Research Article

Myogenic regulatory transcription factors regulate growth in rhabdomyosarcoma

Cancer Center, and Regenerative Medicine, Massachusetts General Hospital, United States
Harvard Stem Cell Institute, United States
Abel Salazar Biomedical Sciences Institute, University of Porto, Portugal
Greehey Children's Cancer Research Institute, United States
Pediatric Oncology Branch, Advanced Technology Center, National Cancer Institute, United States
Faculdade de Medicina, Universidade de Lisboa, Portugal
University of Washington, United States
Massachusetts General Hospital, United States

Jan 12, 2017

https://doi.org/10.7554/eLife.19214

Open access
Copyright information

Abstract
Introduction
Results
Discussion
Materials and methods
References
Article and author information
Metrics

Abstract

Rhabdomyosarcoma (RMS) is a pediatric malignacy of muscle with myogenic regulatory transcription factors MYOD and MYF5 being expressed in this disease. Consensus in the field has been that expression of these factors likely reflects the target cell of transformation rather than being required for continued tumor growth. Here, we used a transgenic zebrafish model to show that Myf5 is sufficient to confer tumor-propagating potential to RMS cells and caused tumors to initiate earlier and have higher penetrance. Analysis of human RMS revealed that MYF5 and MYOD are mutually-exclusively expressed and each is required for sustained tumor growth. ChIP-seq and mechanistic studies in human RMS uncovered that MYF5 and MYOD bind common DNA regulatory elements to alter transcription of genes that regulate muscle development and cell cycle progression. Our data support unappreciated and dominant oncogenic roles for MYF5 and MYOD convergence on common transcriptional targets to regulate human RMS growth.

https://doi.org/10.7554/eLife.19214.001

Introduction

Continued tumor growth and relapse are driven by molecularly defined tumor propagating cells (TPCs). These TPCs share molecular and functional properties with non-transformed stem cells. For example, molecularly defined TPCs can divide to produce daughter cells with similar functional characteristics as the parental TPC, akin to the process of self-renewal found in normal stem cells. TPCs can also divide to produce mature, specialized cells that have specific functions within the developing cancer. Given that TPCs drive continued tumor growth, it is not surprising that these cells are often retained following treatment and ultimately drive refractory, metastatic, and relapse disease (Reya et al., 2001). TPCs have been identified in a wide range of cancers, including both zebrafish and human rhabdomyosarcoma (RMS), a devastating pediatric cancer of muscle (Langenau et al., 2007; Chen et al., 2014; Ignatius et al., 2012). Yet, to date, the molecular mechanisms driving TPC frequency and sustained tumor growth in RMS have not been fully defined. Moreover, it is unknown the extent to which normal muscle stem cell programs play a role in regulating RMS growth.

Rhabdomyosarcoma is a common sarcoma characterized by impaired muscle differentiation. These tumors express myogenic regulatory transcription factors (MRFs) including Myogenic factor 5 (MYF5) and Myoblast determination protein D (MYOD) (Clark et al., 1991; Kumar et al., 2000; Parham, 2001; Sebire and Malone, 2003) and are used in the clinical diagnosis of RMS. RMS is comprised of two molecular subtypes. Alveolar RMS (ARMS) harbor Pax7-FOXO and Pax3-FOXO genomic fusions (Sorensen et al., 2002) and have few additional recurrent genomic changes (Chen et al., 2013b; Shern et al., 2014). By contrast, 90% of human embryonal rhabdomyosarcoma (ERMS) have RAS pathway activation and a higher mutation burden when compared with ARMS (Chen et al., 2013b; Langenau et al., 2007; Shern et al., 2014). Common mutations found in ERMS include inactivation of Tp53 and activating mutations of FGFR4, PDGFA, and NOTCH1 (Chen et al., 2013b; Shern et al., 2014). Yet, roles for these pathways in regulating TPC number and proliferation have not been reported. In fact, to date, only the Sonic-Hedgehog and canonical WNT/B-catenin signaling pathways have been shown to regulate TPC function in a subset of human RMS (Chen et al., 2014; Satheesha et al., 2016). Understanding additional underlying mechanisms of TPC growth and function will be important for defining new therapies to treat pediatric RMS.

Despite the similarity of RMS cells with embryonic and regenerating muscle and well-known roles for the myogenic regulatory transcription factors MYF5 and MYOD in regulating these processes, their role in driving RMS growth has yet to be reported. Rather, it has been suggested that activation of the myogenic gene regulatory programs likely reflect the target cell of transformation and may not be required for continued RMS growth (Keller and Guttridge, 2013; Kikuchi et al., 2011; Macquarrie et al., 2013b; Rubin et al., 2011). Despite MYF5 and MYOD being highly expressed in human and animal models of RMS (Langenau et al., 2007; Rubin et al., 2011), exerting important roles in muscle development and stem cell self-renewal in regeneration (Buckingham and Rigby, 2014), and being able to reprogram fibroblasts into proliferating myoblasts (Braun et al., 1989; Tapscott et al., 1988); a functional requirement for these transcription factors in regulating RMS growth has gone unexplored since their discovery over two decades ago.

Transgenic zebrafish models have become a powerful tool to uncover new biological insights into human cancer (Langenau et al., 2003, 2007; Le et al., 2007; Park et al., 2008; Patton et al., 2005; Sabaawy et al., 2006; Yang et al., 2004; Zhuravleva et al., 2008). In the setting of ERMS, we have developed a mosaic transgenic zebrafish that express human kRAS^G12D under control of the rag2 minimal promoter, which is expressed in lymphoid cells (Jessen et al., 2001; Langenau et al., 2003) and muscle progenitor cells (Langenau et al., 2007). Thus, when kRAS^G12D was expressed under control of this promoter, 20–40% mosaic injected fish developed ERMS (Langenau et al., 2007). Because 10–20 transgene copies are commonly integrated into the genome (Langenau et al., 2008), one can inject multiple transgenes into one-cell stage embryos with stable integration and expression being observed in developing tumors. Using this mosaic transgenic approach, we can deliver transgenic expression of kRAS^G12D, a fluorescent label to mark ERMS cells, and a modifying gene to assess synergies in regulating tumor initiation (Langenau et al., 2008). Importantly, the zebrafish model accurately mimics many of the molecular underpinnings of the human disease and has been used to uncover important genes and pathways relevant to human cancer (Chen et al., 2013a, 2014; Ignatius et al., 2012; Langenau et al., 2007; Le et al., 2013). The model has also been used to identify functional heterogeneity in molecularly defined cell types, including isolation of myf5:GFP+ TPCs (Ignatius et al., 2012). In total, the zebrafish kRAS^G12D ERMS model has emerged as one of the most relevant for discovering pathways that drive cancer growth in human RMS (Chen et al., 2013a, 2014; Ignatius et al., 2012; Kashi et al., 2015; Langenau et al., 2007, 2008; Le et al., 2013; Storer et al., 2013; Tang et al., 2016)

Here we show that Myf5 is not only a marker of TPCs in the zebrafish ERMS model (Ignatius et al., 2012), but was sufficient to impart tumor propagating potential to differentiated ERMS cells in vivo. Myf5 re-expression also lead to tumors that initiated earlier, had higher penetrance, and were larger than kRAS^G12D-alone expressing ERMS. Experiments in human RMS uncovered significant inter-tumoral heterogeneity of MRF expression with high MYF5 or MYOD defining largely mutually exclusive groups of tumors. Functional studies showed that both MYF5 and MYOD are required for continued RMS proliferation, likely acting redundantly with one another to regulate common molecular programs found in normal muscle development and regeneration. Consistent with this interpretation, ChIP-seq analysis identified common binding sites of MYF5 and MYOD in promoter and enhancer regions of genes that regulate cell cycle and muscle differentiation. A subset of these same genes were confirmed to be downregulated upon MYF5 or MYOD knockdown. Finally, we show that MYF5 and MYOD are also required for efficient human RMS tumor growth in vivo. Our data supports a previously unappreciated role for MYF5 and MYOD in regulating growth, proliferation, and TPC activity in rhabdomyosarcoma.

Results

Re-expression of myf5 in zebrafish ERMS cells accelerated tumor onset and increased penetrance

We have uncovered that myf5 is highly expressed in undifferentiated, molecularly defined TPCs in zebrafish kRAS^G12D-induced ERMS (Langenau et al., 2007; Ignatius et al., 2012). Remarkably, this TPC fraction shares molecular and functional properties with non-transformed muscle satellite stem cells. For example, cell transplantation and direct live cell imaging has revealed that myf5:GFP+/myosin-negative progenitor cells drive tumor growth and specifically label TPCs in this animal model (Ignatius et al., 2012; Chen et al., 2014). To assess roles for myf5 in regulating ERMS growth, we transgenically expressed myf5 under control of the differentiated myosin light chain muscle promoter (mylpfa). This transgene faithfully drives expression in terminally-differentiated muscle cells in both transient and stable transgenic fish (Xu et al., 1999; Langenau et al., 2007; Ignatius et al., 2012; Storer et al., 2013; Chen et al., 2014) and has been used to identify zebrafish ERMS cell subfractions that lack myf5, have low proliferative capacity, cannot self-renew, and do not sustain ERMS growth in vivo (Ignatius et al., 2012). Here, rag:kRAS^G12D was co-injected with mylpfa:myf5 into one-cell-stage zebrafish and analyzed for tumor onset.

Histological analysis was performed on ERMS tumors arising in rag2-kRAS^G12D;mylpfa-myf5 AB-strain transgenic fish and compared with those that express only kRAS^G12D (Figure 1A–F, Figure 1—figure supplement 1). Tumors were histologically staged based on differentiation (Storer et al., 2013; Figure 1—figure supplement 2). As reported previously, primary kRAS^G12D-induced ERMS were comprised of 50% undifferentiated stage 1 ERMS (N = 5 of 10, Figure 1B,C and Figure 1—figure supplement 2), which harbored mostly small round blue cells. By contrast, mylpfa:myf5 expressing primary ERMS contained only 7.7% stage 1 ERMS (N = 2 of 26, p=0.015, Chi-square test, Figure 1E–F), with the remaining tumors being highly differentiated stage 2 and 3 ERMS (Figure 1F and Figure 1—figure supplement 2). These tumors had large numbers of rhabdomyoblasts and cells with fibrous and spindle cell morphology. Transcriptional profiling of bulk tumor cells by qRT-PCR confirmed that mylpfa:myf5 expressing ERMS cells had high myf5 transgene expression, were more differentiated, and yet also had elevated expression of TPC-associated markers including c-met and cadherin 15 (Figure 1G, Langenau et al., 2007; Ignatius et al., 2012). These gene markers are also commonly expressed in zebrafish muscle progenitor and satellite cells (Siegel et al., 2013; Gurevich et al., 2016). Collectively, these data show that re-expression of myf5 in myosin-expressing ERMS cells leads to tumors with differentiated morphology and are consistent with the re-activation of muscle stem cell programs in differentiated cell types.

Figure 1 with 3 supplements see all

Download asset Open asset

Transgenic *myf5* elevates tumor cell differentiation, increases tumor size, and accelerates time to primary tumor-onset when expressed in myosin-expressing ERMS cells.

