A New Immuno-, Dystrophin-Deficient Model, the NSG-mdx4Cv Mouse, Provides Evidence for Functional Improvement Following Allogeneic Satellite Cell Transplantation


  • Robert W. Arpke,

    1. Lillehei Heart Institute, University of Minnesota, Minneapolis, Minnesota, USA
    2. Department of Pediatrics, University of Minnesota, Minneapolis, Minnesota, USA
    Search for more papers by this author
  • Radbod Darabi,

    1. Lillehei Heart Institute, University of Minnesota, Minneapolis, Minnesota, USA
    2. Department of Medicine, and University of Minnesota, Minneapolis, Minnesota, USA
    Search for more papers by this author
  • Tara L. Mader,

    1. Program in Physical Therapy and Rehabilitation Science, University of Minnesota, Minneapolis, Minnesota, USA
    Search for more papers by this author
  • Yu Zhang,

    1. Lillehei Heart Institute, University of Minnesota, Minneapolis, Minnesota, USA
    2. Department of Pediatrics, University of Minnesota, Minneapolis, Minnesota, USA
    Search for more papers by this author
  • Akira Toyama,

    1. Lillehei Heart Institute, University of Minnesota, Minneapolis, Minnesota, USA
    2. Department of Pediatrics, University of Minnesota, Minneapolis, Minnesota, USA
    Search for more papers by this author
  • Cara-lin Lonetree,

    1. Lillehei Heart Institute, University of Minnesota, Minneapolis, Minnesota, USA
    2. Department of Pediatrics, University of Minnesota, Minneapolis, Minnesota, USA
    Search for more papers by this author
  • Nardina Nash,

    1. Lillehei Heart Institute, University of Minnesota, Minneapolis, Minnesota, USA
    2. Department of Pediatrics, University of Minnesota, Minneapolis, Minnesota, USA
    Search for more papers by this author
  • Dawn A. Lowe,

    1. Program in Physical Therapy and Rehabilitation Science, University of Minnesota, Minneapolis, Minnesota, USA
    Search for more papers by this author
  • Rita C.R. Perlingeiro,

    1. Lillehei Heart Institute, University of Minnesota, Minneapolis, Minnesota, USA
    2. Department of Medicine, and University of Minnesota, Minneapolis, Minnesota, USA
    Search for more papers by this author
  • Michael Kyba

    Corresponding author
    1. Lillehei Heart Institute, University of Minnesota, Minneapolis, Minnesota, USA
    2. Department of Pediatrics, University of Minnesota, Minneapolis, Minnesota, USA
    • University of Minnesota, Nils Hasselmo Hall, 312 Church Street SE, Minneapolis 55455, Minnesota, USA===

    Search for more papers by this author
    • Telephone: + 1-612-626-5869; Fax: + 1-612-624-8118

  • Author contributions: R.W.A. and R.D.: data collection, analysis, and manuscript writing; T.M., Y.Z., A.T., N.N.: data collection and analysis; C.L.: data collection; D.A.L. and R.C.R.P.: study design, manuscript writing, and financial support; M.K.: conception and design, manuscript writing, and financial support.


Transplantation of a myogenic cell population into an immunodeficient recipient is an excellent way of assessing the in vivo muscle-generating capacity of that cell population. To facilitate both allogeneic and xenogeneic transplantations of muscle-forming cells in mice, we have developed a novel immunodeficient muscular dystrophy model, the NSG-mdx4Cv mouse. The IL2Rg mutation, which is linked to the Dmd gene on the X chromosome, simultaneously depletes NK cells and suppresses thymic lymphomas, issues that limit the utility of the SCID/mdx model. The NSG-mdx4Cv mouse presents a muscular dystrophy of similar severity to the conventional mdx mouse. We show that this animal supports robust engraftment of both pig and dog muscle mononuclear cells. The question of whether satellite cells prospectively isolated by flow cytometry can confer a functional benefit upon transplantation has been controversial. Using allogeneic Pax7-ZsGreen donors and NSG-mdx4Cv recipients, we demonstrate definitively that as few as 900 FACS-isolated satellite cells can provide functional regeneration in vivo, in the form of an increased mean maximal force-generation capacity in cell-transplanted muscles, compared to a sham-injected control group. These studies highlight the potency of satellite cells to improve muscle function and the utility of the NSG-mdx4Cv model for studies on muscle regeneration and Duchenne muscular dystrophy therapy. STEM Cells 2013;31:1611–1620


