SEARCH

SEARCH BY CITATION

Keywords:

  • Rhabdomyosarcoma;
  • Fusion;
  • IL-4;
  • Satellite cells

Abstract

  1. Top of page
  2. A
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgments
  9. Disclosure of Potential Conflicts of Interest
  10. References
  11. Supporting Information

Tumor cells of the muscle-related cancer alveolar rhabdomyosarcoma (aRMS) have dysregulated terminal myogenic differentiation that is characterized by continuous proliferation, decreased capacity to express markers of terminal differentiation, and inability of tumor cells to fuse to one another in the manner seen for normal myoblasts. Whether aRMS tumor cells can fuse with normal myogenic progenitors such as skeletal muscle stem cells (satellite cells) or myoblasts is unknown, as is the biological effect of fusion events if the phenomenon occurs. To study this possibility, we isolated primary satellite cells harboring a lacZ Cre-LoxP reporter gene for coculture with murine aRMS primary tumor cells expressing Cre. Results of in vitro and in vivo experiments demonstrated tumor cell—muscle cell progenitor fusion events as well as accelerated rates of tumor establishment and progression when satellite cells and derived muscle progenitors were coinjected with tumor cells in an orthotopic allograft model. Interleukin 4 receptor (IL-4R) blocking antibody treatment reversed fusion events in vitro and blocked tumor initiation and progression in vivo. Taken together, this study supports a potential role of tumor cell—host cell fusion and the strong therapeutic potential of IL-4R blockade to prevent the establishment of RMS tumors at new anatomical sites. Stem Cells 2013;31:2304–2312


Introduction

  1. Top of page
  2. A
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgments
  9. Disclosure of Potential Conflicts of Interest
  10. References
  11. Supporting Information

Rhabdomyosarcoma (RMS) is a malignancy presumed to derive from the skeletal muscle lineage but having a phenotype consistent with only the earliest stages of myogenic differentiation [1]. RMS is divided into two major histological subtypes, embryonal and alveolar. Whereas embryonal RMS (eRMS) recapitulates the phenotypical and biological features of embryonic muscle, alveolar RMS (aRMS) is a more loosely organized tumor displaying poor muscle differentiation [2]. We have reported that aRMS primary cell cultures from a well-characterized transgenic mouse model lack fusion potential between neighboring tumor cells [3]. Whether aRMS tumor cells can fuse with normal myogenic progenitors such as skeletal muscle stem cells (satellite cells [SCs]) or myoblasts is unknown, but elegant studies of normal myogenic progenitors suggest that myoblast fusion to myotubes is an Interleukin 4 (IL-4) and IL-4 receptor (IL-4R)-dependent process [4]. Furthermore, recent studies in other cancers suggest that cell fusion between tumor cell and normal host cells can lead to tumor initiation [5], that fusion can be an origin of both cancer stem cells and recurrent cancer stem cells [6], that fusion can make tumor cell acquire drug resistance [5], and that fusion can facilitate metastasis [7, 8]. An intriguing aspect of RMS is the heterogeneity of cells at a histological level, which includes the frequent presence of rhabdomyoblasts that not only exhibit features of myodifferentiation (Table 1, Fig. 1A, 1B, 1D, 1E) but can be proliferative whether mononuclear or multinucleated (Fig. 1C, 1F).

Table 1. Features of rhabdomyoblasts
• Giant nuclei or giant cells (many are multinucleated)
• Prominent nucleoli
• Rosette or wreath-shaped arrangement of nuclei
• Highly differentiated and containing cross-striations
• Pleomorphic
• Abundant and intensely eosinophilic cytoplasm
• High mitotic rate
image

Figure 1. Rhabdomyoblasts in human and mouse alveolar rhabdomyosarcoma. (A, B): Human multinucleated and mononuclear rhabdomyoblasts are indicated with black arrowheads. (C): Human multinucleated rhabdomyoblasts positive for the cell proliferation marker, ki67, by immunohistochemistry. (D, E): Murine multinucleated and mononuclear rhabdomyoblasts from the genetically engineered mouse model [9-11] are indicated with black arrowheads. (F): Murine multinucleated rhabdomyoblasts positive for the cell proliferation marker, ki67. Scale bars = 50 μm.

