If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Address reprint requests to Karl Rouger, Ph.D.; correspondence to Karl Rouger, Ph.D., or Yan Chérel, Ph.D., INRA, UMR 703, École Nationale Vétérinaire, Agroalimentaire et de l'Alimentation Nantes-Atlantique (Oniris), Route de Gachet, B.P. 40706, 44307 Nantes, France
INRA, UMR 703 “Développement et Pathologie du Tissu Musculaire,” Nantes, FranceLUNAM Université, École Nationale Vétérinaire, Agro-alimentaire et de l'alimentation Nantes-Atlantique (Oniris), Nantes, France
INRA, UMR 703 “Développement et Pathologie du Tissu Musculaire,” Nantes, FranceLUNAM Université, École Nationale Vétérinaire, Agro-alimentaire et de l'alimentation Nantes-Atlantique (Oniris), Nantes, France
INRA, UMR 703 “Développement et Pathologie du Tissu Musculaire,” Nantes, FranceLUNAM Université, École Nationale Vétérinaire, Agro-alimentaire et de l'alimentation Nantes-Atlantique (Oniris), Nantes, France
INRA, UMR 703 “Développement et Pathologie du Tissu Musculaire,” Nantes, FranceLUNAM Université, École Nationale Vétérinaire, Agro-alimentaire et de l'alimentation Nantes-Atlantique (Oniris), Nantes, France
INRA, UMR 703 “Développement et Pathologie du Tissu Musculaire,” Nantes, FranceUnité “Histopathologie Humaine et Modèles Animaux,” Département Infection et Epidémiologie, Institut Pasteur, Paris, France
INRA, the UMR 707 “Immunologie-Endocrinologie Cellulaire et Moléculaire,” Nantes, FranceLUNAM Université, École Nationale Vétérinaire, Agro-alimentaire et de l'alimentation Nantes-Atlantique (Oniris), Nantes, France
INRA, UMR 703 “Développement et Pathologie du Tissu Musculaire,” Nantes, FranceLUNAM Université, École Nationale Vétérinaire, Agro-alimentaire et de l'alimentation Nantes-Atlantique (Oniris), Nantes, France
INRA, UMR 703 “Développement et Pathologie du Tissu Musculaire,” Nantes, FranceLUNAM Université, École Nationale Vétérinaire, Agro-alimentaire et de l'alimentation Nantes-Atlantique (Oniris), Nantes, France
INRA, UMR 703 “Développement et Pathologie du Tissu Musculaire,” Nantes, FranceLUNAM Université, École Nationale Vétérinaire, Agro-alimentaire et de l'alimentation Nantes-Atlantique (Oniris), Nantes, France
INRA, UMR 703 “Développement et Pathologie du Tissu Musculaire,” Nantes, FranceLUNAM Université, École Nationale Vétérinaire, Agro-alimentaire et de l'alimentation Nantes-Atlantique (Oniris), Nantes, France
INRA, UMR 703 “Développement et Pathologie du Tissu Musculaire,” Nantes, FranceLUNAM Université, École Nationale Vétérinaire, Agro-alimentaire et de l'alimentation Nantes-Atlantique (Oniris), Nantes, France
INRA, UMR 703 “Développement et Pathologie du Tissu Musculaire,” Nantes, FranceLUNAM Université, École Nationale Vétérinaire, Agro-alimentaire et de l'alimentation Nantes-Atlantique (Oniris), Nantes, France
INRA, UMR 703 “Développement et Pathologie du Tissu Musculaire,” Nantes, FranceLUNAM Université, École Nationale Vétérinaire, Agro-alimentaire et de l'alimentation Nantes-Atlantique (Oniris), Nantes, France
INRA, UMR 703 “Développement et Pathologie du Tissu Musculaire,” Nantes, FranceLUNAM Université, École Nationale Vétérinaire, Agro-alimentaire et de l'alimentation Nantes-Atlantique (Oniris), Nantes, France
INRA, UMR 703 “Développement et Pathologie du Tissu Musculaire,” Nantes, FranceLUNAM Université, École Nationale Vétérinaire, Agro-alimentaire et de l'alimentation Nantes-Atlantique (Oniris), Nantes, France
Address reprint requests to Karl Rouger, Ph.D.; correspondence to Karl Rouger, Ph.D., or Yan Chérel, Ph.D., INRA, UMR 703, École Nationale Vétérinaire, Agroalimentaire et de l'Alimentation Nantes-Atlantique (Oniris), Route de Gachet, B.P. 40706, 44307 Nantes, France
INRA, UMR 703 “Développement et Pathologie du Tissu Musculaire,” Nantes, FranceLUNAM Université, École Nationale Vétérinaire, Agro-alimentaire et de l'alimentation Nantes-Atlantique (Oniris), Nantes, France
Duchenne muscular dystrophy (DMD) is a genetic progressive muscle disease resulting from the lack of dystrophin and without effective treatment. Adult stem cell populations have given new impetus to cell-based therapy of neuromuscular diseases. One of them, muscle-derived stem cells, isolated based on delayed adhesion properties, contributes to injured muscle repair. However, these data were collected in dystrophic mice that exhibit a relatively mild tissue phenotype and clinical features of DMD patients. Here, we characterized canine delayed adherent stem cells and investigated the efficacy of their systemic delivery in the clinically relevant DMD animal model to assess potential therapeutic application in humans. Delayed adherent stem cells, named MuStem cells (muscle stem cells), were isolated from healthy dog muscle using a preplating technique. In vitro, MuStem cells displayed a large expansion capacity, an ability to proliferate in suspension, and a multilineage differentiation potential. Phenotypically, they corresponded to early myogenic progenitors and uncommitted cells. When injected in immunosuppressed dystrophic dogs, they contributed to myofiber regeneration, satellite cell replenishment, and dystrophin expression. Importantly, their systemic delivery resulted in long-term dystrophin expression, muscle damage course limitation with an increased regeneration activity and an interstitial expansion restriction, and persisting stabilization of the dog's clinical status. These results demonstrate that MuStem cells could provide an attractive therapeutic avenue for DMD patients.
Duchenne muscular dystrophy (DMD) is a progressive, fatal, X-linked recessive disorder of skeletal and cardiac muscles. It is the most common muscular dystrophy, affecting one in 3500 male births,
Based on this feature, they appeared as a natural candidate for DMD cell therapy. Several studies revealed that the transfer of myoblasts (ie, in vitro descendants of activated satellite cells) could restore dystrophin-expressing myofibers in X-linked muscular dystrophy (mdx) mice and DMD patients.
which demonstrated a robust participation in muscle regeneration and satellite cell pool re-population, revealing that in vitro expansion highly contributes to the impaired engraftment capability of satellite cells. Based on their self-renewal and differentiation ability into different specialized cell types, including myogenic cells, the characterization of adult stem cells in a large number of tissues has led to new proposals of cell-based therapy approaches for genetic diseases such as DMD. These stem cells included side population (SP) cells,
Intramuscular or intra-arterial injection of genetically corrected CD133+ cells, isolated from peripheral blood or muscles of DMD patients, resulted in significant recovery of muscle morphology, function, and dystrophin expression in scid/mdx mice.
Flow cytometric characterization of myogenic cell populations obtained via the preplate technique: potential for rapid isolation of muscle-derived stem cells.
More recently, studies confirmed that adult skeletal muscle contains nonadherent stem cells that are capable in vivo to contribute to the repair of injured muscle.
In this report, we describe the characterization and the potential clinical use of a poorly adherent muscle-derived cell type that we called MuStem cells (muscle stem cells). These cells, isolated from dog skeletal muscle after serial replatings, were defined by an extensive proliferation capacity associated with atypical division modalities by generating two morphologically distinct cells. They had an ex vivo multilineage differentiation potential even though they appeared to be committed to the myogenic lineage as evidenced by their ability to spontaneously differentiate into myotubes. In the GRMD dog, which represents the clinically relevant animal model for DMD,
we showed that MuStem cells can regenerate muscle fibers, allowed dystrophin recovery, and relocated the satellite cell niche. When intra-arterially delivered, they contributed to a partial muscle tissue remodeling with an increase of the fiber regeneration activity and a limitation of the interstitial expansion. In addition, a striking and persistent clinical stabilization was reported for the transplanted GRMD dogs that were defined by an improved fatigability and a low intensity of limb stiffness and ankylosis. Altogether, these data reveal a potential therapeutic application for the MuStem cells.
