University of Southern California Dissertations and Theses

The role of LGR5 in the pathogenesis of Ewing sarcoma: a marker of aggressive disease and a contributor to the malignant phenotype

(USC Thesis Other)

The role of LGR5 in the pathogenesis of Ewing sarcoma: a marker of aggressive disease and a contributor to the malignant phenotype

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content THE ROLE OF LGRS IN THE P ATHOGENEIS OF EWING SARCOMA: A MARKER OF AGGRESSIVE DISEASE AND A CONTRIBUTOR TO THE MALIGNANT PHENOTYPE by Christopher Scannell A Dissertation Presented to the FACULTY OF THE USC GRADUATE SCHOOL UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfillment of the Requirements for the Degree DOCTOR OF PHILOSOPHY (SYSTEMS BIOLOGY AND DISEASE) Acknowledgements I want to thank the many people who shaped my early scientific aspirations and motivated me to pursue my PhD. I owe the greatest thanks to my father and mother who encouraged my love of science from an early age. Without their love and support, I would not be the man or scientist that I am today. I am also indebted to my high school science teachers Paul Spangenberg and Sharon Masse who exposed me to the wondrous worlds of cell and molecular biology, respectively. They imparted me with a love of the life sciences that set my graduate career in motion. My decision to go into research was in large part influenced by my college mentors Drs. Billlie Swalla, Dick Kocan, Paul Hersherberger, and Claire Canon. I had the privilege of working with each of these individuals in intimate research environments where they taught me the rigors of the scientific method. Under their tutelage, I learned both the pleasure of rejecting the null hypothesis and the disappointment of having to accept it. They allowed me to succeed and fail with my experiments and gave me the confidence to pursue even my craziest of scientific notions. My graduate education would not have been possible without the backing of faculty and staff involved with the MD/PhD and Systems Biology and Disease (SBD) Programs at USC. Sandy Mosteller, Roland Rapanot, and Dr. Robert Chow in the MD/PhD Program provided administrative support and guidance to overcome the challenges of starting medical school. Dawn Burke and Drs. Alicia McDonough and Martin Kast in SBD Program helped me to transition into the PhD years and stay in touch with Program activities even after I moved to the University of Michigan (UM). I am especially indebted to members of my thesis committee: Drs. Shahab Asgharzadeh, Alan 11 Epstein, Amir Goldkorn, Martin Pera, and Clive Taylor. They were supportive of my move to Michigan and have provided excellent research mentorship during my PhD. I have spent the majority of my PhD years in a laboratory environment working with as few as one other person and as many as ten. These people have acted as mentors, collaborators, confidants, and friends. In particular, I worked with two groups that either joined the lab while I was at USC or were recruited once we arrived at UM. I would like to thank Aaron Cooper, Ynnez Gwye, Long Hung, and Darren Russell, and Drs. Jessie Hsu, Connie and Gregor von Levetzow, and John van Doorninck who were part of the lab when I joined at USC and made our corner of the tenth floor of the Smith Tower the most exciting place to work at Children's Hospital Los Angeles (CHLA). At UM, I would like to thank Natashay Bailey, Ashley Harris, Melanie Krook, Elisabeth Pedersen, Matthew Thayer, and Drs. Merlin Airik, Jack Mosher, Lauren Nicholls, and Laurie Svoboda who challenged me creatively on a daily basis and made Michigan my second home. In addition to those who provided intellectual help, I would like to thank those who contributed material support. All of the Ewing' sarcoma (ES) samples that I used during my project were either provided by CHLA or the Children's Oncology Group. The researchers affiliated with these organizations collected samples from patients with ES with the goal of better understanding the disease and finding more effective treatments. Most importantly, these samples would not have been provided without the generosity and sacrifice of patients with ES and their families. I hope my own research will in some small way benefit these people and give them hope for an eventual cure. 111 I would finally like to acknowledge the incredible mentorship of my thesis advisor, Dr. Elizabeth Lawlor. I was passionate about research when I entered her lab, but I did not yet possess the technical expertise or organizational skills to fully design and implement my own research project. Dr. Lawlor helped me cultivate these skills and set myself up for a career in research. She has always been willing to listen to my ideas, give critical feedback, teach thorough experimental design, and challenge my ability to effectively communicate my scientific findings. She has also been an inspiration in finding new avenues of stem cell and cancer research to explore. We spent many hours in front of the white board drawing out elaborate biochemical pathways or hotly debating new scientific papers and how to incorporate the findings into my project. She has been my greatest scientific mentor and I am indebted for the positive impact she has had on my life. Thank you. IV Table of Contents Acknowledgements List of Figures Common abbreviations Abstract Chapter I: Background and significance Chapter 2: Discovery of LGR5 in high risk Ewing sarcoma Introduction 11 VI Vll V111 I 6 6 Results 10 Table I: Developmental functions are highly represented among 13 genes that are differentially expressed by primary drug resistant ES Discussion 21 Chapter 3: LGR5 is a marker of tumor development in Ewing sarcoma 24 Introduction 24 Results 28 Discussion 40 Chapter 4: LGR5 enhances Wnt signaling in Ewing sarcoma and promotes 44 malignant phenotypic properties Introduction 44 Results 47 Discussion 65 Chapter 5: Biologic and clinical implications of LGR5 expression in ES 70 Chapter 6: Material and methods 7 5 Table 2: SYBR primers for qRT-PCR 77 Bibliography 85 Appendix: Peer reviewed publication in Frontiers in Pediatric Oncology (2013) 96 LGR5 is expressed by Ewing sarcoma and potentiates Wnt/~-catenin signaling v List of Figures Figure 1: In silica data analysis of primary drug resistant case of 12 Ewing sarcoma (ES) Figure 2. Leucine-rich repeat-containing G-protein coupled receptor 5 (LGR5) 15 is highly expressed in primary chemoresisant ES. Figure 3. LGR5 is expressed byES 16 Figure 4. LGR5 is associated with poor event free survival (EFS) 19 Figure 5: LGR5 is increased in aggressive disease 20 Figure 6. LGR5 is expressed by neural crest stem cells (NCSC) 29 Figure 7: LGR5 is increased under serum free conditions 31 Figure 8. LGR5 is a marker of chemoresistance 34 Figure 9: LGR5 is increased in putative cancer stem cells 38 Figure 10: LGR5 is expressed by discrete subpopulations in CHLA25 cells 39 Figure 11. No correlation exists between LGR5 expression and Wnt/B-catenin 50 activity in ES cells in standard culture Figure 12. Exogenous Wnt3a induces nuclear localization of ~-catenin and 51 is potentiated by RSP02. Figure 13. RSP02 potentiates Wnt/~-catenin signaling in LGR5-high ES cells 52 Figure 14. The role of LGR5 in Wnt/PCP signaling in ES cells 54 Figure 15. LGR5 does not promote ES cell proliferation 58 Figure 16. LGR5 does not promote chemoresistance in ES cells in vitro 60 Figure 17. LGR5 promotes anchorage-independent growth in ES cells 62 Figure 18. LGR5 promotes chemotaxis in LGR5-high ES cells 64 Figure 19. Proposed mechanism ofWnt/~-catenin signaling during initiation and 73 progression of ES tumors VI ALDH BM-MSC EFS ES esc ESC ISH LGRS MSC NCSC NC-MSC PCP PROMl QRT-PCR RSPO Common abbreviations Aldehyde dehydrogenase Bone marrow-derived mesenchymal stem cell Event free survival Ewing sarcoma Cancer stem cell Embryonic stem cell In situ hybridization Leucine-rich repeat-containing G protein-coupled receptor 5 Mesenchymal stem cell Neural crest stem cell Neural crest-derived mesenchymal stem cell Planar cell polarity Prominin 1 (encodes CD133) Quantitative real-time polymerase chain reaction R-spondin Vll Abstract Ewing sarcoma (ES) is an aggressive bone and soft tissue tumor of putative mesenchymal (MSC) and/or neural crest (NCSC) stem cell origin that predominantly occurs in children and young adults. Although most patients with localized ES can be cured with intensive therapy, the clinical course is variable and up to one third of patients relapse following initial remission. Unfortunately, little is yet known about the biologic features that distinguish low-risk from high-risk disease or the mechanisms of ES disease progression. In this study, we set out to identify potential molecular biomarkers and understand how they contribute to the pathogenesis ofES. Using whole genome micro array analysis to profile a primary drug resistant case of ES before and after induction therapy, we identified the gene LGR5 as a candidate marker of drug resistant ES cells. We then found LGR5 was variably expressed in ES tumors and cell lines and was increased in cases with poor outcome, indicating it was a potential prognostic biomarker for ES. LGRS (leucine-rich repeat-containing G-protein coupled