Rhabdomyosarcoma, a malignancy showing features of skeletal muscle differentiation, is the most common soft tissue sarcoma of childhood. The identification of distinct clinical presentation patterns, histologic tumor types, and risk groups suggests that rhabdomyosarcoma is a collection of highly related sarcomas rather than a single entity. In an effort to understand this seemingly heterogeneous malignancy, we constructed a genetically defined but malleable model of rhabdomyosarcoma by converting less differentiated human skeletal muscle cell precursors (SkMC) and committed human skeletal muscle myoblasts (HSMM) into their malignant counterparts by targeting pathways altered in rhabdomyosarcoma. Whereas the two cell types were both tumorigenic, SkMCs gave rise to highly heterogeneous tumors occasionally displaying features of rhabdomyosarcoma, whereas HSMMs formed rhabdomyosarcoma-like tumors with an embryonal morphology, capable of invasion and metastasis. Thus, despite introducing the same panel of genetic changes, altering the skeletal muscle cell of origin led to different tumor morphologies, suggesting that cell of origin may dictate rhabdomyosarcoma tumor histology. The ability to now genetically induce human rhabdomyosarcoma-like tumors provides a representative model to dissect the molecular mechanisms underlying this cancer.

Malignant tumors resembling skeletal muscle, collectively known as rhabdomyosarcomas, are the most common soft tissue sarcoma of childhood, with a 3-year failure-free survival of high-risk patients of only 20% (1). Tumor histology plays a significant role in the prognosis of rhabdomyosarcoma (1), and combined with variations in clinical group and stage, suggest that the etiology of rhabdomyosarcoma is heterogeneous. There are only a few in vivo experimental systems to study the variable molecular events leading to rhabdomyosarcoma. A number of human rhabdomyosarcoma tumor cell lines have been established, but these lines represent the final stages of rhabdomyosarcoma development and hence are not amenable to dissecting early events. Rhabdomyosarcoma tumors appear at low or variable incidence in a variety of transgenic mouse backgrounds (2, 3), and recently, mouse models have been generated for both the embryonal and alveolar histologic variants of rhabdomyosarcoma (4, 5). However, tumorigenesis can be different between humans and rodents (6, 7); hence, there is value in studying cancer in human cells. To elucidate the cellular mechanisms underlying human rhabdomyosarcoma, we therefore sought to define the molecular events sufficient to drive normal human skeletal muscle cell precursors towards a cancerous fate.

Rhabdomyosarcoma shares a number of changes common to other human malignancies. Specifically, the p53 tumor suppressor pathway is impaired in up to 50% of rhabdomyosarcoma tumors and cell lines (2, 8, 9) and rhabdomyosarcoma can occur in children with germ line inactivation of p53 (10). The RB tumor suppressor pathway also seems dysregulated in rhabdomyosarcoma through amplification of cyclin-dependent kinase CDK4 and/or deletion of the tumor suppressor p16INK4A (2, 9, 11). The MYCN proto-oncogene is up-regulated in rhabdomyosarcoma (2, 12), targeting 40% of genes similarly activated by c-Myc (13). Telomere stabilization and correspondingly cell immortalization is illegitimately restored in rhabdomyosarcoma by reactivation of the hTERT catalytic subunit or through alternative telomere lengthening mechanisms (1416). Lastly, activation of the Ras pathway (17) occurs in rhabdomyosarcoma through point mutations in Ras (2, 18, 19), activation of upstream tyrosine kinase receptors (2), or loss of the negative regulator neurofibromin (20). To recapitulate the genetic alterations of rhabdomyosarcoma in a human model system, we therefore disrupted these pathways (21) in human primary cells. The cell of origin of rhabdomyosarcoma is unknown but suggested to be from cells developing at any point along the skeletal muscle cell axis (2). Hence, we chose to introduce these alterations in human primitive fetal skeletal muscle cell precursors (SkMC) as well as human postnatal skeletal muscle myoblasts (HSMM) already committed to the skeletal muscle lineage. We used expression of the SV40 large T oncoprotein to inactivate the tumor suppressors p53 and RB, small t oncoprotein to inactivate PP2A, leading to stabilization of the c-Myc oncprotein (22), hTERT to impart immortalization (23), and oncogenic (V12G) Ras to provide self-sufficiency in growth signals (17).