Materials and Methods
Animals
GRMD dogs display an A→G mutation in the acceptor splice site of intron 6 of the dystrophin gene. Skipping of exon 7 disrupts the mRNA reading frame and results in premature termination of translation.
Golden retriever crossbred dogs from a GRMD colony maintained in the Boisbonne Center for Gene Therapy of Oniris, Nantes-Atlantic College of Veterinary Medicine, Food Sciences and Engineering were studied. Affected dogs, which have progressive clinical dysfunction similar to that of DMD boys, as previously described,
Development of a snapback method of single-strand conformation polymorphism analysis for genotyping Golden Retrievers for the X-linked muscular dystrophy allele.
The animal experiments were approved by the French National Institute for Agronomic Research and were performed according to the guidelines of the Institute. Investigations done in GRMD and healthy dogs are reported in Table 1.
Table 1Summary of Investigations Performed on Dogs
Dog number
Genotypic status
Age (onset of experiment)
Specific experiment
Nature of injected cells
Immune suppression
1 to 7
WT
2-month-old
MuStem cell and myoblast isolation
None
8
GRMD
2.5-month-old
IM injection, tissue distribution
nls-lacZ MuStem cells and myoblasts
Cyc A, MMF
9 to 11
GRMD
8-month-old
IM injection, tissue distribution
nls-lacZ MuStem cells
Cyc A, MMF
12, 13
WT
2.5-month-old
IM injection, tissue distribution
nls-lacZ MuStem cells
Cyc A, MMF
14, 15
GRMD
7-month-old
IF injection, tissue distribution
nls-lacZ MuStem cells
Cyc A, MMF
16
GRMD
2-month-old
IF injection, clinical follow-up
MuStem cells
Cyc A, MMF
17
GRMD
3-month-old
IF injection, clinical follow-up
MuStem cells
Cyc A, MMF
18
GRMD
4-month-old
IF injection, clinical follow-up
MuStem cells
Cyc A, MMF
19
GRMD
1.5-month-old
Clinical follow-up
Cyc A, MMF
20
GRMD
3-month-old
Clinical follow-up
Cyc A, Pred
21
GRMD
3-month-old
Clinical follow-up
Cyc A, Aza
22, 23
GRMD
3-month-old
Clinical follow-up
None
24
GRMD
1-month-old
Clinical follow-up
None
A sequential number defines different dogs. Their age at the onset of the experiments and genotypic status are mentioned (GRMD or WT, wild-type). Nature of injected cells and mode of delivery are indicated (IM, intra-muscular injection; IF, intra-femoral injection).
Muscle-derived cells were obtained independently from seven 2-month-old healthy dogs from a pool of hind limb muscles (gluteus medius and superficialis, semitendinosus, semimembranosus, biceps femoris, vastus lateralis and medialis, sartorius cranialis and caudalis, gracilis, tibialis cranialis, flexor digitorum superficialis, and gastrocnemius lateralis and medialis muscles), as previously described.
Cells were placed in a growth medium [44% DMEM (VWR, Strasbourg, France), 44% M199 (VWR), 10% fetal calf serum (Sigma, St. Louis, MO), 1% penicillin/streptomycin/fungizon (Sigma), and 1% l-glutamine (Sigma)], seeded at 105 cells/cm2 on gelatin-coated flasks (Sigma), and submitted to an adaptation of the preplating technique.
After 1 hour, floating cells were collected and replated on new flasks for 24 hours. This procedure was repeated daily for 4 days, after which time, floating cells were placed at 5.103 to 104 cells/cm2 in new flasks and maintained for another 3 days without medium change. Adherent cells were then expanded in medium (37% DMEM, 2.5 g/L glucose, 37% M199, 10% fetal calf serum, 10% horse serum, 1% penicillin/streptomycin/fungizon, 20 mg/mL insulin) containing human recombinant factors [10 ng/mL basic fibroblast growth factor, 50 ng/mL epidermal growth factor, and 25 ng/mL stem cell factor (PromoCell, Heidelberg, Germany)]. Myoblasts, corresponding to a pool of cells collected from preplatings 2 to 4, were expanded in growth medium.
In Vitro Proliferation Analysis
Clonal cultures were obtained by limiting dilution and were performed for MuStem cells and myoblasts that served as a control. After 8 days, clones were fixed in 4% paraformaldehyde (PFA) and incubated 1 hour at room temperature with mouse monoclonal antibody (mAb) against desmin (1:50; Dako, Glostrup, Denmark) or Pax7 [1:10; Developmental Studies Hybridoma Bank (DSHB), Iowa City, IA] in combination with biotinylated goat anti-mouse Ig (30 minutes, room temperature; Dako) that was revealed by peroxidase-diaminobenzidine staining (Dako). Proliferation was determined by counting the nuclei number in each desmin+ or Pax7+ single cell–derived colony stained with Giemsa. In addition, population doubling level was examined on four MuStem cell–derived primary cultures at each passage as previously described.
To assess the differentiation potential of MuStem cells, primary bulk cultures (ie, culture of all single cells) were maintained in standard growth medium until confluence, after which they were incubated in specific cell-type differentiation media. For myogenic differentiation, 10% fetal calf serum was replaced by 2% horse serum in medium. After 2 days, differentiation was assessed on the basis of cell morphology and the developmental isoform of myosin heavy chain (MyHCd) expression. Cultures were fixed in 4% PFA, treated with 0.5% Triton X-100/20% (w/v) goat serum in PBS, and incubated 1 hour with human MyHCd mAb (Novocastra Laboratories, Newcastle on Tyne, UK). Immunolabeling was revealed as described above. Osteogenic and adipogenic differentiation were induced and characterized as described previously.
For flow cytometry, four MuStem cell samples and three myoblast samples were resuspended in PBS/5% dog serum and incubated (30 minutes, 4°C) with fluorochrome-conjugated antibodies (Ab) to the following antigens: CD14, CD34, CD44, CD49d, CD62L, CD90 (BD Biosciences, Franklin Lakes, NJ), CD5, CD21, CD45 (AbD Serotec, Düsseldorf, Germany), CD56 (Dako), Bcrp1 (eBiosciences, Montrouge, France). CD11b (AbD Serotec) labeling was performed according to a classic two-step protocol using fluorochrome-conjugated secondary Ab (AbD Serotec). To validate labelings, preliminary experiments were conducted on canine peripheral blood cells and bone marrow cells. Surface antigens were evaluated in at least 200,000 viable cells using a FACSAria flow cytometer and analyzed using Diva v6 1.2 software (BD Biosciences). Isotype-matched Ab were used as negative controls for gating and analyses. For cytological immunolabelings on cytospin preparations and Lab-Tek chamber slides (Nalge Nunc International, Rochester, NY), three MuStem cell samples and three myoblast samples were fixed in 2% PFA (10 minutes) and treated with 0.5% triton X-100 (30 minutes), except for CD31 Ab. After incubation (1 hour, room temperature) in blocking buffer (2% goat serum in PBS), cells were incubated with Ab: CD31 (1:50; Dako), Pax7 (1:25; DSHB), Myf5 (1:500; Santa Cruz Biotechnology, Santa Cruz, CA), MyoD (1:25, Dako), desmin (1:50; Dako), and β1-integrin (1:50; DSHB) (1 hour, room temperature for CD31, Pax7, Myf5; overnight, 4°C for MyoD, desmin, β1-integrin). The slides were incubated with Alexa fluor 488 or 555 secondary Ab (1:500; Invitrogen, Carlsbad, CA) (1 hour, room temperature) and DRAQ5 red fluorescent cell-permeable DNA probe (Biostatus, Loughborough, UK) (15 minutes, room temperature). More than 550 cells were counted per sample for cytospin preparations, whereas at least 118 round cells and 214 spindle-shaped cells were considered for Lab-Tek chamber slides. Data were presented as the mean ± SD of independent experiments.