Muscle regeneration is accomplished by a quiescent population of muscle stem cells referred to as satellite cells, which reside between the muscle fiber plasma membrane and its basement membrane [1, 2]. Satellite cells in adult skeletal muscle are defined by the expression of the transcription factor Pax7 [3] as well as by a number of surface antigens [4–6]. When transplanted into injured muscle, satellite cells not only generate new muscle fibers but also contribute to the satellite cell pool [7, 8]. By single cell transplantation, individual satellite cells have been shown capable of extensive proliferation, fiber generation, as well as self-renewal in vivo [9]. Considering these properties, it stands to reason that satellite cell transplantation would be beneficial for the treatment of any muscle disease, particularly in the replacement of genetically defective muscle fibers such as those lacking dystrophin in patients with Duchenne muscular dystrophy. The gold standard in this regard would be the improvement in muscle strength following satellite cell transplantation. The dystrophin-deficient mdx mouse [10] represents an excellent model system to test such therapies. Although improvements in muscle force generation have been reported following transplantation of embryonic stem cell (ESC)/induced pluripotent stem cell (iPSC)-derived myogenic progenitors into mdx mice [11, 12], or transplantation of intact myofibers [13], to date only one report has evaluated force-generating capacity of muscle following transplantation of satellite cells prospectively isolated from adult hosts [14]. This report remains controversial because as a group, transplanted muscles did not show a statistically significant improvement over control nontransplanted muscles, rather a correlation was observed between cases with the greatest fold difference in force (between transplanted and contra-lateral control) and engraftment efficiency. It has been suggested that this correlation was due to unusually low force measured in the control leg rather than elevated force produced by the transplanted leg [15]. Therefore, it remains to be demonstrated that satellite cell transplantation can improve the force generation of dystrophin-deficient muscles.

Transplantation of satellite cells, like transplantation of any tissue, is limited by the host immune system. Two immune-deficient dystrophin-deficient models have been used in transplantation studies: SCID/mdx [16] and mdx/nude [17]. In both cases, although T-cell-mediated responses are ablated, functional natural killer (NK) cells remain. The combination of the severe combined immunodeficient (SCID) mutation with deletion of the interleukin-2 (IL2) receptor common gamma chain (IL2Rg) results in much improved engraftment of human hematopoietic stem cells in mice [18]. This is attributed to the ablation of NK cells which recognize foreign cells not by the presence of foreign antigens but by their lack of self major histocompatibility complex (MHC) [19]. Even in cases of isogenic transplantation, the ablation of NK cells can improve engraftment, particularly where donor cells do not express high levels of MHC class I molecules [20, 21]. In order to facilitate evaluation of satellite cell transplantation from any mouse strain, including strains bearing various genetic modifications, which tend to be of mixed background, as well as transplantations from xenogeneic sources, we have sought to develop an improved immune-deficient mdx recipient by incorporating the IL2Rg mutation.

Here we describe the NSG-mdx4Cv mouse, a novel immune-deficient, dystrophin-deficient mouse model that can serve as a transplantation recipient for muscle cells from either xenogeneic or allogeneic sources. Using allogeneic donors, and without immune suppression, we demonstrate conclusively that fluorescence-activated cell sorting (FACS)-isolated adult satellite cells not only engraft robustly after transplant but can also impart a statistically significant physiological improvement in strength in a muscle stricken with muscular dystrophy.

Materials and Methods


Breeding pairs of mdx (C57BL/10ScSn), NOD.Cg-Prkdcscid Il2rgtm1Wjl/SzJ NOD/SCID;IL2Rg, and B6Ros.Cg-Dmdmdx-4Cv/J mice were purchased from Jackson Laboratories (Bar Harbor, ME, http://www.jax.org). All procedures were carried out in accordance with approved protocols by the University of Minnesota Institutional Animal Care and Use Committee.

Genotyping for Mdx4Cv

Polymerase chain reaction (PCR) using the following primers was used to identify the mdx4Cv mutation: Mdx4CvF:TGCCCACAAGTAAGTGCTGAGGT and Mdx4CvR: AGCTTGCCTCTGACCTGCCCT. Cycle conditions were 94°C for 3 minutes, followed by 35 cycles of 94°C for 30 seconds, 55°C for 30 seconds, and 72°C for 1 minute, followed by an extension of 72°C for 10 minutes. One half of the PCR product volume was digested with Bsm AI (New England Biolabs, Ipswich, MA, https://www.neb.com) at 55°C overnight. Digested PCR product was run alongside the undigested PCR product on a 2% agarose gel (band size(s): mutant, 476 bp; WT, 325 bp and 150 bp.

Antibodies for FACS Analysis

For characterization of the peripheral blood cells, we used anti-mouse CD3ε Phycoerythrin (PE) (145-2C11), anti-mouse CD4 Allophycocyanin (APC)-eFluor 780 (RM4-5), anti-mouse CD8a (Ly-2) APC-Cy7 (53-6.7), anti-mouse Ly-6G (Gr-1) fluorescein isothiocyanate (FITC) (RB6-8C5), anti-mouse NK1.1 FITC (PK136), and anti-mouse pan-NK cells (CD49b) (DX5) from eBioscience (San Diego, CA, http://www.ebioscience.com) and anti-mouse CD19 PE-Cy7 (1D3) was from BD Pharmingen (San Diego, CA, http://www.bdbiosciences.com). FACS analysis was performed on a FACSAria (BD Biosciences, San Diego, CA.