receptor 5) is a somatic stem cell marker that functions as an oncogene in several human cancers, most notably colorectal carcinoma. Recently, LGRS has also been established as a receptor for the R-spondin (RSPO) family of ligands and RSPO-mediated activation of LGRS potentiates Wnt/~-catenin signaling, contributing to stem cell proliferation and self renewal. To investigate LGR5 expression in relation to the putative stem cell origin of ES, we measured expression in neuro-mesenchymal stem cell populations and found that neural crest-derived stem cells express LGR5 at comparable levels to ES cell lines. We also examined whether LGR5 is a marker of ES cells with aggressive features and found LGR5 was increased in cells grown under physiologic and chemotherapeutic stress V111 conditions and in putative cancer stem cell populations. To explore the signaling mechanism ofLGRS in ES, we investigated whether LGRS could potentiate Wnt/~ catenin signaling as has been established in epithelial malignancies. Although ES cells had low levels of basal Wnt/B-catenin signaling activity, these cells exhibited a dramatic increase in Wnt/B-catenin signaling when grown in the presence of Wnt3a and RSP02. We further demonstrated this increase was LGR5-dependent. LGRS has also been reported to regulate non-canonical Wnt signaling and we found a decrease in Wnt/PCP target gene expression with LGR5 knockdown. Finally, we went on to elucidate the function of LGR5 in promoting various malignant phenotypes in ES. Although we did not find a role for LGR5 in regulating proliferation or chemoresistance, we found LGR5 promotes both anchorage-independent growth and chemotaxis in ES cell lines. These data indicate that LGR5 is both a marker of cells with aggressive features and directly contributes to an aggressive clinical phenotype. These findings have potential prognostic and therapeutic applications in the treatment of ES. IX Chapter 1 Background significance Ewing sarcoma (ES) is the second most common malignant bone tumor occurring in children and adolescents. These tumors can also present in soft tissues and affect older adults. The overall five-year survival rate for patients with localized ES is 75%, but patients with metastatic or relapsed disease only have a five-year event-free survival rate of 10-20% (Balamuth and Womer, 2010). Unfortunately, there are no clinical or pathologic criteria for predicting cure or relapse in newly diagnosed patients apart from the presence of metastases. The standard of treatment for ES is a combination oflocal control with surgery and radiotherapy and systemic control with chemotherapy. The use of chemotherapy has helped to drastically increase survival rates for ES over the past five decades, but is associated with considerable future health risks in surviving patients. At present, most ES patients receive the same intensive multi-agent chemotherapy regimen for both localized and metastatic disease and there is no risk stratification based on expected patient response (van Maldegem eta!., 2012). Thus, there is a pressing need to identify markers of aggressive disease in ES to help better predict outcome and to judiciously apply treatment protocols that will provide maximum benefit with fewer complications. The understanding of the molecular and cellular origin of ES tumors is highly speculative. Although most ES tumors are genetically defined by the t(l1;22) chromosomal translocation creating the novel fusion oncogene EWS-FLJJ (Balamuth and Womer, 201 0), expression of EWS-FLI 1 by itself is not sufficient to induce malignant transformation in most primary cells (Deneen and Denny, 2001;Lessnick eta!., 2002). 1 This finding suggests that additional mutations or epigenetic changes may be required for transformation in cells expressing EWS-FLil. Several reports indicate deregulation of stem cell-associated genes and pathways may mediate such epigenetic changes and contribute to ES pathogenesis. In particular, the polycomb proteins EZH2 and EMil are highly expressed in ES and are thought to maintain an undifferentiated, stem cell-like phenotype (Douglas eta!., 2008;Richter et a!., 2009). The cell of origin for ES also remains unknown. ES tumors are characterized by a primitive neuroectodermal histology and present in diverse tissue types, suggesting a relatively undifferentiated cell of origin (Meltzer, 2007). Current studies indicate mesenchymal stem cells (MSC), neural crest stem cells (NCSC), or their early progenitors as leading candidates (Staege eta!., 2004;Tirode eta!., 2007;Riggi eta!., 2008;von Levetzow eta!., 2011). Both human MSC and NCSC can tolerate expression of EWS-FLI1 and ectopic expression initiates changes to an ES-like state (Riggi eta!., 2008;von Levetzow eta!., 2011). The proposed stem origin of ES and the importance of stem cell pathways in its pathogenesis raise the question of whether stem cell markers could have potential prognostic and therapeutic applications for ES. Leucine-rich repeat-containing G-protein coupled receptor 5 (LGRS) is a seven transmembrane spanning receptor that has recently been identified as a somatic stem cell marker and plays key functional roles in both normal development and cancer. Mouse studies have demonstrated that Lgr5 is widely expressed during embryonic development but expression in postnatal tissues is limited to discrete stem cell populations (Barker et a!., 2007). Such stem cells can be found in the small and large intestine, stomach, hair follicles, mammary gland, kidney, and liver (Barker eta!., 2007;Jaks eta!., 2008;Barker 2 eta!., 2012;Plaks eta!., 2013). The self-renewal and differentiation capabilities ofLgr5+ stem cells have been studied most extensively in the mouse intestinal crypt. Lgr5+ intestinal stem cells (IS C) proliferate rapidly and can give rise to all epithelial lineages within the crypt. In the context of cancer, Lgr5+ ISC have also been shown to play a critical role in the pathogenesis of intestinal tumors. Lineage tracing experiments have shown that Lgr5+ ISC are a cell of origin for mouse intestinal adenomas and fuel the growth of these tumors once established. Studies of human cancer cell lines have now confirmed that LGRS promotes the growth and/or survival in colorectal and basal cell carcinoma (McClanahan eta!., 2006;Tanese eta!., 2008), glioblastoma (Nakata eta!., 2013), and neuroblastoma (Balamuth eta!., 2010). These data indicate that LGRS can act as an oncogene in tumors of both epithelial and neural origin. Retrospective studies with various tumor types have also shown increased LGRS expression is associated with poor survival (Becker eta!., 2010;Wu eta!., 2012;Nakata eta!., 2013). In the case of colorectal carcinoma, LGRS is expressed by a subpopulation of cells with stem cell-like properties (i.e. cancer stem cells or CSC) (Kemper eta!., 2012;Kobayashi eta!., 2012). LGRS+ colorectal esc have increased clonogenic and tumorigenic potential compared to bulk tumor cells and lose expression upon in vitro differentiation (Kemper eta!., 2012). Gene expression profiling experiments using Lgr5+ ISC have also shown the gene signature derived from these cells predicts disease relapse in colorectal cancer patients (Merlos Suarez eta!., 2011). Thus, there is compelling evidence in both human and murine intestinal tumors that LGRS+ stem cells contribute to cancer initiation and progression and that high LGRS expression is associated with worse clinical outcome. 3 The ligand and signaling mechanism for LGRS have only recently been discovered and these new findings have shed light on how it promotes stem and tumor cell functions. LGRS was originally cloned in 1998 and sequence analysis of the encoded protein showed similarity to glycoprotein hormone receptors. For many years LGRS was considered an orphaned G-protein coupled receptor, but in 20 II three groups independently identified LGRS and closely related homologues LGR4 and LGR6 as the cognate receptors for R-spondin (RSPO) family of secreted proteins (Carmon eta!., 20ll;de Lau eta!., 20ll;Glinka eta!., 2011). RSPOs function as Wnt agonists and play pivotal roles as regulators of normal embryonic development and stem cell proliferation (de Lau eta!., 2012). These groups demonstrated that RSPO-LGRS engagement mediates both canonical Wnt/~-catenin and non-canonical Wnt signaling. When RSPO binds to LGRS the receptor associates with the Frizzled/LRP complex to increase ~-catenin activation and potently enhance downstream TCF reporter activity. The potentiation of Wnt/~-catenin signaling in the presence of RSPO is now believed to mediate the self renewal and proliferation of LGRS+ stem cells, both normal and malignant. LGRS+ intestinal and liver stem cells can be isolated and maintained in long-term culture with the addition ofRSPOI (Sato eta!., 2009;de Lau eta!., 2012). Similarly, the self-renewal capacity of LGRS+ colorectal CSC are thought to be maintained by autocrine or paracrine production ofRSPO (Kemper eta!., 2012). In addition to regulating Wnt/~-catenin signaling, LGRS also mediates Wnt/planar cell polarity (PCP) signaling (Glinka eta!., 2011). This pathway controls tissue polarity and cell motility during development and aberrant activation leads to adhesion, invasion, and metastasis in cancer (Katoh, 2005). Glinka eta!. found that knocking down either LGR4 or LGRS reduced ATF2 reporter 4 activity, a readout ofWnt/PCP signaling, and reduced LGR4levels resulted in gastrulation defects (Glinka eta!., 2011). These studies provide a link between LGRS, Wnt signaling, development, and cancer and suggest a mechanism by which LGRS can promote malignant stem cell-associated properties. The central hypothesis of this thesis is that LG R5 is a marker of aggressive disease and promotes a malignant phenotype in ES. We first show that LGRS is associated with high-risk disease using an in silica approach followed by expression studies to demonstrate LGRS is enriched in neural crest-derived stem cells and CSC-like cells. Finally, we use functional and mechanistic assays to demonstrate that it promotes a malignant phenotype and mediates both canonical and non-canonical Wnt signaling. Together these studies suggest LGRS plays an important role in the origin and pathogenesis of ES and highlight its potential clinical application as a prognostic marker. 