Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Mol Ther. Author manuscript; available in PMC 2009 Dec 1.
Published in final edited form as:
PMCID: PMC2678899

Direct muscle delivery of GDNF with human mesenchymal stem cells improves motor neuron survival and function in a rat model of familial ALS


Amyotrophic Lateral Sclerosis (ALS) is a fatal neurodegenerative disease in which there is a progressive loss of motor neurons and their connections to muscle leading to paralysis. To maintain muscle connections in a rat model of familial ALS, we performed intramuscular transplantation with human mesenchymal stem cells (hMSC) as “Trojan horses” to deliver growth factors to the terminals of motor neurons as well as the skeletal muscles. hMSC engineered to secrete glial cell line derived neurotrophic factor (hMSC-GDNF) were transplanted bilaterally into three muscle groups. The cells survived within the muscle, released GDNF, and significantly increased the number of neuromuscular connections and motor neuron cell bodies in the spinal cord at mid stages of the disease. Furthermore, intramuscular transplantation with hMSC-GDNF could ameliorate motor neuron loss within the spinal cord which connected to the limb muscles with transplants. While disease onset was similar in all animals, hMSC-GDNF significantly delayed disease progression, increasing overall lifespan by up to 28 days, which is one of the longest effects on survival noted for this rat model of familial ALS. This pre-clinical data provides a novel and practical approach towards ex vivo gene therapy for ALS.


Amyotrophic lateral sclerosis (ALS) is a fatal neurodegenerative disease characterized by motor dysfunction that leads to eventual paralysis [1, 2]. The majority of ALS cases are of a sporadic nature, while ∼10% maintain a familial component. The cause of sporadic ALS remains unclear. However, a clear genetic link to point mutations in the cytosolic Cu2+/Zn2+ superoxide dismutase 1 (SOD1) has been shown in a small group of familial ALS (FALS) patients [3, 4]. Both mouse [5] and rat [6] models that over-express the G93A, G82R, or G37R mutations have been developed and show a similar disease phenotype and progression to that seen in both the familial and sporadic form of human cases. Degeneration of motor neurons is clearly a complex process and probably involves multiple pathways including formation of protein aggregates, axonal transport defects, oxidative damage, mitochondrial defects, alterations in calcium homeostasis, and finally cell death [7, 8]. Interestingly, there is mounting evidence that axonal atrophy and withdrawal from the muscle may occur early in the disease process [9].

Glial cell line derived neurotrophic factor (GDNF) has been shown to protect motor neurons in a number of different models [10-13], is present at high levels in embryonic limb and muscle at the time of innervation and necessary for normal neuromuscular development [13, 14]. GDNF also increases neural sprouting and prevents cell death [14-16]. Healthy motor neurons express GDNF receptor alpha (GFRα) and c-Ret, the heterodimer receptor system of GDNF, and can bind, internalize, and transport the protein in both antero- and retrograde directions in a receptor-dependent manner [17-19]. In transgenic mice over-expressing GDNF in muscle, a hyperinnervation of muscle by motorneurons has been reported [20]. However, delivery of GDNF to either motor neurons in the spinal cord or muscle end plates in adult animals is difficult due to the poor penetrance of this factor from the blood to body tissues. Recent studies have used gene therapy approaches to deliver GDNF and other growth factors directly to the muscle in a mouse model of familial ALS with encouraging results [11, 21]. However, there may be difficultly in translating these studies to larger species such as the rat and monkey. Following disease progression in ALS, loss of motor neurons leads to muscle atrophy, leading eventually to respiratory failure and death. In an attempt to prevent muscle atrophy and death of spinal motor neurons, an ex vivo gene therapy targeting skeletal muscles may be effective. This idea has successfully been proven by one previous report which showed that GDNF delivery to muscle using genetically modified myoblasts can produce protective effects on motor neurons in the spinal cord [11]. However, for clinical trials, it may be difficult to amplify cell number by passaging and obtain sufficient cells for biological significance after injection back into the muscle [22].

Human mesenchymal stem cells (hMSC) are found in bone marrow as well as in other mesenchymal tissues, are easy to harvest and can be expanded ex vivo to clinically relevant numbers while retaining their normal karyotype and differentiation capacity [23-25]. They appear to have a significant effect on disease progression in a number of animal models of human disease including heart damage, stroke, and Parkinson's disease, while the mechanism of this protective effect remains elucidate, they may include growth factor release or increased angiogenesis [26]. In the present study we show that combining the specific administration of GDNF with the therapeutic effects of hMSC transplants leads to significant improvement of both motor neuron survival and function in a well established rat model of familial ALS. As such it represents a unique, targeted, and practical therapeutic approach for this devastating disease.


hMSCGFP-GDNF survived and released GDNF in the skeletal muscle of SOD1G93A rats

hMSC were isolated from neonatal bone marrow aspirates from healthy donors after informed consent and confirmed by immunophenotyping and multilineage differentiation [27-29]. This line was then genetically modified to express green fluorescent protein (GFP) using retrovirus (hMSCGFP) [27]. hMSCGFP were infected with a lentiviral construct encoding GDNF under the control of phosphoglycerol kinase (pgk) promoter (Fig. 1a). GDNF protein was detected by ELISA in the conditioned medium and shown to be released at a rate of 480 pg/24 hours/million cells (Fig. 1b). Non-infected hMSCGFP did not release detectable amounts of GDNF. Immunocytochemical analysis using hMSCGFP-GDNF titer showed that approximately 98% and 95 % of total cells expressed GFP and GDNF protein, respectively (Fig. 1c).