Download figure to PowerPoint

We recently reported that RMS tumor cells express IL-4R [12] and that IL-4R blockade prevents establishment of metastatic tumors at lymphatic and pulmonary sites. Given that normal myogenic progenitors use IL-4R as a mechanism for cell-cell fusion [4], but that aRMS do not fuse to themselves, we hypothesized that a subpopulation of aRMS tumor cells may have the capacity to fuse with SCs and the fusion events change the biological behavior of tumor progression in vivo. In this study, we investigate this potential phenomena and the role of IL-4R.

Materials and Methods

  1. Top of page
  2. A
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgments
  9. Disclosure of Potential Conflicts of Interest
  10. References
  11. Supporting Information

Mice

All animal procedures were conducted in accordance with the guidelines for the care and use of laboratory animals and were approved by the institutional animal care and use committee at the Oregon Health & Science University (OHSU). We previously described a Myf6Cre,Pax3:Foxo1a,p53 conditional mouse model of aRMS that faithfully recapitulates the human disease [9-11]. Mouse primary cell cultures (U48484, U21459) were established from mouse aRMS tumors as previously described [13] and used at low passage (<p7). The tumor cells harbor a luciferase reporter allele [14]. Primary aRMS tumor cells harbor an active cre recombinase expression. Skeletal muscle SCs were isolated from 4 to 6-week-old Rosa26tm1Sor Cre/loxP lacZ reporter mice [15]. Orthotopic allograft studies used SCID Hairless Outbred (SHO) mice purchased from Charles River Laboratory (Crl:SHO-PrkdcscidHrhr) (Wilmington, MA, www.criver.com).

Cell Lines

The C2C12 mouse myoblast cell line was purchased from ATCC (Manassas, VA, www.atcc.org) and maintained in the same culture conditions as primary tumor cell cultures.

SCs Isolation

Primary SCs were isolated from Rosa26tm1Sor Cre/loxP lacZ reporter mice as previously described [16]. Briefly, cells were cultured in collagen coated 24-well plate (cat#354408, ThermoFisherScientific, Waltham, MA, www.thermofisher.com) with Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum (FBS) and 1% penicillin-streptomycin in 5% CO2 at 37°C. On culture day 3, the medium was changed once and cells were then used for further experiments at day 7.

Myodifferentiation Assays and Immunocytochemistry

Three separate populations of cells which included the C2C12 cell line, aRMS primary tumor cell culture, and isolated primary SCs were cultured in DMEM with 10% calf serum. When cells reached confluence in vitro, the media were changed to DMEM with 2% horse serum after which cells were incubated for three additional days. Myotubes that formed were quantified by counting myosin heavy chain (MHC)-positive cells. For immunocytochemistry, the MHC antibody (cat# ab49457, Cambridge, MA, www.abcam.com) was used at a titer of 1:200 for immunostaining of three populations of cells according to the manufactures protocol.

In Vitro Cocultures

U48484 aRMS primary culture cells and freshly isolated SCs from Rosa26tm1Sor Cre/loxP reporter mouse cells were cultured in vitro for 7 days in 10% FBS DMEM then trypsinized and counted. For coculture studies, 5 × 103 SCs and 50 aRMS cells were seeded per well in a 96-well plate (ratio 100:1). Culture medium was changed every 2–3 days. X-gal immunofluorescence and chemical enzymatic staining were used at day 10 to identify the β-galactosidase (β-gal) positive (fused) cells following a previously described procedure [17].

Orthotopic Allograft Studies for Fusion Events

For in vivo experiment, 1 × 105 SCs from Rosa26tm1Sor Cre/loxP lacZ reporter mice and 1 × 104 aRMS cells were coinjected into gastrocnemius (GN) muscle of SHO mice. Tumors were harvested 10 days after injection. X-gal immunofluorescence and chemical enzymatic staining were used to identify the lacZ-positive fusion cells in vivo by methods previously described [17].