Cell lines. Low-passage normal human fetal SkMCs or normal HSMMs from a teenage donor (Clonetics Cell Systems, Cambrex Corp., East Rutherford, NJ) were sequentially infected with amphotrophic retroviruses derived from pBABE-neo-T/t-Ag, pBABE-bleo-FLAG-H-RasV12G (7), pBABE-hygro-FLAG-hTERT, pBABE-puro-FLAG-H-RasV12G (21), or the corresponding empty vectors and sequentially selected for 7 to 10 days in medium supplemented with 0.25 μg/mL puromycin (Sigma Chemical Co., St. Louis, MO), 50 μg/mL hygromycin B, 250 μg/mL G418, or 800 μg/mL Zeocin (all from Life Technologies Invitrogen, Carlsbad, CA). Cells were verified to be of skeletal muscle lineage by expression of one or more skeletal muscle markers: SkMCs were desmin positive, skeletal muscle–specific actin, and myoglobin negative, whereas HSMMs were desmin and skeletal muscle–specific actin positive and myoglobin negative (data not shown).

Detection of gene products. For Western blotting, 100 μg of whole cell lysates were separated and immunoblotted with antibodies anti–T Ag SC-147, anti-actin SC-8432 (Santa Cruz Biotechnology, Santa Cruz, CA), anti-FLAG-M2 (Sigma Chemical), anti-desmin M0760, anti–muscle-specific actin M0635, or anti-myoglobin A0324 (DakoCytomation, Carpinteria, CA), using established protocols (7).

For immunohistochemistry, 5-μm sections of paraffin-embedded tissue were subject to heat-induced antigen retrieval using the steam method and incubated with antibodies (DakoCytomation) against myoglobin (A0324, 1:3,000), muscle-specific actin (M0635, 1:200), desmin (M0760, 1:50), myogenin (M3559, 1:50), and MyoD1 (M3512, 1:50) at 37°C for 45 minutes. Biotinylated secondary antibody (Vector Laboratories, Burlingame, CA) was incubated for 20 minutes, and the tertiary detection system was ABCElite Complex (Vector Laboratories), with a 3,3′-diaminobenzidine chromagen substrate (Innovex Biosciences, Richmond, CA). Slides were lightly counterstained with hematoxylin. Normal human skeletal muscle and isotype-specific antibodies were used as positive and negative controls, respectively. Pathologists with experience in the evaluation of pediatric solid tumors (S.Q. and R.C.B.) evaluated slides. Standard H&E stained sections were also prepared.

For telomerase assays, 0.5 μg cellular lysates were assayed for activity using a PCR-based telomere repeat amplification assay, as described previously (21).

For reverse transcription-PCR, 2 μg of total RNA, prepared using the RNAzol B reagent (TEL-TEST, Friendswood, TX), was reverse transcribed using the Ominscript RT kit (Qiagen, Valencia, CA) with OligodT primer (Life Technologies Invitrogen), after which 4 μL of each reaction was PCR amplified using primers specific for insulin-like growth factor-II mRNA (5′-ATCGTT-GAGGAGTGCTGTTTCC-3′; 5′-ATTGCTGGCCATCTCTGG-3′) and glyceraldehyde-3-phosphate dehydrogenase (21).

For transmission electron microscopy, tumor samples were processed as described (24) except that glutareldehyde was 4%, OsO4 buffer was 0.2 mol/L cacodylate, stain/counterstain was uranyl acetate/lead citrate, and images were captured on a Philips EM410 electron microscope.

Transformation and tumorigenesis. Soft agar growth was assessed after 4 weeks as described previously (7). For xenograft assays, 1 × 107 cells were suspended in 50 μL Matrigel (BD Biosciences, San Jose, CA) and injected s.c. into the flank of a severe combined immunodeficient/Beige mouse as previously described (7). For orthotopic assays, 3 × 105 cells suspended in 50 μL PBS were injected into the right gastrocnemius muscle. For metastasis assays, 4 × 106 cells suspended in 200 μL medium were injected into the tail vein. Each cell line was tested in quadruplicate. All experiments were done under the Duke Institutional Animal Care and Use Committee–approved protocols.