Figure 1
Preparation of human mesenchymal stem cells to release GDNF by lentiviral infection

We next asked whether hMSCGFP-GDNF could be transplanted into the skeletal muscle of pre-symptomatic ALS rats (SOD1G93A rats). However, following direct implantation to intact muscle, there was very poor survival and integration of the cells (data not shown). It has previously been demonstrated that hMSC could integrate into the skeletal muscle of nude mice but required focal muscle injury using cardiotoxin prior to transplantation for survival [30]. The injury appears to have created a better environment for cell survival and integration, perhaps through the release of cytokines and other growth factors. In the present study we used intramuscular injection of local anesthetics, bupivacaine hydrochloride (BVC) to induce focal injury prior to transplantation [31, 32]. Female SOD1G93A rats (80 days, pre-symptomatic) were immunosuppressed with cyclosporine and then BVC was bilaterally injected into the tibialis anterior (TA) muscle (Supplementary Fig. S1). Twenty-four hours, one week and two weeks later hMSCGFP-GDNF (120,000 cells in 30 μl) were injected into the same muscle. All animals were and left until end point defined as the inability to right themselves after 30 seconds.

We first checked whether hMSC could survive transplantation in the muscle of SOD1G93A rats with prior toxin lesions. Even at disease end point, many grafted cells could be detected within the muscle using immunostaining with anti-GFP antibody (Fig. 2a, b) and GFP expression (Fig. 2c). There were no adverse effects in any of the animals and no sign of tumor formation. Immunostaining for laminin antibody to identify the basal lamina (Fig. 2d) revealed that GFP–positive hMSC resided both within and between the laminin-positive basal lamina and muscle fibers (designated by arrows and arrow head in Fig. 2d). The GFP signal did not extend to the sarcolemma of the adjacent myofibers, indicating that the cells had not fused with rat muscle fibers. Double staining with the human specific marker (hNUC) and GFP also showed surviving hMSC in the muscles (Fig. 2e). Additionally, our preliminary data showed that hMSC could also survive in the muscles after only a single transplantation after 24 hours of BVC injection. However, the number of surviving cells was increased with more injections (data not shown). We also performed RT-PCR on muscle tissues with primers specific for human or rat cDNA. At 2 days and 8 weeks post-transplantation, we detected human β-actin in the TA muscles injected with hMSCGFP-GDNF (Fig. 2f), while these mRNAs were not detected in the TA muscle without cells. Furthermore, the expression of human myosin heavy chain IIx/d (MyHC-IIx/d) gene was also detected and increased in the transplanted muscles, suggesting that some of the MSC transplanted in the muscles may acquire the skeletal muscle phenotype. To establish whether GDNF was being released by the hMSC, TA muscles from transplanted or control animals were dissected and then processed for immunohistochemistry using GFP and GDNF antibody (Fig. 2g). Significant amounts of GDNF could be detected in the areas with hMSCGFP cells, particularly between the basal lamina and muscle fibers (Fig. 2g). We also confirmed visual expression of GDNF with protein expression using ELISA which showed that hMSCGFP-GDNF released significant amounts of this growth factor into the skeletal muscle (Fig. 2h). Together this data show that adult human mesenchymal cells can survive well within rat muscle tissues and release GDNF providing a minor toxic lesion is present.

Figure 2
Survival of hMSC transplanted into the skeletal muscle of the SOD1G93A transgenic rats

hMSCGFP-GDNF transplants ameliorate denervation of neuromuscular junctions in hSODG93A rats

We next analyzed neuromuscular junctions (NMJs) within the region of the transplant in order to establish whether the GDNF had any effect on endplate innervation. For these studies, another cohort of 80 day old female SOD1G93A rats (n=13) were unilaterally injected with BVC in their TA muscle and after 24 hours, 1 week and 2 weeks transplanted with hMSCGFP-GDNF (n=5), hMSCGFP (n=4), or just received vehicle infusions (control; n=4). The level of innervation in the transplanted muscles was estimated by using double staining for axons and motor endplates based on the degree of acetylcholine receptor (AChR) post synaptic clusters (Fig. 3a-c). We have previously shown that while over 80% of endplates were innervated in the pre-symptomatic SOD1G93A rats up to 80 days old, this number gradually decreased to the point where all endplates were denervated by end-stage [33]. In the current study, approximately 75% of endplates were denervated in the muscle of non-transplanted control SOD1G93A rats at 122 days old (Fig. 3e, f). The number of denervated endplates tended to be reduced in hMSCGFP transplanted rats, although this difference did not reach statistical difference (Fig. 3e, f). However, there was a significant decrease in endplate denervation within hMSCGFP-GDNF group when compared with the non grafted control (Fig. 3e, f; P<0.05).

Figure 3
Neuromuscular junction innervation following hMSCGFP-GDNF transplantation

The localization AChR clusters at the endplate requires the expression of agrin, a large proteoglycan in the synaptic cleft that plays an important role in the maintenance of the molecular architecture of the post synaptic membrane [34]. A previous report showed that agrin expression was significantly reduced in the skeletal muscle of symptomatic SODG93A mice [35]. The number of agrin-positive endplates was reduced in non-transplanted SOD1G93A rats (23.1 ± 4.8% per total endplates) compared to hMSCGFP-WT (46.6 ± 15.8%) or GDNF (55.6 ± 18.3%) animals (Fig. 3d), further underscoring a role of hMSCGFP-GDNF in the maintenance of muscle innervation.