(A–F) Primary ERMS developing in *myf5:GFP/mylpfa:mCherry* AB-strain zebrafish. Transgenic *kRAS^G12D*-expressing ERMS (A–C) compared with those that express both *kRAS^G12D*and *mylpfa:myf5* (D–F). Animals imaged at 35 dpf (A,D). Hematoxylin and Eosin-stained sections of representative tumors (B,E) and quantification of differentiation within individual tumors (C,F; 1-less differentiated and 3-most differentiated). Asterisk denotes p=0.015 by Chi-square test. (G) Quantitative real-time PCR gene expression performed on bulk ERMS cells, confirming high *myf5* expression, increased differentiation, and high expression of TPC associated genes in ERMS that co-express *kRAS^G12D*and *mylpfa:myf5* (K+M, N = 5). Endogenous *myf5* was assessed using primers specific to the 3’UTR and total *myf5* assessed by primers that amplify the coding sequence (cds). *cadherin 15* (*cdh15*) and *myogenin* (*myog*). *kRAS^G12D* alone expressing ERMS (K, N = 4). Average gene expression with 50% confidence intervals denoted by box. Mean, maximum, and minimum also denoted. (H) Relative tumor size of primary ERMS at 30 days post fertilization (dpf). Box shows 50% confidence interval. Mean, maximum, and minimum denoted. Asterisk denotes p=0.0108, Student’s t-test. (I) Kaplan-Meijer analysis denoting time-to-tumor onset (p<0.001, Log-rank Statistic, N = 494 fish analyzed for K and N = 470 for K+M). Scale bars equal 2 mm (A,D) and 50 μm (B,E). Asterisks in panels G-H denote *p<0.05; **p<0.01; ***p<0.001 by Student’s t-test.

https://doi.org/10.7554/eLife.19214.002

Tumors arising in double transgenic rag2:kRAS^G12D; mylpfa:myf5 expressing ERMS were also larger by 30 days postfertilization than those that expressed only kRAS^G12D (Figure 1H, p=0.0108, Student’s t-test). However, apoptosis was not altered following re-expression of myf5 (Figure 1—figure supplement 3). Proliferation was assessed by both phospho-histone H3 staining and EDU incorporation following intra-peritoneal injection and assessed at 6 hr. From this analysis, we uncovered wide variation in proliferation between tumors, with a trend toward increased proliferation in mylpfa:myf5 expressing ERMS when assessed by EDU incorporation (Figure 1—figure supplement 3D). Together, our data support a model where mylpfa:myf5 expressing ERMS initiate earlier and with higher penetrance than those that express only kRAS^G12D (Figure 1I, p<0.001, log-rank Mantel-Cox test), likely reflecting a dominant role for transgenic myf5 in transforming a wider range of cell types and to a lesser degree on elevating proliferation.

To confirm that differentiation changes were confined to fully transformed ERMS cells, we next assessed the histology of ERMS following transplantation into immune-deficient rag2^E450fsrecipient fish (Figure 2A–F; Tang et al., 2014). kRAS^G12D-expressing ERMS were comprised exclusively of undifferentiated stage one tumors (Figure 2B,C and Figure 2—figure supplement 1, n=10 transplanted fish arising from four independent tumors). By contrast, ERMS that re-expressed myf5 had differentiated histology and were comprised exclusively of stage 2 and 3 tumors (Figure 2E,F and Figure 2—figure supplement 1, n=15 transplanted fish from four independent tumors, p<,0.001, Chi-square test). Consistent with our histological evaluation, flow cytometric analysis revealed that differentiated, mylpfa:mCherry-positive (R+) tumor cells were greatly expanded in ERMS that aberrantly express myf5 (Figure 2G–I, p=0.006, Student’s t-test). These same tumors had reduced numbers of myf5:GFP (G+) and double-positive (G+R+) cells. As was seen in primary ERMS, mylpfa:myf5 expressing ERMS also initiated earlier and with higher penetrance when engrafted into rag2^e450fsrecipient animals (Figure 2J, 2.5 × 10⁵ cells/animal, p=0.046, Mantel-Cox log-rank statistic). These tumors also had a trend toward being larger when assessed at 30 days post-transplantation (Figure 2K). Effects on ERMS differentiation were confirmed in transplanted CG1 strain syngeneic animals, showing that mylpfa-myf5 expressing ERMS were more differentiated based on morphology (Figure 3—figure supplement 1A–F, p<0.01, Chi-square test) and contained larger numbers of differentiated, mylpfa-mCherry-positive (R+) ERMS cells (Figure 2L, p<0.001, Student’s t-test). These transplanted tumors also had significant reductions in myf5-GFP (G+) and double-positive (G+R+) ERMS cells. Together, these data confirm that mylpfa:myf5 expressing ERMS were fully transformed and exhibited a more differentiated cellular phenotype when compared with kRAS^G12D alone expressing ERMS (Figure 2E,F and Figure 2—figure supplement 1).

Figure 2 with 1 supplement see all

Download asset Open asset

Tumors that transgenically express *myf5* are fully transformed and retain a differentiated phenotype following engraftment into recipient animals.

(A–F) Analysis of ERMS arising in transplanted fish. *kRAS^G12D* expressing ERMS arising in *rag2^E450fs* transplant recipient fish (A–C) compared with those that express both *kRAS^G12D* and *mylpfa:myf5* (D–F). Tumors were created in stable transgenic *myf5:GFP/mylpfa:mCherry* transgenic, AB-strain zebrafish and imaged following engraftment into recipient fish at 30 days post transplantation (dpt). Hematoxylin and eosin stained sections of representative tumors (B,E) and quantification of differentiation within individual ERMS (C,F; 1-less differentiated and 3-most differentiated). Asterisks denote p<0.001 by Chi-square test. (G,H) Representative flow cytometry analysis of fluorescently-labeled ERMS cells isolated from transplanted *rag2^E450fs* zebrafish. (I) Graphical summary of ERMS cell sub-fractions that grow following engraftment into immune-deficient *rag2^E450fs* recipients. Individual tumors are represented as separate bars with the proportion of G+ (green), G+R+ (yellow) and R+ (red) sub-populations denoted. **p=0.006. (J) Kaplan-Meijer analysis showing time-to-tumor onset in transplanted ERMS arising in *rag2^E450fs* zebrafish (p=0.046, Log-rank Statistic, 2 × 10⁵ cells/fish, N > 12 animals per arm, representing ≥3 independently-arising primary ERMS). (K) Relative tumor size at 30 days post engraftment (same animals analyzed as in J). (L) ERMS cells were also more differentiated following engraftment of *myf5:GFP/mylpfa:mCherry* ERMS cells into syngeneic recipient fish (p<0.001, Student’s T-test, N ≥ 3 independently arising primary ERMS and assessed in n ≥ 2 animals per transplanted tumor). Scale bars equal 2 mm (A,D) and 50 μm (B,E).

https://doi.org/10.7554/eLife.19214.006

Myf5 reprograms differentiated ERMS cells into TPCs

Because endogenous myf5 expression labels molecularly defined TPCs in zebrafish kRAS^G12D-induced ERMS (Ignatius et al., 2012; Chen et al., 2014), we next questioned if TPC frequency might be altered in mylpfa:myf5 expressing ERMS. Specifically, rag2:kRAS^G12D was co-injected with or without mylpfa:myf5 into one-cell stage, CG1 syngeneic myf5:GFP/mylpfa:mCherry transgenic animals (Mizgireuv and Revskoy, 2006; Ignatius et al., 2012). Following tumor growth in primary transplant recipients (Figure 3A–C), cell subpopulations were isolated by FACS and transplanted into secondary syngeneic recipient fish at limiting dilution (Figure 3D–I, Figure 3—figure supplement 1G–P, 1x10³−10 cells/animal, purity >85%, and >95% viability). As previously reported (Ignatius et al., 2012), only the myf5:GFP+ (G+) single-positive ERMS cells from kRAS^G12D-alone expressing ERMS could efficiently engraft tumors into CG1-strain syngeneic recipient animals (Figure 3J and Table 1; N = 3 independent tumors). By contrast, both the myf5:GFP+ single-positive (G+) and differentiated myf5:GFP+; mylpfa:mCherry+ double positive (G+R+) ERMS cells could engraft disease when isolated from mylpfa:myf5 expressing ERMS (N = 3 tumors analyzed, Table 1 and Figure 3L; p=0.0002, ELDA analysis). Importantly, engrafted tumors displayed similar differentiated histology following engraftment with sorted cells when compared with primary tumors (Figure 3B,E,H). Quantitative real-time PCR of sorted cell fractions also showed largely similar myogenic gene expression in ERMS cell subfractions isolated from either kRAS^G12D-expressing or kRAS^G12D+ myf5 expressing ERMS, confirming that our cell lineage labeling approach identified similar molecularly-defined subpopulations of ERMS cells in these tumors. One notable exception was myf5, which was also highly expressed in the G+R+ population of mylpfa:myf5 transgenic tumor as expected (Figure 3K,M). Taken together, these data show that re-expression of myf5 can lead to acquisition of tumor propagating potential in differentiated mylpfa-expressing ERMS cells in the zebrafish model.

Figure 3 with 1 supplement see all

Download asset Open asset

Limiting dilution cell transplantation shows that *myf5* can confer tumor-propagating ability to differentiated *myf5:GFP+/mylpfa:mCherry+* cells.

Tumors were generated in *myf5:GFP/mylpfa:mCherry* CG1-strain syngeneic zebrafish. Representative tumors arising in primary transplanted fish (1°T, A–C) or secondary transplanted fish following engraftment with highly purified *myf5*:GFP+, *mylpfa*:mCherry-negative (2°T G+, D–F) or *myf5*:GFP+, *mylpfa*:mCherry+ ERMS cells (2°T G+R+, G–I). Sort purity following FACS is noted in the lower left panels of D and G and was >92% for each population. These cells were used for cell transplantations and data provided in D-I. Cell viability was >95%. (J,L) Graphical summary of tumor engraftment following limiting dilution cell transplantation using highly purified sorted ERMS cells. Data is combined from all tumors shown in Table 1. ***p<0.0002 by ELDA analysis. (K,M) Relative gene expression analysis of sorted G+ or G+R+ ERMS cells from representative *kRAS^G12D* (K) or *kRAS^G12D; mylpfa:myf5* (M) expressing ERMS (Standard Deviation, n = 3 technical replicates per PCR condition). *p<0.05; **p<0.01 and ***p<0.001 by Student’s t-test.

https://doi.org/10.7554/eLife.19214.008

Table 1

myf5 confers tumor-propagating ability to differentiated myf5-GFP+/mylpfa-mCherry+ ERMS cells. Engrafted animals per cell dose are noted. Experiments for three independent tumors are shown. G+ (myf5-GFP+/mylpfa-mCherry-), G+R+ (myf5-GFP+/mylpfa-mCherry+), R+ (myf5-GFP-/mylpfa-mCherry+), DN (myf5-GFP-/mylpfa-mCherry-). Not applicable (NA); tumor-propagating cell frequency (TPC Freq.); 95% confidence interval (95% CI). Lower panel denotes cumulative TPC frequency for all three ERMS analyzed per genotype. Asterisk denotes p=0.0002 by ELDA analysis.

https://doi.org/10.7554/eLife.19214.010

*kRAS^G12D Tumor #1*					*kRAS^G12D + mylpfa:myf5 Tumor #*1
Cell #	G⁺	G⁺R⁺	R⁺	DN	Cell #	G⁺	G⁺R⁺	R⁺	DN
1000	6 of 6	2 of 7	0 of 6	0 of 7	1000	2 of 3	4 of 5	0 of 6	0 of 6
100	5 of 9	0 of 9	0 of 8	0 of 10	100	6 of 10	2 of 10	0 of 8	0 of 7
10	0 of 8	0 of 8	0 of 9	0 of 7	10	3 of 10	1 of 10	0 of 10	0 of 8
TPC Freq.	1 in 140	1 in 3561	NA	NA	TPC Freq.	1 in 81	1 in 477	NA	NA
95% CI	59–329	872–13740	NA	NA	95% CI	40–165	201–1129	NA	NA
*kRAS^G12D Tumor #2*					*kRAS^G12D + mylpfa:myf5 Tumor #2*
Cell #	G⁺	G⁺R⁺	R⁺	DN	Cell #	G⁺	G⁺R⁺	R⁺	DN
1000	6 of 6	0 of 6	0 of 6	0 of 6	1000	2 of 3	1 of 2	0 of 7	0 of 7
100	4 of 7	2 of 10	0 of 10	0 of 10	100	1 of 6	3 of 6	0 of 7	0 of 10
10	1 of 8	0 of 9	0 of 10	0 of 8	10	0 of 8	0 of 9	0 of 10	0 of 8
TPC Freq.	1 in 109	1 in 3495	NA	NA	TPC Freq.	1 in 809	1 in 467	NA	NA
95% CI	44–270	808–15120	NA	NA	95% CI	244–2685	137–1589	NA	NA
*kRAS^G12D Tumor #3*					*kRAS^G12D + mylpfa:myf5 Tumor #3*
Cell #	G⁺	G⁺R⁺	R⁺	DN	Cell #	G⁺	G⁺R⁺	R⁺	DN
1000	2 of 3	0 of 2	0 of 3	0 of 4	1000	2 of 3	3 of 5	0 of 3	0 of 3
100	8 of 9	0 of 8	0 of 8	1 of 8	100	3 of 10	1 of 10	0 of 9	0 of 10
10	1 of 8	0 of 9	0 of 9	0 of 9	10	0 of 10	0 of 10	0 of 10	0 of 10
TPC Freq.	1 in 159	NA	NA	1 in 4840	TPC Freq.	1 in 530	1 in 1080	NA	NA
95% CI	63–401	NA	NA	632–37094	95% CI	194–1445	395–2957	NA	NA
*Cumulative TPC frequency kRASG12D*					*Cumulative TPC frequency kRAS^G12D + mylpfa:myf5*
Cell #	G⁺	G⁺R⁺	R⁺	DN	Cell #	G⁺	G⁺R⁺	R⁺	DN
TPC Freq.	1 in 146	1 in 4206	NA	NA	TPC Freq.	1 in 377	1 in 639*	NA	NA
95% CI	87–245	1550–11409	NA	NA	95% CI	212–670	363–1125	NA	NA