Detection of LacZ-Positive Cells by Fluorescence-Activated Cell Sorting

In vitro individual cell cultures or 100:1 cocultures of Rosa26tm1Sor Cre/loxP lacZ reporter mouse-derived SCs and aRMS primary tumor cells were cultured described above were trypsinized at day 10 and analyzed for lacZ gene product (β-gal) activity by fluorescence-activated cell sorting (FACS) using the substrate fluorescein di-β-d-galactopyranoside (FDG) (cat#F1179, Invitrogen/life Technologies, Grand Island, NY, www.invitrogen.com) according to the manufacture protocol. Briefly, 3 × 105 SCs, U48484 aRMS cells, or mixed cells were collected and each population was divided into two 250 μl aliquots. Each aliquot was treated either with an equal volume of FDG or vehicle (sterile water). Cell suspensions were incubated in dark for 1-minute at 37°C after which time 1 ml 10% FBS DMEM was added to the cell suspension to wash the cells. Cells were centrifuged at 800 rpm for 3 minutes and propidium iodine was added to a final concentration of 5 μg/ml. Cells were kept on ice until FACS sorting. At least 100,000 cells were counted per sample, each in triplicate.

Orthotopic Tumor Engraftment Studies

A mixture of 30 × 103 SCs and 1 × 103 U48484 aRMS cells (ratio 30:1), 30 × 103 SCs alone, or 1 × 103 U48484 cells alone were injected into right GN muscle of SHO mice (n = 5 per group). Tumor growth was serially monitored in live mice by bioluminescent detection of luciferase activity of aRMS cells at days 14, 21, 28, and 35 as previously described [14]. Tumor sizes were measured at days 28 and 35 and then tissues were harvested at necropsy on day 35 and frozen for further analysis.

Optical Imaging for Luminescence

In vivo bioluminescence imaging of live mice was performed using Xenogen IVIS-Illumina system (PerkinElmer; Alameda, CA, www.perkinelmer.com). Animals were maintained under inhaled anesthesia using 2% isoflurane in 100% oxygen at the rate of 2.5 l/minute. For imaging of the firefly luciferase reporter harbored by tumor cells, intraperitoneal injection with a single luciferin dose 150 mg/kg b.wt. (PerkinElmer) was administered 15–20 minutes prior to imaging. The image acquisition parameters were 1-minute exposure time, binning 8, 12.5 cm field of view, and f/stop of 1. Data were acquired and analyzed using the manufacturer's proprietary Living Image software.

Proliferation Assay

Populations of SCs (n = 3,000), U48484 aRMS tumor cells (n = 30), or cocultured cells (n = 3,000 SCs and n = 30 tumor cells; ratio 100:1) were seeded into the wells of a 96-well plate in triplicate. Cells were cultured for 9 days and their proliferation potential was measured using CellTiter 96 non-radioactive cell proliferation assay (cat#G4000, Promega, Madision, WI, www.promega.com).

Immunoblotting

Cell lysates were prepared using cell lysis buffer (cat#9803, Cell Signaling Technology, Danvers, MA, www.cellsignal.com) supplemented with protease and phosphatase inhibitors (Sigma-Aldrich, St. Louis, MI, www.sigmaaldrich.com). Immunoblotting was performed then visualized with chemiluminescent signal using Super Signal Western Pico or Dura Substrate (Pierce Biotechnology, Rockford, IL, www.piercenet.com). Antibodies used for immunoblotting were IL-4Rα (cat#sc-686, Santa Cruz Biotechnology, Dallas, TX, www.scbt.com), IL-13Rα1 (cat#sc-25849, Santa Cruz Biotechnology), and IL-2Rγ (cat#sc-668, Santa Cruz Biotechnology). All antibodies were used at dilution (1:1,000). ß-Actin (1:10,000, Sigma-Aldrich) was used as the control.

IL-4R Blocking Antibody Treatment In Vitro

IL-4R neutralizing antibody (cat#552288, BD Pharmingen, San Jose, CA, www.bdbiosciences.com) was added to SCs and U48484 aRMS tumor cell cocultures at either 1 μg/ml or 5 μg/ml. X-gal staining in situ was used at day 7 to identify the lacZ-positive cell in the cultured wells. The number of these lacZ-positive cells was counted by microscopy and compared with vehicle only group.