5 Chapter2 Discovery of LGR5 in high risk Ewing sarcoma Introduction Ewing sarcoma (ES) is an aggressive bone and soft tissue tumor with few clinical or pathologic features available for risk stratification. The only broadly used criteria for subdividing ES patients is the presence of metastases at diagnosis or disease recurrence. Patients with localized disease have benefitted immensely from the introduction of multi agent chemotherapeutic regimens, which have increased survival rates from 10% to 75% (Balamuth and Womer, 2010). Regrettably, no such progress has been made for treating patients with metastatic or recurrent disease (van Mal de gem eta!., 2012). Attempts at increasing doses of specific chemotherapeutic drugs or myeloablative chemotherapy with total-body irradiation for patients with multifocal or relapsed ES have thus far failed to improve patient survival (Balarnuth and Womer, 2010). Given the lack of more effective treatment options for aggressive disease, almost all ES patients receive the same chemotherapeutic regimen. This "one size fits all" treatment strategy results in a situation where patients are often receiving treatment that is ineffective or unnecessary. Patients that do survive are then at risk for developing late effects such as secondary neoplasms, chronic cardiac and pulmonary conditions, and reduced fertility (Ginsberg eta!., 2010). Consequently, there is a need to find more informative prognostic biomarkers for ES that will help in the design of rational treatments based on a patient's likelihood of response. The term biomarker refers to a "characteristic that is objectively measured and evaluated as an indicator of normal biologic processes, pathogenic processes, or pharmacologic responses to a therapeutic intervention" (Atkinson eta!., 2001). The first 6 biomarker studies for ES focused on a single marker or subset of markers that could be detected in patient blood or tumor samples. Blood sample markers have ranged from blood cell counts, angiogenic factors, cytokines, and the enzyme lactate dehydrogenase, which is thought to reflect tumor cell turnover. Reports looking at the association between these biomarkers and survival have mostly been inconsistent or have not been validated in independent studies(van Maldegem eta!., 2012). Different genetic lesions have also been identified in ES with varying degrees of prognostic significance. The first and most important recurrent genetic lesion discovered in ES was the t(ll ;22) translocation event creating the EWS-FLJJ fusion oncogene (Delattre et a!., 1992). The most common form of this fusion involves the in-frame joining of exon 7 of EWS to exon 6 of FLJJ, called a type I EWS-FLJJ fusion. Early work demonstrated a survival advantage for patients with tumors harboring a type I fusion compared to all non-type I fusions (Zoubek eta!., 1996;de Alava eta!., 1998). Two large prospective studies, however, found this advantage has been erased with current treatment protocols (Le De ley eta!., 2010;van Doorninck eta!., 2010). Secondary mutations in addition to EWS FLJJ have also been identified in ES. The most common secondary mutations include loss of function of the tumor suppressor genes p53 or pl6 present in 25% of tumors (Huang eta!., 2005). Alterations in these genes are associated with poor overall survival and chemoresponse. Clinically though, the mutational status of these genes is not used to predict response to therapy. Thus, many of the early biomarker studies for ES identified markers that have not been validated or have not been incorporated into standard clinical practice. These studies were limited in part by their focus on small groups of markers that 7 did not account for the global gene expression and signaling patterns found within tumors. Recent efforts to identify new biomarkers in ES have relied upon powerful genome-wide profiling tools to assess the association between genetic and epigenetic parameters and survival. These tools have included methods for measuring transcription levels, methylation status, copy number variation, and protein binding for thousands of genetic loci. The large datasets generated by these platforms can then be filtered according to clinical or biologic criteria to find genes of interest. As an early demonstration of applying microarray technology to identify potential diagnostic and therapeutic markers in ES, Khan et a!. developed computer-based algorithms to correctly distinguishES microarray expression profiles from other pediatric tumors (Khan eta!., 2001 ). Several studies have since used gene expression micro arrays to identify markers of aggressive disease. By profiling tumors grouped into high and low-risk survival groups, CDHJJ andMTAJ were found to be elevated in samples derived from patients with disease progression within five years (Ohali eta!., 2004). Likewise, expression profiles from tumors classified based on extent of spread and sensitivity to chemotherapy revealed metastatic and chemo-response gene signatures, respectively (Schaefer et a!., 2008). Two independent groups also identified a negative association between genes involved in glutathione metabolism and survival (Luo eta!., 2009;Scotlandi et a!., 2009). Furthermore, these groups used chemical or genetic means to inhibit glutathione synthesis and increased the sensitivity to chemotherapeutic agents. These studies were able to simultaneously identify markers of aggressive disease and demonstrate how such makers could be molecularly targeted for therapeutic purposes. Although the use 8 microarray technology for biomarker discovery in ES is still at an early stage, the technology is quickly evolving and leading to the identification of promising new prognostic markers. The goal of this study was to utilize new transcriptional profiling methods to identify novel prognostic biomarkers in ES. We used Affymetrix human exon (HuEx) arrays to measure gene expression in a primary drug resistant case of ES both before and after treatment with chemotherapy to find genes that contributed to progressive disease. Based on a comparison of transcriptional profiles derived pre- and post-treatment from this patient's tumor to a representative panel of other ES tumors, we generated a list of differentially expressed genes that showed enrichment for many developmental functions. In particular, we focused on the stem cell marker LGR5 which was highly expressed both pre and post-treatment. We further demonstrate that LGR5 is variably expressed in ES, but generally elevated in patients with poor prognosis. In summary, we demonstrate LGR5 is a potential marker of chemoresistant disease and has potential prognostic applications in ES. 9 Results Stem cell-associated genes are highly expressed in a primary drug resistant case ofES To discover new potential prognostic biomarkers in ES, we utilized biopsy specimens from a case of primary drug resistant disease diagnosed at Children's Hospital Los Angeles. The patient was a 9 year-old male who was originally diagnosed with localized ES of the humerus. The patient received neoadjuvant chemotherapy before tumor resection, but did not show a histologic response. The patient eventually died 13 months after diagnosis. Biopsy samples were obtained from the patient both before receiving neoadjuvant chemotherapy and afterwards. These samples were then profiled using Aftfymetrix HuEx arrays to generate transcriptional profiles. The HuEx expression data from the pre- and post-treatment specimens formed the basis oflooking for genes that were differentially expressed in progressive disease. The transcriptional profiles of both the pre- and post-treatments samples was first compared against a panel often representative ES samples to identify genes with a 2-fold change in expression in the primary chemoresistant case. This comparison generated two gene lists that were then compared against each other to find genes with increased differential expression post-treatment. In other words, the goal was to identify genes that were up- or down-regulated in the pre-treatment sample and then were further up- or down-regulated in the post-treatment sample. Based on this criterion, we identified 130 genes with increased differential expression post-treatment. To understand how these genes might be biologically related, we used Ingenuity Pathway Analysis (IPA) to find biochemical and functional categorizations within the gene list. In particular, we used the network analysis feature to find highly interconnected groups of genes based on gene-gene and gene- 10 protein interaction data found in the literature. This analysis yielded 17 gene networks, 8 of which were involved in developmental functions (Table 1). These functions included "Nervous System Development and Function", "Cardiovascular System Development and Function", "Cellular Development", "Skeletal and Muscular System Development and Function", among others. The high representation of gene networks involved in developmental functions lead us to focus on stem cell-associated genes as possible prognostic biomarkers. 