GDNF released by hMSC can maintain large cholinergic motor neurons in the ventral horn

In order to establish whether hMSCGFP-GDNF could ameliorate motor neuron loss, we next counted Nissl stained or choline acetyltransferase (ChAT) positive cells within the specific lumbar spinal cord region (L2-4) that projected the transplanted hind limb muscles. We previously confirmed that the TA muscle transplanted hMSC received projections from the spinal cord region analyzed [33]. While there was a significant loss of ChAT positive motor neurons in non transplanted SOD1G93A rats (Fig. 4a, c, e), this was prevented by transplants of either hMSCGFP or hMSCGFP-GDNF (Fig. 4g). However, the rate of protection was larger for the hMSCGFP-GDNF group (P<0.01 vs. control) when compared to the hMSCGFP(P<0.05 vs. control). Interestingly, the hMSCGFP-GDNF cells appeared to specifically protect the larger motor neuron pool, or increased the size of smaller surviving motor neurons based on detailed cell counting combined with size measurements (Fig. 4h).

Figure 4
Motor neurons are protected by GDNF secreting hMSC transplants in the skeletal muscle of SOD1G93A rats

hMSCGFP-GDNF transplants do not affect host glial cell activation

Activation of host glial cells is one hallmark of ALS, and increases with disease progression in the SOD1G93A rats which may be either protective or contribute to motor neuron death. Significant astrogliosis was detected in the ventral horn of the lumbar spinal cord in non-transplanted SOD1G93A rats (Supplementary Fig. S2). The numbers of host GFAP reactive astrocytes tended to be reduced in both hMSCGFP and hMSCGFP-GDNF transplanted rats at 6 weeks post transplantation, although that difference did not reach statistical significance. This suggested that the protective effects of hMSCGFP-GDNF were not mediated entirely through modulation of host reactive astrogliosis. We also determined the levels of microglial activation by using immunostaining for microglial marker OX-42 (CD11b). As seen for GFAP staining, there was no obvious difference in the level of microglial activation in the lumbar spinal cord regions which connected to the transplanted hind limb muscles, although the size and branching of many microglia were smaller in the transplanted animals (Supplementary Fig. S2)

hMSCGFP-GDNF transplants increased the survival of SOD1G93A rats

We next wanted to establish whether intramuscular hMSCGFP-GDNF transplantation could prolong survival period in SOD1G93A rats. First, we showed that partial muscle injury with BVC at 80 days old did not change disease onset and progression compared with non-treated (i.e. non partial injury using BVC) control SOD1G93A rats (data not shown). The efficacy of intramuscular transplantation was then tested using GDNF secreting hMSC in a new cohort of SOD1G93A rats. For these studies, we used two different cohorts of rats and the variation with regard to disease onset and progression in our SOD1G93A rats was observed as previously described [36] In a third cohort of SOD1G93A rats with early disease progression (n=9), hMSCGFP-GDNF were bilaterally transplanted into the TA, forelimb triceps brachii, the long muscles of the dorsal trunk muscles (Supplementary Fig. S1). The survival period in the hMSCGFP-GDNF transplanted rats was significantly prolonged by 18 days (P<0.05; Fig. 5a, b) compared to the control SOD1G93A littermates (n=9).

Figure 5
Prolonged survival by hMSC transplantation in two different colonies of SOD1G93A rats

We next asked whether this effect was reproducible in another colony of our SODG93A rats with slow disease progression [36] and further dissociate the effects of GDNF release from the effects of the cells themselves. A cohort of female SODG93A rats in a colony showing slow disease progression were transplanted at 80 days with either hMSCGFP(n=13), hMSCGFP-GDNF (n=14) or BVC-lesioned control (n=14). There was no difference in disease onset in either hMSCGFP or hMSCGFP-GDNF group when compared to control animals (Fig. 5c). However, hMSCGFP-GDNF transplantation significantly prolonged survival in the hSOD1G93A through slowing disease progression compared with lesioned control animals (P<0.05). Although hMSCGFP transplanted animals showed a delay in progression when compared to control animals, this did not reach statistical significance (P=0.08; Fig. 5d, e). The survival period in the hMSCGFP-GDNF transplanted rats was significantly prolonged by approximately 28 days compared to the control SOD1G93A littermates (Fig. 5e). This is one of the longest effects on survival noted for this rat model of FALS. We also tested limb function using the Basso-Beatti-Bresnahan (BBB) locomotor rating test. Motor dysfunction progressed significantly slower in both hMSCGFP and hMSCGFP-GDNF transplanted rats when compared to the control rats (P<0.05; Fig. 5f).


In the present study we used hMSC as long-term ‘mini pumps’ to deliver GDNF directly into the skeletal muscle of the SOD1G93A rats. The major advantage of using this ex vivo gene therapy approach (cellular delivery of growth factor), rather than direct in vivo gene therapy (force host cells to express growth factor through direct viral injection) is that the new cells provide the growth factor rather than host muscle cells that are undergoing degeneration due to the disease process [37, 38]. Furthermore, it is clear that hMSC have other poorly characterized effects on the local environment through either release of growth factors, reduction in acute inflammation or enhancement of angiogenesis [26]. Recent preclinical and clinical studies demonstrated a number of potential therapeutic applications of hMSC targeting for various diseases such as myocardial infarction, stroke, and graft-versus-host disease [26]. It was of great interest to us that the hMSC on their own had a strong trend towards increasing motor neuron number and improving functional performance. However, only when combined with GDNF did this reach significance.