Retroviral Infection
Recombinant nuclear-localizing site nls-lacZ retroviral particles were used to label MuStem cells and myoblasts with a nuclear lacZ expression, as previously described.
A control of retroviral infection efficiency was performed by determining the percentage of lacZ+ nuclei (always more than 85%).
Immunosuppressive Treatment
GRMD dogs were immunosuppressed with 32 mg/kg/day of oral cyclosporine (Neoral; Novartis, Rueil-Malmaison, France) in combination with 6 mg/kg mycophenolate mofetil (CellCept; Roche, Paris, France). Ten mg/kg of ketoconazole (Nizoral; Janseen-Cilag, Issy-les-Moulineaux, France) was also added daily to decrease cyclosporine catabolism. Blood levels of cyclosporine were controlled twice a week and maintained between 250 and 300 ng/mL. The immunosuppressive regimen was started 1 week before cell administration and maintained throughout the experiment. One mock-treated GRMD dog received the same immunosuppressive regimen while the second received 2 mg/kg/day prednisolone (Megasolone; Merial, Lyon, France) in place of mycophenolate mofetil.
Intramuscular Injection
Gluteus superficialis muscle, triceps brachii muscle, and semitendinosus muscle of a 2.5-month-old GRMD dog (#8 in Table 2) were surgically exposed and injected with 2·106 viable nls-lacZ–transduced cells suspended in 250 μL of 0.9% NaCl/2.5% homologous serum: the left muscles received MuStem cells, whereas the right counterparts were injected with myoblasts. Alternatively, MuStem cells were injected in the triceps brachii muscle of three 8-month-old GRMD dogs (#9 to #11) and in the Biceps femoris muscle of two 2.5-month-old dogs (#12, #13). Four weeks later, injected muscles were biopsied.
Table 2Distribution of lacZ+ Nuclei in GRMD Dog Muscles after MuStem Cell Delivery
Dog number
Muscle
Number of lacZ+ nuclei
Tissue localization of lacZ+ nuclei
Subbasal position
Centronuclear position
Interstitial tissue
8
Gluteus superficialis
3205
2445 (76.3%)
320 (10.0%)
440 (13.7%)
Triceps brachii
2999
1865 (62.2%)
693 (23.1%)
441 (14.7%)
Semitendinosus
2540
1952 (76.9%)
351 (13.8%)
237 (9.3%)
Total
8744
6262
1364
1118
Percentage
71.6%
15.6%
12.8%
95% confidence interval
70.7–72.6
14.9–16.4
12.1–13.5
9
Triceps brachii
549
510 (92.9%)
38 (6.9%)
1 (0.2%)
10
Triceps brachii
641
598 (93.3%)
42 (6.5%)
1 (0.2%)
11
Triceps brachii
536
503 (93.8%)
32 (6.0%)
1 (0.2%)
Total
1726
1611
112
3
Percentage
93.3%
6.5%
0.17%
95% confidence interval
92.2–94.5
5.3–7.7
0–0.4
Dog number
Muscle
Number of lacZ+ nuclei
Tissue localization of lacZ+ nuclei
Below plasma membrane
Above basal membrane
Between both membrane
12
Biceps femoris
326
238 (73.0%)
35 (10.7%)
53 (16.3%)
13
Biceps femoris
314
217 (69.1%)
43 (13.7%)
54 (17.2%)
Total
640
455
78
107
Percentage
71.1%
12.2%
16.7%
95% confidence interval
67.6–74.6
9.7–14.7
13.8–19.6
Dog number
Muscle
Number of lacZ+ nuclei
Tissue localization of lacZ+ nuclei
Subbasal position
Centronuclear position
Interstitial tissue
14
Semitendinosus
127
100 (78.7%)
0 (0%)
27 (21.3%)
Gracilis
239
191 (79.9%)
0 (0%)
48 (20.1%)
15
Semitendinosus
104
67 (64.4%)
14 (13.5%)
23 (22.1%)
Gracilis
251
249 (99.2%)
0 (0%)
2 (0.8%)
Total
721
607
14
100
Percentage
84.2%
1.9%
13.9%
95% confidence interval
81.5–86.9
0.9–2.9
11.4–16.4
Tissue localization of lacZ+ nuclei was determined on several skeletal muscles of eight dogs, 4 weeks after MuStem cell injection. Six dogs received intramuscular injection (#8 to #13), whereas two others received intra-arterial delivery (#14 and #15).
MuStem cells were suspended at 2·106 cells/mL in 0.9% NaCl/2.5% homologous serum/50 U/mL heparin. A 2-cm-long segment of the femoral artery was surgically exposed through an inguinal incision and a 26-gauge catheter (1.9 cm long; Terumo, Leuven, Belgium) was totally inserted in a retrograde direction. Consequently, its extremity was not advanced as deep as the aortoiliac bifurcation, and so cells were consistently injected unilaterally in the left femoral artery. Five injections of 1·107 MuStem cells/kg and three injections of 2·107nls-lacZ–transduced MuStem cells/kg were performed respectively in three (#16 to #18) and two (#14, #15) GRMD dogs at 2- to 4-week intervals, using laminar flow at a rate of 5 mL/min. Intra-arterial injections were always performed on GRMD dogs aged from 2 to 6 months old.
Muscle Biopsy
Biopsies of nls-lacZ–transduced MuStem cell–injected muscles were divided into two parts for immunohistochemistry analysis (cryopreserved) and lacZ histochemistry (paraffin-embedded) using an in toto enzymatic technique, as previously described.
Small fragments (0.5 cm3) of biceps femoris and/or tibialis cranialis muscle were collected from healthy dogs, mock-immunosuppressed GRMD dogs, and MuStem cell–injected GRMD dogs at various time points and divided into two parts for histological and molecular analysis. Semitendinosus and gracilis muscle biopsies were done 8 weeks after the nls-lacZ–transduced MuStem cell systemic administration and processed as described above for nls-lacZ–transduced MuStem cell–injected muscles.
RT-PCR Analysis
Total RNA was isolated with the TRIzol method (Invitrogen) and transcribed into cDNA using a M-MLV (Moloney Murine Leukemia Virus) Reverse Transcriptase (Invitrogen) (1 hour, 37°C) and a primer specific for the canine dystrophin mRNA (5′-GTGATGATGTTGTTCTGATACTCCAGCCAG-3′). Because the first AG in exon 7 acts as an alternate acceptor splice site and generates a rare dystrophin transcript including exon 7 in which only the first five bases are missing,
a reaction specific for the wild-type canine dystrophin mRNA was performed using Ampli Tag Gold DNA polymerase (Ambion, Foster City, CA) with primers at the junction of exon 6/7 (5′-TCTCATCCACAGTCATAGGCCAG-3′) and in exon 9 (5′-AATGCTGTGAAGGAAGTGGGCTC-3′). The PCR cycle consisted of: initial denaturation (5 minutes, 95°C) followed by 40 cycles (30 seconds, 94°C; 30 seconds, 63°C; 1 minute, 72°C), and a final extension (10 minutes, 72°C). An internal control reaction was performed to detect the sequence of exon 1 to exon 3 (5′-GGGATCACTCACTTTCCCCTTAC-3′/5′-AAAGGTCTAGGAGGCGTCTCCC-3′). The PCR cycle was: initial denaturation (5 minutes, 95°C) followed by 40 cycles (30 seconds, 94°C; 30 seconds, 60°C; 1 minute, 72°C), and a final extension (10 minutes, 72°C). The reactions generated, respectively, a 455-bp amplicon and a 374-bp amplicon that were analyzed using agarose gel electrophoresis and ethidium bromide staining.