IL-4R Blocking Antibody Treatment In Vivo

Twenty-five previously uninjured SHO mice were divided into five cohorts: phosphate-buffered saline (PBS) injection only, SCs injection only (30 × 103 cells per mouse), U48484 aRMS cell injection only (1 × 103 cells per mouse), mixed cells injections (30 × 103 SCs plus 1 × 103 U48484 cells per mouse) treated with isotype IgG2a control antibody (cat#554687, BD Pharmingen), or mixed cells injection (30 × 103 SCs plus 1 × 103 U48484 cells per mouse) treated with IL-4R blocking Ab (cat#552288, BD Pharmingen). Antibody doses were 125 μl of 1 μg/μl IL-4R Ab or IL-4R Ab isotype IgG2a per 45-g mouse administered intraperitoneally twice weekly for 5 weeks. The tumor luciferase activity in vivo was monitored at days 14, 21, 28, and 35 as described above. Tumor sizes were measured by calipers at days 28, 35, 42, and 49. All mice were sacrificed and tissue samples were harvested at day 49 and then frozen for subsequent studies.

Immunohistochemical Staining

Hematoxylin and eosin or lacZ/eosin staining was used to identify the histological structure of tumors according to established protocols [17]. Luciferase antibody (Abcam, ab21176) and ß-galactosidase antibody (FITC) (Abcam, ab6641) were used for immunofluorescences at a titer of 1:200. CD163 (Vector lab, VP-c374) were used for immunofluorescences at a titer of 1:100. Ki67 staining was performed at the OHSU Pathology Core Resource Facility.

Statistical Analysis

Statistical significance between groups was determined using Student's unpaired t test and p < .05 was considered significant. All numbers provided are mean ± SD.

Results

  1. Top of page
  2. A
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgments
  9. Disclosure of Potential Conflicts of Interest
  10. References
  11. Supporting Information

Tumor Cells Show Lack of Fusion Capacity

When compared with C2C12 myoblasts and SCs, U48484 aRMS primary tumor cell cultures showed near-complete lack of fusion or myodifferentiation potential with less than 1% of myotubes able to express MHC or undergo tumor cell-tumor cell fusion under low serum differentiation conditions (Fig. 2A, 2B). This result is consistent with our previous reports [3].

Cell Fusion Events Occur Between Tumor Cells and SCs In Vitro and In Vivo

A schematic of the experimental design for in vitro and in vivo studies are shown in Supporting Information Fig. S1A. Primary tumor cell cultures expressing Cre under the control of the Myf6 promoter were admixed with freshly isolated primary SCs from LacZ conditional reporter mice whose LacZ allele is silent unless a fusion event facilitates Cre-mediated recombination of a Rosa26-Lox-Stop-lox-LacZ allele into an active Rosa26-Lox-LacZ allele that creates the LacZ gene product, β-gal. In cocultures, most β-gal+ cells were multinucleated at day 7 but some cells also appeared mononuclear by day 10 (Fig. 2C, 2D, respectively). From a 10-day coculture of 5 × 103 SCs and 50 aRMS cells (ratio 100:1) seeded in replicates in vitro, 174.7 ± 28.5 (1.2%) β-gal-positive cells per coculture were observed by either immunofluorescence (Fig. 2E, subpanels i–iv). Refinement of x-gal staining (Fig. 2E, subpanel v) led to the observation that in vitro β-gal positivity could be found in a punctate pattern. For in vivo experiments, 1 × 105 SCs and 1 × 104 aRMS cells were coinjected orthotopically into the GN muscle of SHO mice. In our experience, primary aRMS cell cultures can create orthotopic allograft tumors, but only in the setting of preconditioning the implantation site by cardiotoxin-mediated muscle injury (Supporting Information Fig. S1B). However, in the nonpreconditioned mice coinjected with SCs and aRMS cells, tumors developed and were harvested 10 days later, revealing β-gal-positive cells in harvested tumor tissue (Fig. 2E, subpanels vi, vii). β-gal was found to also exhibit a punctate pattern of staining, similar to that which was observed in vitro (Fig. 2E, subpanels v).