Satellite cells and myoblasts represent the natural first choice in ex vivo therapeutics for skeletal muscle because of their intrinsic myogenic commitment, and have previously been used to deliver GDNF to the muscle in a mouse model of familial ALS with some success [11, 39]. However, for clinical trials, primary myoblasts are recovered in low number from biopsies, poorly expandable in vitro, and rapidly undergo senescence [22]. In contrast to human myoblasts, hMSC derived from the bone marrow can be easily expanded to clinically relevant numbers and differentiate, either in vivo or in vitro, into bone, stroma, cartilage, ligament/tendon, fat, and muscle [23-25, 30]. In the current study there was very poor survival of hMSC following direct injection hMSC into intact muscle. However, after mild injury the cells survived well and integrated suggesting that local injury induced cytokine release or other factors are required to encourage cell survival, as suggested in previous studies [30, 40-42]. We showed that partial injury did not change disease onset and progression in SODG93A rats, perhaps due to the fact that muscle repairs very rapidly. While BVC was used to induce muscular injury in these experiments, other types of milder muscular injury models such as toxic and mechanical methods (such as extreme exercise) or freezing may be sufficient as an alternative method to increase hMSC grafting in the skeletal muscles. This would clearly be more clinically relevant and is the focus of ongoing studies.

The underlying mechanisms of protective effects on neuromuscular connections and motor neuron survival remain to be determined. A recent observation, which used promoter specific expression of GDNF in either the muscle or spinal cord of the SOD1G93A mice, showed that muscle expression of GDNF was able to slow disease progression and onset, but the that expression of GDNF in the spinal cord had no effect [43]. Interestingly, GDNF is known to work as a synaptotrophic modulator for both pre- and postsynaptic differentiation to strengthen the functional and structural connections between nerve and muscle, and contribute to the maintenance and plasticity of neuromuscular synapses. In transgenic mice overexpressing GDNF, there is an increase in both axonal branching of motor neurons [14, 20] and the size, amplitudes and frequency of spontaneous synaptic currents [13]. Recent data also indicated that Ret tyrosine kinase receptor, which mediates the transmembrane signaling by GDNF, is localized specifically to the presynaptic membrane of both embryonic and adult neuromuscular synapse [44], suggesting that GDNF released from the muscle and/or terminal Schwann cells can promote both presynaptic maturation. This study using mouse primary cultured muscle cells showed that GDNF promotes the insertion and stabilization of postsynaptic AChR [45]. Interestingly, in the current study the level of agrin expression, which is necessary for the localization of AChR clusters at the endplate and the maintenance of the postsynaptic membrane in the synaptic cleft, was maintained in the hMSCGFP-GDNF transplanted muscle (Fig. 3d). Furthermore, it is also possible that indirect action either through distal activation of endogenous other unknown factors, or through synergistic effect with other trophic factors secreted by naïve hMSC such as IGF-I and VEGF, as hMSCGFP secret detectable level of IGF-I and VEGF in the conditioned medium (Suzuki and Svendsen, unpublished observation; Supplementary Fig. S3).

Transplantation of different types of stem cells into the spinal cord of rodent models of ALS have previously been shown to result in some motor neuron protection and functional improvement (reviewed in [38, 46]). Our own studies have also shown that transplantation of hNPC releasing GDNF directly into the lumbar spinal cord results in robust cellular migration into degenerating regions, efficient delivery of GDNF and remarkable preservation of host motor neurons [33, 47]. Surprisingly, however, this robust motor neuron survival was not accompanied by continued innervation of muscle end plates and thus resulted in no improvement in ipsilateral hind limb use [33]. Interestingly, increased survival of motor neurons but loss of functional connection to the muscle has been shown in several other models such as mice deficient in the proapoptotic protein Bax [48] and mutant mice displaying motor neuron degeneration (progressive motor neuropathy; pnm) [49]. Previous studies, using transgenic mice and viral vectors over expressing neuroprotective growth factors, showed that the level of growth factor-mediated rescue of motor neurons correlated with the level of decreased astrogliosis and microglial activation at different stages of disease [43, 50]. However, in the present study, hMSCGFP-GDNF transplants to the muscle did not reduce glial activation in the spinal cord. Our previous results transplanting human cortical neural progenitor cells releasing GDNF into he spinal cord of this rat model of ALS also had no effect host glial activation, but could protect motor neurons [33]. Therefore, we suggest that while reduction of glial activation may be associated with motor neuron survival in some cases, it is not a required event.

As muscle is an accessible tissue, the present study points to a very straightforward approach to treating ALS with a cell type and growth factor that have both been used in human clinical trials. At present we have only targeted three muscle groups, but it is possible that further injections of cells may enhance this effect on survival even further. In particular, targeting muscles of the diaphragm might be of great importance for protecting respiratory motor neurons. Furthermore, it might be that combining stem cell and growth factor delivery targeting to both the skeletal muscles (i.e. nerve terminals of motor neurons) and spinal cord (i.e. cell body) would provide the optimum combination for trying to slow the progression of this devastating disease [38].

Materials and Methods


Sprague Dawley SOD1G93A rats, which had been originally generated by Howland et al. [6], were obtained from Taconic (Hudson, NY), and colonies were developed by crossing male founders with wild-type female rats. Heterozygous SOD1G93A progeny were identified with genomic PCR of tail DNA with primers specific for hSOD1. We used two different cohorts of rats and the variation with regard to disease onset and progression in our SOD1G93A rats was observed as previously described [36]. They were maintained in a room with controlled illumination (lights on 0600-1800 h) and temperature (23 ± 1 °C), and given free access to laboratory chow and tap water. Rats were considered end-stage (endpoint) when they no longer exhibited reflexes allowing them to right themselves within 30 seconds. All the procedures in the present study were carried out in accordance with the guidelines for University of Wisconsin-Madison and NIH standards of animal care.