MYF5 and MYOD are required for continued tumor growth in human RMS

To explore the role of MRFs in human RMS, we next assessed MYF5 and MYOD transcript expression in human primary tumor samples. Analysis of microarray gene expression (N = 133 samples) (Davicioni et al., 2009) and RNA-sequencing (RNA-seq) data sets (N = 98 samples) (Shern et al., 2014) uncovered that MYOD and MYF5 were expressed along with specific muscle genes and defined two distinct gene regulatory modules in human RMS. One gene module included the co-expression of MYF5, MYF6 and PAX7 while the other expressed MYOD and higher levels of CDH15, and MYOG (Figure 4A,B). This correlation in gene expression was seen in comparison of all human RMS (Figure 4A,B) or within specific RMS subtypes (Figure 4—figure supplement 1), suggesting that MYF5 and MYOD likely sit atop a transcriptional hierarchy to regulate muscle-specific gene programs in RMS. We next assessed a panel of human RMS cell lines for expression of MYF5 and MYOD following Western blot analysis. Remarkably, we found that the expression of these proteins was largely mutually exclusive in human RMS cell lines (N = 7, Figure 4C), suggesting that these proteins may act redundantly to regulate human RMS growth. This analysis also uncovered that only the Rh18 ERMS cells expressed MYF5 in our panel of human cell lines. MYF5 and MYOD expression were also assessed at the single cell level through immunofluorescence and verified that MYF5 and MYOD were mutually exclusively expressed in Rh18 and RD cells (Figure 4—figure supplement 2). Collectively, our data show significant inter-tumoral heterogeneity in the expression of myogenic factors in human RMS and suggests convergence of these transcription factors on regulating a common set of genes that are likely required for RMS growth.

Figure 4 with 4 supplements see all

Download asset Open asset

MYF5 and MYOD are required for human ERMS proliferation and growth.

(A–B) Pearson correlation for gene expression of myogenic genes in primary human RMS as assessed by microarray (A) or RNA-sequencing (B). Heatmap represents correlation coefficients. (C) Western blot analysis for MYF5 and MYOD in human RMS cell lines. (D–I) Rh18 ERMS cells following MYF5 knockdown with siRNA (D–F) or shRNA (G–I). (J–O) RD ERMS cells following MYOD knockdown with siRNA (J–L) or shRNA (M–O). Western blot analysis following knockdown at 48 hr (D,J) and 72 hr (G,M). EdU and Propidium Iodide (PI) cell cycle analysis assessed by flow cytometry at 48 hr (E,F,K,L) and 72 hr (H,I,N,O). Standard Deviation denoted in FACS plots and graphs. Analysis shown in D-O was completed as technical replicates and completed ≥3 independent times with similar results. ***p<0.001 by Student’s t-test.

https://doi.org/10.7554/eLife.19214.011

Human Rh18 ERMS cells express high levels of MYF5 and were utilized in loss-of-function studies to assess roles in regulating proliferation, growth, and apoptosis. MYF5 protein expression was effectively reduced following si-MYF5 mediated knockdown (Figure 4D) and resulted in significant impairment of cell proliferation as assessed by EdU-incorporation and flow cytometry (Figure 4E–F, p<0.001, Student’s t-test). For example, si-MYF5 treated cells showed a remarkable 70% reduction in S-phase cycling cells following 48 hr of treatment (p<0.001, Student’s t-test, Figure 4E–F). Apoptosis was not increased in si-MYF5 treated cells at 72 hr but lead to increased numbers of apoptotic cells by 96 hr (Figure 4—figure supplement 3B). These results were independently confirmed using stable knockdown with three independent lentiviral shRNAs specific to MYF5 (Figure 4G–I, protein knockdown ranged from 50–95%). All sh-MYF5 knockdown cells showed a remarkable cell cycle arrest with a virtual abrogation of S-phase cycling cells (Figure 4H–I and Figure 4—figure supplement 3A; p<0.001, Student’s t-test). This phenotype was not associated with an overall increase in apoptosis (Figure 4—figure supplement 3C,D) and yet lead to a significant 60% decrease in cell number as assessed by manual nuclei counts performed at 96 hr and compared with shRNA control treated cells (Figure 4—figure supplement 3E). We conclude that MYF5 loss results in impaired cell cycle and secondarily elevates apoptosis.

To assess if MYOD can also drive continued tumor growth and proliferation in human RMS, we next performed MYOD knockdown in human RD ERMS cells. siRNA knockdown resulted in reduced proliferation and a striking reduction in S-phase cycling cells (Figure 4J–O; p<0.001, Student’s t-test). In keeping with our model that apoptotic cell death was induced secondary to cell cycle defects, we found that si-MYOD treatment had no effect on viability after 72 hr of knockdown, while analysis at 96 hr post-treatment resulted in elevated cell death of RD knockdown cells (Figure 4—figure supplement 3G). shRNA knockdown showed similar effects on suppressing cell cycle progression and growth (Figure 4M–O and Figure 4—figure supplement 3F,H). In fact, RD cell number was reduced >30% following stable sh-MYOD knockdown (Figure 4—figure supplement 3H). In support of MYOD having important roles in regulating TPC number, sphere colony formation was also greatly reduced in RD cells following shRNA knockdown (p<0.01, Figure 4—figure supplement 3I,J). Sphere colony formation is an in vitro surrogate for quantifying TPC number and correlates well with in vivo limiting dilution cell transplantation experiments (Walter et al., 2011; Satheesha et al., 2016). Finally, MYOD knockdown also impaired cell cycle and growth in additional RMS cell lines including ERMS cell lines 381T and RMS559 and the ARMS cell line Rh3. These cells lines all had significant reductions in S-phase cycling cells following siRNA treatment with variable effects on apoptosis at the time points analyzed (Figure 4—figure supplement 4A–L). Importantly, specificity of MYF5 and MYOD knockdown was confirmed in multiple cells lines, showing that siRNA knockdown effects were specific to each MRF and that myogenic factors were not redundantly re-expressed following knockdown (Figure 4—figure supplement 4M–T). In total, our data show that MYF5 and MYOD are individually expressed in different RMS tumor cells and yet have similar roles in regulating cell cycle progression and proliferation in RMS.

Myogenic transcription factors are required for continued xenograft growth

Given the prominent role MYF5 had in regulating cell growth in human ERMS cells in vitro and imparting tumor propagating potential to differentiated zebrafish ERMS cells, we next wanted to assess if MYF5 was required for ERMS maintenance and growth in vivo. Rh18 cells were infected with shRNAs and harvested at 72 hr post-infection. MYF5 knockdown was confirmed by Western blot analysis (Figure 5A). Luciferase-mKate expressing Rh18 shRNA cells were transplanted into the flanks of NOD/SCID/IL2rg null mice (1 × 10⁶ viable cells in matrigel per site). Non-targeting control shRNA cells were implanted subcutaneously into the left flank and MYF5 knockdown cells into the right (N = 6 animals, two independent shRNAs). 5 hr after injection, mice were injected with luciferin and bioluminescence was measured, confirming that the same amount of control and knockdown cells had been injected into recipient mice (Figure 5B). Serial bioluminescence imaging showed that tumor volume was reduced in MYF5 knockdown cells while control cells continued to grow (p<0.05, Student’s t-test; Figure 5—figure supplement 1A). MYF5 knockdown cells were largely undetected at late time points (p<0.05, Student’s t-Test; Figure 5B,C). Analysis of mice at necropsy revealed that only 2 of 12 mice had tumors derived from MYF5-deficient Rh18 cells and that overall tumor weight was greatly reduced when compared with control Rh18 knockdown cells (p<0.01; Student’s t-test; Figure 5D–E). For these rare ERMS that developed from MYF5 knockdown Rh18 cells, they retained ERMS histology (Figure 5—figure supplement 1D,E). Taken together, these data show that MYF5 is required for efficient xenograft tumor cell growth in vivo.

Figure 5 with 1 supplement see all

Download asset Open asset

MYF5 and MYOD are required for human ERMS xenograft growth.

Xenograft growth in Rh18 (A–E), RD (F–J), and RMS559 (K–O) following knockdown with scramble control shRNA (sh-SCR) or shRNAs specific to MYF5 or MYOD. (A,F,K) Western blot analysis of shRNA expressing cells harvested for transplantation at 72 hr after lenti-viral shRNA knockdown. Percent knockdown compared to shRNA control is shown. (B,G,L) Luciferase bioluminescent imaging of a representative animal at the time of implantation (left panel) or at later time points (right panel). Control shRNA cells were implanted into left flank and knockdown cells into the right (N = 6 mice per shRNA). Intensity represents total luminescence units measured per region of interest (L.U.) (C,H,M) Quantification tumor volume when assessed by luciferase imaging. Relative luminescence units (R.L.U). (D,I,N) Representative images of mice at the time of necropsy, with excised tumors shown in lower panels. (E,J,O) Quantification of tumor weight at the time of necropsy. Tumors that could not be identified at time of necropsy were assigned a value of zero for this analysis. Standard Error of the Mean are denoted in graphs. **p<0.01; ***p<0.001 by Mann-Whitney non-parametric test. Scale bar equals 1 cm in D,I, and N.

https://doi.org/10.7554/eLife.19214.016

Next, we assessed if MYOD was also important for continued xenograft growth of human ERMS cells. Using the same approach outlined above, we found that shRNA knockdown of MYOD resulted in efficient knockdown prior to tumor cell implantation (Figure 5F,K) and reduced xenograft growth of both RD and RMS559 cells when assessed by total body luciferase imaging completed overtime (p<0.01; Figure 5F–H,K–M and Figure 5—figure supplement 1B,C). MYOD knockdown tumors were also smaller at the time of necropsy (Figure 5I,N) and weighed significantly less than control shRNA expressing tumors (p<0.01, Student’s t-test; Figure 5J,O). Unlike our Rh18 experiments, MYOD knockdown cells continued to grow in transplant recipient animals, albeit at greatly reduced levels. These data suggest both important similarities for MRF factors in driving proliferation and growth and yet, also suggest that additional molecular mechanisms likely contribute to continued tumor growth in MYOD-expressing ERMS. As with the Rh18 experiments, tumors that formed following shMYOD knockdown retained similar RMS morphology when compared with control treated cells (Figure 5—figure supplement 1F–I).