image

Figure 2. Fusion events between aRMS tumor cells and SCs occur in vitro and in vivo. (A): Comparative myodifferentiation and homogenous cell-cell fusion potentials displayed by C2C12 myoblasts, U48484 aRMS cells, and SCs as demonstrated by immunocytochemistry for MHC and DAPI for multinucleation, respectively. (B): Both C2C12 cells and SC demonstrated significantly higher fusion potential than U48484 cells (**, p < .01). Fusion index represents the frequency of multinucleated cells. (C, D): Cocultured myogenic progenitors with aRMS primary tumor cells. At day 7, most β-gal+ cells were multinucleated (C) By day 10, a subset of β-gal+ cells were mononuclear (D). (E): In vitro immunostaining of cocultured cells showed dual expression of β-gal (panel E, subpanel ii, green) and MHC (panel E, subpanel iii, red), although cells were mononuclear. The cell with DAPI staining (panel E, subpanel i) and bright-field view colocalized with immunofluorescence is shown in panel E, subpanel iv (panel E, subpanel v). Depending upon the incubation time of β-gal substrate, the β-gal-positive cells suggestive of fusion in vitro exhibited extranuclear blue x-gal staining in a punctuate pattern. Cell margin and boundaries of two nuclei are shown with dotted lines. (Panel E, subpanel vi) A fused cell in vivo is identified with x-gal staining (also a punctuate pattern) and eosin counterstaining of tumor tissue generated by injection of mixed aRMS cells and SCs. The comparatively strong eosin counterstaining of the β-gal-positive cell is consistent with enriched myofibrillary protein expression. The same cell can be also observed by hematoxylin and eosin staining of an adjacent histological section (panel E, subpanel vii). At this level of the cross-section, only a single central nucleus (n) is observed. Scale bars = 50 μm. Abbreviations: aRMS, alveolar rhabdomyosarcoma; MHC, myosin heavy chain; SC, satellite cell.

Download figure to PowerPoint

Flow Cytometry Analysis of β-Gal-Positive Cells Reveals a Moderate Frequency of Cell Fusion

When U48484 aRMS primary tumor cells and Rosa26-Lox-Stop-lox-LacZ mouse primary SCs were cultured separately or cocultured in vitro for 10 days, β-gal-positive cells that had undergone fusion were quantified by flow cytometry using a fluorescence x-gal substrate, FDG. Using our gating parameters, background fluorescence cell frequency was 0.401% and 0.186% of cells for pure primary SCs or pure aRMS tumor cells, but 2.44% for cocultured cells (Fig. 3A), suggesting at least a ∼2% frequency of tumor cell fusion to a primary SC or daughter myoblast, which is slightly higher than the 1.2% of cells observed by in situ methods (Fig. 2E) and may be a result of the increased sensitivity of the flow cytometry based method.

image

Figure 3. In vivo mixed aRMS cells and SCs showed more malignant biological behavior than aRMS cells alone. (A): Using fluorescence-activated cell sorting-FDG technique, fusion events in heterogeneously cocultures of aRMS cells and SCs can be quantified as a result of β-gal expression when Cre-expressing aRMS cells fuse with silent LacZ harboring SCs. (B): Mice with mixed aRMS cells and SCs injections demonstrated a higher frequency of luciferase positivity (threshold 1 × 104 photons/second/cm2/steradian) than aRMS cells injection alone. (C): Representative photograph at day 35 after injection of an intramuscular mixed cell aRMS+SC injection yielding a substantially larger tumor compared to aRMS tumor cell injection alone. (D): Luciferase activity of from the same mice as panel C. Min and Max range are 1.02 × 105 to 2 × 106 photons/second/cm2/steradian. (E): Quantification of tumors volumes by caliper confirms that mixed cells injections yield larger size tumor when compared with aRMS-only cell injections (*p < .05). Abbreviations: aRMS, alveolar rhabdomyosarcoma; FDG, fluorescein di-β-d-galactopyranoside; SC, satellite cell.