Immunohistochemistry
Transverse cryosections were incubated (overnight, 4°C) with the primary Ab against β-galactosidase (1:3000; Chemicon, Euromedex, Mundolsheim, France), dystrophin (1:50; Novocastra; 1:50; Santa Cruz Biotechnology), utrophin (1:50; Novocastra), β-sarcoglycan (1:50; Novocastra), γ-sarcoglycan (1:50; Novocastra), β-dystroglycan (1:50; Novocastra), MyHCd (1:100; Novocastra), Pax7 (1:10; DSHB), laminin (1:1000; Sigma). For triple immunolabelings, Alexa fluor (488, 555, or 633) conjugated goat anti-mouse or goat anti-rabbit IgG (1:300; Invitrogen) (1 hour, room temperature) were used. For CD4 (1:400; Serotec, Kidlington, UK), CD8 (1:400; Serotec), CD11b (1:300; Serotec), and CD79 (1:500; Dako), sections were fixed in acetone and 4% PFA, respectively, treated with 10% H2O2 in methanol (10 minutes, room temperature), blocked with buffer (0.2% PBS/Tween, 20% goat serum) (30 minutes, room temperature), and incubated (overnight, 4°C) with the primary Ab. The sections were incubated with biotinylated goat anti-mouse (1:300; Dako) or goat anti-rat IgG (1:400; Invitrogen) (1 hour, room temperature) and streptavidin horseradish peroxidase (15 minutes, room temperature) that was revealed using 3,3′-diaminobenzidine (DAB) chromogen (10 minutes, room temperature). For localization of lacZ+ nuclei, paraffin sections previously submitted to enzymatic technique were treated with 0.1% trypsin (10 minutes, room temperature), 3% H2O2 in methanol (10 minutes, room temperature), and with blocking buffer (0.2% PBS/Tween, 5% goat serum; 30 minutes, room temperature). Sections were incubated with rabbit polyclonal Ab against dystrophin (1:25; Chemicon) (1 hour, room temperature) followed by biotinylated goat anti-rabbit (1:300; Vector Laboratories, Burlingame, CA) (30 minutes, room temperature) and streptavidin horseradish peroxidase (30 minutes, room temperature) that was revealed using DAB chromogen (15 minutes, room temperature). Sections were then incubated with mouse mAb against laminin (1:500; DSHB) (1 hour, room temperature) followed by biotinylated goat anti-mouse (1:300; Vector Laboratories) (30 minutes, room temperature) and streptavidin alkaline phosphatase (30 minutes, room temperature) that was revealed using fuchsin (15 minutes, room temperature). Immunofluorescence labelings were observed with a laser scanning confocal microscope (Nikon C1; Champigny, France). For dystrophin labeling, all acquisitions were performed with the same signal amplification resulting from identical detector gain value. With this value, no fluorescent signal was detected on control slides corresponding to cell-injected GRMD dog muscle sections incubated with immunoglobulin isotype control or in GRMD dog muscle sections incubated with dystrophin mAb. Blinded examination of the dystrophin labeling was always performed by at least two persons. To determine the proportion of dystrophin+ fibers, a total of 1000 laminin+ fibers were counted in separate sections from the biceps femoris muscle and tibialis cranialis muscle of MuStem cell–injected GRMD dogs (n = 2), and the percentage of fibers expressing dystrophin was determined.
Histomorphometry
Biceps femoris muscle samples of 7-month-old dogs (healthy, mock-immunosuppressed GRMD and MuStem cell–injected GRMD; n = 3 per group) were processed in 8-μm-thick cryosections. Morphometric analysis was done using a digital camera (Nikon DXM 1200; Nikon Instruments, Badhoevedorp, the Netherlands) combined with image-analysis software (NIS; Nikon). Microscopic fields were randomly selected on hematoxylin-eosin-safranin–stained sections using intermediate magnification to observe at least 100 fibers (160 ± 31 per sample). The minimal Ferret diameter was used to determine fiber size distribution. Necrotic muscle fibers were determined on 10 high-magnification fields randomly selected on Gomori trichrome–stained sections and the percentage of necrotic fibers was calculated considering the total number of fibers. Fibrosis was determined as the ratio of areas rich in collagen on the total muscle area in an overall cross section, as described elsewhere.
Endomysial space thickness was measured among two high-magnification fields using Gomori trichrome staining. Foci of calcification, revealed by Alizarin Red staining, were measured on 10 low-magnification fields. To determine the percentage of MyHCd+ fibers, at least 500 fibers (640 ± 84) were numbered on two randomly selected microscopic fields. For each measurement, reproducibility was above 92%.
Immunoblotting
Membrane-enriched fraction (KCL-washed microsomes) was isolated from muscle biopsies by ultracentrifugation at 4°C, as previously described.
Protein concentration was determined using bicinchoninic acid protein assay (Pierce, Rockford, IL) with bovine serum albumin as standard. Proteins were separated by 6% SDS-polyacrylamide gel electrophoresis (PAGE) and transferred to a protran BA83 nitrocellulose membrane (Whatman, Maidstone, UK) by electroblotting with a Mini Trans-Blot Cell (Bio-Rad, Marne-la-Coquette, France). The membranes were blocked (overnight, 4°C) with Tris-buffered saline [20 mmol/L Tris-HCl, 500 mmol/L NaCl (pH 7.5)]/0.1% Tween 20/5% nonfat dry milk and incubated (3 hours, room temperature) with dystrophin mAb (1:20, DYS1; Novocastra) or with myosin mAb (1:2000, MF20; DSHB) in blocking buffer. After washes in TBS, the membranes were incubated (1 hour, room temperature) with Alexa Fluor 680 conjugated goat anti-mouse IgG (1:10,000 in blocking buffer; Invitrogen). The fluorescence emitted by the protein bands was monitored using the Odyssey Infrared Imaging system (Li-COR Biosciences, Lincoln, NE).
Clinical Follow-Up
A clinical evaluation was performed weekly by the same D.V.M. observer on GRMD dogs (n = 3), mock-immunosuppressed GRMD dogs (n = 2), and MuStem cell–injected ones (n = 3), using an extended version of a published grid.
The observer always followed the same protocol on animals walking around in a quiet room, and scoring items were always observed in the same order. For practical reasons, it was not possible to perform this evaluation blindly. In addition to the previously described 11 locomotion criteria, 6 items related to the general health status (dysphagia, ptyalism, hypertrophy of the base of the tongue, mouth opening, global activity, and breathing) were added. Each item was scored from 0 to 2, with 0 corresponding to a normal appearance, 1 to an intermediate phenotype, and 2 to a severe alteration. Data related to validation of the clinical evaluation method were already published
Cordazzo M-C:[Development of Tools for the Clinical Evaluation of Dogs Suffering from Muscular Dystrophy] (doctoral thesis). [Maisons-Alfort, France]: École Nationale Vétérinaire d'Alfort
and available at http://theses.vet-alfort.fr/telecharger.php?id=1015. The clinical score was expressed as the complement of a healthy dog score of 100% and a tendency curve (mobile means order 3) was built to represent the score evolution. Serum levels of creatine kinase and aspartate aminotransferase were measured weekly from 1 week before the first MuStem cell administration.
Statistics
All data were reported as means ± SD. Mean fiber size and endomysial thickness were compared among different dog groups with analysis of variance followed by Fisher PLSD tests and creatine kinase levels with analysis of covariance, using StatView software (Brain Power, Calabasas, CA). Means were compared using an unpaired Student's t-test for the size of colonies between myoblasts and MuStem cells. Percentages of MyHC+ fibers were compared between MuStem cell–injected and mock-immunosuppressed GRMD dogs using a Mann-Whitney test with a two-tailed P value. A value of P < 0.05 was considered to be statistically significant.
Results
MuStem Cells Exhibit High Proliferation Rate and Atypical Division Pattern
When healthy dog skeletal muscle–derived cells were grown in vitro, a marginal fraction of nonadherent cells (representing 1.2% ± 0.5% of total extracted cells; n = 7) was isolated among myoblasts that firmly adhered to the coated plastic. These refringent rounded cells, named MuStem cells, were isolated on day 4 using serial platings; they required three additional days to anchor slightly to a collagen matrix, and initially grew by forming microspheroid colonies. The colonies rapidly became composed of a large number of superposed cells and scattered to generate a majority of spindle-shaped cells while others remained round (Figure 1A). These two cell phenotypes were maintained after several passages (Figure 1B), with some round cells that divided into one round cell and one spindle-shaped myoblast (Figure 1B). Round cells represented 17.8% ± 1.1%, 10.2% ± 1.9%, and 10.6% ± 0.8% of all cells at passage 1 (P1), P3, and P6, respectively (n = 3500 cells counted per passage). Originally, when cultured under nonadherent condition, MuStem cells proliferated as clusters of rounded cells termed myospheres containing many hundreds of cells (Figure 1C). Myospheres maintained the ability to spontaneously give rise to a mixed population of spindle-shaped and round cells when replaced in an adherent condition (Figure 1C), demonstrating that MuStem cells adopt distinct behavior depending on the environment.