Preparation of human mesenchymal cells expressing GDNF

Human mesenchymal cells were established using methods previously described [27-29]. These cells were derived from neonatal bone marrow aspirates from healthy donors after informed consent. Blood tissue collection for research purposes was approved by the Research Ethics Committees of Hammersmith Hospital and Queen Charlotte's Hospital. Single-cell suspensions of neonatal bone marrow were prepared by collecting the bone marrow cells in the humeri and femurs using a syringe and 22-gauge needle. Unselected nucleated cells were plated in 100-mm dishes at a density of 105/ml and cultured in culture medium into culture medium [Dulbecco modified Eagle medium (DMEM; Sigma-Aldrich, St. Louis, MO) supplemented with 20% fetal bovine serum (FBS; Hyclone, Logan, UT), 2 mM L-glutamine, 50 U/ml penicillin, and 50 mg/ml streptomycin (Invitrogen, Carlsbad, CA)]. After three days, non-adherent cells were removed and the medium was replaced. When colony size exceeded >500 cells, single colonies were trypsinized and collected using clonal cylinders. The mesenchymal nature of adherent nucleated cells was confirmed by immunophenotyping and multilineage differentiation [27-29]. In the current study, we used a hMSC line genetically modified to express green fluorescent protein (GFP) using retrovirus (N97-GFP; hMSCGFP) [27]. This line was transduced with retroviral vector containing enhanced GFP using second passage hMSC.

To the preparation of hMSC expressing GDNF, we used our experience with neural progenitor cells. Our previous work has shown that both rat and human NPC can be efficiently infected with specific lentiviral constructs [33, 47]. We used a same viral construct for constitutive expression of GDNF under the control of the mouse phosphoglycerate kinase 1 (PGK) promoter [33]. hMSCGFP cultures were gently dissociated with 0.05% Trypsin-EDTA to single cell suspension, washed extensively, counted and diluted to 500 cells/μl in the culture medium, and plated for 24 hours in a 24 well plate. The cells were then incubated for 24 hrs in medium containing various virus titers (0-100 ng/p24/million cells). GDNF concentration in the conditioned medium was determined by ELISA kits (R&D Systems, Minneapolis, MN).

Intramuscular hMSC transplantation

According to our pilot experiments, we need to induce focal muscular injury before hMSC transplantation. Thus, we used intramuscular injection of local anesthetics, bupivacaine hydrochloride to induce focal injury. Similarly to cardiotoxin, bupivacaine has broadly been used to make a partial muscular lesion [31, 32]. SOD1G93A female rats at 80 days of age were anesthetized and injected with a local anesthetic, bupivacaine hydrochloride (0.35 mg; Sensorcaine-MPF, AstraZeneca, Miami, FL), bilaterally into the tibialis anterior (TA), forelimb triceps brachii, the long muscles of the dorsal trunk muscles (Supplementary Fig. S1). A 30-gauge needle connected to a 1 ml syringe was used for injections. Twenty-four hours later, culture-expanded hMSCGFP (120,000 cells in 30 μl) were injected into the same muscles using a 33-gauge needle connected to a 100 μl Hamilton syringe. The injection of hMSC was repeated twice with a one-week interval. All animals were immunosuppressed with cyclosporine (10 mg/kg/day) and left until end point.


The animals were transcardially perfused with chilled 0.9% saline and 4% paraformaldehyde-phosphate buffered saline (PFA-PBS), and the spinal cords were collected. To analyze hMSC survival and neuromuscular junctions, tibalis anterior (TA) muscles of some animals were removed prior to fixation and flash-frozen in supercooled isopentane. Muscles were sectioned at 20 μm using a cryostat and placed on glass slides for staining. The sections were fixed with 4% PFA-PBS and labeled with human nuclei (hNUC, mouse monoclonal, 1: 100, Millipore, Billerica, MA), human GDNF (goat polyclonal, 1:250, R&D Systems Inc.), GFP (mouse monoclonal, 1: 100, Invitrogen), and laminin (rabbit polyclonal, 1: 500, Sigma-Aldrich). For motor neuron counting [33], spinal cords were cryoprotected in 30% sucrose with 0.1% sodium azide, and the lumbar regions (L2-4) were cut in 50 μm sections using a cryostat. One in six sections were processed for Nissl staining or immunostained for human nuclei (hNUC, mouse monoclonal, 1: 100, Chemicon), human GDNF (goat polyclonal, 1:250, R&D Systems Inc.), choline acetyl transferase (ChAT, goat polyclonal, 1:100, Chemicon), glial fibrillary acidic protein (GFAP, rabbit polyclonal 1:2000, Dako, Glostrup, Denmark), and OX-42 (CD11b, 1:500, MorphoSys, Planegg, Germany). Primary antibodies were followed by secondary antibodies conjugated to Cy3 or Alexa Fluor 488 (anti-IgG, 1:1000, Jackson ImmunoResearch Laboratories, West Grove, PA) or biotinylated secondary antibody (1:200, Jackson ImmunoResearch Laboratories) followed by DAB with avidin–biotin method (Vectors Laboratories Inc. Burlingame, CA). Nickel ammonium sulfate enhancement was used for some sections labeled with GFP or GDNF (Fig. 2g). All images were captured by using a fluorescence microscope (Nikon, Tokyo, Japan) with a digital camera (SPOT II) or a laser scanning confocal microscope (Nikon). To quantify astrocyte activation the number of GFAP positive cells on both sides of the spinal cord was counted using Metamorph Imaging software (Molecular devices, Downingtown, PA), presented as a percentage of the total number of counted cells. To quantify astrocyte and microglial activation the number of GFAP positive cells and the mean intensity of OX-42 fluorescence on both sides of the spinal cord was determined as described previously. [33]


Total RNA isolation and RT-PCR were performed using TRIzol reagent (Invitrogen) and GeneAmp RNA PCR kit (Roche Molecular Systems, Basel, Switzerland). For RT-PCR, the following primer pairs were used: human myosin heavy chain IIx/d (MyHC-IIx/d) [30], human β-actin [30], and rat GAPDH (R&D Systems Inc.). The amplification profile was 94 °C for 0.5 min, 58 °C for 1 min, and 72 °C for 1 min. All samples were incubated for 28 cycles.