Histology and Immunofluorescence

Tibialis anterior (TA) muscle was removed for analysis 4 weeks after transplantation. Samples were placed in O.C.T. Compound, frozen in liquid nitrogen-cooled 2-methylbutane (Sigma-Aldrich, St Louis, MO, http://www.sigmaaldrich.com), and stored at −80°C. 10-μm cryosections were cut on a Leica CM3050 S cryostat (Leica Microsystems, Buffalo Grove, IL, http://www.leica.com). For dystrophin and laminin staining, cryosections were fixed with cold acetone for 5 minutes. After air drying for 10 minutes followed by 10 minutes Phosphate Buffered Saline (PBS) rehydration, the sections were blocked for 1 hour with 3% bovine serum albumin (BSA) in PBS. After primary antibody staining with a rabbit polyclonal antibody to dystrophin (Abcam, Cambridge, MA, http://www.abcam.com) and a mouse monoclonal antibody to laminin (Sigma-Aldrich) or a mouse monoclonal antibody to dystrophin (Sigma-Aldrich) at RT for 1 hour, sections were incubated with secondary antibodies conjugated with Alexa Fluor 555 goat anti-rabbit and Alexa Fluor 488 goat anti-mouse (Invitrogen, Grand Island, NY, http://www.invitrogen.com). 4′,6-Diamidino-2-Phenylindole (Santa Cruz Biotechnology, Santa Cruz, CA, http://www.scbt.com) was used to counter-stain nuclei. Coverslips were mounted using Immu-Mount (Thermo Scientific, Kalamazoo, MI, http://www.thermoscientific.com). Slides were photographed with a Zeiss Axio Imager M1 Upright microscope with an AxioCam HRc camera. The number of donor (dystrophin +) muscle fibers and total muscle fibers were quantified in cross-sections from the transplanted TA.

In Vivo Assessment of Anterior Crural Muscles

Mice were anesthetized via an intraperitoneal injection of fentanyl citrate (0.2 mg/kg body mass [BM]), droperidol (10 mg/kg BM), and diazepam (5 mg/kg BM) and the left foot placed in a foot plate attached to the shaft of a servomotor (Model 300 B-LR; Aurora Scientific, Aurora, Ontario, Canada, http://www.aurorascientific.com) as described previously [22, 23]. Two platinum electrodes were inserted percutaneously on either side of the peroneal nerve for stimulation of the anterior crural muscles (TA, extensor digitorum longus (EDL), and extensor hallicus longus muscles). The voltage was adjusted from 5.0 to 8.0 V until maximal isometric torque was achieved (preinjury torque). Susceptibility to contraction-induced injury was assessed by an eccentric contraction protocol [23]. For this, the foot was passively moved by the servomotor to 19° of ankle dorsiflexion and then muscles were stimulated for 150 ms and 250 Hz while muscles were simultaneously lengthened as the foot moved to 19° of ankle plantarflexion at an angular velocity of 2,000°/second. There was a 10-second rest between each of the 100 eccentric contractions and then 5 minutes following the last, maximal isometric torque was remeasured (postinjury torque). Percent torque loss was calculated as an indicator of physiological injury ([preinjury torque − postinjury torque]/preinjury torque × 100). Mice were returned to their cages and 7 days later a subgroup of mice was anesthetized again and maximal isometric torque was remeasured to evaluate recovery from contraction-induced injury.

In Vitro Assessment of EDL Muscle

The EDL muscle from right leg (i.e., leg that was not tested in vivo) was dissected and assessed for force generating capacity and susceptibility to eccentric contraction-induced injury [23]. Muscles were mounted on a dual-mode lever system (300B-LR; Aurora Scientific) with 5-0 suture in a 1.5-mL bath filled with Krebs-Ringer buffer at 25°C and adjusted to anatomical optimal length (Lo). Maximal twitch and isometric tetanic forces were measured. Five eccentric contractions (10% Lo change at 0.75 Lo/second) were completed and then peak isometric tetanic force was remeasured to calculate percent force loss. All contractions were separated by 3-minute rest.

Plasma Creatine Kinase Activity

Blood from the retroorbital sinus was collected in heparinized tubes from a subset of mice. Plasma was prepared, stored at −80°C, and then later analyzed for creatine kinase (CK) activity on Vitros CK/CPK slides and spectrometer (Ortho Clinical Vitros DT 60 II Dry Slides and System; Rochester, NY, http://www.orthoclinical.com).

