Clinical studies were approved by the ethical committee of the School of Veterinary Medicine and Animal Science of São Paulo University. The animals were identified by numbers and experimental procedures were described in Table 1. DMD genotyping was done from blood genomic DNA extracted with GFX Genomic Blood DNA Purification Kit (GE Healthcare, Piscataway, NJ – USA/Canada) and established as previously reported 18. DMD diagnosis was confirmed by restriction digestion of PCR products of the dystrophin gene with Sau96I and by elevated serum creatine kinase (CK).

Normal mice were used to investigate the ability of hIDPSC migration by intraperitoneal injections and engraftment capacity of these cells without immunosuppression.

Cells were obtained and characterized as described previously 13. Shortly, dental pulp was extracted from normal exfoliated human deciduous teeth of 5- to 7-year-old children under local anesthetic. Tissue explant of dental pulp was used to isolate the cells. Human IDPSC cultures were maintained in DMEM/F12 (Dulbecco's-modified Eagle's medium/Ham's F12, 1:1, Invitrogen, Carlsbad, CA) supplemented with 15% fetal bovine serum (FBS, Hyclone, Washington), 100 units/ml penicillin, 100 units/ml streptomycin, 2 mM L-glutamine, and 2 mM nonessential amino acids. Cells were maintained semi-confluent to prevent their differentiation and replaced every 4 or 5 days. Medium was replaced daily. Human IDPSC were incubated at 37°C in a high humidity environment with 5% CO2.

Monoclonal anti-human SH2 (CD105, Serotec, Oxford, UK), SH3 and SH4 (CD73) (Case Western Reserve University, Cleveland, Ohio, USA) antibodies against cell surface molecules and their respective isotype controls were used in flow cytometry analysis. About 10⁶ cells were incubated with primary antibody for 30 minutes at 4°C and washed in PBS with 2% FBS and 1 M sodium azide (buffer) followed by addition of secondary anti-mouse-PE – conjugated antibody according to manufacturer's instructions (Becton Dickinson, NJ, USA; Guava Technologies, CA). Negative controls were also performed incubating the cells in PBS followed by incubation with respective secondary antibody only. Results were analyzed using Guava Express^®Plus software (Guava Technologies, Hayward, CA, USA). Flow cytometry was analyzed using a fluorescence-activated cell sorter (FACS, Becton, Dickinson, San Jose, CA) with CELL Quest program (Becton, Dickinson, San Jose, CA).

Cells were cultured in DMEM-HG medium supplemented with 10% fetal bovine serum (FBS, Hyclone, Washington), 5% horse serum, 0.1 M dexamethasone, 50 nM hydrocortisone (Sigma Aldrich, São Paulo, SP), 1% antibiotic (100 units/ml penicillin and 100 mg/ml streptomycin; Invitrogen, São Paulo, SP) for 60 days.

The culture cell was washed twice in calcium and magnesium-free Dulbecco's phosphate-buffered solution (DPBS, Invitrogen) and dissociated with 0.25% trypsin/EDTA solution (Invitrogen). The suspension was centrifuged and the cell pellet was ressuspended in DMEM (Invitrogen) with 10% FBS (Invitrogen) containing the fluorescent dye (Vybrant CM-Dil Cell-Labelling Solution; Molecular Probes, Invitrogen). Cells were incubated for 15 minutes at 37°C, washed twice in DPBS immediately prior to intraperitoneal injection in mice.

Biceps femoralis biopsy samples were taken before the hIDPSC transplantation experiment on day 0 (d 0) and after 47, 107 and 117 days in different animals as summarized in Table 1. A second biopsy was taken from the male under systemic SC, delivery (MT1-S), after one year of treatment. Each biopsy from biceps femuralis was divided into 3 fragments for histological, immunofluorescence/WB and FISH analyses. For IF analysis, tissue samples were embedded in Jung Tissue Freezing Medium (Leica Mycrosystems Nussloch GmbH, Nussilosh, Germany) and frozen in liquid nitrogen. For WB analysis, 2 mm² fragments were frozen directly in liquid nitrogen in small flasks. For histological analysis routine hematoxylin and eosin (HE) staining was applied.