Figure 1Growth modalities of MuStem cells. MuStem cells were isolated from muscle-derived cells after six successive platings (n = 7, independent experiments). A: Morphological examination in phase contrast revealing progressive formation of microspheroid colonies that generated adherent spindle-shaped cells (arrow) and round cells (arrowhead). B: After several passages, cultures displayed spindle-shaped cells (arrow) and round cells (arrowhead). Right panels: The round cells divided by generating round cells and spindle-shaped ones (arrow). C: Under nonadherent conditions, MuStem cells proliferate by generating myospheres composed of hundreds of round cells. Replated on gelatin-coated matrix, myospheres gave rise to spindle-shaped cells (arrow) and poorly adherent round cells (arrowhead). D: Clonal culture of MuStem cells. Right panel: Colonies were characterized by the presence of the two morphologically distinct cell types. Scale bars: 10 μm (B, right panels); 25 μm (A); 50 μm (B, C, and D, right panel); and 100 μm (D).
Clonal culture analyses showed that MuStem cells displayed clonogenic ability (Figure 1D). The average nucleus number per colony was 360 ± 325 compared to 231 ± 265 for myoblasts after 8 days (n = 161 clones), indicating that MuStem cells have a higher proliferation capacity than myoblasts (P < 0.05). In addition, we showed that MuStem cells are able to make 20.4 ± 1.6 population-doubling levels in 36 days of primary culture without reaching senescence. Importantly, as described for the original primary cultures, presence of both spindle-shaped cells and round ones was detected in MuStem cell–derived colonies (Figure 1D), which demonstrated atypical division modalities for the MuStem cells.
MuStem Cells Are Mainly Early Myogenic Progenitors with Oligopotency
Fluorescence-activated cell sorting analysis and immunocytochemistry on cytospin preparation showed that 81% ± 4% and 59% ± 10% of the MuStem cells were positive for the satellite cell markers CD56 and β1-integrin, respectively; 46% ± 4% and 42% ± 3% of the cells expressed the paired box transcription factor Pax7 that is required for specification of myogenic cells and the early myogenic regulatory factor Myf5, respectively. Expression of the key regulator of myoblast differentiation MyoD and the intermediate filament desmin was detected in 49% ± 2% and 34% ± 4%, respectively (Figure 2A and data not shown). MuStem cells were uniformly negative for surface markers CD45 and CD34, typically expressed by hematopoietic and endothelial lineage cells, and for CD49d, CD62L, and Bcrp1. Adhesion molecule CD44 was detected in all MuStem cells, whereas 2% to 7.4% of cells consistently expressed the cell-surface glycoprotein Thy-1/CD90 (Figure 2A). Endothelial marker CD31 and blood lineage markers such as CD5, CD11b, CD14, and CD21 were not expressed by MuStem cells or myoblasts (data not shown). In addition, immunofluorescence analysis in fixed cultured cells showed that Pax7, Myf5, and MyoD were expressed by 73% ± 18%, 45% ± 15%, and 36% ± 7% of the round cells, whereas they were present in 56% ± 1%, 47% ± 4%, and 44% ± 5% of the spindle-shaped cells, respectively, revealing a mild expression for both cell types (Figure 2B). Compared with myoblasts (see Supplemental Figure S1 at http://ajp.amjpathol.org), these data demonstrate that MuStem cells mainly correspond to committed muscle cells at an early stage of the myogenic lineage.
Figure 2Phenotype of MuStem cells. A: Surface labeling for CD56, CD45, CD34, CD49d, CD62L, Bcrp-1, CD44, and CD90 was determined for expanded MuStem cells (n = 4). Shaded areas, control immunoglobulins; black lines, specific antibodies. RIgG2b and mIgG1 corresponded to control immunoglobulins for CD44 and CD90, respectively. B: Cells were cultured on Lab-Tek chambers slides and subjected to immunolabelings for myogenic markers. Expression of Pax7, MyoD, and Myf5 by round cells and spindle-shaped cells. Arrows and arrowheads showed positive and negative cells, respectively. Scale bar = 20 μm.
Using appropriate differentiation media, we demonstrate that MuStem cells are able to differentiate into myocytes, osteocytes, and adipocytes. After myogenic differentiation, MuStem cell–derived cultures displayed numerous multinucleated myotubes that were highly positive for the developmental isoform of MyHCd (Figure 3A). Osteogenic differentiation was demonstrated by the formation of multiple layers of dense cells, a large proportion of which became positive for alkaline phosphatase (ALP) and by massive calcium depositions as revealed by Alizarin Red staining (Figure 3B). After adipogenic induction, almost all cells presented extensive accumulation of small neutral lipid vesicles in their cytoplasm after staining with Oil Red O (Figure 3C).
Figure 3In vitro multilineage differentiation of MuStem cells. A: Myogenic differentiation. Before and 2 days after treatment with low serum medium, cells were labeled for MyHCd. B: Osteogenic differentiation. Before and 21 days after treatment with osteogenic medium, cells were stained with ALP and Alizarin Red for calcium deposition and mineralized nodules. C: Adipogenic differentiation. Before and 14 days after treatment with adipogenic medium, cells were stained with Oil Red O for lipid droplets (n = 2 per group). Scale bars: 100 μm (A and B); 50 μm (C).
MuStem Cells Participate in Muscle Fiber Formation and Restore Dystrophin
To determine whether MuStem cells could regenerate fibers in highly damaged muscles, nls-lacZ–transduced MuStem cells were injected into skeletal muscles of a 2.5-month-old, immunosuppressed GRMD dog. As a control, nls-lacZ–transduced myoblasts were injected into contralateral muscles. When analyzed 4 weeks later, each MuStem cell–injected muscle displayed many lacZ+ nuclei, which dramatically contrasted with the absence of lacZ+ nuclei in the myoblast-injected muscles (Figure 4A). The tissue distribution of the lacZ+ nuclei is presented in Table 2 (dog #8). The vast majority of the nuclei (71.6%) were found in a peripheral position, whereas the remaining ones were found either centrally located (15.6%) or in the endomysial tissue (12.8%) (Figure 4B). Similar results were obtained when MuStem cells were injected into the triceps brachii muscle of three 8-month-old GRMD dogs (#9 to #11). To precisely locate lacZ+ nuclei with a peripheral position, double immunolabeling of dystrophin and laminin was performed on MuStem cell–injected muscle of two 2.5-month-old dogs (#12, #13). We determined that 71.1% had a subplasma membrane position, 12.2% were found above the basal membrane, and 16.7% were found between the plasma and the basal membrane, ie, in the satellite cell niche (Figure 4C). Of interest is that Pax7 expression could be demonstrated for these latter nuclei, indicating that MuStem cells could acquire satellite cell identity (Figure 4D) and supplement the pool of endogeneous satellite cells in dystrophic context. To document the myogenic potential of the MuStem cells that did not fuse with host fibers (ie, those located in endomysial tissue or in satellite cell niche), nls-lacZ–transduced MuStem cells were injected in the Biceps femoris muscle of two 2.5-month-old dogs (#12, #13). Four weeks later, mononucleated cells were isolated from the injected muscles and seeded in primary culture. As shown in Figure 4E, lacZ+ nuclei were observed in several myotubes that resulted from their fusion with non-lacZ+ nuclei, demonstrating that MuStem cells in muscle-resident positions maintain their myogenicity. Altogether, these results provide strong evidence that MuStem cells are effective in muscle fiber formation, either directly by fusing with host fibers or by generating myogenic-resident cells.