Genetic transformation of skeletal muscle cell precursors. Unlike adult carcinomas, which derive from the malignant transformation of epithelial cells, the histogenesis of pediatric mesenchymal tumors such as rhabdomyosarcoma is less clear. It has been suggested that rhabdomyosarcoma tumors may derive from the transformation of cells developing at any point along the skeletal muscle cell axis (2), generically termed here as “skeletal muscle cell precursors.” Although historically these precursors were presumed to be satellite cell myoblasts located beneath the basement membrane of the skeletal myofiber, skeletal muscle cell precursors are also believed to include multipotential stem cells, which derive from the bone marrow but reside in skeletal muscle tissue, and possibly skeletal muscle myonuclei, which may be stimulated to reenter the cell cycle after specific cues (25). In an effort to model rhabdomyosarcoma, we postulated that transformation of an unselected primary skeletal muscle cell population might yield tumors akin to rhabdomyosarcoma. Therefore, a heterogeneous population of skeletal muscle cell precursors derived from human fetal muscle, termed SkMC cells, was serially infected and selected for antibiotic resistance with amphotrophic retroviruses encoding T/t-Ag, hTERT, and H-RasV12G, or a combination thereof, in which one or more transgenes was substituted with an empty vector. Stable polyclonal cell lines representing all possible combinations of T/t-Ag (T), and/or hTERT (H), and/or H-RasV12G (R), and/or vector (V), were generated and confirmed to appropriately express the desired transgenes (Fig. 1A-B). A cell line expressing 5-fold lower levels of H-RasV12G was also generated (Fig. 1A; M-THRLo) to address the effect of oncogenic Ras expression on rhabdomyosarcoma tumorigenicity, as low oncogenic Ras expression has been found to limit tumor growth in human mammary tumors (26).

Figure 1.

Evaluation of transgene expression, cell growth, and in vitro tumorigenic capacity of primary human fetal SkMCs. Skeletal muscle cell precursors (M) were serially infected with amphotrophic retroviruses encoding T/t-Ag (T), hTERT (H), and H-RasV12G (R), or empty vectors (V). Polyclonal populations capable of sustained growth were analyzed for (A) T-Ag, H-RasV12G, and actin loading control expression by immunoblotting with monoclonal antibodies anti-T-Ag, anti-FLAG, and anti-actin. B, telomerase activity restored by ectopic hTERT expression, as shown by a six nucleotide ladder. Internal standard (IS); water (W), positive (+), and negative (−) control cell lines. C, cell growth in population doublings (pd) versus time of M-THV (▴), M-THR (▪), M-TVV (▵), M-TVR (□), M-VHV (○), and M-VVV (⋄). D, ability to grow in soft agar. Cells (5 × 104) stably expressing the indicated transgenes were seeded into soft agar in triplicate. Colonies visible to the naked eye were counted after 4 weeks. Representative result of one of three independent experiments. Bars, SD.

Figure 1.

Evaluation of transgene expression, cell growth, and in vitro tumorigenic capacity of primary human fetal SkMCs. Skeletal muscle cell precursors (M) were serially infected with amphotrophic retroviruses encoding T/t-Ag (T), hTERT (H), and H-RasV12G (R), or empty vectors (V). Polyclonal populations capable of sustained growth were analyzed for (A) T-Ag, H-RasV12G, and actin loading control expression by immunoblotting with monoclonal antibodies anti-T-Ag, anti-FLAG, and anti-actin. B, telomerase activity restored by ectopic hTERT expression, as shown by a six nucleotide ladder. Internal standard (IS); water (W), positive (+), and negative (−) control cell lines. C, cell growth in population doublings (pd) versus time of M-THV (▴), M-THR (▪), M-TVV (▵), M-TVR (□), M-VHV (○), and M-VVV (⋄). D, ability to grow in soft agar. Cells (5 × 104) stably expressing the indicated transgenes were seeded into soft agar in triplicate. Colonies visible to the naked eye were counted after 4 weeks. Representative result of one of three independent experiments. Bars, SD.