Mouse Satellite Cell Isolation

Satellite cells were isolated from Pax7-ZsGreen mice [24] on a C57BL/6 background. Briefly, hind limb muscle was removed. Using a razor blade parallel to the muscle fiber, forceps were used to separate the fibers. Muscle was then incubated shaking with 0.2% collagenase type II (Gibco, Grand Island, NY, http://www.invitrogen.com; 17101-015) in high glucose Dulbecco's modified Eagle's medium containing 4.00 mM l-glutamine, 4,500 mg/L glucose, and sodium pyruvate (HyClone, Logan, UT, http://www.hyclone.com; SH30243.01) at 37°C for 75 minutes. Using Rinsing Solution (F-10 + ), Ham's/F-10 medium (HyClone, SH30025.01) supplemented with 10% horse serum (Gibco, 26050-088), 1% 1 M HEPES buffer solution (Gibco, 15630-106), and 1% Pen/Strep (Gibco, 15140-122), samples were washed two times, poured into a petri dish, and mechanically scraped with a sheared Pasteur pipette. The sample was centrifuged and washed again with F-10 + . After aspiration, the sample was resuspended in F-10 + containing collagenase II and dispase (Gibco, 17105-041), vortexed, and incubated shaking at 37°C for 30 minutes. After incubation, the sample was again vortexed, drawn, and released four times with a 16-gauge needle, then with an 18-gauge needle to dislodge cells from the muscle fibers before applying the sample to a 40-μm cell strainer. The cell suspension was centrifuged at 1,500 rpm for 5 minutes, resuspended in F-10 +, drawn and released four times again with an 18-gauge needle, and applied to a new 40 μm cell strainer. After centrifugation, the sample was resuspended in FACS staining medium: PBS (HyClone, SH30256.01) containing 2% fetal bovine serum (HyClone). Cell sorting and analysis was performed on a Cytomation MoFlo cytometer (Dako, Carpinteria, CA, http://www.dako.com).

Muscle Injury and Transplantation

Two days prior to intramuscular injection of cells, 4-month-old NSG-mdx4Cv mice were anesthetized with ketamine and xylazine and both hind limbs were subjected to 1,200 cGy dose of irradiation using an X-RAD 320 Biological Irradiator (Precision X-Ray, Inc., Branford, CT, http://www.pxinc.com). Mice were covered with a lead shield that permitted exposure only to the hind limbs. One day prior to transplantation, 15 μL of cardiotoxin (10 μM, Calbiochem, Philadelphia, PA, http://www.emdbiosciences.com) was injected into the left and right TA muscle of each mouse to induce muscle injury. 24 hours later, various numbers of sorted Pax7-ZsGreen positive mouse hind limb satellite cells or 500,000 freshly isolated pig or dog mononuclear cells were resuspended in 15 μL sterile saline and injected into the left TA of each mouse using a Hamilton syringe. Four weeks after transplantation, TAs from both experimental and control legs were harvested, sectioned, and stained, and counted as described above to detect muscle fiber contribution of transplanted cells. For the dilution study, each group consisted of six independent transplantations.

In Vitro Assessment of TA Muscle

As described previously [11], mice were anesthetized using intraperitoneal injection of Avertin (250 mg/kg i.p.), and intact TA muscles were carefully dissected out and transferred into an organ bath filled with mammalian Ringer solution containing: 120.5 mM NaCl, 20.4 mM NaHCO3, 10 mM glucose, 4.8 mM KCl, 1.6 mM CaCl2, 1.2 mM MgSO4, 1.2 mM NaH2PO4, 1.0 mM pyruvate, adjusted to pH 7.4. The organ bath was perfused continuously with 95% O2/5% CO2 and the temperature was adjusted to 25°C. Then the isolated TA muscles were stimulated by an electric field generated between two platinum electrodes placed longitudinally on either side of the TA muscles (using square wave pulses at 25 V amplitude, 0.2 ms in duration, 150 Hz). Muscles were maintained at the optimum length (Lo) during the determination of isometric twitch force with a 5 minute recovery period between stimulations. Optimal muscle length (Lo) and stimulation voltages (25 V) were chosen based on muscle length manipulation and a series of twitch contractions that generated the maximum isometric twitch force. For fatigue time measurement, muscles were stimulated for 1 minute and the time for force to decline to 30% of Fo was determined. After adjusting the optimal muscle length (Lo) and measuring the maximum isometric tetanic force, total muscle cross-sectional area (CSA) was calculated by dividing muscle mass (mg) by the product of muscle length (mm) and 1.06 mg/mm3, the mammalian skeletal muscle density. Specific force (sFo) was then calculated by normalizing maximum isometric tetanic force to CSA.


Data were analyzed by two-tailed t tests. Data are reported as means with standard errors and significance was accepted at the α < 0.05 level.


We generated the NSG-mdx4Cv mouse by combining three mutations: SCID (NOD/SCID) mutation at the protein kinase, DNA activated, catalytic polypeptide (Prkdcscid) locus, which results in the complete absence of mature B and T cells [25], the IL2Rg mutation, ablating NK cells [26], and an allele of Dmd named mdx4Cv [27]. IL2Rg and Dmd are both X-linked. The mdx4Cv mutation is a C to T transition at position 7,916 (Fig. 1A) in exon 53 resulting in a premature stop codon [28]. This single transition destroys a BsmAI restriction site facilitating genotyping (Fig. 1A, see methods). Although all mdx mice lack dystrophin, rare dystrophin + muscle fibers are observed. These also occur in Duchenne patients and are thought to represent somatic reversion or suppression of the mdx mutation [29]. The 4Cv mutant was reported to have approximately 10-fold fewer revertants than mdx at both 2 and 6 months of age in quadriceps cross-sections [30].

Figure 1.