11 Figure 1. In silico data analysis of a primary drug resistant case of Ewing sarcoma (ES). Flow chart of the approach we used to identify potential prognostic markers in a primary drug resistant case of ES. This approach allowed us to identify the stem cell marker LGR5 as a gene of interest. Identified 1 case of primary drug resistant ES l Case profiled using Affymetrix HuEx arrays before and after treatment l Compared profiles of 10 chemosensitive samples of ES l Identified 130 genes with increased differential expression post-treatment Identified the stem cell marker LGR5 / Ingenuity Pathway Analysis showed enrichment of developmental functions / 12 Table 1. Developmental functions are highly represented among genes that are differentially expressed by primary drug resistant ES. Network analysis was performed using Ingenuity Pathway Analysis (IP A) software on 130 genes that were at least 2-fold differentially expressed at diagnosis and for which differential expression increased further following induction chemotherapy. The table shows the top ten gene networks generated using IP A. The Focus Molecules column indicates the number of genes used from the list of 130 to generate each network and the Score provides a corresponding index to rank the networks based on the number of genes used. © 2000-2009 Ingenuity Systems, Inc. All rights reserved. ID Score Focus Molecules [I' op Functions ~ell Death, Cancer, Nervous System Development and 1 50 24 IF unction ~ardiovascular System Development and Function, 2 27 15 ~ellular Movement, Lipid Metabolism ~ardiovascular System Development and Function, 3 25 14 !Amino Acid Metabolism, Small Molecule Biochemistry ~ell Signaling, Molecular Transport, Vitamin and 4 23 13 ~ineral Metabolism Pene Expression, Cellular Development, Cellular 5 18 11 Prowth and Proliferation ~ardiovascular Disease, Renal Hydronephrosis, Renal 6 18 11 ~nd Urological Disease ~ellular Development, Nervous System Development 7 17 11 ~nd Function, Cellular Assembly and Organization PNA Replication, Recombination, and Repair, Nervous 8 2 1 System Development and Function, Tissue Morphology ~ellular Compromise, Nervous System Development 9 2 1 ~nd Function, Cell Morphology ~ndocrine System Disorders, Immunological Disease, 10 2 1 ~etabolic Disease 13 LGR5 is expressed byES Among the most highly differentially expressed genes in the chemoresistant case was the stem cell marker LGR5. We foundLGR5 was up-regulated 17-fold pretreatment and 21-fold post-treatment (Figure 2). LGR5 encodes a cell surface protein that is expressed by various somatic and cancer stem cell populations (Barker et a!., 2007;Jaks eta!., 2008;Barker eta!., 2012;Kobayashi eta!., 2012;Plaks eta!., 2013) and its expression and function had yet to be characterized in any pediatric sarcoma. To address this knowledge gap, we analyzed the expression of LGR5 in a cohort of 49 primary ES tumor samples and 15 ES cell lines. Unfortunately, the specificity of commercially available antibodies for studies of human LGRS remains inadequate for immuno detection studies so we limited our evaluation of LGR5 expression to quantitative RT PCR analysis. As shown, LGR5 was widely detectable in ES tumors and cell lines, however, the level of expression was highly variable (Figures 3A and 3B). 14 Figure 2. Leucine-rich repeat-containing G-protein coupled receptor 5 (LGRS) is highly expressed in primary chemoresistant ES. Affymetrix human exon data showing LGR5, a stem cell marker, was up-regulated in a primary chemoresistant case ofES at diagnosis compared to a panel of 10 representative ES tumors. It was further up-regulated after treatment. ..- 10 N 0) • • 0 .._ • • ~ en 8 • c JB .c co c 6 _C) • en • I.C.> a:: t (!) -.J 4 Resistant case Sensitive Pre Post I 16.6x up 21.3x up 15 Figure 3. LGR5 is expressed byES. qRT -PCR analysis revealed low to very high-level expression of LGR5 in (A) 49 primary ES and (B) 15 ES cell lines. For primary tumors levels of expression relative to GAPDH were <0.001% in 1 tumor, 0.001-0.01% in 7 tumors, 0.01-0.1% in 17 tumors, 0.1-1% in 12 tumors, 1-10% in 9 tumors, and 10-100% in 3 tumors. (A) [!! 0 E .3 1 0 'It Primary tumors LGR51evel (%rei to GAPDH) 16 LGR5 is increased in high-risk ES Given the wide range of expression of LGR5 in ES tumors, we wanted to determine if this variability was related to clinical outcome. Previously, we profiled 56 localized samples using Affymetrix HuEx arrays (kindly provided by the Children's Oncology Group). We used the expression data for LGR5 from these transcriptional profiles for a clinical correlates analysis. Although there was no association between LGR5 and overall survival (Figure 4A), there was a trend toward a significant association between expression and event free survival (p-value ~ .0784; Figure 4B). In contrast, there was no association between PROM I (also known as CD133), another stem cell marker up-regulated in the primary drug resistant case, and any measurement of survival (Figures 4C and 4D). Thus, LGR5 became our lead gene to pursue in further expression and survival studies. To further probe the connection between LGR5 and outcome in ES, we analyzed the expression of LGR5 in cell lines and tumor samples stratified by stage or survival time. We first comparedLGR5 expression in the CHLA9 and CHLAlO cell lines, which were derived from the same patient prior to (CHLA9) and after (CHLAl 0) chemotherapy. It is noteworthy that the CHLAl 0 cell line was derived from a metastatic focus and therefore represents a progressive state of disease. As shown, LGR5 levels were found to be 10-fold higher in CHLAlO than CHLA9 cells (Figure SA). To evaluate whether clinically aggressive ES express higher levels of LGR5 in vivo we interrogated whole genome expression data that were generated from primary human tumor samples. LGR5 expression levels from diagnostic biopsies were compared in 4 patients who succumbed to rapidly progressive disease survival in 5 to 13 months and 10 patients who 17 remained disease free for at least 48 months. LGR5 expression was found to be extremely high in 3 of 4 patients with aggressive disease and mean expression was significantly higher than in long-term event survivors (Figure SB). Interestingly, 2 of the 3 patients with the highest expression of LGR5 had primary drug-resistant tumors. In summary, these results provide preliminary evidence in support of the hypothesis that over expression of LGR5 in primary tumors may be associated with an aggressive drug resistant clinical phenotype. 18 Figure 4. LGRS is associated with poor event free survival (EFS). Clinical correlates analysis of(A) and (B) LGR5 and (C) and (D) PROMJ expression in primary tumors derived from 56 ES patients. We did not find an association between LGR5 and (A) overall survival, but we did observe a trend towards a significant association between higher expression and (B) poor EFS. PROMllevels were not association with either (C) overall survival or (D) EFS. (A) (C) Survival by LGR5 L o.,--..,.,----------'-------,== ~c-----· ~ - ~--]___ __ _ '-------1 08 06 00-L...,.--~-~-~-~--..,....-----J 06 Sww .oalt1me(years) Bdm\· median I Surviv•l by PROM I !._ ___ .., I I I •----------- 10 Belm\• ined!an ! (B) EFS by LGR5 10 EFS t•mc( years) (D) ~\ , ~L..._ ___ _ 06 -, ----------1 L ___________ _ 0.2 EFS ume(years) AbO\'Cilledian Bdcwo · nl«<w• l 19 Figure 5. LGRS is increased in aggressive disease. (A) Metastatic tumor-derived CHLAlO cells express higher levels of LGR5 than CHLA9 cells, which were derived from the primary tumor at diagnosis. (B) LGR5 expression was found by microarray analysis to be increased in tumors from 4 patients with rapidly progressive and fatal primary ES (DOD-dead of disease) compared to 10 patients with at least 48 months disease free survival (LTS-long term survivors). (A) 1 (B) 10 ~ ••• · cn 8 _a;- c • Q)<( 2 >_J c 2I ro 6 •• l()() • c O::o 0:: Ol · cn 4 p=.001 -J~ l() =·· a:: • (!) 