Close modal

Tumorigenic conversion of skeletal muscle cell precursors by expression of T/t-Ag, hTERT, and H-RasV12G. To explore the effects of these genetic changes on the tumorigenic process, the resulting cell populations were assayed in vitro for immortalization and anchorage-independent growth, common features of cancer cells, and in vivo for tumor growth in immunocompromised mice. As regards immortalization, cells lacking T/t-Ag entered a permanent growth arrest between population doublings 13 and 15 with morphologic features consistent with senescence (Fig. 1C; uninfected, M-VVV, M-VHV), in agreement with the results of others (27). The stable expression of H-RasV12G alone or in combination with hTERT resulted in cell death (Fig. 1C; M-VVR, M-VHR), presumably reflecting a cellular response against Ras up-regulation in the presence of an intact p53 pathway (28). As expected (29), expression of T/t-Ag alone or in the presence of oncogenic Ras extended the proliferative life span of the cells by up to 60 population doublings, at which time the cells entered crisis and perished (Fig. 1C; M-TVV, M-TVR). As with other cell types (21), independent of RasV12G expression, only T/t-Ag in conjunction with hTERT greatly extended cellular life span (Fig. 1C; M-THR, M-THV). Thus, combined ectopic expression of T/t-Ag and hTERT endows human SkMCs with an infinite life span, in agreement with the observed dysregulation of the p53, RB, Myc, and hTERT pathways in rhabdomyosarcoma tumors and cell lines (2, 1416).

As regards transformation, each of the above-described cell lines was also assayed for growth in soft agar, one of the most stringent assays for transformation in vitro. Expression of all four transgenes, and not any less, was necessary and sufficient for anchorage-independent growth of SkMCs (Fig. 1D; M-THR). This growth depended upon increased oncogenic Ras expression, as both the number of colonies and the heterogeneity of colony sizes decreased upon lowering oncogenic Ras expression 5-fold (Fig. 1D; other data not shown). Thus, Ras expression is required in addition to T/t-Ag and hTERT for anchorage-independent growth of SkMCs.

The most telling assay of tumorigenesis is tumor growth itself. SkMCs expressing all four transgenes, or control SkMCs expressing two or three of these transgenes, were injected s.c. into immunocompromised mice and monitored for tumor growth. As in soft agar, expression of all four transgenes was required for tumor growth (Table 1). This growth was sensitive to Ras expression, as a decrease in expression of this oncogene delayed or abolished tumor growth (Table 1), suggesting an important role for the stimulation of the Ras pathway in rhabdomyosarcoma. In sum, SkMCs minimally must undergo dysregulation of the p53, RB, Myc (and possibly other targets of PP2A), Ras, and hTERT pathways to form tumors in vivo.

Table 1.

Xenograft tumor formation in immunocompromised mice

Parental cell typeCell lineInjection routeMice developing tumors/mice injectedEarliest time to palpable tumor
SkMC M-TVV SQ 0/4 — 
 M-THV SQ 0/8 — 
 M-TVR SQ 0/4 — 
 M-THRLo SQ 6/8 11 wks 
 M-THR SQ 4/4 4.5 wks 
HSMM MY-THR SQ 5/5 2 wks 
 MY-THR IV 4/5 NA 
 MY-THR IM 4/4 8 wks 
Parental cell typeCell lineInjection routeMice developing tumors/mice injectedEarliest time to palpable tumor
SkMC M-TVV SQ 0/4 — 
 M-THV SQ 0/8 — 
 M-TVR SQ 0/4 — 
 M-THRLo SQ 6/8 11 wks 
 M-THR SQ 4/4 4.5 wks 
HSMM MY-THR SQ 5/5 2 wks 
 MY-THR IV 4/5 NA 
 MY-THR IM 4/4 8 wks 

NOTE: Polyclonal cell lines derived from normal human SkMCs or normal human myoblasts (HSMM) stably expressing SV40 T/t-Ag (T), and/or hTERT (H), and/or H-RasV12G (R), and/or empty vector (V) were injected s.c., i.v., or i.m. to assay for tumor growth. “Lo” is a cell line expressing low levels of H-RasV12G.

Abbreviations: NA, not applicable (as mice injected via tail vein developed lung nodules); SQ, s.c.; IV, i.v.; IM, i.m.

SkMC-derived tumors ranged from highly undifferentiated small round blue cell tumors with large nuclei (Fig. 2A) and scant cytoplasm to tumors with foci of spindle-shaped cells (Fig. 2C). No classic embryonal or alveolar histology could be identified by light microscopy, nor were there cytoplasmic cross-striations to confirm skeletal muscle differentiation. Immunohistochemical staining for desmin, muscle-specific actin, and myoglobin, markers used in clinical practice to evaluate for rhabdomyosarcoma (1), showed some biochemical evidence of skeletal muscle differentiation. Staining for these markers ranged from undetectable (Fig. 2B) to positive for all three antigens in the tumor that showed foci of spindle-shaped cells (Fig. 2D-F). Staining for myogenin and MyoD1, muscle-specific transcription factors expressed in the nuclei of rhabdomyosarcomas (1), was negative (data not shown). All tumors showed brisk mitotic activity and areas of necrosis, consistent with high-grade sarcomas. Thus, introducing genetic changes characteristic of rhabdomyosarcoma in cultures of human SkMCs led to a broad spectrum of sarcomas, ranging from undifferentiated small round blue cell tumors (sarcomas, not otherwise specified) to tumors exhibiting some differentiation markers characteristic of rhabdomyoblasts, but lacking frank histopathologic features of either embyronal or alveolar rhabdomyosarcoma.