MYF5 and MYOD control common transcriptional targets to regulate proliferation and myogenic state in human RMS

Our work uncovered that MYF5 and MYOD are largely mutually exclusively expressed in RMS and that each is required for proliferation. MYF5 and MYOD also bind directly to enhancers of well-known muscle regulated genes in development, including myogenin and CDH15 (m-cadherin) (Conerly et al., 2016). These data suggest that these transcription factors likely regulate a common set of transcriptional targets that lock RMS cells in a proliferative myogenic state. To further explore this hypothesis, we next performed ChIP-seq for MYF5 in Rh18 cells and compared these results with ChIP-seq data performed for MYOD in RD cells (MacQuarrie et al., 2013a) (Figure 6 and Figure 6—figure supplement 1). This analysis uncovered a common set of promoter and enhancer regions bound by both MYOD and MYF5 (Figure 6A). 86% of commonly bound genomic DNA regions were confined to enhancer regions as defined by H3K27-acetylation occupancy (Figure 6A,B). Unbiased analysis of commonly bound target genes using GREAT (McLean et al., 2010), Supplementary file 2) revealed an enrichment of genes that regulate cell cycle and myogenic cell fate (Figure 6C). Enrichment of GO terms included ‘cyclin-dependent protein kinase holoenzyme complex’ (Supplementary file 3), ‘skeletal muscle tissue development’ and ‘embryonic skeletal system development’ (Supplementary file 4; binomial, p<1x10⁻⁹). Signal tracks of ChIP-seq and RNA-seq independently confirmed common binding of MYF5 and MYOD to genes that regulate cell cycle and myogenic cell fate (Figure 6D), including cyclin-dependent kinase cyclin D2 (CCND2), myogenin (MYOG), and cadherin 15 (m-Cadherin, CDH15, Figure 6D and Figure 6—figure supplement 1B).

Figure 6 with 2 supplements see all

Download asset Open asset

MYF5 and MYOD bind common promoter and enhancer regions and induce genes involved in muscle development and cell cycle.

(A–B) ChIP-seq analysis showing genomic regions bound by both MYOD in RD cells and MYF5 in RH18 cells. H3K27 acetylation (H3K27ac). (C) Gene ontology enrichment of gene regions bound by both MYOD in RD cells and MYF5 in RH18 cells. GO Biological Processes, GO Cellular Component predictions, and binomial p-values denoted. (D) Signal tracks for ChIP-seq and RNA-seq surrounding *MYOG* (top) and *CCND2* (bottom). Numbers to the right indicate reads per million mapped reads. (E) Quantitative real-time PCR gene expression analysis of RH18 (top) and RD cells (bottom). Cells were assessed following siRNA-mediated knockdown at 2 days (2dpt, blue bars) or 3 days post-transfection (three dpt, red bars). Error bars denote standard deviation. Student’s t-test; *p<0.05, **p<0.01, ***p<0.001.

https://doi.org/10.7554/eLife.19214.018

In order to show that MYF5 and MYOD are regulators of cell fate, we next performed qRT-PCR gene expression analysis for muscle differentiation genes following si-MYF5 or si-MYOD knockdown in Rh18 and RD cells, respectively (Figure 6E). As expected if myogenic transcription factors regulate muscle cell fate, both Myogenin and CDH15 were downregulated following MYF5 or MYOD knockdown at either 48 or 72 hr (Figure 6E). Additionally, the cell cycle regulatory gene, CCND2 was also reduced following MYF5 or MYOD knockdown (Figure 6E). Moreover, the zebrafish orthologue ccnd2a transcript showed a trend toward higher expression in mylpfa:myf5 expressing zebrafish ERMS (Figure 6—figure supplement 2), correlating well with elevated myf5 expression in these tumors. CCND2 is a CDK4/6-associated cyclin that is amplified in a subset of human RMS and is required for cell proliferation and viability in human RMS (Chen et al., 2013a). Importantly, CCND2 is a predicted direct target of MYF5 and MYOD binding (Conerly et al., 2016). Together, our data show that MYF5 and MYOD regulate common gene programs that lock cells in an arrested myogenic fate and are required for sustained proliferation of RMS cells.

Discussion

Rhabdomyosarcomas express b-Helix-loop-Helix (bHLH) myogenic regulatory transcription factors (MRFs), including MYF5 and MYOD (Clark et al., 1991; Kumar et al., 2000; Sebire and Malone, 2003) but fail to activate terminal muscle differentiation programs. Several mechanisms have been shown to play a role in this differentiation arrest. These include disruption of the balance of MRF-E12 heterodimers and inhibitor complexes (Macquarrie et al., 2013b; Yang et al., 2009), presence of inhibitory miRNAs (Macquarrie et al., 2012), and deregulation of cell cycle (Fiddler et al., 1996). These data have therefore led to the suggestion that MRFs do not have a role in RMS transformation or in sustained tumor growth, but are merely retained from the target cell of transformation (Keller and Guttridge, 2013). While MYOD overexpression fails to differentiate ERMS cells (Yang et al., 2009), both MYF5 and MYOD potently reprogram fibroblasts into proliferating muscle cells (Braun et al., 1989; Tapscott et al., 1988). Moreover, MYF5 and MYOD are commonly re-expressed in experimental animal models of RMS irrespective of the cell of origin (Hettmer et al., 2011, 2015; Ignatius et al., 2012; Langenau et al., 2007; Rubin et al., 2011; Storer et al., 2013) suggesting roles for these transcription factors in driving tumor growth and TPC function. Our experiments have shown that Myf5 can impart tumor-propagating potential to ERMS cells in the zebrafish model, suggesting important roles for myogenic regulatory transcription factors in regulating growth. These data were confirmed by loss-of-function studies in human RMS where MYF5 or MYOD knockdown suppressed RMS proliferation and reduced viability in vitro. Similar effects were also observed in xenograft studies, where MYF5- and MYOD-knockdown ERMS cells failed to grow efficiently in vivo. Remarkably, despite over 25 years of study into MYOD and MYF5, loss-of-function studies in RMS have not been reported and thus roles for these factors in regulating RMS growth have gone unappreciated.

It is becoming increasingly noted that developmental transcription factors and pathways are commonly co-opted by cancer to regulate growth and tumor-propagating activity. For example, the TAL1/SCL bHLH transcription factor is required for hematopoietic stem cell specification and self-renewal during development. TAL1/SCL is overexpressed in 60% of T-cell acute lymphoblastic leukemia (T-ALL) (Ferrando et al., 2002) and can reprogram thymocytes into self-renewing, pre-leukemic cells (Gerby et al., 2014). This same paradigm has also been seen in brain tumors. For example, the bHLH transcription factor OLIG2 is required for self-renewal of normal neural progenitor cells (Imayoshi and Kageyama, 2014) and is also a marker of glioblastoma TPCs (Beyeler et al., 2014; Ligon et al., 2007; Suvà et al., 2014; Trépant et al., 2015). Our results in RMS parallel those outlined for T-ALL and glioblastoma, showing that MRFs are capable of reprogramming differentiated RMS cells into TPCs and are required for sustained proliferation and viability of human RMS. Our data also suggests, that like T-ALL and glioblastoma, the MYF5 and MYOD bHLH proteins regulate common molecular pathways in self-renewal and growth of both normal and malignant muscle. Importantly, transcription factors have recently been therapeutically targeted (Bhagwat and Vakoc, 2015; Roe et al., 2015), raising hope that developing drugs that inhibit MYF5 and MYOD cancer cell dependencies could be efficacious in treating RMS patients in the future.

Our molecular analysis also uncovered that MYF5 and MYOD are mutually-exclusively expressed in human RMS. Loss-of-function studies uncovered important roles for either MYF5 or MYOD in regulating RMS growth and muscle cell fate. This data contrasts starkly with co-expression of these factors and functionally overlapping roles in development. For example, MYF5 and MYOD are well-known to act redundantly in muscle development to regulate muscle specification and differentiation (Rudnicki et al., 1993). This same redundancy of Myf5 and MyoD in development and muscle injury has now been reported in zebrafish (Hinits et al., 2009; Siegel et al., 2013), Drosophila (Abmayr and Keller, 1998) and Xenopus (Chanoine and Hardy, 2003), showing a high conservation of the MYOD/MYF5 transcriptional machinery in regulating muscle specification and development throughout evolution. These same MRFs are also required for self-renewal of adult muscle satellite cells (Cooper et al., 1999; Ustanina et al., 2007; Yablonka-Reuveni et al., 1999), yet roles for these factors in individually regulating muscle fate and self-renewal are now just emerging in the literature. For example, a subset of muscle progenitors are specified by MYOD without the contribution of MYF5 (Haldar et al., 2008, 2014). Moreover, a subset of adult muscle progenitors express MYF5 and then MYOD sequentially during their specification with both being required for muscle regeneration following injury (Comai et al., 2014). Our data suggest that either MYOD or MYF5 are uniquely expressed within human RMS and are each individually sufficient to drive tumor growth. These data are consistent with similar roles for either Myf5 or MyoD to reprogram fibroblasts into muscle cell fates (Braun et al., 1989; Tapscott et al., 1988) and a high degree of overlap in binding of common enhancer and promoter targets with normal myoblasts (Conerly et al., 2016).

Finally, our work uncovered downstream pathways required for RMS growth that were regulated by MYOD and MYF5, including muscle specification programs and cell cycle. For example, m-cadherin (CDH15) and myogenin (MYOG) were transcriptionally regulated by MYF5 and MYOD in human RMS, consistent with regulation of these same factors by MYF5 and MYOD in myoblasts (Conerly et al., 2016). Our work also uncovered roles for MRFs in regulating cyclin D2 (CCND2) in human RMS. Importantly, the CDK4/6-associated cyclin D2 (CCND2) complex is required for both myoblast proliferation and human RMS growth, is highly expressed in primary human RMS and is amplified in a subset of human and zebrafish RMS (Chen et al., 2013a; Saab et al., 2006; Webster and Fan, 2013). Collectively, our data suggest that MYOD and MYF5 likely exert import roles in regulating muscle cell identity and cell cycle regulation, both of which are required for sustained tumor growth and likely shared with normal muscle to regulate stem cell self-renewal.

Materials and methods

Animals and protocol approvals

Request a detailed protocol

Studies were approved by the Massachusetts General Hospital Subcommittee on Research Animal Care under the protocol #2011 N000127 (zebrafish) and #2013 N000038 (mouse). Biosafety lentiviral work was approved by the Partners IBC under protocol #2013B000039. Zebrafish used in this work include: CG1 strain (Mizgireuv and Revskoy, 2006), myf5-GFP (Chen et al., 2007) and mylpfa-mCherry (previously mylz2-mCherry)(Xu et al., 1999) transgenic zebrafish lines and rag^E450fs (ZFIN IND rag2^fb101) homozygous fish (Tang et al., 2014; Tenente et al., 2014). myf5-GFP/mylpfa-mCherry double transgenic fish (AB strain) were outcrossed 10 times into CG1-strain zebrafish to generate compound syngeneic transgenic zebrafish (Ignatius et al., 2012). 6-week-old NOD/SCID/Il2rg null female mice were used in this work.

Micro-injection and ERMS generation in transgenic zebrafish

Request a detailed protocol

rag2-kRAS^G12D and mylpfa-mCherry constructs were described previously (Langenau et al., 2007; Smith et al., 2010). The mylpfa-myf5 construct was obtained by gateway cloning using a zebrafish myf5 ORF from 24hpf zebrafish embryo cDNA (http://tol2kit.genetics.utah.edu). rag2-kRAS^G12D and mylpfa-myf5 constructs were linearized with XhoI, phenol:chloroform-extracted, ethanol-precipitated, re-suspended in 0.5× Tris-EDTA + 0.1 M KCl, and injected into one-cell stage embryos of the respective backgrounds, as previously described (Langenau et al., 2007; Tenente et al., 2014). We and others have used the mylpfa promoter to drive transgene expression in differentiated muscle cells both in the stable and mosaic transgenic setting (Ju et al., 2003; Ignatius et al., 2012; Storer et al., 2013; Tang et al., 2016) confirming that mylpfa transgene expression is confined to differentiated ERMS cells.

Quantification of zebrafish RMS size, tumor onset, and penetrance

Request a detailed protocol

Zebrafish were monitored every 3–4 days for time-to-tumor onset using an epi-fluorescent stereomicroscope. Animals were imaged at 10 days postfertilization until 55 days postfertilization. Primary tumor size was quantified from 6.3x or 10x photomicrographs taken at 30 postfertilization and calculated by multiplying fluorescence intensity by 2D pixel area using the ImageJ software package as previously described (Chen et al., 2014). Kaplan-Meier tumor onset and penetrance analysis was performed using Graphpad Prism Software and statistically analyzed using the Log-rank statistic.