2 -I 0 CHLA9 CHLA10 DOD LTS Patient Outcome 20 Discussion In this study, we have for the first time identified LGR5 as a putative marker for high-risk ES. We used HuEx array profiling to demonstrate LGR5 was highly expressed in a primary drug resistant case and in patients with a poor prognosis. This strategy for finding potential biomarkers in ES was based on previous studies using microarray technology to assess genome-wide expression levels in tumors samples stratified by measures of aggressive disease. Our approach was unique in utilizing biopsy samples derived from the same patient pre- and post-chemotherapeutic treatment as part of the initial screen for prognostic markers. These samples allowed us to control for genetic differences between patients and identify specific changes in gene expression based on disease progression. With the rapid proliferation of genome-wide profiling tools and increased emphasis on personalized medicine for treating patient's individual tumors, such an approach may become more widely used in the future for finding prognostic and therapeutic targets in ES. We were able to identify LGR5 as a marker of interest because of the over representation of genes involved in developmental functions in the primary drug resistant case. Previous studies have shown poorly differentiated tumors often co-opt developmental programs and express a stem cell-like gene expression signature that is associated with poor prognosis. For example, poorly differentiated breast tumors, glioblastoma, and bladder carcinomas all demonstrated an embryonic stem cell-like gene expression signature (Ben-Porath eta!., 2008). Breast cancer patients with this embryonic stem cell signature also had a significantly higher mortality rate and the signature provided prognostic information beyond histologic grade. These observations indicate 21 that poorly differentiated tumors arising in different tissues from distinct cells of origin often display a similar developmental program. In progressive disease, various cancers types can also have a gene expression profile similar to their respective cell-of-origin. Both poorly differentiated breast tumors and mammary stem cells show a high degree of concordantly up- and down-regulated genes (Pece eta!., 2010). Additionally, gene expression signatures derived from mammary stem cells and hematopoietic stem cells are associated with worse prognosis in breast cancer and acute myeloid leukemia, respectively (Eppert eta!., 2011). These findings show stem cell gene expression signatures have prognostic potential and this guided our selection of stem cell-associated genes as a candidate markers for follow-up studies. Expression studies demonstrated LGR5 was present at highly variable levels in ES tumor samples and cell lines, leading us to question whether this variability was relevant to prognosis. We found that LGR5 was associated with poor survival in large cohort of patients with localized disease. This finding supports previous studies that have shown LGR5 is a prognostic marker in various tumor types. Studies of colorectal and gastric carcinoma, glioblastoma, and esophageal adenocarcinoma have all demonstrated heterogeneity of LGRS expression and shown that high LGR5 levels are associated with worse outcomes ((Becker eta!., 2010;Simon eta!., 2012;Wu eta!., 2012;Nakata eta!., 2013 ). Likewise, several studies have found a connection between high LGR5 expression and chemoresistance (Bauer eta!., 2012) and metastasis (Uchida eta!., 2010;Takahashi et a!., 2011;Valladares-Ayerbes eta!., 2012;Wu eta!., 2012) in gastrointestinal malignancies. Interestingly, recent studies in mouse models of neuroblastoma (Balamuth eta!., 2010) and medulloblastoma (Kawauchi eta!., 2012) also discovered increased 22 expression of Lgr5 in the most aggressive tumors. Thus, there is now substantial evidence in tumors of both epithelial and neural origin to implicate LGRS as a marker of an aggressive clinical phenotype. Although we have identified LGR5 as a potential prognostic biomarker in ES, this finding will need to be validated and further analyzed to develop future clinical applications. The first step in validating LGR5 as prognostic biomarker will be to perform a clinical correlates analysis with an independent cohort of archived ES tumor samples. The second step will be to conduct a prospective clinical study, the gold standard by which newly identified prognostic biomarkers are evaluated. This study would be designed to measure LGR5 levels at diagnosis in ES patients and then the clinical course of each patient would be monitored to determine whether LGR5 is predictive of outcome. Questions regarding the function of LGR5 in ES pathogenesis will also need to be addressed to understand how LGR5 can serve as a marker of aggressive disease. In particular, whether LGR5 directly promotes a malignant phenotype or is an indirect marker of cells with such properties. If LGR5 directly contributes to malignant phenotypic properties, it may also serve as a candidate marker for molecularly targeted therapies. There still remains a great deal to be discovered about the prognostic significance and function of LGR5 in ES, but these initial findings highlight the potential clinical applications for understanding the role of LGR5 in this highly aggressive, childhood disease. 23 Chapter3 LGR5 is a marker of tumor development in Ewing sarcoma Introduction Ewing sarcoma (ES) is an aggressive bone and soft tissue tumor that has many stem cell features. Histologically, tumors are characterized by an undifferentiated small round blue cell phenotype with features of primitive neuroectodermal cells. Although predominantly a bone and connective tissue tumor, clinically ES can present in multiple organs and tissue types throughout the body, suggesting a relatively undifferentiated and potentially highly migratory cell of origin (Meltzer, 2007). Indeed, current evidence supports the hypothesis that ES arises from either mesenchymal stem cells (MSC) or neural crest stem cells (NCSC) or their early progenitors transformed by an EWS-ETS fusion (Staege eta!., 2004;Tirode eta!., 2007;Riggi eta!., 2008;von Levetzow eta!., 2011). Putative cancer stem cells (CSC) have also been isolated in ES based on CD133 expression and aldehyde dehydrogenase activity (Suva eta!., 2009;Awad eta!., 2010;Jiang eta!., 2010). Compared to bulk tumor cells, CSC show enrichment for tumorigenic potential and often demonstrate increased survival under physiologic stress and drug treatment conditions (Vis vader and Lindeman, 20 12). These cells can in turn contribute to metastasis and chemoresistance, the major causes of mortality among cancer patients (Longley and Johnston, 2005;Chaffer and Weinberg, 2011). The putative stem cell origin and CSC subpopulations of ES, thus, offer insight into the different stages of tumor development. The identification of stem cell markers that are expressed by these cell populations offer a way of following the development of ES from its earliest stages 24 to the later acquisition of aggressive clinical features that contribute to cancer-related mortality. Although the stem cell origin and CSC hierarchy of many malignancies remains uncertain, there are several cancer types where these cells have a shared marker phenotype that can be used to track disease progression. Given the accessibility of peripheral blood and bone marrow samples, leukemia is the best characterized of these cancer types. Both chronic myeloid leukemia (CML) and acute myeloid leukemia (AML) can originate from hematopoietic stem cells (HSC) with a defined CD34+ /CD38. surface phenotype (Becker and Jordan, 2011). Likewise, the CD34+/CD38. surface phenotype can be used to prospectively isolate leukemic stem cells (LSC) from CML and AML patients. The LSC gene expression signature also been shown to be associated with worse prognosis in AML patients (Gentles eta!., 2010). In the case of breast cancer, mammary stem cells, a putative cell of origin, and breast CSC can express similar markers. Pece et a!. demonstrated that mammary stem cells and breast CSC share a CD49F+/DLLhigh/DNERhigh surface phenotype and that poorly differentiated tumors have a greater percentage of these cells (Pece eta!., 2010). Most notably, LGRS has emerged as a leading marker for the cell of origin and as a CSC marker for colorectal carcinoma. Lgr5 is highly expressed by intestinal stem cells (IS C), which have been identified as the cells of origin for murine intestinal tumors and tumor-maintaining CSC in established adenomas (Barker eta!., 2009;Schepers eta!., 2012). LGRS+ CSC have also been identified for human colorectal cancer (Kemper eta!., 2012;Kobayashi eta!., 2012). In addition, gene expression profiling experiments have demonstrated that the Lgr5-stem cell gene signature predicts disease relapse in colorectal cancer patients (Merlos-Suarez 25 eta!., 2011). Thus, there is compelling evidence that stem cell markers such as LGRS can be used to study cancer initiation and progression and that high levels of these markers are associated with worse clinical outcomes. Changes in expression of LGR5 have been already used to study aggressive clinical features of cancer. Increased LGRS is associated with increased invasion and/or metastasis in gastrointestinal malignancies (Uchida eta!., 2010;Takahashi eta!., 2011;Valladares-Ayerbes eta!., 2012;Wu eta!., 2012). Expression was also found be increased after chemotherapeutic treatment in biopsy samples taken from patients with gastric carcinoma (Bauer eta!., 2012). The association between high levels of LGR5 and these aggressive features may be due to extracellular stress factors that induce expression or select for the presence of LGRS+ cells. Both presence of reactive oxygen species and anchorage-independent growth conditions have been found to enrich for LGR5 in colorectal cancer cells (Kanwar eta!., 2010;Kim eta!., 2012). The expression of LGR5 in these conditions acts as marker for probing how tumor cells respond to stressors in the microenvironment and furthers our understanding of the growth of established tumors. Based on the stem cell features of ES and our earlier finding that LGR5 is a potential prognostic biomarker, we hypothesized it may act as a marker for tumor initiation and progression as has been shown for colorectal carcinoma. We first analyzed expression in neuro-mesenchymal stem cells to determine if LGR5 may serve as a marker for the cell of origin. To understand the role of the microenvironment on LGR5levels, we also cultured ES cells under physiologic and chemotherapeutic stress conditions to determine how expression changed. Finally, we measured expression in putative ES CSC and directly probed for the existence of an LGR5-high subpopulation in ES cell lines to 26 account for elevated expression under stress conditions. Cumulatively, these results show that LGR5 can be used as a marker to follow disease initiation and progression and they provide biologic and clinical insights into ES pathogenesis. 