Differentiated cells were fixed in 4% paraformaldehyde for 2 h at 4°C, washed three times in wash buffer (150 mM NaCl, 1 mg/ml BSA, 0.5% Nonidet P- 40, 50 mM Tris pH 6.8) and membranes were permeabilized with two 10 min incubations in RIPA (150 mM NaCl, 1% Nonidet P-40, 0.5% sodium deoxycolate, 0.1% SDS, 1 mM EDTA, 50 mM Tris pH 8.0). Cells were subsequently re-fixed in 4% PFA for 30 min at 4°C, blocked in wash buffer for 1 h at 4°C and incubated overnight at 4°C with primary antibodies. Cy3-conjugated goat anti-(mouse IgG) secondary antibody and FITC-conjugated goat anti-(rabbit IgG) secondary antibody, normal mouse IgG and normal rabbit IgG were used. The primary antibody was omitted from some slides to serve as a negative control.

Myogenic differentiation was characterized using mouse anti-human muscle α-actinin antibody and rabbit anti-human muscle myosin antibody (Santa Cruz Biotechnology, CA, USA) and by RT-PCR using MyoD1(MD1) forward primer 5'AAGCGCCATCTCTTGAGGTA3' and reverse primer 3'GCCTTTATTTTGATCACC5' following the protocol described previously13.

Tissues samples were excised, immediately frozen in liquid nitrogen and then stored at -80°C. Cryostat tissue sections with 5 μm thickness were mounted on a glass slide. According to the antibody used, slides were fixed and dehydrated with cold methanol (Merck, Darmstadt, Germany). To detect the presence of hIDPSC mouse, anti-hIDPSC antibody previously obtained 13 was used at a 1:500 dilution, whereas mouse anti-human nuclei antibody (Monoclonal, Chemicon International, California, USA) was used at a 1:100 dilution followed by FITC-conjugated secondary antibodies of respective isotype at a 1:1000. Dystrophin analysis was done using IF by applying the methodologies from our group 19. Human specific anti-dystrophin monoclonal Mandys106 2C6 antibody was kindly provided by Dr. Glenn E. Morris at Center for Inherited Neuromuscular Diseases, Oswestry, UK 20. Antibody DYS2 (Vector Laboratories – Burlingame, CA) against the C-terminal domain of dystrophin and anti-mouse IgG Cy3- conjugated secondary antibody were also used to confirm the presence of some positive dystrophin fibers. Microscope slides were mounted in Vectashield mounting medium (Vector Laboratories) with or without 4',6-Diamidino-2-phenylindol (DAPI). IF analysis was done in a Zeiss Imager Z1 Apotome microscope with epi-fluorescence, or using an argon ion laser scan microscope LSM 410 (Zeiss – Jena, Germany).

Confocal microscopy: An argon ion laser set at 488 nm for FITC and at 536 for rodamine excitation were used. The emitted light was filtered with a 505 nm (FITC) and 617 nm (rodamine) long pass filter in a laser scan microscope (LSM 410, Zeiss – Jena, Germany). Sections (5 mm) were taken approximately at the mid – height level of tissues. Photo-multiplier gain and laser power were kept constant throughout each experiment.

FISH analysis was done with specific centromere probes for human chromosomes X (green) and Y (red) (Aquiarius-Cytocell, Cambridge, UK). Tissue samples were fixed in 4% paraformaldehyde, embedded in Jung Tissue Freezing Medium and 5 μm sections were made.

The slides with sections were then fixed with cold methanol (Merck, Darmstadt, Germany) and dehydrated in series of ethanol. FISH reactions were done according to the manufacturer's protocol. Microscope slides were mounted in antifade (Vectashield mounting medium) with or without Propidium Iodide (PI, Vectashield mounting medium/PI). FISH analysis was made using confocal microscope as described above.

Dogs underwent periodic veterinary examinations during treatment. Blood samples were collected monthly by jugular venipuncture. Hematological and serum biochemical testing were performed with an automated cell counter (Baker System 9000, Serno-Baker Diagnostics Inc, Allentown, Pennsylvania). Urea, alanine aminotransferase (ALT), alkaline phosphatase (ALP), creatine kinase (CK) and creatinine were measured in serum with an automated analyzer (Labtest^® – LABTEST Diagnóstica S.A. – Lagoa Santa, MG). Data are shown in Table 2.