Zebrafish histology, immunohistochemistry and EdU incorporation

Request a detailed protocol

Paraffin embedding, sectioning and immunohistochemical analysis of zebrafish sections were performed as described (Chen et al., 2013a, 2014; Ignatius et al., 2012). Antibodies used for immunohistochemistry included: phospho-H3 (1:6000, Santa Cruz Biotechnology, Dallas, Texas) and cleaved-caspase3 (CC3, 1:250, Cell Signaling Technology, Danvers, MA). All histopathology procedures were performed at the MGH and BWH DF/HCC Research Pathology Cores. Slides were imaged using a transmitted light Olympus BX41 microscope. Pathology review and staging were completed by G.P.N. Tumor histology classification was assigned as described in Figure 1—figure supplement 2 and Figure 2—figure supplement 1 with stage one being the least differentiated with tumors being comprised of only small round blue cells. Stage 2 and 3 ERMS were assigned based on the preponderance of rhabdomyoblast cells, fibrous and spindle cell morphology, with a low proportion of interspersed smaller round blue cells. EdU was injected intraperitoneally into live tumor-bearing zebrafish and incubated for 6 hr prior to fixation as described previously (Ignatius et al., 2012). Animals were cryosectioned and stained using the Click-iT Alexa Fluor 647 imaging kit (Invitrogen, Carlsbad, CA). Images were acquired using a Zeiss 710 Confocal microscope(Zeiss, Oberkochen, Germany).

Zebrafish ERMS cell transplantation and FACS

Request a detailed protocol

FACS analysis and RMS cell transplantation by intra-peritoneal injection were completed essentially as described (Chen et al., 2014; Ignatius et al., 2012; Langenau et al., 2007; Smith et al., 2010). Freshly isolated RMS tumor cells were stained with DAPI to exclude dead cells and sorted twice using a Laser BD FACSAria II Cell Sorter. Sort purity and viability were assessed after two rounds of sorting when possible, exceeding 85% and 95% respectively. Fish were monitored for tumor engraftment from 10 to 120 days post transplantation. Tumor-propagating cell frequency was quantified following transplantation into CG1 syngeneic recipient fish using the Extreme Limiting Dilution Analysis software (http://bioinf.wehi.edu.au/software/elda/). A subset of transplanted fish were fixed in 4% PFA in PBS, sectioned, stained with Hematoxylin and Eosin (H and E), and staged for differentiation score.

Gene expression analysis

Request a detailed protocol

Total RNA was isolated from AB-strain embryos 6 and 24 hr postfertilization, FAC-sorted ERMS cell subpopulations, bulk unsorted primary zebrafish ERMS or human RMS samples. Quantitative real-time PCR utilized gene-specific PCR primers (Supplementary file 1), and expression was normalized to 18S controls (zebrafish samples) or GAPDH (human samples) to obtain relative transcript levels using the 2^-ddCT method. Technical triplicates were completed for all qRT-PCR reactions and data presented as average expression ±1 standard deviation. For zebrafish RMS sub-populations, relative gene expression was normalized within individual samples, and cumulative transcript expression across the two ERMS cell subpopulations was set to 50. Samples were assessed in relation to 24 hr postfertilization embryos to ensure that results for 2^-ddCT results for any given gene were not lower than 10-fold expression found in normal development, as previously described (Ignatius et al., 2012). For zebrafish bulk primary ERMS gene expression analysis, samples were also assessed in relation to 24 hr postfertilization embryos. For human RMS gene expression analysis, knockdown samples were assessed in relation to si-SCR controls.

Human RMS cell lines

Request a detailed protocol

The human RD cell line was obtained from ATCC’s cell biology collection (Manassas, Virginia). SMS-CTR, 381T, Rh3, Rh5 and Rh30 cell lines were kindly provided by Dr. Corrine Linardic (Duke University, North Carolina), the Rh18 cell line (fusion-negative) by Dr. Peter Houghton (Ohio State University, now at UTHSCSA) and RMS559 by Dr. Jonathan Fletcher (Brigham and Women’s Hospital, Massachusetts). All RMS cell lines were authenticated by STR profiling and were mycoplasma tested. Cell lines used in this work are not commonly misidentified based on the International Cell Line Authentication Committee. The human MB1208-1 human skeletal myoblast cell line was kindly provided by Dr. Louis Kunkel (Boston Children’s Hospital, Massachusetts). Characteristics of these human RMS (Hinson et al., 2013; Sokolowski et al., 2014) and skeletal myoblast (Alexander et al., 2011) cell lines have been reported previously.

Western blot analysis

Request a detailed protocol

Total cell lysates from human RMS cell lines were obtained following lysis in 2%SDS lysis buffer supplemented with protease inhibitors (Santa Cruz Biotechnology, Dallas, Texas). Samples were boiled, vortexed and homogenized through a 28G syringe. 20–40 μg of protein was loaded in 4–20% Mini-Protean TGX gels (Biorad, Hercules, CA) and transferred onto PVDF membranes. Western blot analysis used primary antibodies: rabbit a-MYF5 (1:5000, Abcam ab125078, Cambridge, MA), mouse a-MYOD1 (1:1000, Abcam ab16148, Danvers, MA), rabbit a-MYOD1 (1:1000, Abcam ab133627), rabbit a-GAPDH (1:2000, Cell Signaling 2118), mouse a-TUBULIN (1:2500, Abcam ab4074) and secondary antibodies: HRP anti-rabbit (1:2000, Cell Signaling 7074) or HRP anti-mouse (1:3000, GE Healthcare NA93IV, Marlborough, MA). Blocking was completed using 5% skim milk/TBST. Membranes were developed using an ECL reagent (Western Lightening Plus-ECL, Perkin Elmer, Waltham, MA or sensitive SuperSignal West phemto Maximum Sensitivity Substrate, Thermo Scientific, Waltham, MA).

MYF5 and MYOD siRNA knockdown and immunofluorescence

Request a detailed protocol

Gene-specific smart-pool or control siRNAs (Dharmacon, GE Life Sciences, Marlborough, MA) (0.01 μM) were reverse-transfected into cells using RNAiMax lipofectamine transfection reagent (Life Technologies, Waltham, MA) in flat clear bottom 96 well plates. Cells were then fixed at 72 hr post transfection in 4% PFA/PBS, washed in x1 PBS and permeabilized in 0.5% TritonX-100/PBS. Antibodies used were rabbit a-Myf5 (1:400, Abcam ab125078) and mouse a-MyoD (1:200, Abcam ab16148) in 2% goat serum/PBS, Alexa 488 goat anti-mouse (1:1000, Invitrogen A11029) and Alexa 594 goat anti-rabbit (1:1000 Invitrogen A11037). Cells were incubated with DAPI (1 μg/ml), and imaged at 200x using a LSM710 Zeiss Laser scanning confocal microscope. Images were processed in ImageJ and Adobe Photoshop.

For EdU and AnnexinV assays, gene-specific smart-pool or control siRNAs (Dharmacon, GE Life Sciences) (5 μM) were added to Rh18, RD, 381T, RMS559 and Rh3 cells in a 6-well plate and incubated for 48–96 hr prior to analysis.

MYF5 and MYOD lentiviral shRNA knockdown

Request a detailed protocol

Non-targeting scrambled (SCR) control shRNA and MYF5 or MYOD specific shRNAs were delivered on the pLKO.1-background vector (from MGH Molecular Profiling Laboratory) and packaged using 293T cells (Supplementary file 1). RMS cells were infected with viral particles for 24 hr at 37°C with 8 μg/ml of polybrene (EMD Millipore, Billerica, MA).

Human RMS in vitro assays

Request a detailed protocol

Nuclei counts were performed following incubation with NucBlue Live ReadyProbes Reagent (Life Technologies). Cells were imaged at 100X and 400X magnification using an inverted fluorescent microscope. Manual cell counts were performed using the ImageJ software. Three fields were counted per well and completed in triplicate. Cell cycle analysis was performed using the EdU Click-iT plus EdU Flow Cytometry-AlexaFluor 647 picol azide assay (Life Technologies) following 2 hr (RD, 381T, Rh3, RMS559) or 6 hr incubation (Rh18) with 10 μM EdU. Apoptosis was assessed using the AnnexinV-AlexaFluor 647/PI or 7AAD assays (Life Technologies and BD Biosciences, San Jose, CA). Flow cytometric analysis was performed using the SORP4 Laser BD LSRII Flow Cytometer and processed with the FlowJo Software. All experiments were run as technical triplicates and repeated ≥3 independent times. Sphere formation assays using RD cells were completed essentially as previously described (Walter et al., 2011).

Primary human RMS gene expression and correlation analysis

Request a detailed protocol

Previously published microarray gene expression data were processed and normalized using Robust Multichip Average (RMA) normalization (Davicioni et al., 2006) (raw data was obtained from the NCI Cancer Array Database). Previously published RNA-seq gene expression data from human RMS were processed and normalized using a standard Tuxedo pipeline (Shern et al., 2014; Trapnell et al., 2012). The resulting expression values from the microarray and RNA-seq datasets were then log2 transformed. Pearson correlation was determined for the following genes: CDH15, MYF5, MYF6, MYOD1, MYOG, PAX3, and PAX7. The correlation heatmap was plotted using the R package ‘fheatmap’ (Fantastic Heatmap. R package version 1.0.1. http://CRAN.R-project.org/package=fheatmap) and processed using Adobe Photoshop.

ChIP-seq of human RMS cell lines

Request a detailed protocol

Chromatin immunoprecipitations were performed on Rh18 cells using the Chip-IT High sensitivity kit (Active Motif, Carlsbad, CA) and anti-H3K27ac (Active Motif) or anti-MYF5 (C-20, Santa Cruz) antibodies. Resultant purified immune-precipitated DNA was used for library preparation using the TruSeq ChIP sample preparation kit (Illumina, San Diego ,CA) without modifications. 11–18 library preps were mixed for multiplexed single read sequencing using the NextSeq500 (Illumina).Previously published MYOD ChIP-seq data from RD was downloaded and processed in parallel with the newly generated sequencing data (GSE50415, GSE84630) (MacQuarrie et al., 2013a). Reads were aligned to the hg19 reference using BWA. ChIP-seq peaks were identified using MACS 2.1 (Zhang et al., 2008). Gene ontology was performed using GREAT (McLean et al., 2010). Differential peak calling between RD and Rh18 was performed using bedtools v2.25.0 and visualized using NGS plot (Shen et al., 2014). Genomic regions were visualized using IGV v2.3.40.

Mouse xenografts, luciferase imaging, necropsy and histological analysis

Request a detailed protocol

Rh18, RD and RMS559 ERMS cells were co-infected with pLKO.1-shRNA lentivirus and pLKO.1-luc-mKate (gift from Drs. Matthijssens and Van Vlierberghe, Ghent University, Belgium). At 3 days post-infection, cells were collected and counted. An aliquot of cells was analyzed using the SORP4 Laser BD LSRII Flow Cytometer to determine viability following DAPI staining. Separate aliquots of cells were harvested and used for Western blot analysis. Equal numbers of viable cells were then embedded into Matrigel (Corning Life Sciences, Tewksburg ,MA) at a final concentration of 1 × 10⁶ of viable cells per 200 μl. Six-week-old NOD/SCID/IL2rg null female mice were anesthetized by isofluorane and transplanted with scramble-shRNA/mKate-luc cells subcutaneously into the left flank (N = 6 animals/shRNA construct) whereas sh-MYF5 or sh-MYOD/mKate-luc cells were injected on the right (200 μl/flank injection). Tumor growth was monitored by bioluminescence imaging following subcutaneous injection into the loose tissue over the neck of 75 mg/kg D-luciferin (Perkin Elmer, Waltham, MA) in 100 μl of PBS. Imaging was completed and analyzed using the IVIS Lumina II (Caliper Life Science, Hopkinton, MA). At time of necropsy, mouse brightfield images were acquired using a regular camera and tumors were excised, weighed, processed and stained using Hematoxylin and Eosin. Comparisons of tumor size and weight between groups used the Student’s t-test after a normality test was performed, otherwise a Mann-Whitney analysis was performed, as indicated in the figure legends.