Figure 2.

S.c. tumor xenografts derived from SkMC cells stably expressing T/t-Ag, hTERT, and H-RasV12G show variable rhabdomyoblastic morphology and immunohistochemical staining. H&E (A) and (B) desmin staining of the least differentiated tumor xenograft. H&E (C), (D) desmin, skeletal muscle–specific actin (E), and myoglobin (F) staining of the most differentiated tumor containing areas of spindle-shaped cells. Immunoreactivity (brown). Elongate cell with biopolar cytoplasmic processes forming a “spindle” shape (open arrow). Small foci of immunoreactivity (closed arrows). Magnification, 400×.

Figure 2.

S.c. tumor xenografts derived from SkMC cells stably expressing T/t-Ag, hTERT, and H-RasV12G show variable rhabdomyoblastic morphology and immunohistochemical staining. H&E (A) and (B) desmin staining of the least differentiated tumor xenograft. H&E (C), (D) desmin, skeletal muscle–specific actin (E), and myoglobin (F) staining of the most differentiated tumor containing areas of spindle-shaped cells. Immunoreactivity (brown). Elongate cell with biopolar cytoplasmic processes forming a “spindle” shape (open arrow). Small foci of immunoreactivity (closed arrows). Magnification, 400×.

Close modal

Transformation of human skeletal muscle myoblasts generates an embryonal rhabdomyosarcoma phenotype. Given the variable tumor histology resulting from a mixed population of SkMCs, we hypothesized that the skeletal muscle cell-of-origin might underlie differences in rhabdomyosarcoma tumor histology. As recent data suggests that embryonal rhabdomyosarcoma might derive from satellite cell myoblasts (30), which should be present in the SkMC population, an enriched population of HSMMs was infected with retroviruses encoding T/t-Ags, hTERT, and H-RasV12G, generating a new polyclonal myoblast MY-THR cell line. These cells were tumorigenic in mice (Table 1). However, in contrast to SkMC cells, transformed HSMM cells showed rhabdomyosarcoma morphology, characterized by large numbers of rhabdomyoblasts with abundant, deeply eosinophilic cytoplasm (Fig. 3A), and spindle-shaped cells in a myxoid background (Fig. 3B-C). Immunohistochemical staining showed skeletal muscle differentiation, with diffuse and strong desmin staining, focally positive muscle-specific actin, and diffusely positive myoglobin (Fig. 3D-F). The tumors were focally positive for MyoD1 and myogenin (Fig. 3G-H) and expressed IgF2 (Fig. 3I), a fetal growth factor overexpressed in rhabdomyosarcoma tumors (reviewed in ref. 2). When examined by electron microscopy, the most rigorous test for establishing skeletal muscle origin, these tumors had cytoplasmic myofilaments (Fig. 3J), and myofilaments with z-band material attempting to form sarcomeres (Fig. 3K). Thus, genetic changes observed in rhabdomyosarcoma can convert HSMMs to tumors resembling human rhabdomyosarcoma. Moreover, given the presence of embryonal rhabdomyosarcoma-specific findings (nuclear pleomorphism, myxoid change with spindling, focal staining of MyoD1, and myogenin) and the absence of alveolar findings (nuclear monotony, diffuse immunostaining with MyoD1 and myogenin, collagenous septae lined by rhabdomyoblasts with associated nesting of tumor cells), these tumors are most consistent with an embryonal rhabdomyosarcoma histology.

Figure 3.