Behavioral testing

Body weight measurements and all behavioral testing began at an age of 65 days and continued until endpoint. Rats were considered end-stage (endpoint) when they no longer exhibit reflexes allowing them to right themselves within 30 seconds. To analyze motor function in SOD1G93A rats, the Basso-Beattie-Bresnahan (BBB) locomotor rating test and a beam walking test were performed twice a week. Open field locomotor scores was obtained in a small enclosure using the BBB locomotor rating scale as described previously [33, 36, 47]. Briefly, each rat was allowed to walk around in a wading pool with a textured floor while we observed hind and fore limb movements for approximately 3 to 5 minutes. Each hind limb score was based on the 21 point scoring scale from no movement (0) to normal locomotion (21). Scoring takes into account paw rotation, toe clearance, weight support, the frequency of each, and the amount of movement occurring from each joint. Each score was then recorded and summed for each hind limb to determine that limb's score. The total score was obtained by adding both hind limb scores. Disease onset was estimated by using the BBB rating score of 17 or lower [36].

Assessment of neural muscular junction (NMJs) innervation

Endplate innervation was analyzed as described previously [33]. Tibalis anterior muscles were dissected and flash-frozen in supercooled isopentane. Muscles were sectioned at 20 μm using a cryostat and placed on glass slides for staining. The sections were fixed with 4% PFA-PBS and labeled with alpha-bungarotoxin conjugated with fluorescence marker Alexa Fluor 594 (1:1000, Invitrogen), anti-neurofilament 160 (mouse monoclonal, 1:40, Sigma-Aldrich) and anti-synaptophysin (rabbit polyclonal, 1:40, Dako) antibodies for 1 hour at room temperature. For agrin staining, the sections were double-labeled with Alexa Fluor 594-conjugated bungarotoxin and anti-rat agrin antibody (mouse monoclonal, 1:500, StressGen). After washing with PBS, the sections were incubated with both anti-mouse and rabbit Alexa Fluor 488-conjugated secondary antibodies (1:1,000, Jackson Laboratories) for 1 hour at room temperature. Following washing, the sections were covered with cover glasses using aqueous mounting medium (Immuotech, Marseilles, France). Axons and NMJs were imaged using a fluorescence microscope (Nikon) and Metamorph Imaging software (Universal Imaging Corporation). For quantitative analysis, we classified NMJs into three groups based on degree of innervation of postsynaptic receptor plaques by nerve terminals [33]. Endplates were scored as “innervated” if there was overlap with the axon terminal, or “denervated” if the end-plate was not associated with an axon [33]. About 50 endplates were exhaustively analyzed. The number of endplates of each category was presented as a percentage of the total number of counted NMJs.

Statistical Analysis

Prizm software (Graphpad software Inc., La Jolla, CA) was used for all statistical analyses. All counting data from immunohistochemical analysis and survival periods of the animals were expressed as means ± SEM and analyzed by two-tailed t-test. Statistical analysis of disease onset and survival was performed using the Kaplan-Meier method with log rank test. Differences in BBB score were calculated using two-way ANOVA with Bonferonni post hoc test. Differences were considered significant when P<0.05.

Supplementary Material


This work was supported by the grant from the ALS association, NIH/NINDS [1PO1NS057778 (C.N.S.) and 1R21NS06104 (M.S.)], the University of Wisconsin Foundation, and the Les Turner ALS foundation. We gratefully acknowledge the lentiviral production by Dr. Romain Zufferey and Ms. Viviane Padrun.