References

1. Abmayr SM
2. Keller CA
(1998) Drosophila myogenesis and insights into the role of nautilus
Current Topics in Developmental Biology 38:35–80.

https://doi.org/10.1016/S0070-2153(08)60244-6
- PubMed
- Google Scholar
1. Alexander MS
2. Casar JC
3. Motohashi N
4. Myers JA
5. Eisenberg I
6. Gonzalez RT
7. Estrella EA
8. Kang PB
9. Kawahara G
10. Kunkel LM
(2011) Regulation of DMD pathology by an ankyrin-encoded miRNA
Skeletal Muscle 1:27.

https://doi.org/10.1186/2044-5040-1-27
- PubMed
- Google Scholar
1. Beyeler S
2. Joly S
3. Fries M
4. Obermair F-J
5. Burn F
6. Mehmood R
7. Tabatabai G
8. Raineteau O
(2014) Targeting the bHLH transcriptional networks by mutated E proteins in experimental glioma
Stem Cells 32:2583–2595.

https://doi.org/10.1002/stem.1776
- Google Scholar
1. Bhagwat AS
2. Vakoc CR
(2015) Targeting transcription factors in cancer
Trends in Cancer 1:53–65.

https://doi.org/10.1016/j.trecan.2015.07.001
- PubMed
- Google Scholar
(1989)
A novel human muscle factor related to but distinct from MyoD1 induces myogenic conversion in 10t1/2 fibroblasts

The EMBO Journal 8:701–709.
- PubMed
- Google Scholar
1. Buckingham M
2. Rigby PW
(2014) Gene regulatory networks and transcriptional mechanisms that control myogenesis
Developmental Cell 28:225–238.

https://doi.org/10.1016/j.devcel.2013.12.020
- PubMed
- Google Scholar
1. Chanoine C
2. Hardy S
(2003) Xenopus muscle development: from primary to secondary myogenesis
Developmental Dynamics 226:12–23.

https://doi.org/10.1002/dvdy.10206
- PubMed
- Google Scholar
1. Chen EY
2. DeRan MT
3. Ignatius MS
4. Grandinetti KB
5. Clagg R
6. McCarthy KM
7. Lobbardi RM
8. Brockmann J
9. Keller C
10. Wu X
11. Langenau DM
(2014) Glycogen synthase kinase 3 inhibitors induce the canonical WNT/β-catenin pathway to suppress growth and self-renewal in embryonal rhabdomyosarcoma
PNAS 111:5349–5354.

https://doi.org/10.1073/pnas.1317731111
- PubMed
- Google Scholar
1. Chen EY
2. Dobrinski KP
3. Brown KH
4. Clagg R
5. Edelman E
6. Ignatius MS
7. Chen JY
8. Brockmann J
9. Nielsen GP
10. Ramaswamy S
11. Keller C
12. Lee C
13. Langenau DM
(2013a) Cross-species array comparative genomic hybridization identifies novel oncogenic events in zebrafish and human embryonal rhabdomyosarcoma
PLoS Genetics 9:e1003727.

https://doi.org/10.1371/journal.pgen.1003727
- PubMed
- Google Scholar
1. Chen X
2. Stewart E
3. Shelat AA
4. Qu C
5. Bahrami A
6. Hatley M
7. Wu G
8. Bradley C
9. McEvoy J
10. Pappo A
11. Spunt S
12. Valentine MB
13. Valentine V
14. Krafcik F
15. Lang WH
16. Wierdl M
17. Tsurkan L
18. Tolleman V
19. Federico SM
20. Morton C
21. Lu C
22. Ding L
23. Easton J
24. Rusch M
25. Nagahawatte P
26. Wang J
27. Parker M
28. Wei L
29. Hedlund E
30. Finkelstein D
31. Edmonson M
32. Shurtleff S
33. Boggs K
34. Mulder H
35. Yergeau D
36. Skapek S
37. Hawkins DS
38. Ramirez N
39. Potter PM
40. Sandoval JA
41. Davidoff AM
42. Mardis ER
43. Wilson RK
44. Zhang J
45. Downing JR
46. Dyer MA
47. St. Jude Children’s Research Hospital–Washington University Pediatric Cancer Genome Project
(2013b) Targeting oxidative stress in embryonal rhabdomyosarcoma
Cancer Cell 24:710–724.

https://doi.org/10.1016/j.ccr.2013.11.002
- PubMed
- Google Scholar
1. Chen YH
2. Wang YH
3. Chang MY
4. Lin CY
5. Weng CW
6. Westerfield M
7. Tsai HJ
(2007) Multiple upstream modules regulate zebrafish myf5 expression
BMC Developmental Biology 7:1.

https://doi.org/10.1186/1471-213X-7-1
- PubMed
- Google Scholar
1. Clark J
2. Rocques PJ
3. Braun T
4. Bober E
5. Arnold HH
6. Fisher C
7. Fletcher C
8. Brown K
9. Gusterson BA
10. Carter RL
(1991) Expression of members of the myf gene family in human rhabdomyosarcomas
British Journal of Cancer 64:1039–1042.

https://doi.org/10.1038/bjc.1991.461
- PubMed
- Google Scholar
(2014) Variations in the efficiency of lineage marking and ablation confound distinctions between myogenic cell populations
Developmental Cell 31:654–667.

https://doi.org/10.1016/j.devcel.2014.11.005
- PubMed
- Google Scholar
1. Conerly ML
2. Yao Z
3. Zhong JW
4. Groudine M
5. Tapscott SJ
(2016) Distinct activities of Myf5 and MyoD indicate separate roles in skeletal muscle lineage specification and differentiation
Developmental Cell 36:375–385.

https://doi.org/10.1016/j.devcel.2016.01.021
- PubMed
- Google Scholar
(1999)
In vivo satellite cell activation via Myf5 and MyoD in regenerating mouse skeletal muscle

Journal of Cell Science 112 (Pt 17):2895–2901.
- PubMed
- Google Scholar
(2009) Molecular classification of rhabdomyosarcoma--genotypic and phenotypic determinants of diagnosis: a report from the Children's Oncology Group
The American Journal of Pathology 174:550–564.

https://doi.org/10.2353/ajpath.2009.080631
- PubMed
- Google Scholar
(2006) Identification of a PAX-FKHR gene expression signature that defines molecular classes and determines the prognosis of alveolar rhabdomyosarcomas
Cancer Research 66:6936–6946.

https://doi.org/10.1158/0008-5472.CAN-05-4578
- PubMed
- Google Scholar
1. Ferrando AA
2. Neuberg DS
3. Staunton J
4. Loh ML
5. Huard C
6. Raimondi SC
7. Behm FG
8. Pui CH
9. Downing JR
10. Gilliland DG
11. Lander ES
12. Golub TR
13. Look AT
(2002) Gene expression signatures define novel oncogenic pathways in T cell acute lymphoblastic leukemia
Cancer Cell 1:75–87.

https://doi.org/10.1016/S1535-6108(02)00018-1
- PubMed
- Google Scholar
(1996) Amplification of MDM2 inhibits MyoD-mediated myogenesis
Molecular and Cellular Biology 16:5048–5057.

https://doi.org/10.1128/MCB.16.9.5048
- Google Scholar
1. Gerby B
2. Tremblay CS
3. Tremblay M
4. Rojas-Sutterlin S
5. Herblot S
6. Hébert J
7. Sauvageau G
8. Lemieux S
9. Lécuyer E
10. Veiga DF
11. Hoang T
(2014) SCL, LMO1 and Notch1 reprogram thymocytes into self-renewing cells
PLoS Genetics 10:e1004768.

https://doi.org/10.1371/journal.pgen.1004768
- PubMed
- Google Scholar
1. Gurevich DB
2. Nguyen PD
3. Siegel AL
4. Ehrlich OV
5. Sonntag C
6. Phan JM
7. Berger S
8. Ratnayake D
9. Hersey L
10. Berger J
11. Verkade H
12. Hall TE
13. Currie PD
(2016) Asymmetric division of clonal muscle stem cells coordinates muscle regeneration in vivo
Science 353:aad9969.

https://doi.org/10.1126/science.aad9969
- PubMed
- Google Scholar
1. Haldar M
2. Karan G
3. Tvrdik P
4. Capecchi MR
(2008) Two cell lineages, myf5 and myf5-independent, participate in mouse skeletal myogenesis
Developmental Cell 14:437–445.

https://doi.org/10.1016/j.devcel.2008.01.002
- PubMed
- Google Scholar
1. Haldar M
2. Karan G
3. Watanabe S
4. Guenther S
5. Braun T
6. Capecchi MR
(2014) Response: Contributions of the Myf5-independent lineage to myogenesis
Developmental Cell 31:539–541.

https://doi.org/10.1016/j.devcel.2014.11.010
- PubMed
- Google Scholar
(2015) Distinct malignant behaviors of mouse myogenic tumors induced by different oncogenetic lesions
Frontiers in Oncology 5:1–5.

https://doi.org/10.3389/fonc.2015.00050
- PubMed
- Google Scholar
1. Hettmer S
2. Liu J
3. Miller CM
4. Lindsay MC
5. Sparks CA
6. Guertin DA
7. Bronson RT
8. Langenau DM
9. Wagers AJ
(2011) Sarcomas induced in discrete subsets of prospectively isolated skeletal muscle cells
PNAS 108:20002–20007.

https://doi.org/10.1073/pnas.1111733108
- PubMed
- Google Scholar
(2009) Differential requirements for myogenic regulatory factors distinguish medial and lateral somitic, cranial and fin muscle fibre populations
Development 136:403–414.

https://doi.org/10.1242/dev.028019
- PubMed
- Google Scholar
1. Hinson AR
2. Jones R
3. Crose LE
4. Belyea BC
5. Barr FG
6. Linardic CM
(2013) Human Rhabdomyosarcoma cell lines for Rhabdomyosarcoma research: utility and pitfalls
Frontiers in Oncology 3:183.

https://doi.org/10.3389/fonc.2013.00183
- PubMed
- Google Scholar
1. Ignatius MS
2. Chen E
3. Elpek NM
4. Fuller AZ
5. Tenente IM
6. Clagg R
7. Liu S
8. Blackburn JS
9. Linardic CM
10. Rosenberg AE
11. Nielsen PG
12. Mempel TR
13. Langenau DM
(2012) In vivo imaging of tumor-propagating cells, regional tumor heterogeneity, and dynamic cell movements in embryonal rhabdomyosarcoma
Cancer Cell 21:680–693.

https://doi.org/10.1016/j.ccr.2012.03.043
- PubMed
- Google Scholar
1. Imayoshi I
2. Kageyama R
(2014) bHLH factors in self-renewal, multipotency, and fate choice of neural progenitor cells
Neuron 82:9–23.

https://doi.org/10.1016/j.neuron.2014.03.018
- PubMed
- Google Scholar
1. Jessen JR
2. Jessen TN
3. Vogel SS
4. Lin S
(2001) Concurrent expression of recombination activating genes 1 and 2 in zebrafish olfactory sensory neurons
Genesis 29:156–162.

https://doi.org/10.1002/gene.1019
- PubMed
- Google Scholar
1. Ju B
2. Chong SW
3. He J
4. Wang X
5. Xu Y
6. Wan H
7. Tong Y
8. Yan T
9. Korzh V
10. Gong Z
(2003) Recapitulation of fast skeletal muscle development in zebrafish by transgenic expression of GFP under the mylz2 promoter
Developmental Dynamics 227:14–26.

https://doi.org/10.1002/dvdy.10273
- PubMed
- Google Scholar
(2015) Probing for a deeper understanding of rhabdomyosarcoma: insights from complementary model systems
Nature Reviews Cancer 15:426–439.

https://doi.org/10.1038/nrc3961
- PubMed
- Google Scholar
1. Keller C
2. Guttridge DC
(2013) Mechanisms of impaired differentiation in rhabdomyosarcoma
FEBS Journal 280:4323–4334.

https://doi.org/10.1111/febs.12421
- PubMed
- Google Scholar
(2011) Developmental origins of fusion-negative rhabdomyosarcomas
Current Topics in Developmental Biology 96:33–56.

https://doi.org/10.1016/B978-0-12-385940-2.00002-4
- PubMed
- Google Scholar
1. Kumar S
2. Perlman E
3. Harris CA
4. Raffeld M
5. Tsokos M
(2000) Myogenin is a specific marker for Rhabdomyosarcoma: an immunohistochemical study in paraffin-embedded tissues
Modern Pathology 13:988–993.