Figure 4In vivo behavior of MuStem cells after intramuscular injection. Four weeks after intramuscular injection of nls-LacZ–transduced MuStem cells, muscles were biopsied and investigated. A: Kernechtrot stain of representative sections treated for lacZ expression. B: Tissue distribution analysis (n = 6 muscles on four dogs: #8 to #11) revealing different localization of lacZ+ nuclei: peripheral (left), centronuclear (middle), or in an interstitial (right) position. C: Immunolabelings for lacZ (red), plasma membrane (dystrophin+, green), and basal membrane (laminin+, blue) showing the presence of peripheral lacZ+ nuclei below the plasma membrane of fibers (left), above the basal membrane (middle), or between both membranes (right). D: LacZ+ nuclei (left, red) located above the plasma membrane (left, dystrophin+: green), was Pax7+ (middle, blue); merged image (right). E: Primary culture of cells isolated from muscle previously injected with MuStem cells was assayed for lacZ expression (blue) to reveal the presence of lacZ+ nuclei in myotubes. Scale bars: 50 μm (A); 10 μm (B and E: inset); 20 μm (C and D); 25 μm (E).
In addition, we determined that all fibers containing lacZ+ nuclei were dystrophin+ (Figure 5A) and also expressed γ-sarcoglycan (γ-SG), β-sarcoglycan (β-SG), and β-dystroglycan (β-DG) throughout the fiber membrane, where they down-expressed utrophin (Figure 5B). These results establish that MuStem cells could restore the dystrophin–glycoprotein complex in GRMD dog fibers.
Figure 5Dystrophin and dystrophin-associated glycoprotein expression after intramuscular injection of MuStem cells. A: Muscle sections were assayed for lacZ expression (blue) and immunolabeled for dystrophin (brown), to reveal lacZ+/dystrophin+ fibers. B: In serial sections, immunolabelings for laminin (lam), lacZ (red) and dystrophin (dys), γ-SG, β-SG, β-DG, utrophin (green) were performed (n = 3). Clusters of lacZ+/dystrophin+ muscle fibers expressing the different dystrophin-associated glycoproteins were observed (asterisks). LacZ+ nuclei (arrows) and dystrophin+/utrophin− fibers (arrowheads) were indicated. Scale bars: 100 μm (A); 50 μm (B).
Systemic Delivery of MuStem Cells Leads to Clinical Stabilization of GRMD Dogs
The potential use of MuStem cells as a clinical tool for cell therapy would be reinforced if they are shown to be able to reach their muscle target following systemic delivery. To check this possibility, nls-lacZ MuStem cells were intra-arterially injected in two immunosuppressed 7-month-old GRMD dogs (#14, #15). Eight weeks later, several hundreds of lacZ+ nuclei were observed in hind limb muscles of each dog with a tissue localization similar to those observed after intramuscular injection (Table 2): 84.2% had a subbasal position, 1.9% were centralized nuclei, and 13.9% displayed an endomysial position. This positive result prompted us to perform a more complete analysis.
Five systemic injections of 107 wild-type MuStem cells/kg were realized on three immunosuppressed GRMD dogs (#16 to #18) at intervals of 2 to 4 weeks. Six untreated dogs (#19 to #24) displayed a progressive clinical impairment with a course distributed in three phases (Figure 6A). Before the age of 14 weeks, the dogs exhibited only few signs characteristic of muscular dystrophy, the most prominent being palmigrade/plantigrade stances (Figure 6B, inset) and increased splaying of the digits. Their clinical score remained above 70% of that obtained by the healthy dogs. Between 14 and 26 weeks, a rapid decline of their walking ability was observed with progressive weakness, abnormal stiff limbs, short strides, and marked weight transfer (Figure 6B). Meanwhile, their score decreased to less than 40% of the healthy dog score. After the age of 26 weeks, GRMD dogs showed unchanged global clinical status (see Supplemental Video S1 at http://ajp.amjpathol.org). Three mock-immunosuppressed dogs (#19 to #21) displayed a similar clinical course compared to the three non-immunosuppressed ones (#22 to #24) (Figure 6B). Importantly, the GRMD dog that had received MuStem cells earlier (during phase 1) remained at a clinical score of about 90% 9 months after the first administration (Figure 6A). The two other GRMD dogs, which were injected at the beginning of phase 2, displayed a stabilization of their scores that was maintained up to 70% of that of the healthy dogs. A statistical difference between mock-immunosuppressed GRMD dogs and MuStem cell–injected ones was determined from 17 weeks to 50 weeks of age (repeated measures analysis of variance; P = 0.014). More than 6 months after the last MuStem cell injection, the three treated dogs still walked well and were active, in striking contrast with the mock-treated ones (see Supplemental Video S2 at http://ajp.amjpathol.org). The most obvious corrected criteria were the palmigrade/plantigrade stances, the weight transfer (Figure 6B), and the ease of standing up (see Supplemental Video S3 at http://ajp.amjpathol.org). One of the dogs injected in phase 2 showed a mild decrease of its score due to moderate ankylosis and limb stiffness. Creatine kinase levels did not differ between mock-immunosuppressed GRMD dogs and MuStem cell–injected ones, but depended on cyclosporinemia (P = 0.031, analysis of covariance), as illustrated in Supplemental Figure S2 at http://ajp.amjpathol.org, This tight correlation between creatine kinase levels and cyclosporinemia should preclude the use of creatine kinase as a biological marker of treatment efficacy in case of immunosuppression. Aspartate aminotransferase, another enzyme released by damaged muscle fibers, showed an overall similar pattern (data not shown). Collectively, these results demonstrate that systemic delivery of MuStem cells allows global and persistent stabilization of the GRMD dog's clinical status.
Figure 6Clinical evaluation of GRMD dogs. The clinical score was determined weekly and expressed as a percentage of a theoretical healthy dog score. A: Clinical course of muscular dystrophy on GRMD dogs (n = 3), mock-immunosuppressed GRMD dog (n = 2; brown lines), and MuStem cell–injected ones (n = 3). The first cell administration (arrowheads) and the time when dogs were excluded for ethical reasons (asterisks) were noted. B: Right lateral view of a mock-treated 36-week-old GRMD dog (top, dark brown line in A). Anterior weight transfer and plantigrady (inset) were visible. Right lateral view of a treated 36-week-old GRMD dog (bottom, red line, littermate with dog presented above). Note the roughly normal posture of the animal and the straightness of the limbs (inset).
Systemic Delivery of MuStem Cells Allows Dystrophin Recovery in GRMD Dog Muscles
To document dystrophin expression in muscles after systemic delivery of MuStem cells, muscle biopsies were obtained at various time points and subjected to RT-PCR analysis. One month after the first injection, wild-type dystrophin RNA was present in skeletal muscles of the left limb, which is the side that was injected, indicating that a single injection of 107 MuStem cells/kg is sufficient to allow dystrophin synthesis in muscles downstream from the injection site (Figure 7A). One month after the last injection, dystrophin RNA was detected in the biceps femoris muscle of both limbs. More important, dystrophin RNA persisted in muscles of both limbs by 4 months after the last cell injection. In addition, a large number of muscle fibers expressing dystrophin were demonstrated in cross sections, not only of the left muscles, but also of the right muscles (Figure 7B and Figure 8). It should be noted that dystrophin expression identified isolated fibers as well as clusters of fibers and that labeling was characterized by a low level compared to that observed in healthy dog muscle. Four months after the last injection, dystrophin+ fibers ranged from 20% to 25% and 25% to 30% in the left biceps femoris and tibialis cranialis muscles of GRMD dogs, respectively, whereas “revertant” fibers represented less than 0.2% of fibers in untreated GRMD dog muscles. Western blot analysis of muscle biopsies collected on two MuStem cell–injected GRMD dogs 4 and 7 months after the last injection confirmed the presence of dystrophin in treated muscles (see Supplemental Figure S3 at http://ajp.amjpathol.org). Even though the dystrophin expression level was much lower than that observed in healthy dog muscles, these results demonstrate that systemic delivery of MuStem cells allows an efficient homing of these cells to the muscle, resulting in long-term dystrophin expression.
Figure 7Dystrophin expression after systemic delivery of MuStem cells. A: RT-PCR analysis revealed the presence of wild-type dystrophin mRNA (PCR “exon 6/7 to exon 9”) on muscle biopsies collected at various time points after the first cell injection. The PCR “exon 1/exon 3” was shown as an internal control. B: Immunolabeling for laminin (red) and dystrophin (green) showed the presence of numerous scattered fibers with dystrophin expression in the whole muscle section, at the different time points of the protocol (n = 3 per muscle and time point). Scale bar = 50 μm. ND, non-determined; sm, size markers.
