Figure 1.
Immunohistochemical analysis of skeletal muscle tumors.
A: Biopsies from rhabdomyoma and the 3 rhabdomyosarcoma subtypes, alveolar, pleomorphic and embryonal, were analyzed for expression of Dlk1, Pax7 and Myogenin. Dlk1 was strongly expressed by all skeletal muscle tumor types and can thus be considered a solid marker. Expression of Pax7 and Myogenin in all tumors confirmed the skeletal muscle origin. Scalebar representative for all images: 50 µm. B: Dlk1 expression was detected specifically in cells co-expressing either Pax7 or Myogenin, shown as double staining. Two examples of rhabdomyosarcomas with these co-expression patterns are shown. Scalebar representative for all images: 50 µm.
Table 1.
Immunohistochemical profile of rhabdomyosarcomas (RMS).
Figure 2.
Immunohistochemical analysis of tumors with adipogenic or osteogenic origin.
A: A biopsy from a liposarcoma with an adipogenic and a myogenic metastasis was investigated for expression of Dlk1, Desmin, Myogenin, Pax7 and NCAM. All myogenic markers were predominantly detected in the myogenic metastasis and not in the main liposarcoma or the adipogenic metastasis. Accordingly, Dlk1 was primarily expressed in the myogenic metastasis. Few Dlk1 positive cells could, however, be detected in the adipogenic metastasis, where few cells were found to express Desmin and Myogenin as well. This supports the observation that Dlk1 is a strong marker for tumors with skeletal muscle phenotype. Scalebar representative for all images: 50 µm. B: Biopsies from tumors with osteogenic origin were analyzed for expression of Dlk1 and all types analyzed (osteosarcoma, osteochondroma and osteoblastoma), were negative for Dlk1 protein expression. Scalebar representative for all images: 50 µm.
Table 2.
Immunohistochemical profile of liposarcomas.
Table 3.
Dlk1 expression in osteogenic lesions.
Figure 3.
Dlk1 transgenic mice display an initially enhanced formation of myotubes following injury.
A morphological analysis of skeletal muscle regeneration in Dlk1 transgenic mice (TG) and littermate control mice (LC) after knife stab injury was performed. Days 3, 4, 5, and 7 post injury is represented for each genotype in a sirius, NCAM (Neural Cell Adhesion Molecule) and dystrophin staining. Dlk1 TG mice display an early presence of newly formed myotubes at day 3 of which most express NCAM and few express dystrophin (arrows). In comparison, the LC control mice only have few myotubes at day 3 as indicated by sirius and NCAM (arrows), and the myotubes have no dystrophin in the membranes until day 5 (arrows). The regenerative response remains accelerated at days 4 and 5 in the Dlk1 TG mice compared to the controls (arrows). Even though the LC control mice display structured muscle tissue at day 5 with expression of both NCAM and dystrophin, the muscle architecture continues to appear more mature in the Dlk1 TG mice at this time point. Scalebar: 30 µm.
Figure 4.
Analysis of skeletal muscle regeneration and the myostatin regulatory pathway in Dlk1 transgenic (TG) and littermate control (LC) mice by qPCR and morphometrics.
A: qPCR analysis of c-met, Mef2a, Myf5, and Pax7, Myod mRNA levels, during regeneration of Dlk1+ TG and LC control mice. The relative expression levels were normalized to expression of the reference genes 18s rRNA, Hprt1, Pgk1 and Tbp and the mRNA levels were calculated as fold-change to uninjured LC control muscle (calibrator = 1) for both genotypes. The qPCR represents a pool of m. gastrocnemius harvested from 3 mice for each time point and genotype and was run in triplicates. Black squares: Dlk1 TG mice +/− SD. Open circles: LC control mice +/− SD. B: Morphometric analysis of Pax7, ki67, myogenin and p27 protein expression during regeneration of Dlk1 TG and LC control mice (n = 3 in each group for each genotype). Black squares: Dlk1 TG mice +/− SD. Open circles: LC control mice +/− SD. C: Presence of centrally localized nuclei in Dlk1 TG muscle compared to LC controls is shown both as transverse and longitudinal sections representing 2 different individuals for each genotype. D: qPCR analysis of Myostatin, Follistatin, TGFβ1, Smad7 and Decorin mRNA expression during regeneration of Dlk1 transgenic mice and littermate controls. The qPCR is presented as fold-change to uninjured LC control muscle (calibrator = 1). The relative levels have been determined by normalizing the data to 18s rRNA, Gapdh, Hprt1, Pgk1 and Tbp as described in materials and methods. Black bars: Dlk1 TG mice +/− SD. White bars: LC control mice +/− SD.
Figure 5.
Ectopic ovine Dlk1 inhibits expression of endogenous Dlk1 and affects myostatin expression pattern.