https://doi.org/10.1038/modpathol.3880179
- PubMed
- Google Scholar
1. Langenau DM
2. Keefe MD
3. Storer NY
4. Guyon JR
5. Kutok JL
6. Le X
7. Goessling W
8. Neuberg DS
9. Kunkel LM
10. Zon LI
(2007) Effects of RAS on the genesis of embryonal rhabdomyosarcoma
Genes and Development 21:1382–1395.

https://doi.org/10.1101/gad.1545007
- PubMed
- Google Scholar
1. Langenau DM
2. Keefe MD
3. Storer NY
4. Jette CA
5. Smith AC
6. Ceol CJ
7. Bourque C
8. Look AT
9. Zon LI
(2008) Co-injection strategies to modify radiation sensitivity and tumor initiation in transgenic zebrafish
Oncogene 27:4242–4248.

https://doi.org/10.1038/onc.2008.56
- PubMed
- Google Scholar
1. Langenau DM
2. Traver D
3. Ferrando AA
4. Kutok JL
5. Aster JC
6. Kanki JP
7. Lin S
8. Prochownik E
9. Trede NS
10. Zon LI
11. Look AT
(2003) Myc-induced T cell leukemia in transgenic zebrafish
Science 299:887–890.

https://doi.org/10.1126/science.1080280
- PubMed
- Google Scholar
1. Le X
2. Langenau DM
3. Keefe MD
4. Kutok JL
5. Neuberg DS
6. Zon LI
(2007) Heat shock-inducible cre/Lox approaches to induce diverse types of tumors and hyperplasia in transgenic zebrafish
PNAS 104:9410–9415.

https://doi.org/10.1073/pnas.0611302104
- PubMed
- Google Scholar
1. Le X
2. Pugach EK
3. Hettmer S
4. Storer NY
5. Liu J
6. Wills AA
7. DiBiase A
8. Chen EY
9. Ignatius MS
10. Poss KD
11. Wagers AJ
12. Langenau DM
13. Zon LI
(2013) A novel chemical screening strategy in zebrafish identifies common pathways in embryogenesis and rhabdomyosarcoma development
Development 140:2354–2364.

https://doi.org/10.1242/dev.088427
- PubMed
- Google Scholar
1. Ligon KL
2. Huillard E
3. Mehta S
4. Kesari S
5. Liu H
6. Alberta JA
7. Bachoo RM
8. Kane M
9. Louis DN
10. Depinho RA
11. Anderson DJ
12. Stiles CD
13. Rowitch DH
(2007) Olig2-regulated lineage-restricted pathway controls replication competence in neural stem cells and malignant glioma
Neuron 53:503–517.

https://doi.org/10.1016/j.neuron.2007.01.009
- PubMed
- Google Scholar
1. MacQuarrie KL
2. Yao Z
3. Fong AP
4. Diede SJ
5. Rudzinski ER
6. Hawkins DS
7. Tapscott SJ
(2013a) Comparison of genome-wide binding of MyoD in normal human myogenic cells and rhabdomyosarcomas identifies regional and local suppression of promyogenic transcription factors
Molecular and Cellular Biology 33:773–784.

https://doi.org/10.1128/MCB.00916-12
- PubMed
- Google Scholar
(2013b) Genome-wide binding of the basic helix-loop-helix myogenic inhibitor musculin has substantial overlap with MyoD: implications for buffering activity
Skeletal Muscle 3:26.

https://doi.org/10.1186/2044-5040-3-26
- PubMed
- Google Scholar
1. Macquarrie KL
2. Yao Z
3. Young JM
4. Cao Y
5. Tapscott SJ
(2012) miR-206 integrates multiple components of differentiation pathways to control the transition from growth to differentiation in rhabdomyosarcoma cells
Skeletal Muscle 2:7.

https://doi.org/10.1186/2044-5040-2-7
- PubMed
- Google Scholar
1. McLean CY
2. Bristor D
3. Hiller M
4. Clarke SL
5. Schaar BT
6. Lowe CB
7. Wenger AM
8. Bejerano G
(2010) GREAT improves functional interpretation of cis-regulatory regions
Nature Biotechnology 28:495–501.

https://doi.org/10.1038/nbt.1630
- PubMed
- Google Scholar
1. Mizgireuv IV
2. Revskoy SY
(2006) Transplantable tumor lines generated in clonal zebrafish
Cancer Research 66:3120–3125.

https://doi.org/10.1158/0008-5472.CAN-05-3800
- PubMed
- Google Scholar
1. Parham DM
(2001) Pathologic classification of rhabdomyosarcomas and correlations with molecular studies
Modern Pathology 14:506–514.

https://doi.org/10.1038/modpathol.3880339
- PubMed
- Google Scholar
1. Park SW
2. Davison JM
3. Rhee J
4. Hruban RH
5. Maitra A
6. Leach SD
(2008) Oncogenic KRAS induces progenitor cell expansion and malignant transformation in zebrafish exocrine pancreas
Gastroenterology 134:2080–2090.

https://doi.org/10.1053/j.gastro.2008.02.084
- PubMed
- Google Scholar
1. Patton EE
2. Widlund HR
3. Kutok JL
4. Kopani KR
5. Amatruda JF
6. Murphey RD
7. Berghmans S
8. Mayhall EA
9. Traver D
10. Fletcher CD
11. Aster JC
12. Granter SR
13. Look AT
14. Lee C
15. Fisher DE
16. Zon LI
(2005) BRAF mutations are sufficient to promote nevi formation and cooperate with p53 in the genesis of melanoma
Current Biology 15:249–254.

https://doi.org/10.1016/j.cub.2005.01.031
- PubMed
- Google Scholar
(2001) Stem cells, cancer, and cancer stem cells
Nature 414:105–111.

https://doi.org/10.1038/35102167
- PubMed
- Google Scholar
1. Roe JS
2. Mercan F
3. Rivera K
4. Pappin DJ
5. Vakoc CR
(2015) BET bromodomain inhibition suppresses the function of hematopoietic transcription factors in acute myeloid leukemia
Molecular cell 58:1028–1039.

https://doi.org/10.1016/j.molcel.2015.04.011
- PubMed
- Google Scholar
1. Rubin BP
2. Nishijo K
3. Chen HI
4. Yi X
5. Schuetze DP
6. Pal R
7. Prajapati SI
8. Abraham J
9. Arenkiel BR
10. Chen QR
11. Davis S
12. McCleish AT
13. Capecchi MR
14. Michalek JE
15. Zarzabal LA
16. Khan J
17. Yu Z
18. Parham DM
19. Barr FG
20. Meltzer PS
21. Chen Y
22. Keller C
(2011) Evidence for an unanticipated relationship between undifferentiated pleomorphic sarcoma and embryonal rhabdomyosarcoma
Cancer Cell 19:177–191.

https://doi.org/10.1016/j.ccr.2010.12.023
- PubMed
- Google Scholar
(1993) MyoD or Myf-5 is required for the formation of skeletal muscle
Cell 75:1351–1359.

https://doi.org/10.1016/0092-8674(93)90621-V
- PubMed
- Google Scholar
1. Saab R
2. Bills JL
3. Miceli AP
4. Anderson CM
5. Khoury JD
6. Fry DW
7. Navid F
8. Houghton PJ
9. Skapek SX
(2006) Pharmacologic inhibition of cyclin-dependent kinase 4/6 activity arrests proliferation in myoblasts and rhabdomyosarcoma-derived cells
Molecular Cancer Therapeutics 5:1299–1308.

https://doi.org/10.1158/1535-7163.MCT-05-0383
- PubMed
- Google Scholar
1. Sabaawy HE
2. Azuma M
3. Embree LJ
4. Tsai HJ
5. Starost MF
6. Hickstein DD
(2006) TEL-AML1 transgenic zebrafish model of precursor B cell acute lymphoblastic leukemia
PNAS 103:15166–15171.

https://doi.org/10.1073/pnas.0603349103
- PubMed
- Google Scholar
1. Satheesha S
2. Manzella G
3. Bovay A
4. Casanova EA
5. Bode PK
6. Belle R
7. Feuchtgruber S
8. Jaaks P
9. Dogan N
10. Koscielniak E
11. Schäfer BW
(2016) Targeting hedgehog signaling reduces self-renewal in embryonal rhabdomyosarcoma
Oncogene 35:2020–2030.

https://doi.org/10.1038/onc.2015.267
- PubMed
- Google Scholar
1. Sebire NJ
2. Malone M
(2003) Myogenin and MyoD1 expression in paediatric rhabdomyosarcomas
Journal of Clinical Pathology 56:412–416.

https://doi.org/10.1136/jcp.56.6.412
- PubMed
- Google Scholar
1. Shen L
2. Shao N
3. Liu X
4. Nestler E
(2014) Ngs.plot: Quick mining and visualization of next-generation sequencing data by integrating genomic databases
BMC Genomics 15:284.

https://doi.org/10.1186/1471-2164-15-284
- PubMed
- Google Scholar
1. Shern JF
2. Chen L
3. Chmielecki J
4. Wei JS
5. Patidar R
6. Rosenberg M
7. Ambrogio L
8. Auclair D
9. Wang J
10. Song YK
11. Tolman C
12. Hurd L
13. Liao H
14. Zhang S
15. Bogen D
16. Brohl AS
17. Sindiri S
18. Catchpoole D
19. Badgett T
20. Getz G
21. Mora J
22. Anderson JR
23. Skapek SX
24. Barr FG
25. Meyerson M
26. Hawkins DS
27. Khan J
(2014) Comprehensive genomic analysis of rhabdomyosarcoma reveals a landscape of alterations affecting a common genetic axis in fusion-positive and fusion-negative tumors
Cancer Discovery 4:216–231.

https://doi.org/10.1158/2159-8290.CD-13-0639
- PubMed
- Google Scholar
(2013) A myogenic precursor cell that could contribute to regeneration in zebrafish and its similarity to the satellite cell
FEBS Journal 280:4074–4088.

https://doi.org/10.1111/febs.12300
- PubMed
- Google Scholar
1. Smith AC
2. Raimondi AR
3. Salthouse CD
4. Ignatius MS
5. Blackburn JS
6. Mizgirev IV
7. Storer NY
8. de Jong JL
9. Chen AT
10. Zhou Y
11. Revskoy S
12. Zon LI
13. Langenau DM
(2010) High-throughput cell transplantation establishes that tumor-initiating cells are abundant in zebrafish T-cell acute lymphoblastic leukemia
Blood 115:3296–3303.

https://doi.org/10.1182/blood-2009-10-246488
- PubMed
- Google Scholar
(2014) Proof-of-concept rare cancers in drug development: the case for rhabdomyosarcoma
Oncogene 33:1877–1889.

https://doi.org/10.1038/onc.2013.129
- PubMed
- Google Scholar
1. Sorensen PH
2. Lynch JC
3. Qualman SJ
4. Tirabosco R
5. Lim JF
6. Maurer HM
7. Bridge JA
8. Crist WM
9. Triche TJ
10. Barr FG
(2002) PAX3-FKHR and PAX7-FKHR gene fusions are prognostic indicators in alveolar rhabdomyosarcoma: a report from the children's oncology group
Journal of Clinical Oncology 20:2672–2679.

https://doi.org/10.1200/JCO.2002.03.137
- PubMed
- Google Scholar
1. Storer NY
2. White RM
3. Uong A
4. Price E
5. Nielsen GP
6. Langenau DM
7. Zon LI
(2013) Zebrafish rhabdomyosarcoma reflects the developmental stage of oncogene expression during myogenesis
Development 140:3040–3050.

https://doi.org/10.1242/dev.087858
- PubMed
- Google Scholar
1. Suvà ML
2. Rheinbay E
3. Gillespie SM
4. Patel AP
5. Wakimoto H
6. Rabkin SD
7. Riggi N
8. Chi AS
9. Cahill DP
10. Nahed BV
11. Curry WT
12. Martuza RL
13. Rivera MN
14. Rossetti N
15. Kasif S
16. Beik S
17. Kadri S
18. Tirosh I
19. Wortman I
20. Shalek AK
21. Rozenblatt-Rosen O
22. Regev A
23. Louis DN
24. Bernstein BE
(2014) Reconstructing and reprogramming the tumor-propagating potential of glioblastoma stem-like cells
Cell 157:580–594.