Characterization of the NSG-mdx4Cv mice. (A): The mdx4Cv point mutation, indicated in red, destroys a recognition site for the restriction enzyme, BsmA1. The recognition motif, underlined, is indicated in green. Cut sites are indicated with arrowheads. (B): Fluorescence-activated cell sorting profiles for wt and NSG-mdx4Cv mice. NK cells are recognized as double-positive for the antibodies NK1.1 and DX5. (C): Cross-sections of tibialis anterior from wt and NSG-mdx4Cv mice aged 6 weeks stained for Dystrophin. Scale bar = 100 μm. (D): Sirius red staining of diaphragm cross-sections from wt, mdx, and NSG-mdx4Cv mice of various ages. Scale bar = 100 μm. (E): Sirius red staining of quadriceps cross-sections from wt, mdx, and NSG-mdx4Cv mice of various ages. Scale bar = 100 μm. Abbreviation: wt, wild-type.

From trans-heterozygous females bred to WT males, 140 pups were screened, and two recombinants were identified. These were intercrossed, and animals homozygous at all loci were identified. To confirm that the NSG-mdx4Cv mice are immune compromised, peripheral blood was evaluated by FACS for the presence of T and B cells as well as NK cells. As expected, the mdx mice had T, B, and NK cells, while the NSG-mdx4cv mice lacked these cell types (Fig. 1B).

Characterization of the Muscular Dystrophy in NSG-Mdx4Cv Mice

TA muscle sections from 6-week-old WT and NSG-mdx4Cv mice were stained for dystrophin (Fig. 1C). As expected, we observed dystrophin + muscle fibers in the WT mice while only rare dystrophin + fibers were observed in the NSG-mdx4Cv mice.

We next compared the muscular dystrophy phenotype of NSG-mdx4Cv animals to conventional mdx mice in a number of ways. Sections from the diaphragm (Fig. 1D) and quadriceps (Fig. 1E) from 6-week-, 4-month- and 1-year-old WT, mdx, and NSG-mdx4Cv mice were stained with Sirius Red to examine levels of fibrosis. More fibrosis was observed in the mdx and NSG-mdx4Cv samples as compared to WT mouse sections at each of the time points examined.

To gauge muscle strength in vivo, we measured the generation of torque by repetitive stimulation of anterior crural (TA, EDL, and extensor hallicus longus) muscles in live mice. Relative to WT mice, maximal isometric torque was 48% lower in the 6-week mice lacking dystrophin (Table 1), and approximately 30% lower when normalized to body mass (p < .001; Fig. 2A). These values were equally low in mdx and NSG-mdx4Cv mice. Maximal eccentric torque generation was also equally low in mdx and NSG-mdx4Cv mice relative to WT (Table 1). We then measured the percent of eccentric torque lost during a damaging series of 100 eccentric contractions. Torque loss after repetitive eccentric contractions was greater in NSG-mdx4Cv mice compared to WT and statistically equivalent to torque loss by mdx mice (measured at eccentric contraction #100, Fig. 2B). Isometric torque loss followed the same pattern although differences between groups did not reach statistical significance (Table 1).

Figure 2.

Similar torque-/force-generating capacity, susceptibility to contraction-induced injury, and rapid recovery in muscles of NSG-mdxCv and mdx mice. (A): Maximal isometric torque generated in vivo by anterior crural muscles normalized by body mass. (B): Torque loss by anterior crural muscles with multiple damaging contractions. Torque is expressed as a percentage of maximal (initial) eccentric torque. (C): Specific force (maximal isometric titanic force normalized to cross-sectional area) of extensor digitorum longus (EDL) muscle. (D): Force loss by EDL muscles. Force expressed as a percentage of initial (maximal) eccentric force. (E): Torque recovery 1 week postinjury. Recovery is improved to preinjury levels in both mdx and NSG-mdx4Cv mice but still reduced in WT mice. Data are means, SE. Nonvisible error bars are contained within the symbol. *, Significantly different from WT.

Table 1. Torque- and force-generating capacities and susceptibility to contraction-induced injury of hind limb muscles, and serum creatine kinase activity from WT, mdx, and NSG-mdx4Cv mice aged 6 weeks or 4 months
 6 weeks4 months 
  • *

    Significantly different than age-matched mdx (p ≤ .05) as determined by two-tailed t tests.

  • Mean (SE).

  • Abbreviations: CK, creatine kinase; EDL, extensor digitorum longus.

In vivo: Anterior crural muscles
Maximal isometric torque (N mm)1.50 (0.18)1.48 (0.12)2.09 (0.24)2.09 (0.17)2.82 (0.21)
Maximal eccentric torque (N mm)2.60 (0.26)2.58 (0.22)3.55 (0.29)4.13 (0.38)5.75 (0.41)
Isometric torque loss (%)69.7 (8.6)62.1 (3.8)54.3 (7.8)58.4 (6.9)41.8 (6.7)
In vitro: EDL muscle
Peak twitch force (mN)52.9 (2.8)40.5 (0.8)*76.1 (11.0)52.1 (5.4)95.1 (5.2)
Peak isometric tetanic force (mN)267.8 (18.5)192.6 (6.5)*378.2 (25.2)269.5 (33.3)*438.7 (19.4)
Peak eccentric force (mN)393.1 (20.7)304.8 (7.8)*536.9 (41.8)408.3 (41.9)657.0 (27.5)
Isometric force loss (%)50.9 (9.0)29.1 (7.3)43.0 (14.4)15.3 (11.3)3.0 (2.2)
Passive stiffness (N/m)12.8 (0.4)12.5 (0.5)13.0 (0.8)16.6 (0.9)*10.3 (0.3)
Plasma CK
CK activity (Units/L)88,000 (18,000)41,000 (12,000)174,000 (15,000)59,000 (15,000)*1,300 (560)