Download figure to PowerPoint

Increased Tumorigenicity In Vivo Occurs for Mixed Cells in Comparison to Tumor Cells Alone

As stated above, activation of SCs is often achieved with injury [18, 19], and this same injury preceding RMS tumor cell injection is often performed to facilitate tumor engraftment [13, 20]. To determine the pathophysiological significance of normal myogenic progenitor-tumor cell interactions, orthotopic GN allografts were performed of a mixture of 30 × 103 SCs and 1 × 103 U48484 aRMS cells (ratio 30:1), 30 × 103 SCs alone, or 1 × 103 U48484 cells injected into uninjured muscle of recipient SHO mice (n = 5 per group). Tumor growth was serially monitored by bioluminescent detection of luciferase reporter activity (harbored by aRMS tumor cells, as previously described [14]) at days 14, 21, 28, and 35. Tumors developing from mixed cell injections showed significantly increased initiation, growth, and progression when compared with aRMS tumor cells alone (Fig. 3B–3E).

Normal Myogenic Precursors Do Not Alter Proliferation Potential of aRMS Tumor Cells

To assess whether progression of tumors was related to tumor establishment versus tumor cell proliferation rate, a cell growth assay was performed. The coculture of U48484 aRMS cells and mouse primary SCs showed a similar (but not increased) viable cell mass compared to the U48484 aRMS cells alone when cultured 10% FBS DMEM in vitro for 9 days (Fig. 4A). Both groups showed higher viable cell mass than SCs grown for the same period of time (p < .01).

image

Figure 4. IL-4R Ab treatment inhibits fusion events between aRMS cells and SCs both in vitro and in vivo. (A): At baseline, no significant difference in proliferation was observed between mixed cells and U48484 cells in vitro at day 9 (p < .05). Both groups showed significantly higher proliferation rate than SC group (**, p < .01). (B): U48484 aRMS cells express or highly over-express IL4Rα and IL-13Rα1. Previously described U21459 aRMS cells were used as a positive control [12]. (C): IL-4R inhibitor treatment inhibits the fusion potential of mixed cells in vitro as measured by X-gal staining in situ and detected by microscopy. The no-treatment group had a significantly frequency of β-gal positive cells when compare to IL-4R Ab (1 μg/ml, **, p < .01) and IL-4R Ab (5 μg/ml, *, p < .05) groups, respectively. For this experiment, 105 cells were used per group (i.e., the number of β-gal+ cells in the no treatment group was 0.08% by in situ staining). (D): Among cohorts for in vivo experiments, mice with mixed aRMS cells and SCs coinjections treated with the control isotype IgG showed a higher frequency of luciferase positivity (threshold 1 × 104photons/second/cm2/steradian) than mixed aRMS cells and SCs coinjections treated with IL-4R blocking antibody or aRMS cells injected alone. n = 5 mice per cohort. (Panel E, subpanels i–iii) Histological results of aRMS cell alone group showed normal hematoxylin and eosin (H&E) staining, negative CD163 immunostaining and negative neonatal MHC immunostaining. Inset represents CD163 staining of a mouse lymph node as a positive control. (Panel E, subpanels iv–vi) Representative histology at day 49 of mixed cell with IgG isotype treatment showed many multiple nucleated giant cells in H&E staining, negative CD163 immunostaining and positive neonatal MHC immunostaining. (Panel E, subpanels vii–ix) Representative histology of mixed cell coinjections treated with IL-4R blocking antibody demonstrated many single nucleated tumor cells in H&E staining, but CD163 and neonatal MHC negativity. Scale bars = 50 μm (Supporting Information Fig. S3). Abbreviations: aRMS, alveolar rhabdomyosarcoma; MHC, myosin heavy chain; SC, satellite cell.