Figure 8Dystrophin expression in GRMD dog muscle 4 months after the last MuStem cell systemic delivery. In serial muscle sections, laminin (red) and dystrophin (green) immunolabelings were done (n = 3). Low magnification showing scattered dystrophin+ fibers over the whole section. Scale bar = 50 μm.
Systemic Delivery of MuStem Cells Acts on the Histopathological Phenotype of GRMD Dogs
Regenerative activity of dystrophic fibers was assessed on 7-month-old dogs, using a specific labeling to the developmental MyHC isoform whose expression is restricted to development and regeneration processes. Although no MyHCd+ fibers were observed in healthy dog Biceps femoris muscle (n = 3), 14.5% ± 4.1% of fibers expressed this isoform in the corresponding GRMD dog muscle (n = 3, Figure 9A). Strikingly, the MyHCd+ fiber represented 33.4% ± 7.5% of the fibers in Biceps femoris muscle of treated GRMD dog more than 4 weeks after the last MuStem cell injection (n = 3). This higher proportion compared to that observed in mock-immunosuppressed animals (P < 0.05), indicates that MuStem cells actively and persistently contribute to fiber regeneration. On the basis of the minimum Ferret diameter, we showed that the mean fiber diameter was 42.4 ± 13.8, 33.4 ± 12.9, and 37.1 ± 14.3 μm for healthy dogs, mock-immunosuppressed GRMD dogs, and treated ones, respectively (Figure 9B). It was significantly higher in treated GRMD dogs than in mock-immunosuppressed ones (P < 0.001). This increased diameter was illustrated by the modal value that was 40 to 60 μm in treated GRMD dog muscles (41.5% ± 2.5%), such as in healthy dog muscles (47.8% ± 6.7%), whereas it corresponded to 20 to 40 μm in mock-immunosuppressed dog muscles (52.7% ± 10.4%). The largest fibers (with diameter >60 μm) represented 12.6% ± 14.6% of all fibers in healthy dog muscles, whereas this percentage was lower in mock-immunosuppressed GRMD dogs (2.0% ± 1.4%) and increased after treatment in GRMD dogs (5.3% ± 1.5%). Fibrosis was determined as the ratio of collagen-positive areas on the total muscle area, using collagen type I immunolabeling. No significant difference was determined between mock-immunosuppressed GRMD dogs and treated ones, probably because of the minor size of the dog group. Measuring the intercellular spaces that only considered the endomysial component of connective tissue and not both endomysial and perimysial tissues, we showed that endomysial thickness was 0.7 ± 0.1, 2.1 ± 0.4, and 1.1 ± 0.1 μm in healthy, mock-immunosuppressed GRMD dogs, and treated ones, respectively (Figure 9C). Treated GRMD dogs exhibited highly reduced endomysial space all across the sections compared to mock-immunosuppressed animals (P < 0.001) (Figure 9D). Other histopathological features of GRMD dog muscles (ie, calcification, necrosis, and inflammation) were found to be unmodified (data not shown). Altogether, systemic delivery of MuStem cells generates a partial, but significant, histological correcting remodeling of the GRMD dog muscle consistent with the clinical output.
Figure 9Histological impact of MuStem cell systemic delivery. Histomorphometric analysis of muscular elementary lesions was performed on the Biceps femoris muscle of 7-month-old healthy dogs, mock-immunosuppressed GRMD dogs, and MuStem cell–injected GRMD dogs (n = 3 per group), ie, more than 4 weeks after the last cell injection. A: Regenerative activity in GRMD dog muscles was assessed by MyHCd labeling (brown). B: Fiber size distribution. Healthy dogs (white bars), GRMD dogs (gray bars), and treated GRMD dogs (black bars). C: Mean endomysial thickness. D: Hematoxylin-eosin-safranin stain of representative muscle sections. Hypertrophic fibers (asterisks), fibrosis (arrowheads), and calcified fibers (c) were indicated. Scale bar = 100 μm.
Different stem cell populations can be isolated from adult skeletal muscles, and it has been suggested that they could represent a promising alternative for cell-based therapy of muscular diseases based on their myogenic regeneration potential in dystrophic mice.
In return, whether MDSC are able to have tissue and clinical impact on a clinically relevant animal model has not been investigated, except for the mesoangioblasts.
Here, we report the reproducible isolation based on delayed adhesion properties of canine MDSC that we named MuStem cells, and demonstrate for the first time that the systemic delivery of these cells in dystrophic dogs allows dystrophin recovery, efficiently prevents muscle deterioration, and contributes to a global and persistent stabilization of the dog's clinical status.
MuStem cells were isolated as initial floating round cells after a similar procedure to the one described by Huard's group.
Originally, we showed that MuStem cells generated a heterogeneous population composed of spindle-shaped flat cells and a low percentage of round cells that remained constant due to the ability of these cells to perform atypical division pattern. Most of cells expressed satellite cell markers Pax7, CD56, and β1-integrin or myogenic regulatory factors Myf5 and MyoD, suggesting that MuStem cells could originate from satellite cell niche and corresponded mainly to early myogenic progenitors. They exhibited ex vivo multilineage differentiation potential into osteocyte and adipocyte cell lineages even though they appeared to be committed to the myogenic lineage as evidenced by their ability to spontaneously differentiate into myotubes. These features distinguished MuStem cells from mice MDSC,
that do not express key myogenic transcription factor Pax7, and/or differentiate into multinucleated myotubes only when co-cultured with primary myoblasts or after transfection with MyoD. MuStem cells were able to expand in suspension, an experimental condition that does not support proliferation of differentiated cells that rapidly die.
In this original proliferation context, MuStem cells gave rise to large clusters of rounded cells termed myospheres, which have been also described for cells freshly isolated from mice
we detected many hundreds of MuStem cells in muscles, whereas no myoblast could be observed. This revealed that MuStem cells were able to survive in the DMD context after in vitro expansion in contrast to cultured myoblasts known to have an extremely poor survival rate after injection in host muscle.
In parallel to fusion with host fibers and dystrophin recovery, MuStem cells generated satellite cells, an essential feature in the context of satellite cell pool exhaustion in muscular dystrophy.
Elevated p21 mRNA level in skeletal muscle of DMD patients and mdx mice indicates either an exhausted satellite cell pool or a higher p21 expression in dystrophin-deficient cells per se.
This data suggested that MuStem cell injection could have a long-term impact on the regenerative potential of dystrophic fibers by their constant recruitment for the host fiber regeneration. Similar contribution to the satellite cell pool has been demonstrated in injured mouse muscles for muscle SP cells,
However, this is the first time that this behavioral feature is described in highly damaged muscles such as those in GRMD dogs. In addition to their participation on fiber regeneration and satellite cell formation, we observed that MuStem cells intriguingly gave rise to interstitial cells. This behavior has been recently described for a new mouse muscle-resident stem cell subpopulation located in the interstitium, the PICs.
Indeed, these PW1+/Pax7− non-satellite cells efficiently contribute to skeletal muscle regeneration after injection in damaged mice muscle tissue as well as generating satellite cells and PICs. Following intramuscular or systemic delivery, an endothelial differentiation of the interstitial MuStem cells was never demonstrated in contrast to blood- and muscle-derived CD133+ cells
VCAM-1 expression on dystrophic muscle vessels has a critical role in the recruitment of human blood-derived CD133+ stem cells after intra-arterial transplantation.
A marked clinical stabilization of GRMD dogs with a major impact on locomotion features was noticed following systemic delivery of MuStem cells. More than 6 months after the last injection, GRMD dogs were lively in contrast to the untreated ones. Similarly, intra-arterial delivery of wild-type canine Mabs generated persistent clinical amelioration of GRMD dogs.
we documented the clinical course of treated GRMD dogs in parallel to that of non-immunosuppressed but also immunosuppressed GRMD dogs to clearly show that the clinical benefit could not be attributed to the immunosuppressive regimen. Taking into account that the clinical courses are quite similar inside the mock-treated and the treated dog groups, and are also dramatically distinct between the two groups, the clinical impact determined in the treated GRMD dogs probably cannot be explained alone by the phenotypic variability known among GRMD dogs.