S.c. tumor xenografts derived from primary human myoblasts (HSMM) stably expressing T/t-Ag, hTERT, and H-RasV12G show morphologic, immunohistochemical, and ultrastructural resemblance to human embryonal rhabdomyosarcoma. Tumor sections were evaluated using (A-C) H&E for morphology, and immunohistochemical staining for the skeletal muscle-specific antigen markers desmin (D), muscle-specific actin (E), myoglobin (F), MyoD1 (G), and myogenin (H). Images (B) and (C) are from different tumors. I, expression of IGF2 in tumor lysates (MY-THR) was measured by semiquantitative reverse transcription-PCR and compared with IGF2-negative (−) NCI-H460 cells (31), and positive (+) control cell lines. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used to control for total RNA input. Tumor sections were also imaged by transmission electron microscopy, revealing disorganized cytoplasmic myofilaments (J, 29,400×) and myofilaments attempting to organize into sarcomeres (K, 21,875×; arrowheads). N, nucleus. Light microscopy images are 100× (B, D, and E), 200× (A, C, G, and H), or 400× (G).

Figure 3.

S.c. tumor xenografts derived from primary human myoblasts (HSMM) stably expressing T/t-Ag, hTERT, and H-RasV12G show morphologic, immunohistochemical, and ultrastructural resemblance to human embryonal rhabdomyosarcoma. Tumor sections were evaluated using (A-C) H&E for morphology, and immunohistochemical staining for the skeletal muscle-specific antigen markers desmin (D), muscle-specific actin (E), myoglobin (F), MyoD1 (G), and myogenin (H). Images (B) and (C) are from different tumors. I, expression of IGF2 in tumor lysates (MY-THR) was measured by semiquantitative reverse transcription-PCR and compared with IGF2-negative (−) NCI-H460 cells (31), and positive (+) control cell lines. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used to control for total RNA input. Tumor sections were also imaged by transmission electron microscopy, revealing disorganized cytoplasmic myofilaments (J, 29,400×) and myofilaments attempting to organize into sarcomeres (K, 21,875×; arrowheads). N, nucleus. Light microscopy images are 100× (B, D, and E), 200× (A, C, G, and H), or 400× (G).

Close modal

Genetically defined embryonal rhabdomyosarcoma tumors are invasive and metastatic. Aside from histologic markers, human rhabdomyosarcoma tumors are characterized by their ability to invade adjacent tissue and metastasize. Therefore, the genetically defined myoblast-derived tumor cells (MY-THR) were introduced into the systemic circulation, after which four of five injected mice developed clinically and anatomically apparent lung metastases by 8 weeks (Table 1), with almost complete obliteration of normal pulmonary alveolar architecture (Fig. 4A). Tumor cells tended to localize around pulmonary arterioles, often encasing pulmonary bronchioles (Fig. 4B). Histologic evaluation was again consistent with an embryonal rhabdomyosarcoma morphology (Fig. 4C). Although the tail vein assay evaluates many steps in metastasis, it does not assay the capacity of tumor cells to be motile or invasive. In this regard, nodules located at the lung periphery were noted to invade locally, eroding through the visceral and parietal lung pleura, and infiltrated through basement membrane into the adjacent chest wall skeletal muscle and ribs (Fig. 4D). Moreover, MY-THR cells injected orthotopically into the gastrocnemius muscle of four mice developed as tumors (Table 1) that displaced the normal tissue (Fig. 4E) and invaded into the surrounding native skeletal muscle (Fig. 4F), although not as extensively as the lung nodules. Intriguingly, local invasion was not observed in s.c. xenografts (data not shown), suggesting that the surrounding microenvironment provides factors necessary to support this more aggressive phenotype. Thus, expression of T/t-Ags, hTERT, and H-RasV12G endowed human myoblasts with the invasive and metastatic phenotypes characteristic of rhabdomyosarcoma tumors.

Figure 4.

Primary human myoblasts stably expressing T/t-Ag, hTERT, and H-RasV12G show metastatic and invasive behavior when introduced systemically through the tail vein or orthotopically into native skeletal muscle. A, low-power view of mouse lung largely replaced by tumor nodules (10×). Dashed box, area magnified for (B). B, single tumor nodule encasing pulmonary bronchus and vein (100×). Asterisks, areas of remaining intact alveoli. C, magnification of metastatic nodule showing embryonal morphology. D, invasion of adjacent chest wall by peripheral pulmonary nodules (10×). Closed arrows, area of local invasion. Rib is cut transversely for orientation. E, low-power view of gastrocnemius muscle largely replaced by tumor xenograft (10×). Hind limb is cut transversely for orientation. Dashed box, region magnified for (F). F, tumor cells (open arrows) streaming into adjacent native skeletal muscle (200×).