Download figure to PowerPoint

U48484 Cells Express IL-4 Receptor and IL-13 Receptor

We have previously reported human and mouse aRMS to express IL-4 and IL-13 receptors [12], which have in turn also been demonstrated to be the receptors necessary for normal myoblast-myoblast fusion during myodifferentiation [4]. By western blotting we affirmed that our U48484 aRMS primary cell culture system expresses as high or higher levels of IL-4Rα and IL-13Rα1 (Fig. 4B) compared to U21459 aRMS cells, which we have previously shown to express these receptors [12].

IL-4R Antibody Inhibit Fusion Between SCs with Tumor Cells In Vitro

To determine whether IL-4R was the molecular mechanism responsible for fusion of tumor cells to normal myogenic progenitors, we cocultured U48484 aRMS cells and primary SCs from Rosa26-Lox-Stop-lox-LacZ mice in the presence of IL-4R neutralizing antibody. The frequency of fused, β-gal-positive cells on the seventh day of culture was reduced more than twofold in the presence of IL-4R neutralizing antibody treatment versus control at 1 or 5 μg/ml (p < .05 and p < .01, respectively). No significant difference was found between two different doses of IL-4R Ab treatment (Fig. 4C). Reciprocally, neither of the IL-4R ligands, IL-4 or IL-13, could induce multinucleation (e.g., fusion of aRMS tumor cell to themselves, nor mitotic dysjunction) as measured by FACS for ≥4N cells (Supporting Information Fig. S2).

IL-4R Blocking Antibody Abrogates the Effect of Normal Myogenic Progenitors on Tumorigenicity In Vivo

To determine whether IL-4R mediates the enhanced tumor initiation observed when primary SCs and their progeny are coinjected with aRMS primary tumor cells orthotopically, live animal studies were performed with the IL-4R neutralizing antibody administered systemically (intraperitoneally). No tumor formation by volumetric measurement was observed for mice which received PBS alone, SCs alone, or U48484 cells alone injection. Live animal bioluminescence imaging showed that coinjected U48484 aRMS cells and SCs group that were treated with IL-4R neutralizing antibody developed fewer tumors when compared with mixed U48484 aRMS cells and SCs group with the IgG isotype control antibody (Fig. 4D). Similarly, by gross examination then necropsy, four out of five mice in the mixed cells cohort-treated isotype IgG control antibody developed visible tumors at day 49 (5/5 had microscopic tumors at histology [Fig. 4E and Supporting Information Fig. S3]) while one out of five mice treated with IL-4R neutralizing antibody developed grossly noticeable tumor (2/5 had microscopic tumors at histology [Supporting Information Fig. S3]). Thus, the luciferase method of detecting tumors and histology more sensitive than gross examination, but the result of IL-4R neutralizing antibody blocking tumor formation was the same.

Eosin-Rich, MHC-Positive Multinucleated Cells Are Significantly Reduced by IL-4R Blocking Antibody for Coinjected aRMS and Normal Myogenic Progenitor Cells

A remarkable histological observation noted for tumors developing from coinjected U48484 aRMS cells and SCs treated only with the isotype control antibody was the high frequency of eosin-rich, MHC-positive multinucleated cells (Fig. 4E, subpanel iv). As mentioned in Introduction, multinucleated cells are often seen in clinical RMS histology and are described as rhabdomyoblasts [21, 22]. However, these cells were reduced or not present in the cohort of mice developing tumors from coinjected U48484 aRMS cells and SCs when treated 4R neutralizing antibody. Immunostaining for CD163 revealed these cells not to be macrophages (Fig. 4E, subpanel v). Neonatal MHC immunostaining revealed that these giant cells have a myogenic phenotype (Fig. 4E, subpanel vi).

Discussion

  1. Top of page
  2. A
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgments
  9. Disclosure of Potential Conflicts of Interest
  10. References
  11. Supporting Information

In this study, we demonstrate that aRMS primary tumor cells lack the capacity to fuse to other tumor cells, but have the ability to fuse to normal primary myogenic progenitors derived from muscle SCs in vitro and in vivo. The coinjection of SC-derived primary myogenic progenitors with aRMS tumor cells improves the rate of tumor establishment without necessarily altering proliferation rate of the tumor cell mass. The enhancement of tumor establishment by myogenic progenitors is mediated to a great extent by IL-4R, as demonstrated by IL-4R neutralizing antibody treatment in vivo. This IL-4R blockade further decreases the presence of myogenic multinuclear cells, which are reminiscent of the rhabdomyoblasts frequently seen in both the alveolar and embryonal subtypes of human RMS [21, 22].