27 Results LGR5 is expressed by neural crest-derived stem cells Given its designation as a stem cell marker in epithelial tissues we assessed whether LGR5 might be expressed by NCSC and/or MSC. To address this we evaluated adult human bone marrow-derived MSC as well as human embryonic stem cell-derived NCSC and their MSC progeny, neural crest-derived MSC (NC-MSC) as previously described by our group (von Levetzow eta!., 2011). Significantly, our analyses revealed that undifferentiated NCSC express relatively high levels of LGR5, whereas the transcript is undetectable in adult bone marrow-derived MSC (Figure 6A). Consistent with this observation, LGR5 expression declined when NCSC were differentiated towards an MSC identity (NC-MSC) (Figure 6A). Interestingly, the level of LGR5 expression in these neural crest-derived stem cells was comparable to that of ES cell lines (Figure 6A). To determine if the observed differential expression of LGR5 by neural crest cells was merely an artifact of the human embryonic stem cell culture system we next interrogated publicly available gene expression data that compared murine bone marrow stem cells of mesodermal origin to bone marrow stem cells of neural crest origin (GEO accession 30419) (Wislet-Gendebien eta!., 2012). Interestingly, consistent with our human cell studies, murine bone marrow-derived stem cells of neural crest origin expressed higher levels of Lgr5 than MSC of mesodermal origin (Figure 6B). 28 Figure 6. LGRS is expressed by neural crest stem cells (NCSC). (A) qRT-PCR analysis ofhuman embryonic stem cells (hESC), hESC-derived NCSC (hNCSC), hNCSC-derived mesenchymal stem cells (hNC-MSC), bone marrow-derived mesenchymal stem cells (hBM-MSC) and human lung embryo fibroblasts (MRC5) showed that undifferentiated hNCSC express the highest levels of LGR5. (B) Murine bone marrow stromal cells of neural crest origin express higher levels of Lgr5 than cells of mesodermal origin (from publically available microarray data GEO accession 791 GSE30419); (Wislet-Gendebien et al., 2012). N=3 ± SEM. (A) 15 s: 10 -Q ~Q • • ~<3 5 ct~ • <.9 0 - ~ 0.10 •• • •• • ._J i 0.15I • 0.05 •• 0.00........_+,·--+---r--t--..,..._""T""_ • -- - .L )... J._ ,. )... r-- 1/) ID c "ID u en w ~ c u u en en w u .r:: z .r:: c §: ~ u u 1.0 en en u ::2: ::2: IY 0 ~ ::2: z aJ .r:: .r:: (B) 8 p ·u; 6 c .ill c ro c 4 0") ·u; '{? 0> 2 ._J 0 • • • Mesodermal (3) •• • Neural crest (3) Bone marrow stromal cells 29 Expression ofLGR5 is increased in ES cells grown under physiologic stress conditions The expression of LGR5 in neural crest-derived stem cell populations provided a snapshot of what the level of expression may be during tumor initiation. We next wanted to assess how LGR5 levels were affected in conditions found in an established tumor microenvironment. The tumor microenvironment is different from normal tissue compartments due to the poor delivery of oxygen and nutrients from disorganized blood vessels (Hanahan and Weinberg, 20 II). Tumor cells also produce energy much less efficiently that normal cells due to their reliance on glycolysis rather than oxidative phosphorylation (i.e. the Warburg effect). Additionally, Ewing cells also have a spheroid like morphology within tumors and form limited adhesions to the extracellular matrix (ECM) (Lawlor et a!., 2002). This growth pattern is called anchorage-independent growth and many non-transformed cells die through a process called anoikis when detached from the surrounding ECM. We analyzed these physiologic stress conditions to determine whether these factors influenced LGR5 levels in ES cells. Three ES cell lines (CHLA25, A673, and TC71) were grown in no serum, low attachment, and hypoxic conditions to mimic the tumor microenvironment. By microscopic inspection, all three cell lines appeared viable under these conditions and were all adherent when grown in no serum or hypoxia and formed spheroids when grown on low binding plates. The level of LGR5 was also unaffected when the cells were grown on low attachment plates or hypoxia when compared to standard culture conditions (Figure 7). The absence of serum, however, increased LGR5 expression 2 to 4-fold in all three cell lines. This effect may represent an upstream transcriptional control of LGR5 or an inherent survival advantage of LGR5-high expressing cells. 30 Figure 7. LGR5 is increased under serum free conditions. qRT-PCR analysis of LGR5 expression in ES cell lines were grown under various stress conditions (no serum, low attachment, and hypoxia) for 48 hours. LGR5 was consistently increased across all three cell lines when grown in media with no serum. Data is from two independent experiments and error bars are SEM. D A673 D CHLA25 • TC71 31 LGR5 is a marker of chemoresistance in ES cell lines We previously demonstrated that LGR5 is elevated in ES cases with poor event free survival and expression is increased in patient-matched samples post chemotherapeutic treatment. These findings indicate that LGR5 may be a marker of chemoresistance, and we investigated this possibly using ES cell lines treated with different chemotherapeutic agents. By first analyzing data from a previous report (Schaefer eta!., 2008), we found that increasing LGR5levels were associated with doxorubicin resistance as indicated by higher half maximal inhibitory concentration (ICSO) values (Figure 8A). We then confirmed this finding using the isogenic cell lines CHLA9 and CHLAlO. CHLAlO cells which have higher basal levels of LGR5 and were derived post-treatment, had a much higher ICSO for doxorubicin than CHLA9 cells (Figure 8B). We also examined whether doxorubicin treatment could directly enrich for LGR5 expression in these cell lines. Compared to no drug treatment, we found that doxorubicin at several concentrations enriched for LGR5 expression in both cell lines (Figure 8C). The level of enrichment was much lower in CHLAl 0 cells though, perhaps reflecting resistance mechanisms that were developed in the original post-treatment tumor and decreasing the dynamic range of LGR5 expression. We found a similar trend in three other ES cell lines that were derived post-treatment and/or had higher basal levels of LGR5 compared to CHLA9 cells (Figure 8D). These cell lines all demonstrated little or no increase in LGR5 levels after treatment with doxorubicin. This trend suggests that ES cells with higher basal levels of LGR5 are more resistant to doxorubicin, whereas cells with lower basal levels are less resistant but show greater enrichment post treatment. In the case of CHLA9 cells, the low basal level of LGR5 may be a reflection of 32 the chemo-naive state of the tumor cells at the time isolation and treatment with doxorubicin may induce a phenotypic shift in the cells creating an expression profile similar to CHLAl 0 cells. This data may also provide supportive evidence for the existence of an LGR5-high subpopulation of cells within some tumors that are selected for in the presence of doxorubicin. 33 Figure 8. LGR5 is a marker of chemoresistance. (A) Comparison of LGR5levels in 10 ES cell lines and corresponding IC50 values for doxorubicin show a significant correlation (Schaefer et al., 2008). (B) Growth ofCHLA9 and CHLAlO cells at increasing doses of doxorubicin show that CHLAlO cells have a much greater IC50 than CHLA9 cells. (C) LGR5 is emiched at a higher level in CHLA9 cells (left) compared to CHLAlO cells (right) when grown in the presence of different doses of doxorubicin. (D) Similar to CHLAlO cells, CHLA25, A673, and TC71 cells shows no or only a small increase in LGR5 when grown at a high dose of doxorubicin. The data in panels (B)-(D) is from 2-3 independent experiments and error bars are SEM. (A) 15 R 2 =0.8 • ~ 10 p<5.0x10· 4 2; 0 LO \2 5 O+Z~~--~--~--~~--, Oi :::J u 0 c .9 (D) o:; 2 · 2 -o g 1 . .9 ! 1. Q) > .5!! 0. Li) a:: (!) ..... 0. 