https://doi.org/10.1016/j.cell.2014.02.030
- PubMed
- Google Scholar
1. Tang Q
2. Abdelfattah NS
3. Blackburn JS
4. Moore JC
5. Martinez SA
6. Moore FE
7. Lobbardi R
8. Tenente IM
9. Ignatius MS
10. Berman JN
11. Liwski RS
12. Houvras Y
13. Langenau DM
(2014) Optimized cell transplantation using adult rag2 mutant zebrafish
Nature Methods 11:821–824.

https://doi.org/10.1038/nmeth.3031
- PubMed
- Google Scholar
1. Tang Q
2. Moore JC
3. Ignatius MS
4. Tenente IM
5. Hayes MN
6. Garcia EG
7. Torres Yordán N
8. Bourque C
9. He S
10. Blackburn JS
11. Look AT
12. Houvras Y
13. Langenau DM
(2016) Imaging tumour cell heterogeneity following cell transplantation into optically clear immune-deficient zebrafish
Nature Communications 7:10358.

https://doi.org/10.1038/ncomms10358
- PubMed
- Google Scholar
1. Tapscott SJ
2. Davis RL
3. Thayer MJ
4. Cheng PF
5. Weintraub H
6. Lassar AB
(1988) MyoD1: a nuclear phosphoprotein requiring a Myc homology region to convert fibroblasts to myoblasts
Science 242:405–411.

https://doi.org/10.1126/science.3175662
- PubMed
- Google Scholar
(2014) Normal and malignant muscle cell transplantation into immune compromised adult zebrafish
Journal of Visualized Experiments e52597.

https://doi.org/10.3791/52597
- Google Scholar
1. Trapnell C
2. Roberts A
3. Goff L
4. Pertea G
5. Kim D
6. Kelley DR
7. Pimentel H
8. Salzberg SL
9. Rinn JL
10. Pachter L
(2012) Differential gene and transcript expression analysis of RNA-seq experiments with TopHat and cufflinks
Nature Protocols 7:562–578.

https://doi.org/10.1038/nprot.2012.016
- PubMed
- Google Scholar
(2015) Identification of OLIG2 as the most specific glioblastoma stem cell marker starting from comparative analysis of data from similar DNA chip microarray platforms
Tumor Biology 36:1943–1953.

https://doi.org/10.1007/s13277-014-2800-5
- Google Scholar
1. Ustanina S
2. Carvajal J
3. Rigby P
4. Braun T
(2007) The myogenic factor Myf5 supports efficient skeletal muscle regeneration by enabling transient myoblast amplification
Stem Cells 25:2006–2016.

https://doi.org/10.1634/stemcells.2006-0736
- PubMed
- Google Scholar
(2011) CD133 positive embryonal rhabdomyosarcoma stem-like cell population is enriched in rhabdospheres
PLoS One 6:e19506.

https://doi.org/10.1371/journal.pone.0019506
- PubMed
- Google Scholar
1. Webster MT
2. Fan CM
(2013) c-MET regulates myoblast motility and myocyte fusion during adult skeletal muscle regeneration
PLoS One 8:e81757.

https://doi.org/10.1371/journal.pone.0081757
- PubMed
- Google Scholar
1. Xu Y
2. He J
3. Tian HL
4. Chan CH
5. Liao J
6. Yan T
7. Lam TJ
8. Gong Z
(1999) Fast skeletal muscle-specific expression of a zebrafish myosin light chain 2 gene and characterization of its promoter by direct injection into skeletal muscle
DNA and Cell Biology 18:85–95.

https://doi.org/10.1089/104454999315655
- PubMed
- Google Scholar
(1999) The transition from proliferation to differentiation is delayed in satellite cells from mice lacking MyoD
Developmental Biology 210:440–455.

https://doi.org/10.1006/dbio.1999.9284
- PubMed
- Google Scholar
1. Yang HW
2. Kutok JL
3. Lee NH
4. Piao HY
5. Fletcher CD
6. Kanki JP
7. Look AT
(2004) Targeted expression of human MYCN selectively causes pancreatic neuroendocrine tumors in transgenic zebrafish
Cancer Research 64:7256–7262.

https://doi.org/10.1158/0008-5472.CAN-04-0931
- PubMed
- Google Scholar
1. Yang Z
2. MacQuarrie KL
3. Analau E
4. Tyler AE
5. Dilworth FJ
6. Cao Y
7. Diede SJ
8. Tapscott SJ
(2009) MyoD and E-protein heterodimers switch rhabdomyosarcoma cells from an arrested myoblast phase to a differentiated state
Genes and Development 23:694–707.

https://doi.org/10.1101/gad.1765109
- PubMed
- Google Scholar
1. Zhang Y
2. Liu T
3. Meyer CA
4. Eeckhoute J
5. Johnson DS
6. Bernstein BE
7. Nusbaum C
8. Myers RM
9. Brown M
10. Li W
11. Liu XS
(2008) Model-based analysis of ChIP-Seq (MACS)
Genome Biology 9:R137.

https://doi.org/10.1186/gb-2008-9-9-r137
- PubMed
- Google Scholar
1. Zhuravleva J
2. Paggetti J
3. Martin L
4. Hammann A
5. Solary E
6. Bastie JN
7. Delva L
(2008) MOZ/TIF2-induced acute myeloid leukaemia in transgenic fish
British Journal of Haematology 143:378–382.

https://doi.org/10.1111/j.1365-2141.2008.07362.x
- PubMed
- Google Scholar

Article and author information

Author details

Inês M Tenente
1. Molecular Pathology, Cancer Center, and Regenerative Medicine, Massachusetts General Hospital, Boston, United States
2. Harvard Stem Cell Institute, Cambridge, United States
3. GABBA Program, Abel Salazar Biomedical Sciences Institute, University of Porto, Porto, Portugal
Contribution
IMT, Conception and design, Acquisition of data, Analysis and interpretation of data, Drafting or revising the article

Competing interests
The authors declare that no competing interests exist.
Madeline N Hayes
1. Molecular Pathology, Cancer Center, and Regenerative Medicine, Massachusetts General Hospital, Boston, United States
2. Harvard Stem Cell Institute, Cambridge, United States
Contribution
MNH, Conception and design, Acquisition of data, Analysis and interpretation of data, Drafting or revising the article

Competing interests
The authors declare that no competing interests exist.
Myron S Ignatius
1. Molecular Pathology, Cancer Center, and Regenerative Medicine, Massachusetts General Hospital, Boston, United States
2. Harvard Stem Cell Institute, Cambridge, United States
3. Molecular Medicine, Greehey Children's Cancer Research Institute, San Antonio, United States
Contribution
MSI, Conception and design, Acquisition of data, Analysis and interpretation of data, Drafting or revising the article

Competing interests
The authors declare that no competing interests exist.
Karin McCarthy
1. Molecular Pathology, Cancer Center, and Regenerative Medicine, Massachusetts General Hospital, Boston, United States
2. Harvard Stem Cell Institute, Cambridge, United States
Contribution
KM, Conception and design, Acquisition of data, Analysis and interpretation of data, Drafting or revising the article

Competing interests
The authors declare that no competing interests exist.
Marielle Yohe

Oncogenomics Section, Pediatric Oncology Branch, Advanced Technology Center, National Cancer Institute, Gaithersburg, United States

Contribution
MY, Conception and design, Acquisition of data, Drafting or revising the article

Competing interests
The authors declare that no competing interests exist.
Sivasish Sindiri

Oncogenomics Section, Pediatric Oncology Branch, Advanced Technology Center, National Cancer Institute, Gaithersburg, United States

Contribution
SS, Analysis and interpretation of data, Drafting or revising the article

Competing interests
The authors declare that no competing interests exist.

"This ORCID iD identifies the author of this article:" 0000-0003-2516-969X
Berkley Gryder

Oncogenomics Section, Pediatric Oncology Branch, Advanced Technology Center, National Cancer Institute, Gaithersburg, United States

Contribution
BG, Conception and design, Analysis and interpretation of data, Drafting or revising the article

Competing interests
The authors declare that no competing interests exist.
Mariana L Oliveira
1. Molecular Pathology, Cancer Center, and Regenerative Medicine, Massachusetts General Hospital, Boston, United States
2. Harvard Stem Cell Institute, Cambridge, United States
3. Instituto de Medicina Molecular, Faculdade de Medicina, Universidade de Lisboa, Lisbon, Portugal
Contribution
MLO, Conception and design, Acquisition of data, Analysis and interpretation of data, Drafting or revising the article

Competing interests
The authors declare that no competing interests exist.
Ashwin Ramakrishnan
1. Molecular Pathology, Cancer Center, and Regenerative Medicine, Massachusetts General Hospital, Boston, United States
2. Harvard Stem Cell Institute, Cambridge, United States
Contribution
AR, Acquisition of data, Analysis and interpretation of data, Drafting or revising the article

Competing interests
The authors declare that no competing interests exist.
Qin Tang
1. Molecular Pathology, Cancer Center, and Regenerative Medicine, Massachusetts General Hospital, Boston, United States
2. Harvard Stem Cell Institute, Cambridge, United States
Contribution
QT, Acquisition of data, Analysis and interpretation of data, Drafting or revising the article

Competing interests
The authors declare that no competing interests exist.

"This ORCID iD identifies the author of this article:" 0000-0002-9487-570X
Eleanor Y Chen

Department of Pathology, University of Washington, Seattle, United States

Contribution
EYC, Conception and design, Acquisition of data, Drafting or revising the article

Competing interests
The authors declare that no competing interests exist.
G Petur Nielsen

Department of Pathology, Massachusetts General Hospital, Boston, United States

Contribution
GPN, Acquisition of data, Analysis and interpretation of data, Drafting or revising the article

Competing interests
The authors declare that no competing interests exist.
Javed Khan

Oncogenomics Section, Pediatric Oncology Branch, Advanced Technology Center, National Cancer Institute, Gaithersburg, United States

Contribution
JK, Conception and design, Acquisition of data, Analysis and interpretation of data, Drafting or revising the article

Competing interests
The authors declare that no competing interests exist.
David M Langenau
1. Molecular Pathology, Cancer Center, and Regenerative Medicine, Massachusetts General Hospital, Boston, United States
2. Harvard Stem Cell Institute, Cambridge, United States
Contribution
DML, Conception and design, Analysis and interpretation of data, Drafting or revising the article

For correspondence
dlangenau@mgh.harvard.edu

Competing interests
The authors declare that no competing interests exist.

"This ORCID iD identifies the author of this article:" 0000-0001-6664-8318

Funding

National Institutes of Health

Inês M Tenente

Fundação para a Ciência e a Tecnologia

Myron S Ignatius
David M Langenau

China Scholarship Council

Qin Tang

Alex's Lemonade Stand Foundation for Childhood Cancer

David M Langenau

Saint Baldrick's Foundation

David M Langenau

National Institutes of Health (R01CA154923)

David M Langenau

National Institutes of Health (U54CA168512)

David M Langenau

NIH Office of the Director (R24OD016761)

David M Langenau

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

Acknowledgements

This work was funded by NIH grants R01CA154923 (DML), R24OD016761 (DML), U54CA168512 (DML), and a St. Baldricks Research Grant (DML). We thank the Specialized Histopathology Services at Massachusetts General Hospital (MGH) and the Dana-Farber/Harvard Cancer Center (P30 CA06516), the MGH Cancer Center/Molecular Pathology Confocal Core and the MGH CNY Flow Cytometry Core and Flow Image Analysis (1S10RR023440-01A1). We thank Matthew Alexander, PhD for helpful suggestions and comments.

Ethics

Animal experimentation: Studies were approved by the Massachusetts General Hospital Subcommittee on Research Animal Care under the protocol #2011N000127 (zebrafish) and #2013N000038 (mouse). Biosafety lentiviral work was approved by the Partners IBC under protocol #2013B000039.

Copyright

This is an open-access article, free of all copyright, and may be freely reproduced, distributed, transmitted, modified, built upon, or otherwise used by anyone for any lawful purpose. The work is made available under the Creative Commons CC0 public domain dedication.