We then measured peak isometric and eccentric forces of dissected EDL muscles. In both NSG-mdx4Cv and mdx mice, these were moderately affected by the lack of dystrophin (Table 1), and since the EDL is hypertrophic in these animals, specific force (absolute force normalized to cross-sectional area) was severely compromised. The generation of specific force was approximately 34% lower in all EDL muscles lacking dystrophin compared to WT (p < .001; Fig. 2C), with no difference between mdx and NSG-mdx4Cv mice at either age investigated. Loss of force-generating capacity after high-force eccentric contractions was measured in the EDL over a series of five eccentric contractions. Force loss during the eccentric protocol was significantly greater in EDLs of both dystrophin-deficient animals compared to WT (p = .025; Fig. 2D), and although mdx showed slightly greater isometric force loss, differences between mdx and NSG-mdx4Cv were not statistically significant at any age (Table 1). We also measured passive stiffness, which reflects resistance against lengthening in a noncontracting muscle due to elements in parallel with force-generation elements, and is typically increased in muscular dystrophy. This parameter was greater in EDL muscles lacking dystrophin compared to WT and again, no significant differences were observed between mdx and NSG-mdx4Cv (Table 1).

An interesting feature of the mdx mouse is that although muscle is more sensitive to contraction-induced injury, it also recovers more rapidly from such injury than does WT muscle [31, 32]. This can be attributed to the fact that mdx muscle is essentially constantly regenerating, and therefore contains a population of activated satellite cells, immediately ready to respond to induced injury. We tested whether this was true also of the NSG-mdx4Cv mouse by re-evaluating anterior crural muscles that were damaged by repetitive eccentric contractions (above) 1 week after the injury. At this early time point, all dystrophic mice had recovered torque to preinjury levels (96%–104% recovery), while WT mice had only recovered to a mean of 80% of preinjury levels (Fig. 2E). Therefore, the NSG-mdx4Cv mouse also shares the feature of accelerated recovery from injury with the mdx mouse.

Plasma CK activity was elevated in mdx and NSG-mdx4Cv mice in both the 6-week- and 4-month-old as compared to WT mice (Table 1). Although still elevated as compared to WT mice, 4-month-old NSG-mdx4Cv had significantly lower CK activity than age-matched mdx mice.

Xenogeneic Transplantations

To test the utility of NSG-mdx4Cv mice to serve as xenogeneic hosts, we tested the transplantation of pig and dog cells. Whole muscle mononuclear cell fractions were isolated from pig or dog gastrocnemius muscle by enzymatic digestion, and 500,000 cells were transplanted by intramuscular injection into the left TA of NSG-mdx4Cv recipients. The right TA was injected with saline as a negative control. Mice were preconditioned by irradiation of the hind limbs (1,200 cGy, 2 days prior) and cardiotoxin injury to both TA muscles (1 day prior). One month after transplantation, dystrophin + fibers were observed in both pig and dog transplants (Fig. 3). With unfractionated cells, we observed 18% fiber contribution from pig cells and 10% fiber contribution from dog cells.

Figure 3.

Transplantation of pig or dog muscle mononuclear cells into NSG-mdx4Cv mice. Sections of tibialis anterior muscles 4 weeks after transplantation, stained with anti-dystrophin and 4′,6-diamidino-2-phenylindole. Contribution of donor cells to muscle is indicated by the presence of dystrophin + fibers. Top panels, dystrophin; bottom panels, dystrophin and DAPI. Scale bar = 100 μm.

Satellite Cell Transplantations Improve Muscle Function

The current evidence that skeletal muscle progenitors prospectively isolated from adult muscle by flow cytometry can impart a physiological improvement in muscle function is controversial [14, 15]. We therefore designed an experiment using the NSG-mdx4Cv mice to test definitively whether prospectively isolated adult skeletal muscle progenitors have the capacity to improve the force-generation capacity of dystrophic muscle. For this study, we used our novel NSG-mdx4Cv recipients and directly compared them to SCID-mdx4Cv mice, as well as conventional mdx animals, with or without immune suppression with tacrolimus. The conventional mdx mouse carries the B10 genetic background, which is MHC-matched with our B6.Pax7-ZsGreen satellite cell donor mice, meaning that cells should not be rejected through an NK-mediated lack of self pathway. We tested immune suppression of mdx mice in order to bypass potential T-cell response against dystrophin or minor B6 antigens.