1. Brown RH., Jr. Amyotrophic lateral sclerosis: recent insights from genetics and transgenic mice. Cell. 1995;80:687–692. [PubMed]
2. Cleveland DW, Rothstein JD. From Charcot to Lou Gehrig: deciphering selective motor neuron death in ALS. Nat. Rev. Neurosci. 2001;2:806–819. [PubMed]
3. Rosen DR, Siddique T, Patterson D, Figlewicz DA, Sapp P, Hentati A, et al. Mutations in Cu/Zn superoxide dismutase gene are associated with familial amyotrophic lateral sclerosis. Nature. 1993;362:59–62. [PubMed]
4. Andersen PM, Sims KB, Xin WW, Kiely R, O'Neill G, Ravits J, et al. Sixteen novel mutations in the Cu/Zn superoxide dismutase gene in amyotrophic lateral sclerosis: a decade of discoveries, defects and disputes. Amyotroph. Lateral. Scler. Other Motor Neuron Disord. 2003;4:62–73. [PubMed]
5. Gurney ME, Pu H, Chiu AY, Dal Canto MC, Polchow CY, Alexander DD, et al. Motor neuron degeneration in mice that express a human Cu,Zn superoxide dismutase mutation. Science. 1994;264:1772–1775. [PubMed]
6. Howland DS, Liu J, She Y, Goad B, Maragakis NJ, Kim B, et al. Focal loss of the glutamate transporter EAAT2 in a transgenic rat model of SOD1 mutant-mediated amyotrophic lateral sclerosis (ALS) Proc. Natl. Acad. Sci. U. S. A. 2002;99:1604–1609. [PMC free article] [PubMed]
7. Julien JP. Amyotrophic lateral sclerosis. unfolding the toxicity of the misfolded. Cell. 2001;104:581–591. [PubMed]
8. Rowland LP, Shneider NA. Amyotrophic lateral sclerosis. N. Engl. J. Med. 2001;344:1688–1700. [PubMed]
9. Fischer LR, Culver DG, Tennant P, Davis AA, Wang M, Castellano-Sanchez A, et al. Amyotrophic lateral sclerosis is a distal axonopathy: evidence in mice and man. Exp. Neurol. 2004;185:232–240. [PubMed]
10. Henderson CE, Phillips HS, Pollock RA, Davies AM, Lemeulle C, Armanini M, et al. GDNF: a potent survival factor for motoneurons present in peripheral nerve and muscle. Science. 1994;266:1062–1064. [PubMed]
11. Mohajeri MH, Figlewicz DA, Bohn MC. Intramuscular grafts of myoblasts genetically modified to secrete glial cell line-derived neurotrophic factor prevent motoneuron loss and disease progression in a mouse model of familial amyotrophic lateral sclerosis. Hum. Gene Ther. 1999;10:1853–1866. [PubMed]
12. Acsadi G, Anguelov RA, Yang H, Toth G, Thomas R, Jani A, et al. Increased survival and function of SOD1 mice after glial cell-derived neurotrophic factor gene therapy. Hum. Gene Ther. 2002;13:1047–1059. [PubMed]
13. Wang CY, Yang F, He XP, Je HS, Zhou JZ, Eckermann K, et al. Regulation of neuromuscular synapse development by glial cell line-derived neurotrophic factor and neurturin. J. Biol. Chem. 2002;277:10614–10625. [PubMed]
14. Keller-Peck CR, Feng G, Sanes JR, Yan Q, Lichtman JW, Snider WD. Glial cell line-derived neurotrophic factor administration in postnatal life results in motor unit enlargement and continuous synaptic remodeling at the neuromuscular junction. J. Neurosci. 2001;21:6136–6146. [PubMed]
15. Blesch A, Tuszynski MH. GDNF gene delivery to injured adult CNS motor neurons promotes axonal growth, expression of the trophic neuropeptide CGRP, and cellular protection. J. Comp Neurol. 2001;436:399–410. [PubMed]
16. Deshpande DM, Kim YS, Martinez T, Carmen J, Dike S, Shats I, et al. Recovery from paralysis in adult rats using embryonic stem cells. Ann. Neurol. 2006;60:32–44. [PubMed]
17. Glazner GW, Mu X, Springer JE. Localization of glial cell line-derived neurotrophic factor receptor alpha and c-ret mRNA in rat central nervous system. J. Comp Neurol. 1998;391:42–49. [PubMed]
18. Leitner ML, Molliver DC, Osborne PA, Vejsada R, Golden JP, Lampe PA, et al. Analysis of the retrograde transport of glial cell line-derived neurotrophic factor (GDNF), neurturin, and persephin suggests that in vivo signaling for the GDNF family is GFRalpha coreceptor-specific. J. Neurosci. 1999;19:9322–9331. [PubMed]
19. von Bartheld CS, Wang X, Butowt R. Anterograde axonal transport, transcytosis, and recycling of neurotrophic factors: the concept of trophic currencies in neural networks. Mol. Neurobiol. 2001;24:1–28. [PubMed]
20. Nguyen QT, Parsadanian AS, Snider WD, Lichtman JW. Hyperinnervation of neuromuscular junctions caused by GDNF overexpression in muscle. Science. 1998;279:1725–1729. [PubMed]
21. Kaspar BK, Llado J, Sherkat N, Rothstein JD, Gage FH. Retrograde viral delivery of IGF-1 prolongs survival in a mouse ALS model. Science. 2003;301:839–842. [PubMed]
22. Cossu G, Mavilio F. Myogenic stem cells for the therapy of primary myopathies: wishful thinking or therapeutic perspective? J. Clin. Invest. 2000;105:1669–1674. [PMC free article] [PubMed]
23. Prockop DJ. Marrow stromal cells as stem cells for nonhematopoietic tissues. Science. 1997;276:71–74. [PubMed]
24. Pittenger MF, Mackay AM, Beck SC, Jaiswal RK, Douglas R, Mosca JD, et al. Multilineage potential of adult human mesenchymal stem cells. Science. 1999;284:143–147. [PubMed]
25. Deans RJ, Moseley AB. Mesenchymal stem cells: biology and potential clinical uses. Exp. Hematol. 2000;28:875–884. [PubMed]