Figure 4.

Primary human myoblasts stably expressing T/t-Ag, hTERT, and H-RasV12G show metastatic and invasive behavior when introduced systemically through the tail vein or orthotopically into native skeletal muscle. A, low-power view of mouse lung largely replaced by tumor nodules (10×). Dashed box, area magnified for (B). B, single tumor nodule encasing pulmonary bronchus and vein (100×). Asterisks, areas of remaining intact alveoli. C, magnification of metastatic nodule showing embryonal morphology. D, invasion of adjacent chest wall by peripheral pulmonary nodules (10×). Closed arrows, area of local invasion. Rib is cut transversely for orientation. E, low-power view of gastrocnemius muscle largely replaced by tumor xenograft (10×). Hind limb is cut transversely for orientation. Dashed box, region magnified for (F). F, tumor cells (open arrows) streaming into adjacent native skeletal muscle (200×).

Close modal

Summary

The presentation of rhabdomyosarcoma is heterogeneous, arguing that treatments may need to be tailored to tumor type. To explore rhabdomyosarcoma etiology, we found that heterogeneous populations of SkMCs could be converted to a broad spectrum of tumors via the corruption of the p53, RB, Myc, telomerase, and Ras pathways, whereas by limiting these changes to committed myoblasts (HSMMs), embryonal-like rhabdomyosarcoma tumors were generated. We therefore argue that cell type may play a key role in rhabdomyosarcoma, and now with the described ability to genetically model rhabdomyosarcoma, it should be possible to dissect the histogenesis and molecular mechanisms underlying this disease using normal primary human cells.

Grant support: NIH grants 5K12-HD043494-03 (C.M. Linardic) and CA94184 (C.M. Counter), Duke Children's Miracle Network (C.M. Linardic), North Carolina Cooperative Extension (C.M. Linardic), and Leukemia and Lymphoma Society (C.M. Counter).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

We thank members of the laboratories of Chris Counter, Xiao-Fan Wang, and Anthony Means, especially Stacey Adam, Kian-Huat Lim, Chaoyu Ma, and Tom Ribar for assistance with mouse experiments; Philip Breitfeld and Darrell Yamashiro for review of the report; and Henry Estrada and Sara Miller for assistance with electron microscopy.