We established a physical exam score (Table 3), adapted from literature 21 that evaluated dogs motility and posture, and which gave values between 0 (normal) to 24 (severely affected).

Herein we showed that hIDPSC were capable to migrate and engraft into GRDM dog muscle as confirmed by FISH and immunohystochemical analyses. Due to our ethical committee rules which is against dogs euthanasia or taking multiple biopsies we did not quantify the engraftment of these cells into different muscles, however all analyzed biopsy fragments demonstrated the presence of hIDPSC. We observed that these cells were able to form chimeric canine/human fibers with recipient muscles, although the expression of human dystrophin was observed only in limited number of the fibers following early, multiple cell transplantation. Additionally, these cells persisted in the host muscle for at least 1 year. Moreover, it has been suggested that thawing might modify the properties of SC 22.

Two independent groups have previously reported the transplantation of different types of canine stem cells in the GRMD model. In the first study, 7 dogs received canine allogenic hematopoietic cell transplantation at ages 4.5 to 5.5 months 10. However, donor cells did not show any significant contribution to the injured skeletal muscle nor significant increase of dystrophin positive fibers, although near-complete or complete donor hematopoietic chimerism was detected in physically active dogs. The authors concluded that allogenic bone marrow cell transplantation is not a viable therapy in its current form due to the absence of clear clinical benefits 10. The second group used canine heterologous wild type (WT) mesoangioblasts and allogenic genetically modified mesoangioblasts 11. The authors reported GRMD muscle function improvement and high level of dystrophin expression in particular after heterologous wild-type (WT) mesoangioblasts transfer, but less with autologous genetically modified SC. Both early and late SC transplantation showed very promising results 11. However, since immunosuppression was applied in both studies; clinical interpretation of the results is difficult because the use of anti-inflammatory molecules may improve muscle function in affected muscular dystrophy patients 12.

The number of animals, which received SC transplantation in present work, was just the same used in previous publications 1011. However, our study differs in several aspects. First, we used in our experiments an exceptional litter with 3 affected females and 3 affected males. Second we transplanted human, not canine SC without the interference of immunosuppression protocols. Third, clinical monitoring of dogs showed no signs of fever, skin eruption, arthralgia or even glottis edema suggesting, that human cells were well accepted by the dog organism.

As canine mesoangioblasts and hIDPSC are derived from different tissues and species, it is difficult to compare their potential benefits after transplantation in the GRMD model. However, both cell types proliferate efficiently and present a good migrating capacity in host tissues 1011. In our experiment, in contrast to the study reported by Cossu group 11 we did not observe significant increase of dystrophin positive fibers in the 4 GRDM dogs after SC transplantation. Although all dogs showed engraftment of hIDPSC in muscles, only one of them (MT1-S) showed human dystrophin expression. However, this dog, which received the largest amount of human stem cells, is still alive and well at 25 months of age after the experiment was over. It showed a slower rate of progression of the dystrophic process than his five litter-mate dogs or to most of the dogs from the Brazilian GRMD colony, since the majority of them died before 18 months of age (unpublished observation).

At the end of the experiment, MT1-S had no difficulties in swallowing or chewing hard food, maintained jaw strength as well as the ability to run while carrying objects. The female FT2-IM had also good clinical scores despite the fact we could not detect dystrophin in her muscle biopsy. It died suddenly of cardiac failure at 16 months of age. Due to females having a more delicate structure of the locomotion apparatus and other anatomical variation in the thorax and muscle structures 23, gender could indeed have interfered with muscle functions and motility performance. Furthermore, since female GRMD dogs are very rare, there are no parameters to evaluate the natural history of their disease, not to mention the absence of studies that use such dogs. FC, FT1-S, FT2-IM and MT2-IM died from dystrophy complications at the ages 10, 8, 16 and 11 months, respectively.

In DMD patients, serum CK is elevated since birth and shows a peak of activity around 2 to 4 years when there is massive muscle degeneration 24. In accordance with Sampaolesi studies 12 we observed that mean serum CK activities in GRMD dogs were higher between 75 to 180 days followed by an apparent stabilization (Table 2).

