerol and Complete and PhosSTOP cocktails (Roche, Basel, Switzerland). Total protein concentration was determined according to Lowry's method. Samples were resolved on 12% polyacrylamide gel as previously described.
Membrane injury and membrane repair monitoring
Human CD133+ cells were isolated from patients' blood. Before they were engineered with LV-FL DYSF, CD133+ cells from both patients were pooled, as the patients' mutations did not affect the expression of full-length dystrophin. We injected high numbers of cells in order to obtain a higher number of transduced fibers, as LV-FL DYSF could not be visualized. We transplanted 2 × 105 cells into the right TA muscle of five scid/blAJ mice. Membrane repair assay after laser wounding was conducted as previously described [7, 59]. Briefly, single muscle fibers (n =100) were isolated from the TA, 1 month after the injection, by treatment with collagenase I. Fibers were washed and resuspended in Dulbecco's NaCl/Pi containing 1 mm Ca2+ (Invitrogen Life Technologies, Grand Island, NY, USA). Single fibers were mounted on a glass slide chamber, and membrane damage was induced in the presence of FM 1-43 dye (2.5 mm) (Invitrogen Life Technologies, Grand Island, NY, USA) using a two-photon excitation technique with a Chameleon Ultra II (Coherent, Santa Clara, CA, USA) Ti : sapphire infrared laser source, directly coupled to the scanning head of a Leica TCS SP5 confocal microscope. To induce damage, a 5-μm2 area of the sarcolemma on the surface of the muscle fiber was irradiated at full power for 6.4 s at t =20 s. Images were captured at 10-s intervals, beginning 20 s before (t =0) and for 3 min after the irradiation. For every image taken, the fluorescence intensity at the site of the damage was measured with imagej imaging software (De Novo Software, Los Angeles, CA, USA).
This work was supported by the Association Monégasque contre les Myopathies (AMM), the Associazione La Nostra Famiglia Fondo DMD Gli Amici di Emanuele, the Associazione Amici del Centro Dino Ferrari, EU's Framework programme 7 Optistem 223098 and Provincia di Trento Fondo 12-03-5277500-01. We thank Nicolas Lévy for technical assistance and helpful discussion. No conflicts of interest exist.