0 20 40 60 80 100 LGR5 signal intensity CHLA9 Doxorubicin concentration No drug High Doxorubicin concentration (B) 1.5 Oi 2 -o 0 1.0 c .9 ~ g 0.5 :.0 "' > 0.0 -10 -8 -6 Log [Doxorubicin (nM)] Oi 2 CHLA10 :::J u 2 0 c .9 ~ Q) 1 > .5!! Li) 0 (3 ..... 0 No drug Low Medium High Doxorubicin concentration 34 LGR5 is enriched in discrete subpopulations ofES cells Many of the results we have presented thus far have suggested the existence of an LGRS+ or LGR5-high subpopulation of cells within ES tumors or cell lines. Our efforts to study such a population over the last several years have been hampered by the lack of anti-LGRS antibodies for immuno-detection. We have carried out a series of studies to overcome this technical hurdle and establish whether such a subpopulation exists in ES. It has been previously reported that LGR5 expression is enriched in cancer cells with stem cell-like properties (Merlos-Suarez eta!., 2011;Kemper eta!., 2012;Kobayashi eta!., 2012;Nakata eta!., 2013). These cells have the unique ability to self-renew and differentiate compared to bulk tumor cells and can contribute to metastasis and drug resistance (Visvader and Lindeman, 2012). Although putative CSC were identified in a small cohort of primary tumors (Suva eta!., 2009) it has been challenging to isolate putative CSC from established ES cell lines. Nevertheless, CD133 surface expression (Jiang eta!., 2010) and high-level aldehyde dehydrogenase (ALDH) activity (A wad eta!., 2010) have been used to successfully enrich for CSC populations in STA-ET-8.2 and MHH-ES and TC71 cells, respectively. Therefore, we used these previously reported ES cell lines and CSC-enrichment assays to determine whether LGR5 expression is up regulated in putative CSC. Q-RT-PCR analysis ofCD133-sorted STA-ET-8.2 populations consistently demonstrated increased expression of LGR5 in the eSC enriched CD133+ fraction (Figure 9A). Similarly, levels of LGR5 were reproducibly higher in the ALHDhigh subpopulation compared to ALDH!ow subpopulation ofMHH-ES cells (Figure 9B). In contrast, data from TC71 were inconsistent. Although increased expression of LGR5 was detected in ALDHhigh CSC populations in one experiment this 35 finding was not reproducible in two other independent experiments (Figure 9B). These findings suggest thatLGR5 is highly expressed by at least some populations ofCSC-like ES cells and could account for the enrichment of LGR5 when ES cell lines are grown under stress conditions. To detect ES subpopulations on the basis of LGR5 expression alone, we used methods to directly indicate the presence of cells with high-level transcription of LGR5. We focused on using CHLA25 cells, which express high levels of LGR5, and would potentially increase the sensitivity of our assays. We first used an LGR5-specific reporter construct comprised of 1200 bases from the promoter region driving GFP expression. We stably transfected CHLA25 cells with this construct and then sorted cells for high and low GFP expression. Comparing the expression of LGR5 in these two subpopulations, we found LGR5 was increased approximately 2-fold in the LGRS-GFPhigh cells (Figure lOA). Additionally, we used in situ hybridization (ISH) to directly stain for cells expressing the LGR5 transcript. We stained CHLA25 cells and found discrete subpopulations of LGR5- high cells (Figure 1 OB). These data support the presence of LGR5-high subpopulations in CHLA25 cells. Recently, Kemper eta!. developed a new series of monoclonal anti-LGRS monoclonal antibodies for the study of colorectal CSC (Kemper eta!., 20 12). Unfortunately, these antibodies have only become commercially available in the last few months and have not allowed for sufficient time to optimize their usage in ES cell lines. Based on initial testing of one of these antibodies in flow cytometry experiments though, we were able to faintly detect an LGRS+ subpopulation in CHLA25 cells constituting approximately 4% of the total cells (Figure lOC). These results confirm our ISH results 36 as well as findings from Kemper et a!. where they were only able to strongly detect LGRS+ cells in a colorectal cancer cell line with LGRS over-expression. In comparison, this signal was greatly diminished in parent cells. Although these subpopulation studies are preliminary, we will further optimize the use of the new anti-LGRS monoclonal antibodies to increase the detection of ES LGRS+ cells. 37 Figure 9. LGR5 is increased in putative cancer stem cells. (A) LGR5 levels are higher in CD133+ compared to CD133- STA-ET-8.2 cells. Data from two independent sorting experiments are shown. (B) LGR5 expression is increased in ALDHhigh compared to ALDH 1 ow MHH-ES and TC71 cells. Data from three independent sorts are shown. The horizontal lines in both panels represent mean values. (A) 4 -d-> 3 QJC') >.- ~0 l()(_) 2 O::o (!)~ -J~ STA-ET-8.2 o~----~----------~----- CD133- CD133+ (B) 4 ~3 - 0 ~· wi -0 (Q <i! 2 (!) 0 -J~ ~1 • •• • • • o~----~----------~----- MHH-ES TC71 38 Figure 10. LGRS is expressed by discrete subpopulations of CHLA25 cells. (A) LGR5 levels are higher in sorted CHLA25 LGR5-GFPhigh cells compared to LGR5- GFP10w cells (B) In situ hybridization results for CHLA25 cells show that subpopulations of LGR5-high cells exist. Black arrowheads indicate the presence of LGR5-high cells. Nuclei were stained with hematoxylin. (C) This finding was confirmed with initial testing of a new anti-LGR5 monoclonal antibody that demonstrated LGR5+ cells cannot be detected in A673 cells (left panel) but can be detected in CHLA25 cells (right panel). Preliminary data from one experiment for each assay is shown. (A) ~ 0 ...., a.. LL <..9 .9 - ~ Qi > ~ ll) n: (!) -J (C) CHLA25 2.0 1.5 1.0 0.5 0.0 GFP-Iow GFP-high A673 o A06A67~LGR5 ~ . Gate: (R1 in all) g~--~~~==~==~ o . M 0 0 o . 0 g. M ~ g U o Ul • oo g o . P2 0.7% .. 1 ,) ,;J .. 4 .,p .JJ .. 7 .2 FL2·A (B) CHLA25 CHLA25 o A12 CHLA25-LGR5 S _ Gate: (R1 in all) ~~--~~~==~==~ M 0 0 o. 0 0 o . M ~ 0 u o Ul o . oo g o . P2 4.2% .. 1 ,) ,;J .. 4 .,p .JJ .. 7 .2 FL2·A 39 Discussion We report here on the characterization of LGR5 as a marker of ES development from its putative stem cell origins to later stages when cells are exposed to adverse conditions present in the tumor microenvironment. By assessing expression in developmentally related-stem cell populations, we found that LGR5 is highly expressed by neural crest-derived stem cells and is a potential marker for the cell of origin. We also found that levels were enriched under physiologic and chemotherapeutic stress conditions. These results suggested the presence of an LGR5-high subpopulation, which we were subsequently able to demonstrate in putative ES CSC populations and directly using a reporter assay and in situ hybridization specific for LGR5 expression. Together these findings indicate that LGR5 is a marker of a stem cell phenotype that persists throughout ES development and that LGR5-high cells may contribute to aggressive clinical features. The mechanism of transcriptional up-regulation of LGR5 in ES cell populations is unknown. Importantly, however, LGR5 is not induced by EWS-FLI1 and unpublished data from our own lab as well a previously published report (see Navarro eta!. supplementary data) suggest thatLGR5 is, in fact, repressed by EWS-FLI1 (Navarro et a!., 2010). Thus, we reasoned that expression of LGR5 in ES cells may instead be a reflection of their putative stem cell origins. Interestingly, we discovered that undifferentiated NCSC expressed the highest levels of LGR5. Expression was still detectable, albeit at lower levels, in NCSC that had undergone epithelial-mesenchymal transition to an MSC-state (von Levetzow eta!., 2011). In contrast, bone marrow-derived MSC did not express detectable levels of LGR5. Intriguingly, it is now established that 40 rare MSC in the bone marrow are derived from the neural crest (Takashima et a!., 2007;Nagoshi eta!., 2008) and gene expression profiling data from mouse MSC (GEO accession GSE30419) showed increased expression of Lgr5 in neural crest- compared to mesoderm-derived MSC populations (Wislet-Gendebien eta!., 2012). Based on these studies we now speculate that at least some ES might arise from neural crest-derived LGRS+ stem cells in the bone marrow. Our next question pertained to how LGR5ieveis changed after tumor initiation and might be affected by extracellular stress factors present in the tumor microenviroment. For physiologic stress factors, we focused on no serum, low attachment, and hypoxic conditions and found that no serum enriched for LGR5 levels. Likewise, we found doxorubicin treatment enriched for LGR5 levels in several ES cell lines with low endogenous expression. One possibility to explain the enrichment of LGR5 is that stress conditions can induce expression in ES cells. LGR5 was originally identified as a downstream target gene ofWnt/~-catenin signaling (Vander Flier eta!., 2007)and more recently as a Hedgehog and AP-I target gene, all of which can be activated under stress conditions (Tanese et a!., 2008;Aguilera eta!., 20 II). Another possibility accounting for enrichment is that LGR5-high ES cells have a selective survival advantage under stress conditions. These cells may then account for a greater percentage of the total cell population after encountering an extracellular