26. Picinich SC, Mishra PJ, Mishra PJ, Glod J, Banerjee D. The therapeutic potential of mesenchymal stem cells. Cell- & tissue-based therapy. Expert. Opin. Biol. Ther. 2007;7:965–973. [PubMed]
27. Campagnoli C, Bellantuono I, Kumar S, Fairbairn LJ, Roberts I, Fisk NM. High transduction efficiency of circulating first trimester fetal mesenchymal stem cells: potential targets for in utero ex vivo gene therapy. BJOG. 2002;109:952–954. [PubMed]
28. Baxter MA, Wynn RF, Jowitt SN, Wraith JE, Fairbairn LJ, Bellantuono I. Study of telomere length reveals rapid aging of human marrow stromal cells following in vitro expansion. Stem Cells. 2004;22:675–682. [PubMed]
29. Wynn RF, Hart CA, Corradi-Perini C, O'Neill L, Evans CA, Wraith JE, et al. A small proportion of mesenchymal stem cells strongly expresses functionally active CXCR4 receptor capable of promoting migration to bone marrow. Blood. 2004;104:2643–2645. [PubMed]
30. De BC, Dell'Accio F, Vandenabeele F, Vermeesch JR, Raymackers JM, Luyten FP. Skeletal muscle repair by adult human mesenchymal stem cells from synovial membrane. J. Cell Biol. 2003;160:909–918. [PMC free article] [PubMed]
31. Steer JH, Mastaglia FL, Papadimitriou JM, Van B,I. Bupivacaine-induced muscle injury. The role of extracellular calcium. J. Neurol. Sci. 1986;73:205–217. [PubMed]
32. Hill M, Wernig A, Goldspink G. Muscle satellite (stem) cell activation during local tissue injury and repair. J. Anat. 2003;203:89–99. [PMC free article] [PubMed]
33. Suzuki M, McHugh J, Tork C, Shelley B, Klein SM, Aebischer P, et al. GDNF secreting human neural progenitor cells protect dying motor neurons, but not their projection to muscle, in a rat model of familial ALS. PLoS. ONE. 2007;2:e689. [PMC free article] [PubMed]
34. McConville J, Vincent A. Diseases of the neuromuscular junction. Curr. Opin. Pharmacol. 2002;2:296–301. [PubMed]
35. Dobrowolny G, Giacinti C, Pelosi L, Nicoletti C, Winn N, Barberi L, et al. Muscle expression of a local Igf-1 isoform protects motor neurons in an ALS mouse model. J. Cell Biol. 2005;168:193–199. [PMC free article] [PubMed]
36. Suzuki M, Tork C, Shelley B, McHugh J, Wallace K, Klein SM, et al. Sexual dimorphism in disease onset and progression of a rat model of ALS. Amyotroph. Lateral. Scler. 2007;8:20–25. [PubMed]
37. Aebischer P, Kato AC. Playng defense against Lou Gehrig's disease. Sci. Am. 2007;297:86–93. [PubMed]
38. Suzuki M, Svendsen CN. Combining growth factor and stem cell therapy for amyotrophic lateral sclerosis. Trends Neurosci. 2008;31:192–198. [PubMed]
39. Grounds MD, White JD, Rosenthal N, Bogoyevitch MA. The role of stem cells in skeletal and cardiac muscle repair 1. J. Histochem. Cytochem. 2002;50:589–610. [PubMed]
40. Labarge MA, Blau HM. Biological progression from adult bone marrow to mononucleate muscle stem cell to multinucleate muscle fiber in response to injury. Cell. 2002;111:589–601. [PubMed]
41. Corbel SY, Lee A, Yi L, Duenas J, Brazelton TR, Blau HM, et al. Contribution of hematopoietic stem cells to skeletal muscle. Nat. Med. 2003;9:1528–1532. [PubMed]
42. Palermo AT, Labarge MA, Doyonnas R, Pomerantz J, Blau HM. Bone marrow contribution to skeletal muscle: a physiological response to stress. Dev. Biol. 2005;279:336–344. [PubMed]
43. Li W, Brakefield D, Pan Y, Hunter D, Myckatyn TM, Parsadanian A. Muscle-derived but not centrally derived transgene GDNF is neuroprotective in G93A-SOD1 mouse model of ALS. Exp. Neurol. 2007;203:457–471. [PubMed]
44. Baudet C, Pozas E, Adameyko I, Andersson E, Ericson J, Ernfors P. Retrograde signaling onto Ret during motor nerve terminal maturation. J. Neurosci. 2008;28:963–975. [PubMed]
45. Yang LX, Nelson PG. Glia cell line-derived neurotrophic factor regulates the distribution of acetylcholine receptors in mouse primary skeletal muscle cells. Neuroscience. 2004;128:497–509. [PubMed]
46. Turner BJ, Talbot K. Transgenics, toxicity and therapeutics in rodent models of mutant SOD1-mediated familial ALS. Prog. Neurobiol. 2008 [PubMed]
47. Klein SM, Behrstock S, McHugh J, Hoffmann K, Wallace K, Suzuki M, et al. GDNF delivery using human neural progenitor cells in a rat model of ALS. Hum. Gene Ther. 2005;16:509–521. [PubMed]
48. Gould TW, Buss RR, Vinsant S, Prevette D, Sun W, Knudson CM, et al. Complete dissociation of motor neuron death from motor dysfunction by Bax deletion in a mouse model of ALS. J. Neurosci. 2006;26:8774–8786. [PubMed]
49. Sagot Y, Tan SA, Hammang JP, Aebischer P, Kato AC. GDNF slows loss of motoneurons but not axonal degeneration or premature death of pmn/pmn mice. J. Neurosci. 1996;16:2335–2341. [PubMed]
50. Dodge JC, Haidet AM, Yang W, Passini MA, Hester M, Clarke J, et al. Delivery of AAV-IGF-1 to the CNS extends survival in ALS mice through modification of aberrant glial cell activity. Mol. Ther. 2008;16:1056–1064. [PMC free article] [PubMed]
PubReader format: click here to try


Save items

Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


  • MedGen
    Related information in MedGen
  • PubMed
    PubMed citations for these articles
  • Substance
    PubChem Substance links

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...