IL-4R and SCs May Be Implicated in Tumor Establishment by One or More Mechanisms

These intriguing results nonetheless raise a number of questions as to the mechanism by which SC-derived myogenic progenitors enhance aRMS tumor establishment. Three potential models are presented in Figure 5. In the first model, IL-4R-mediated fusion creates multinucleated cells that may act as “stemloid” cells with tumor repopulating ability. A previous study recently reported that fibrosarcoma tumor initiation can be triggered by a single multinucleated cancer cell [23]. As a tumor cell-normal cell hybrid, the multinucleated cell may also have metabolic advantages such improved tolerance of reactive oxidative stress, or may have immune evasion properties. In the second model, activated SCs or myoblasts in the mode of muscle regeneration secrete IL-4 (possible along with other cytokines and growth factors) which then promotes tumor initiation independent of any cell fusion event. In the third model, SCs and/or daughter myoblasts secrete IL-4 and/or IL-13 ligands, which in turn activate IL-4R on M2 macrophages that then secrete tumor-promoting growth factors. This last model is less likely, given that CD163+ macrophages were absent from aRMS-SC tumors.

image

Figure 5. Schematic representation of possible mechanism by which satellite cells and IL-4 may mediate alveolar rhabdomyosarcoma tumor establishment and progression. *Satellite cells (SCs) may alternatively recruit tumor-associated fibroblasts or tumor-associated mesenchymal stem cells. See also Discussion.

Download figure to PowerPoint

IL-4R Is a Therapeutic Target of High Potential Value in RMS

While the question remains whether cell fusion is or is not necessary for the IL-4R-mediated tumor promotion effect, these results are consistent with our previous report that RMS tumor cells express IL-4 receptor (IL-4R) [12] and that IL-4R blockade prevents establishment of metastatic tumors at lymphatic and pulmonary sites. Ongoing studies will further explore the mechanism observed here, but the potential for clinical translation in RMS remains, given the tolerability of IL-4R blocking antibodies in Phase II trials for asthma [24].

Conclusions

  1. Top of page
  2. A
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgments
  9. Disclosure of Potential Conflicts of Interest
  10. References
  11. Supporting Information

Taken together, our data indicate that primary RMS tumor cells fuse to normal myogenic progenitors in vitro and in vivo. Furthermore, we show that normal primary SCs (activated by the harvesting process) enhance tumor establishment. Interestingly, blockade of IL-4 receptor prevents both RMS—myogenic precursor fusion and tumor establishment, and thus IL-4 receptor blockade has therapeutic potential for blocking progression of RMS.

Acknowledgments

  1. Top of page
  2. A
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgments
  9. Disclosure of Potential Conflicts of Interest
  10. References
  11. Supporting Information

G.H. was supported by a fellowship from the Scott Carter Foundation, and studies were supported in part by the Patrick M. Callahan Memorial Fund for Pediatric Cancer Research as well as 5R01CA133229. We thank Matthew Svalina for technical assistance in these studies.

Disclosure of Potential Conflicts of Interest

  1. Top of page
  2. A
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgments
  9. Disclosure of Potential Conflicts of Interest
  10. References
  11. Supporting Information

C.K. has sponsored research agreements or research joint ventures with Johnson & Johnson, Novartis, and Regeneron.

References

  1. Top of page
  2. A
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgments
  9. Disclosure of Potential Conflicts of Interest
  10. References
  11. Supporting Information

Supporting Information

  1. Top of page
  2. A
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgments
  9. Disclosure of Potential Conflicts of Interest
  10. References
  11. Supporting Information

Additional Supporting Information may be found in the online version of this article.

FilenameFormatSizeDescription
stem1491-sup-0001-suppinfo.pdf1200KSupporting Information

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.