A limitation of the present study still resides in the minor size of the dog group. To extrapolate the present results to prospective human trials, a more detailed functional phenotype characterization of treated GRMD dogs will be required to complete the clinical grading that corresponds to a semiquantitative approach. The gold standard methods used for clinical assessment of DMD patients, such as the 6-minute walk test,
TREAT-NMD Neuromuscular Network Towards harmonisation of outcome measures for DMD and SMA within TREAT-NMD; report of three expert workshops: TREAT-NMD/ENMC workshop on outcome measures, 12th–13th May 2007 Naarden, The Netherlands; TREAT-NMD workshop on outcome measures in experimental trials for DMD, 30th June–1st July 2007, Naarden, The Netherlands; conjoint Institute of Myology TREAT-NMD meeting on physical activity monitoring in neuromuscular disorders, 11th July 2007, Paris, France.
enable the comparison of the gait between GRMD dogs and healthy ones. Further investigations will be necessary to determine whether they could represent reliable tools to assess the efficacy of MuStem cell therapy in GRMD dogs.
Systemic administration of wild-type MuStem cells promoted the formation of numerous dystrophin+ fibers scattered over the entire section of several muscles. The dystrophin expression level was lower than that observed in a wild-type muscle as well as after intramuscular injections of MuStem cells. One must keep in mind that intramuscular injections generated a high concentration of donor cells in a limited tissue area and allowed fusion of several MuStem cell with host fibers, whereas systemic delivery resulted in a much wider dispersion of donor cells. This may reflect the fact that many more cells have to be injected to obtain a higher dystrophin expression.
In parallel to the dystrophin recovery, we showed that systemic administration of MuStem cells improved the histopathological phenotype of the GRMD dog biceps femoris muscle and demonstrated for the first time that this correcting remodeling comprised a major endomysial thickness reduction and a high increase of fiber regenerative activity. In contrast to fibrosis that results from the cumulative former pathological events occurring in the muscle tissue, fiber necrosis represents a punctual event. Moreover, because the percentages of necrosis or calcium deposits were very low (comprising between 0.5% and 2.4%) in sampled muscles, no difference between mock-treated and treated dog muscles could be demonstrated in the small groups of animals. It will be a critical issue to determine whether this tissue remodeling that appeared sufficient to induce considerable preservation of locomotion in GRMD dog results directly from MuStem cells and/or from paracrine signaling, as determined for stromal stem cells.
Marrow-derived stromal cells express genes encoding a broad spectrum of arteriogenic cytokines and promote in vitro and in vivo arteriogenesis through paracrine mechanisms.
Concerning the regenerative potential, MyHCd+ fibers were observed several weeks after the MuStem cell administration. Interestingly, this contribution to the regenerated fibers, delayed with regard to the systemic delivery, could promote ongoing repair of dystrophic muscle.
Recently, a marked improvement of muscle performance was measured in both respiratory and cardiac muscles of mdx mice following treatment with halofuginone, a collagen synthesis inhibitor that prevented fibrosis.
By demonstrating the key role of fibrosis on muscle function alterations in a dystrophic context, these findings support the hypothesis that the major restriction in endomysial expansion observed in the muscles of our treated GRMD dogs might have a direct impact on their walking ability and largely contribute to their clinical stabilization.
In conclusion, our results support our proposal that MuStem cells may represent a source of cells with therapeutic potential for DMD. Additional experiments are required to validate this proposal, among which one should demonstrate the existence of a human equivalent to the canine MuStem cells and further investigate the spectra of muscles that can be corrected following systemic delivery of the cells.
Acknowledgments
We thank Philippe Moullier (INSERM UMR 649, Nantes, France), Jamel Chelly and Bénédicte Chazaud (Institut Cochin, INSERM U567, CNRS UMR 8104, Paris, France) for helpful discussion and improving the manuscript. We also thank the staff of the Boisbonne Center (Oniris, Nantes, France) for the handling and care of the GRMD dog colony, and François-Loïc Cosset (INSERM U758, Lyon, France) for providing the nls-lacZ MLV retroviral vector.
Phenotype of Myoblasts. A: Surface labeling for CD56, CD45, CD34, CD49d, CD62L, Bcrp-1, CD44 and CD90 was determined for expanded myoblasts (n = 3). Shaded areas, control immunoglobulins; Black lines, specific antibodies. 97.5±1.0% and 8.9±2.8% of myoblasts expressed CD56 and CD49d while CD44 and CD90 were detected in all myoblasts. They were negative for the others markers. B: Cells were deposited on cytospin slides and subjected to immunolabelings for myogenic markers.
Creatine kinase activity and cyclosporinemia on immunosuppressed GRMD dogs. Creatine kinase levels and cyclosporinemia (bold curve: mobile mean order 3) were measured respectively every two months and twice a week. One mock immunosuppressed GRMD dog (A) and one MuStem cell-injected GRMD dog (B) are presented here. The start (arrowhead) of the immunosuppressive regimen was followed by a rapid and profound decrease of the creatine kinase levels. These levels tended to increase in the short time periods (gray areas) when immunosuppression was suboptimal, i.e., when cyclosporinemia was below 250 ng/mL (line).
Dystrophin protein level after systemic delivery of MuStem cells. Western blot analysis of dystrophin (Dys) protein was performed on muscle biopsies collected from healthy dog (lane 1), two MuStem cell-injected GRMD dogs (lanes 2, 3) and from mock immunosuppressed GRMD dog (lane 4). Lane 1 and lanes 2-4 were loaded with 40 μg and 230 μg of proteins, respectively. MyHC was shown as an internal standard in the bottom panel.
Locomotion features of mock immunosuppressed 12-month-old GRMD dog. Walking, exercise and standing up appeared to be difficult with generalized stiffness due to ankylosis and pronounced palmigrady/plantigrady. The dog showed shuffling gait and short strides. Marked weight transfer on forelimbs was also noted.
Locomotion features of immunosuppressed 12-month-old treated GRMD dog. Nine months after the last MuStem cell administration, dog displayed well-preserved locomotion features. He walked, ran and stood up easily. Gait appeared quite normal without weight transfer and minimum palmigrady.
Comparative locomotion features between littermate immunosuppressed GRMD dogs, with (black-coated) and without (black and white-coated) MuStem cell administration. Treated dog was lively in strong contrast with the mock one. He walked quickly and was able to stand up on his posterior limbs to reach a toy.
References
Emery A.E.
Population frequencies of inherited neuromuscular diseases–a world survey.
Flow cytometric characterization of myogenic cell populations obtained via the preplate technique: potential for rapid isolation of muscle-derived stem cells.
Development of a snapback method of single-strand conformation polymorphism analysis for genotyping Golden Retrievers for the X-linked muscular dystrophy allele.
Cordazzo M-C:[Development of Tools for the Clinical Evaluation of Dogs Suffering from Muscular Dystrophy] (doctoral thesis). [Maisons-Alfort, France]: École Nationale Vétérinaire d'Alfort
Elevated p21 mRNA level in skeletal muscle of DMD patients and mdx mice indicates either an exhausted satellite cell pool or a higher p21 expression in dystrophin-deficient cells per se.
VCAM-1 expression on dystrophic muscle vessels has a critical role in the recruitment of human blood-derived CD133+ stem cells after intra-arterial transplantation.
Towards harmonisation of outcome measures for DMD and SMA within TREAT-NMD; report of three expert workshops: TREAT-NMD/ENMC workshop on outcome measures, 12th–13th May 2007.
Marrow-derived stromal cells express genes encoding a broad spectrum of arteriogenic cytokines and promote in vitro and in vivo arteriogenesis through paracrine mechanisms.
00741-3&id=gr1.jpg){kind=link}
00741-3&id=gr2.jpg){kind=link}
00741-3&id=gr3.jpg){kind=link}
00741-3&id=gr4.jpg){kind=link}
00741-3&id=gr5.jpg){kind=link}
00741-3&id=gr6.jpg){kind=link}
00741-3&id=gr7.jpg){kind=link}
00741-3&id=gr8.jpg){kind=link}
00741-3&id=gr9.jpg){kind=link}