We isolated satellite cells by flow cytometry from muscle mononuclear cell fractions obtained by enzymatic dissociation of total hind limb muscle from Pax7-ZsGreen mice [24] on a C57BL/6 background. Using the same irradiation/injury preconditioning, we transplanted approximately 10,000 Pax7-ZsGreen + satellite cells or PBS into the left and right TA muscles, respectively, of eight recipients of each strain. One month after transplantation both muscles were harvested and subjected to in vitro contractility studies. The absolute maximal isometric tetanic force generation capacity was increased in the cell-transplanted TAs relative to the PBS-injected TAs in each of the dystrophic models (Fig. 4A). However, statistically significant improvement was observed only in the immune suppressed mdx and the NSG-mdx4Cv mice. The same was true of specific force (Fig. 4B). Notably, although mean maximal and specific force measurements were increased in the SCID/mdx model, statistical significance was not observed. Besides the lower mean, the main reason for this is that several mice died between transplant and harvest. This underscores the principal weakness of the SCID/mdx model: their fragility and early demise due to thymic lymphoma.

Figure 4.

Transplantation of Pax7-ZsGreen-positive satellite cells into dystrophic mice. Contractile measurements for the PBS- and Pax7-ZsGreen cell-injected tibialis anterior (TA) muscle of mdx, mdx with immune suppression by tacrolimus, scid;mdx4Cv or NSG-mdx4Cv mice. (A): Average maximal isometric tetanic force. (B): Average specific force. (C): Dystrophin staining of a section of a representative (middle-range) engrafted TA from a NSG-mdx4Cv mouse. Merged from 20× magnification images. (D): Collagen I staining of a serial section of a representative (middle-range) engrafted TA from a NSG-mdx4Cv mouse. Merged from 20× magnification images. (E): Average number of donor-derived dystrophin + and total fibers in cell-injected and contralateral PBS-injected control TA muscles of mdx, mdx with tacrolimus, scid;mdx4Cv or NSG-mdx4Cv mice. (F): Average specific force for PBS- and different numbers of Pax7-ZsGreen cell-injected TA muscles. (G): Average number of donor-derived dystrophin + and total fibers in cell-injected and PBS-injected control TA muscles of NSG-mdx4Cv mice. Abbreviations: PBS, Phosphate Buffered Saline; SCID, severe combined immune deficiency.

Following force measurements, muscles from each of the four groups of mice were processed for laminin and dystrophin staining to determine the degree of engraftment with donor-derived fibers. Engraftment typically occurred as a large, contiguous cluster of dystrophin+ fibers, surrounded by host dystrophin-negative muscle (Fig. 4C). We also stained for collagen I to determine whether the new WT muscle was healthier than the host muscle. It is apparent from serial sections that the large contiguous zone of dystrophin cells is also a relatively clear zone of interfiber collagen staining, that is, this zone does not show signs of fibrosis while the surrounding tissue does (Fig. 4D). In comparison to contralateral PBS-injected control TA muscles, satellite cell-transplanted TA muscles showed an increase in muscle fiber number in each of the dystrophic models (Fig. 4E). In the NSG-mdx4Cv model, approximately 1,700 fibers per TA or 49% of the total fibers were dystrophin+. This is in agreement with the finding above that this model was clearly the most effective at demonstrating functional activity of newly formed muscle. Evaluating the other models, an engraftment series is established: NSG-mdx4Cv > SCID/mdx4Cv > tacrolimus-injected mdx > mdx.

A final interesting observation relates to the frequency of spontaneous revertant fibers, which can be evaluated in the injured, PBS-injected contralateral legs. In both mdx groups, this frequency was >1%. In both models incorporating mdx4Cv, the reversion rate was much lower (0.43% in SCID/mdx4Cv and 0.34% in NSG-mdx4Cv). The low spontaneous reversion rate of the mdx4Cv mutation facilitates interpretation of fiber engraftment.

To determine the minimum number of Pax7-ZsGreen + satellite cells which would yield improvement in force-generating contractility, we transplanted mice with a dilution series of satellite cells going down to 100 cells per TA. This demonstrated that as few as 900 satellite cells are capable of providing the muscle with a statistically significant improvement in force generation (Fig. 4F).


Transplantation of muscle tissue into an immunodeficient recipient is an excellent way of assessing the muscle-generating capacity of a cell population. While the immune deficiency is essential for xenogeneic transplants, it also facilitates transplantation from donors of mixed lineage, common in mice carrying multiple genetic modifications. A genetic deficiency is also generally preferable to immune suppression for a number of reasons. The best immune suppressant, tacrolimus, requires daily injection—feasible for short-term studies but a serious impediment to long-term studies. In addition, tacrolimus is an inhibitor of mammalian target of rapamycin (mTOR). It is probably not advisable to block this anabolic pathway during muscle regeneration.

Early muscle transplants made use of the mdx/nude model [17], however this model is fragile and suffers from a reduced lifespan [33]. The SCID/mdx model [16] has become more popular, however, SCID mice are also relatively short-lived because they develop thymic lymphomas, especially after irradiation. By combining SCID with a deletion of the common gamma chain for the IL2 receptor (IL2Rg), not only are NK cells ablated, but thymic lymphomas, which require IL2Rg signaling, do not arise, therefore mice have normal lifespans if kept in sterile cages [34]. We currently have healthy 22-month-old NSG-mdx4Cv mice in our colony. Both the extended lifespan and the absence of NK cells have made the NSG (NOD/SCID;γ-c) mouse the favored model for human hematopoietic stem cell transplantation. By bringing mdx into the NSG model, these benefits are now available in the context of muscle cell transplantation for studies on muscle regeneration and therapies for Duchenne muscular dystrophy.

