Supplemental figure 1: Flow cytometry studies of C57-J1 cells.

FACS analysis of C57-J1 cells using a panel of CD antibodies (CD44, CD45, CD31, CD34, C-kit and FLK-1), CD13, CD49b and CD45).

Supplemental figure 2: Differentiation assays of C57-J1 cells.

C57-J1 cells expressing alkaline phosphatase (purple cells) after BMP-2 treatment (Suppl. Fig. 2a). Red oil staining showing lipidic vacuoles after culture into adipogenic-promoting medium (Suppl. Fig. 2b). Immunofluorescence analysis detecting SMA expression in C57-J1 cells treated with TGF-( Suppl. Fig. 2c). Myosin staining showing no differences in the the differentiation to striated muscle of C57-J1 cells when cultured over Matrigel® coated dishes or when cultured over plastic (Suppl. Fig. 2d).

Supplemental 3: Phenotypic characterization of the new SCID/BlAJmice.

Haematoxylin and eosin staining showing dystrophic changes (central nuclei, variation on fiber size and splitting) in quadriceps biopsies from 6 months old SCID/BlAJ and BlAJ mice but not in C57Bl/6 mice (Suppl. Fig. 3a, top panel). Azan Mallory staining demonstrating increase in the interstitial fibrous tissue (blue) in BlAJ and SCID/BlAJ (Suppl. Fig. 3a, bottom panel). Pictures were acquired by Leica (ASLMD) microscope at objective 40x.Inmunofluorescence demonstrating presence of dysferlin in the membrane of muscular fibers in C57Bl/6 mice, but not in the new Scid/BlAJ mice (Suppl. Fig. 3b). Pictures were acquired with Leica (DMIRE2). Merged pictures show dysferlin staining in green and counterstaining with DAPI in blue. RT-PCR confirming the absence of dysferlin mRNA in Scid/BlAJ (Suppl. Fig. 3c).Lane 1: QA- Scid/BlAJ, Lane 2: QA – C57Bl6, Lane 3: Human muscle, Lane 4: negative control PCR, Lane 5: Marker size (V, from Roche). PCR product amplification length: 116 bp.

Supplemental figure 4: C57-J1 cells cross vessel walls and colonize surrounding dystrophic muscles.

X-gal staining of dystrophic muscles of 5 months old SCID/BlAJ mice injected into the femoral artery with 5x105 C57-J1 cells, previously labelled with nuclear LacZ. Several LacZ+ cells were detected after one monthin all muscles situated downstream of the injection point (Suppl. Fig. 4, panels). Note cells localized throughout the entire area of the muscle section. RT-PCR confirming LacZ expression in all muscles from the injected mice and not in control non transplanted muscles (Suppl. Fig. 4upper right).

Supplemental 5: Treadmill assays did not demonstrate significant differences.

Motor function of wild type, non transplanted and transplanted SCID/BlAJmice evaluated with the treadmill demonstrating non significant differences between groups in running time (top), distance (middle) and Work (bottom). Data are means +/- s.e. (SCID n: 4 mice, SCID/BlAJn:4 mice, treated SCID/BlAJn:4 mice); statistical analysis was done with Student’s t-test considering P< 0.05 as significant.