A. Endogenous Dlk1 expression was detected during regeneration using rabbit-anti-murine FA1 antibody and days 3–9 following stab wound are represented. Anti-human FA1 antibody cross-reacts with the ovine form of Dlk1/FA1 [8]. Days 3–7 following stab wound are represented. A staining of uninjured LC control muscle confirms that the staining of the ovine form is specific. Scalebar: 50 µm. In LC control mice endogenous Dlk1 is expressed in mononuclear cells located within the injury and primarily observed along the periphery of newly formed myotubes (arrows). In Dlk1 TG mice the expression of Dlk1 is strongly inhibited within the muscle injury and only few mononuclear cells expressing Dlk1 can be observed (arrows). Ectopic ovine Dlk1 is expressed only in the TG mice and not by LC mice. The expression is under control by the myosin light chain promoter and is primarily expressed by myotubes and mature muscle fibers (arrows). B. qPCR analysis of endogenous Dlk1 mRNA confirms the down regulation by ectopic ovine Dlk1. Dlk1 was included in the TaqMan™ Low Density Array Card set-up and calculated as described in materials and methods. Black squares: Dlk1 TG mice +/− SD. Open circles: LC control mice. C. RT-PCR analyses of the Dlk1 splice variants present during regeneration. The primers used span exon 5 which contain all splice sites of Dlk1 and theoretically yields bands corresponding to the A (824 bp), B (671 bp), C/C2 (605/599 bp) and D/D2 (545/539 bp) forms. Pituitary gland (PG) is used as positive control. This analysis shows that the splice variants present are full-length A, C/C2 and confirms the down regulation of endogenous Dlk1 by ectopic Dlk, with all splice variants being equally affected. D. Immunohistochemical analysis of myostatin expression in Dlk1 TG and LC mice. Scalebar representative for all images: 20 µm. Myostatin expression is generally patchy and variable both in TG and LC mice, but TG mice present a less intense staining.
Figure 6.
Dlk1 inhibits myoblast differentiation in vitro.
A: C2C12 cells were constructed to constitutively express full-length Dlk1 under control by the CMV promotor, DLK1-C2C12. Cell morphology and myogenic differentiation potential was analyzed by immunocytochemistry. Forced expression of Dlk1 resulted in inhibition of myogenic differentiation observed as no formation of myotubes, lack of myostatin expression and reduced Desmin staining compared to C2C12 parental control cells. Addition of Dlk1 antibody to the culture supernatant during differentiation reverted the effect of forced Dlk1 expression observed as a regained ability to form myotubes and express myogenin compared to control cells. Scalebars: 20 µm in all images. B: DLK1-C2C12 cells (DLK1) and C2C12 parental control cells were cultured under proliferating conditions for 3 days followed by differentiation for another 5 days. Cells were harvested and analyzed at 1, 3, 5, and 8 days in culture. The differentiation process was analyzed by qPCR of Dlk1, the myogenic markers Mef2a, Myf5, Myod, and Myogenin in addition to Myostatin, Follistatin, TGFβ1, Smad7 and Decorin. All qPCR analyses were run in triplicates on duplicate analyses containing triplicate samples (n = 2 for each time point) and presented as relative expression levels with normalization to.Actb, Gusb, Pgk1, Gapdh and Tfrc as described in materials and methods. Black bars: DLK1+ cells; white bars: control cells. C: Dlk1 (13 kDa band) and myostatin (a double band corresponding to 40–50 kDa) protein level was analyzed with western blotting during proliferation and differentiation of DLK1-C2C12 and control cells. Dlk1 protein was only expressed by DLK1-C2C12 cells and not by control cells. Myostatin protein was expressed in control cells during differentiation but not during proliferation of either of the cells lines or during differentiation of DLK1-C2C12 cells. However, addition of Dlk1 antibody resulted in expression of myostatin by the DLK1-C2C12 cells during differentiation. Beta-actin (45 kDa) was used as control.
Figure 7.
Dlk1 and myostatin display an inverse relationship in human myoblasts, and the expression of Dlk1 is observed in human rhabdomyosarcoma cell lines.
A: Expression of Dlk1 and myostatin in two human myoblast cultures during G0 arrest, proliferation before and after G0, and differentiation. We found that myostatin was highly upregulated during proliferation and differentiation while Dlk1 was downregulated, thus an inverse relation was observed. B. Dlk1 (app. 30 kDa product) and myostatin (a double band around 26 kDa, corresponding to the cleavage product) protein level was analyzed with western blotting during proliferation before and after G0, during G0 (cell cycle arrest) and differentiation of primary isolated human muscle cells. Myostatin protein was expressed during proliferation and differentiation but down regulated during G0. Dlk1 protein was predominantly expressed by G0 arrested myoblasts. Ponceau Red staining (a common 70 kDa band is shown) confirmed equal loading. C: The expression of Dlk1, Myogenin and Desmin was studied in two human rhabodomyosarcoma cell lines, RD and A204 after induction in differentiation medium. We found that both cultures were Dlk1 and Desmin positive, while only RD cells were Myogenin positive. None of the cultures were able to form myotubes when cultured in differentiation medium. Scalebar representative for all images: 20 µm. D. RT-PCR analysis of Dlk1 splice variants present in RMS cells. The primers used span exon 5 and should thus provide bands for all possible splice variants present. The analysis shows the presence of 3 bands in both RMS cell lines and in the positive control tissue used, placenta and human muscle. These bands correspond to the full length A, and two smaller products most likely the B and C/C2 variants. All bands were sequenced and verified to be Dlk1.