We elected to use a low-reversion mdx allele, mdx4Cv, for these studies. Numerous methods have been published to genotype mdx mice [35]. A sequencing method for genotyping the point mutation in mdx4Cv mice has also been described [36]. We recognized the loss of a BsmAI restriction site upon the C to T transition at base 7,916 in exon 53 in the mdx4cv allele. Thus, genotyping can be performed rapidly and simply, by digesting the PCR product spanning the mutation. The lower reversion rate was originally quantified in the uninjured quadriceps muscle from mdx4Cv animals [30]. Our work shows that revertant fibers are also less frequent in the irradiated, injured, TA muscle.

The loss of dystrophin produces a much weaker phenotype in mice compared to humans, nevertheless the NSG-mdx4Cv mice display the classic features expected for dystrophin deficiency in the mouse. When compared to the more commonly used mdx mouse, the NSG-mdx4Cv mouse was basically indistinguishable by every metric, including muscle histology, physiology, and response to injury. The NSG-mdx4Cv animals display continual damage/regeneration cycles, severe fibrosis especially of the diaphragm, high susceptibility to contraction-induced injury, and an overall deterioration in the muscle force-generation capacity. The mdx4Cv animals, like the mdx animals, presented extremely elevated CK activity in serum. The mdx animals showed higher serum CK activity than the mdx4Cv animals, however, we do not interpret this as indicative of a more severe disease. Although readings above baseline indicate leaky muscle cell membranes, variations between levels so far above this threshold do not correlate directly with severity. Serum CK levels do increase with leakier muscle cell membranes but will also decline with greater disease progression as more muscle is turned over to fat and fibrosis, leaving less CK available to leak into the serum.

To date there is no cure for Duchenne muscular dystrophy, which progresses relentlessly until death occurs, usually in the third decade. Current approaches in or near clinical trials target the muscle fiber, which like most cells of the body is eventually turned over and replaced, again necessitating periodic readministration. It would be advantageous to restore dystrophin in the cells that are responsible for building and regenerating muscle throughout life, namely the satellite cell pool. Satellite cells are amenable to transplantation, however, engraftment on a per cell basis is much better when whole fibers are transplanted [7] compared to FACS-isolated cell populations [8]. Perhaps because of this, clear evidence of improvement in strength following satellite cell transplantation (in the form of an increased mean force-generation capacity in a transplanted vs. a control group) is lacking.

The available evidence that skeletal muscle progenitors prospectively isolated from adult muscle by flow cytometry can impart a physiological improvement in muscle function is controversial [15]. Data from the only study claiming success showed very high variability, so rather than mean improvement in force generation comparing contralateral to control legs, which was not statistically significant, the authors argued for a functional improvement based on finding a correlation between rate of engraftment and increase in force of the injected versus control muscles [14]. One possible reason for the high variability in engraftment rates from mouse to mouse could be that immune suppression is not sufficient to completely prevent a response to either the MHC mismatch or to the dystrophin that the incoming cells express. Immune responses are mechanistically stochastic and therefore show high variability. Transplantation into an immune-deficient host would bypass this source of variability. Using the NSG-mdx4Cv mouse, we now demonstrate conclusively that FACS-isolated satellite cells have the capacity to effect the regeneration of muscle to greater strength than sham-transplanted controls.

With regard to the clinical potential of this approach, several hurdles would need to be crossed to develop a satellite cell transplantation therapy for Duchenne muscular dystrophy. The first is that donor cell numbers will likely be limited because satellite cell isolation requires destruction of donor muscle. In the experiments described here, whole hind limbs were sacrificed in donors for the benefit of a single muscle in recipients. Since donor muscle is destroyed in the process of satellite cell isolation, this equation would need to be inverted, that is, engraftment on a per cell basis would need to be increased significantly, probably by incorporating ex vivo expansion. The second hurdle is the difficulty of accessing the hundreds of muscles that would need to be treated by satellite cell transplantation. Therefore, while demonstration of force improvements is an excellent starting point, much work remains to unlock the therapeutic potential of satellite cells in muscular dystrophy.


The NSG-mdx4Cv mouse facilitates transplantation of allogeneic and xenogeneic cells and will facilitate studies on muscle regeneration and Duchenne muscular dystrophy. Using this model, we show conclusively that as few as 900 allogeneic FACS-isolated satellite cells have the capacity to rebuild muscle with increased mean force generation.


The project was supported by the NIH Grants R21 AG034370, R01 AR055299, R01 AR055685, K02 AG036827, and P30 AR0507220. R.W.A. was supported by the Minnesota Muscle Training Grant (NIH T32 AR007612). We thank American Preclinical Services, LLC for generously providing pig and dog muscle samples.

Disclosure of Potential Conflicts of Interest

The authors indicate no potential conflicts of interest.