1
Meyer WH, Spunt SL. Soft tissue sarcomas of childhood.
Cancer Treat Rev
2004
;
30
:
269
–80.
2
Merlino G, Helman LJ. Rhabdomyosarcoma: working out the pathways.
Oncogene
1999
;
18
:
5340
–8.
3
Nanni P, Nicoletti G, De Giovanni C, et al. Development of rhabdomyosarcoma in HER-2/neu transgenic p53 mutant mice.
Cancer Res
2003
;
63
:
2728
–32.
4
Sharp R, Recio JA, Jhappan C, et al. Synergism between INK4a/ARF inactivation and aberrant HGF/SF signaling in rhabdomyosarcomagenesis.
Nat Med
2002
;
8
:
1276
–80.
5
Keller C, Arenkiel BR, Coffin CM, El Bardeesy N, DePinho RA, Capecchi MR. Alveolar rhabdomyosarcomas in conditional Pax3:Fkhr mice: cooperativity of Ink4a/ARF and Trp53 loss of function.
Genes Dev
2004
;
18
:
2614
–26.
6
Ghebranious N, Donehower LA. Mouse models in tumor suppression.
Oncogene
1998
;
17
:
3385
–400.
7
Hamad NM, Elconin JH, Karnoub AE, et al. Distinct requirements for Ras oncogenesis in human versus mouse cells.
Genes Dev
2002
;
16
:
2045
–57.
8
Takahashi Y, Oda Y, Kawaguchi K, et al. Altered expression and molecular abnormalities of cell-cycle-regulatory proteins in rhabdomyosarcoma.
Mod Pathol
2004
;
17
:
660
–9.
9
Gordon AT, Brinkschmidt C, Anderson J, et al. A novel and consistent amplicon at 13q31 associated with alveolar rhabdomyosarcoma.
Genes Chromosomes Cancer
2000
;
28
:
220
–6.
10
Birch JM, Hartley AL, Blair V, et al. Cancer in the families of children with soft tissue sarcoma.
Cancer
1990
;
66
:
2239
–48.
11
Maelandsmo GM, Berner JM, Florenes VA, et al. Homozygous deletion frequency and expression levels of the CDKN2 gene in human sarcomas: relationship to amplification and mRNA levels of CDK4 and CCND1.
Br J Cancer
1995
;
72
:
393
–8.
12
Williamson D, Lu YJ, Gordon T, et al. Relationship between MYCN copy number and expression in rhabdomyosarcomas and correlation with adverse prognosis in the alveolar subtype.
J Clin Oncol
2005
;
23
:
880
–8.
13
Boon K, Caron HN, van Asperen R, et al. N-myc enhances the expression of a large set of genes functioning in ribosome biogenesis and protein synthesis.
EMBO J
2001
;
20
:
1383
–93.
14
Kleideiter E, Schwab M, Friedrich U, Koscielniak E, Schafer BW, Klotz U. Telomerase activity in cell lines of pediatric soft tissue sarcomas.
Pediatr Res
2003
;
54
:
718
–23.
15
Terasaki T, Kyo S, Takakura M, et al. Analysis of telomerase activity and telomere length in bone and soft tissue tumors.
Oncol Rep
2004
;
11
:
1307
–11.
16
Henson JD, Hannay JA, McCarthy SW, et al. A robust assay for alternative lengthening of telomeres in tumors shows the significance of alternative lengthening of telomeres in sarcomas and astrocytomas.
Clin Cancer Res
2005
;
11
:
217
–25.
17
Bos JL. Ras oncogenes in human cancer: a review.
Cancer Res
1989
;
49
:
4682
–9.
18
Wilke W, Maillet M, Robinson R. H-ras-1 point mutations in soft tissue sarcomas.
Mod Pathol
1993
;
6
:
129
–32.
19
Yoo J, Robinson RA. H-ras and K-ras mutations in soft tissue sarcoma: comparative studies of sarcomas from Korean and American patients.
Cancer
1999
;
86
:
58
–63.
20
Reed N, Gutmann DH. Tumorigenesis in neurofibromatosis: new insights and potential therapies.
Trends Mol Med
2001
;
7
:
157
–62.
21
Hahn WC, Counter CM, Lundberg AS, Beijersbergen RL, Brooks MW, Weinberg RA. Creation of human tumour cells with defined genetic elements.
Nature
1999
;
400
:
464
–8.
22
Yeh E, Cunningham M, Arnold H, et al. A signalling pathway controlling c-Myc degradation that impacts oncogenic transformation of human cells.
Nat Cell Biol
2004
;
6
:
308
–18.
23
Collins K, Mitchell JR. Telomerase in the human organism.
Oncogene
2002
;
21
:
564
–79.
24
Linardic CM, Jayadev S, Hannun YA. Activation of the sphingomyelin cycle by brefeldin A: effects of brefeldin A on differentiation and implications for a role for ceramide in regulation of protein trafficking.
Cell Growth Differ
1996
;
7
:
765
–74.
25
Grounds MD, White JD, Rosenthal N, Bogoyevitch MA. The role of stem cells in skeletal and cardiac muscle repair.
J Histochem Cytochem
2002
;
50
:
589
–610.
26
Elenbaas B, Spirio L, Koerner F, et al. Human breast cancer cells generated by oncogenic transformation of primary mammary epithelial cells.
Genes Dev
2001
;
15
:
50
–65.
27
Seigneurin-Venin S, Bernard V, Moisset PA, et al. Transplantation of normal and DMD myoblasts expressing the telomerase gene in SCID mice.
Biochem Biophys Res Commun
2000
;
272
:
362
–9.
28
Bringold F, Serrano M. Tumor suppressors and oncogenes in cellular senescence.
Exp Gerontol
2000
;
35
:
317
–29.
29
Mouly V, Edom F, Decary S, Vicart P, Barbert JP, Butler-Browne GS. SV40 large T antigen interferes with adult myosin heavy chain expression, but not with differentiation of human satellite cells.
Exp Cell Res
1996
;
225
:
268
–76.
30
Tiffin N, Williams RD, Shipley J, Pritchard-Jones K. PAX7 expression in embryonal rhabdomyosarcoma suggests an origin in muscle satellite cells.
Br J Cancer
2003
;
89
:
327
–32.
31
Quinn KA, Treston AM, Unsworth EJ, et al. Insulin-like growth factor expression in human cancer cell lines.
J Biol Chem
1996
;
271
:
11477
–83.