Significant immune modulatory function of MSC has been demonstrated in hematopoietic stem cell transplantation. They strongly inhibit alloantigen-induced dendritic cell differentiation, down-regulate alloantigen-induced lymphocyte expansion, decrease alloantigen-specific cytotoxic capacity mediated by either cytotoxic T lymphocytes or NK cells. Interestingly, a more effective suppressive activity on mixed lymphocyte culture-induced T-cell activation was observed when MSC were heterologous rather than autologous. It was suggested that due to these properties, MSC can be used to prevent immune complications related to both hematopoietic stem cells and solid organ transplantation and it was proposed that MSC are "universal" suppressors of immune reactivity. Moreover, it was shown that regulatory CD4+ or CD8+ lymphocytes were generated in co-cultures of peripheral blood mononuclear cells with MSC. This strongly indicates that these regulatory cells may amplify the reported MSC-mediated immunosuppressive effect 252627. In our experiments the dogs did not show any immune reaction to hIDPSC transplantation. This could be explained by the fact that according to previous publication immunosuppressive activity of human DPSC was significantly higher, when compared to those from bone marrow 15. More recently, similar results were observed with different populations of stem cells isolated from dental pulp 141517.

The absence of dogs immune response to hIDPSC transplantation here reported for the first time is an important observation, because anti-inflammatory drugs could occult the benefits of stem cells therapy for dystrophy 28. Indeed, white blood cells counting did not present any important changes in response to cell transplantation and no lymphocytes infiltration was observed in muscle cells. We also did not observe immune reactions using the same cells in experiment on reconstruction of cranial defects in rats 14. This result is also supported by other authors, which demonstrated that human stem cell isolated from dental pulp have immunosuppressive activity 15. In conclusion, we believe that rejection of hIDPSC did not occur since they present all basic characteristic of mesenchymal stem cells 29.

It is very tempting to suggest that the presence of some dystrophin in muscle as a result of repeated systemic hIDPSC transplantation ameliorated the clinical condition of MT1-S since this disease in dogs causes their death usually between 11 and 18 months 930. Interestingly, after transplantation of human DPSC in non immunosuppressed rats with acute myocardial infarction, surviving and engraftment of the cells in ischemic environments have been observed. Although no differentiation into cardiac or smooth muscle as well as into endothelial cells was observed, these cells apparently contributed for the improvement of left ventricular function, induced angiogenesis and reduced infarct size. According to the authors, the benefits observed after DPSC transplantation might be due to secretion of paracrine factors by these cells 17. Since the expression of dystrophin was modest, clinical improvement in this dog might be the result of the immune modulatory effect of hIDPSC transplantation rather than the presence of dystrophin, or rather than the great clinical variability observed among GRMD dogs 31. However, our preliminary results, which should be validated in a larger sample, suggested that multiple systemic transplantations of hIDPSC in GRMD dogs could slow down the progression of the dystrophic process, through similar mechanisms described by Gandia and co-workers 17.

In short, our investigation confirmed that early systemic delivery of SC in GRMD model is more effective than local injections. Although we showed that donor human SC can engraft, differentiate, and persist in the host, it seems that the apparent clinical benefit observed in treated animals is not due to dystrophin expression in the host muscles, but due to their immune modulatory effect of SC from dental pulp 14151732. However, both observations that cell engraftment increases with consecutive transplantation and the absence of immunological response, have very important implications in designing future therapeutic trials. The potential clinical effect of SC in GRMD model should be supported by further investigation involving a larger number of animals and the efficacy of later as compared to early transplantation.

The authors declare that they have no competing interests.

All authors have read and approved the final manuscript. The specific contributions of each author are: DSM, TPG, ACM, MPB assisted clinical support to the dogs, conduce the clinical trial with laboratory analyses and physical scores establishment. EZ, NV conducted genotyping process, material collections and also immunohistochemistry and blotting dystrophy analyses. SASF, CMCM, OASA conduce the stem cells preparation and FISH analyses; AK conducted immunohistochemistry and confocal microscopy; RMC co-conducted the clinical trial and cell applications and assisted with dogs analyses; IK, CEA, MAM and MZ conceived, designed and directed the entire study, interpreted all data and wrote the manuscript.