stress factor. To address whether LGR5 is expressed by a subpopulation of cells with a selective survival advantage in ES, we measured expression in cells enriched for CSC like features. esc have been shown to be resistant to adverse microenvironmental conditions (Vis vader and Lindeman, 20 12) and we hypothesized these cells would show 41 enrichment for LGR5 in ES. We found expression was enriched in both STA-ET-8.2 CD133+ cells and MHH-ES ALDHhigh cells, established CSC subpopulations for ES (A wad et al., 2010;Jiang et al., 2010). Although both CD133 and ALDH activity can be used for isolating CSC from ES cell lines, these markers show varying degrees of enrichment for cells with stem cell activity. Sorting on the basis of ALDH activity has been shown to provide superior enrichment for stem cell activity compared to CD133 expression (A wad et al., 2010). Given thatLGR5levels were elevated in cells sorted for either one of these markers, methods based on directly isolating cells expressing LGR5 may further enrich for stem cell activity. The availability of new monoclonal antibodies targeting LGRS (Kemper et al., 2012) may allow for direct sorting LGRS+ ES cells and characterizing their CSC-like properties. Studying these cells in isolation may also allow for a greater understanding for the developmental relationship between LGRS+ ES cells and neural crest-derived LGRS+ cells. The CSC hierarchy for many cancers is highly plastic with cells interconverting between a differentiated and de-differentiated state without a clear connection to the cell of origin. (Visvader and Lindeman, 2012). Leukemia, on the other hand, tends to have a more static hierarchy and LSC are thought to originate from HSC (Becker and Jordan, 2011 ). Deciphering this relationship will prove useful for understanding the early pathogenesis of ES and comprehending how the stem cell phenotype is maintained in advanced disease. Finally, we were able to directly show that LGR5 is expressed by a subpopulation ofES cells using several different methods. If this subpopulation represents a true CSC population, this finding would represent a unique opportunity for the development of biologically targeted therapies for LGRS+ cells in ES. CSC often possess increased 42 metastatic and chemoresistant potential and specifically targeting these cells using monoclonal antibodies or small molecule inhibitors could suppress these aggressive clinical features. Together, these data indicate LGR5 is a marker of a malignant stem cell phenotype at multiple stages in ES development and provides support for the concept of CSC-based therapies to eradicate this disease. 43 Introduction Chapter4 LGR5 enhances Wnt signaling in Ewing sarcoma and promotes malignant phenotypic properties In the previous study, we demonstrated that LGR5 is expressed in NCSC, a putative cell of origin for ES, and is a marker for cells with aggressive features. The goal of the current study is to address the question of whether LGRS directly promotes these functions and what signaling mechanism might be involved. LGR5 has been studied in a variety of developmental and tumor-related contexts, but little is known about its role in sarcomagenesis. Mouse studies have shown that Lgr5 is important for the self-renewal and differentiation of various epithelial stem cell populations. Work with organoid culture systems have demonstrated that isolated Lgr5+ intestinal and gastric stem have increased colony-forming efficiency compared to LgrS" cells (Sato eta!., 2009;Barker et a!., 2010). Likewise, Lgr5+ mammary stem cells are necessary for mammary gland organogenesis as demonstrated by targeted ablation of these cells with diphtheria toxin (Plaks eta!., 2013). Several reports using human cancer cell lines have shown LGR5 promotes growth and/or survival of colorectal and basal cell carcinoma (McClanahan et a!., 2006;Tanese eta!., 2008), glioblastoma (Nakata eta!., 2013), and neuroblastoma (Balamuth eta!., 2010). In respect to colorectal carcinoma, LGRS has also been established as a cancer stem cell marker (CSC). The LGRS+ subpopulation was shown to possess increased clonogenic and tumorigenic potential compared to bulk tumor cells (Kemper eta!., 2012;Kobayashi eta!., 2012). Knockdown of LGR5 also leads to complete loss of clonogenic potential in the LS 174T colorectal cell line (Kemper eta!., 44 2012). The only evidence thus far ofLGRS promoting sarcoma pathogenesis was a report from Barrentina et al. showing the LGR5 locus is amplified in various soft tissue sarcomas (Barretina et al., 2010). This study sought to identify potential oncogenes and tumor suppressor genes based on gene amplification or deletion, respectively, using copy number variation. LGR5 was one such locus that was amplified, indicating it might serve as an oncogene in soft tissue sarcomas. The mechanism by which LGRS promotes stem cell self-renewal and proliferation has only recently been elucidated. It is now known that, along with its closely related homologues LGR4 and LGR6, LGRS functions to potentiate canonical Wnt/~-catenin signaling (Carmon et al., 20ll;de Lau et al., 20ll;Glinka et al., 20ll;Gong et al., 2012). This potentiation ofWnt signaling is achieved when LGRS is bound by its ligand R spondin (RSPO). RSPOs are a recently described family of secreted proteins that function as Wnt agonists and play pivotal roles as regulators of normal embryonic development and stem cell proliferation (Glinka et al., 2011 ). When RSPO binds to LGRS the receptor associates with the Frizzled/LRP complex to increase the activation of ~-catenin and downstream TCF reporter activity as well as non-canonical Wnt/PCP signaling (Glinka et al., 2011). The potentiation ofWnt/~-catenin signaling is now believed to mediate the self-renewal and proliferation of LGRS+ stem cells, both normal and malignant, and provides a mechanistic link between LGRS, Wnt signaling, cancer stem cells and cancer progression (Clevers and Nusse, 2012). Given the presumed stem cell origin of ES, the presence of RSP02 in developing bone (Friedman et al., 2009) and prior evidence that Wnt signaling is abnormal in these tumors (Uren et al., 2004;Navarro et al., 2010;Vijayakumar et al., 2011) we hypothesized 45 that LGR5 functions as an oncogene in ES and promotes a malignant phenotype through Wnt/~-catenin and/or Wnt/PCP signaling. We investigated this hypothesis withES cell lines in gain and loss of function experiments. Our findings indicate that LGR5 promotes malignant phenotypic properties in ES cell lines and that these cells demonstrate increased Wnt/~-catenin and Wnt/PCP signaling when exposed to RSPO. Together our studies suggest that LGRS-mediated activation ofWnt signaling may be an important contributor to ES initiation and maintenance, especially in RSPO-rich microenvironments like developing bone. 46 Results R-spondin activates Wnt/{3-catenin signaling in an LGR5-dependent manner Given its role as a potentiator of Wnt/~-catenin signaling, we hypothesized that ES cells with high levels of LGR5 expression would demonstrate relatively high Wnt/~ catenin transcriptional activity. To address this we first evaluated basal levels of the Wnt/~-catenin axis in ES cells that were grown in standard tissue culture conditions. To begin we measured the expression of known Wnt/~-catenin target genes in 3 different ES cell lines (A673, CHLA25, and TC71) and discovered that although the levels of gene expression varied significantly among the three cell lines, there was no correlation with LGR5levels (Figure llA). In support of this, we observed no significant or reproducible change in the expression of Wnt/~-catenin target genes following LGR5 knockdown (Figure llB). In particular, loss of LGR5 did not result in down-regulation of Wnt targets suggesting that, in the context of standard tissue culture (i.e. in the absence of exogenous RSPO), LGRS has little impact on Wnt/~-catenin transcriptional activity (Figure llB). Next, we used TCF-promoter luciferase-reporter assays to directly measure the level of TCF transcriptional activity in CHLA25 ES cells before and after LGR5 knockdown. These assays confirmed that LGR5 knockdown had only minimal impact on basal Wnt/~ catenin transcriptional activity in standard culture conditions (Figure 11 C). Thus, in standard culture conditions ES cells show only minimal canonical Wnt activity and this activity is not impacted by LGRS. Next, we investigated whether the Wnt/~-catenin axis is intact in ES cells and capable of activation in the presence of exogenous ligands. In particular, we hypothesized that cells with high levels of endogenous LGR5 expression might be susceptible to 47 RSPO-mediated potentiation of Wnt/~-catenin signaling. To address this we first evaluated subcellular localization of ~-catenin in ES cells before and after exposure to exogenous Wnt3a. Consistent with a previous report (Uren et a!., 2004), ES cells demonstrated little evidence of ~-catenin in their nuclei under basal conditions (Figure 12A). In contrast, exposure ofES cells to Wnt3a conditioned media resulted in robust nuclear translocation of ~-catenin (Figures 12A and 12B). Indeed, Wnt3a induced nuclear translocation of ~-catenin in all three lines irrespective of their endogenous LGR5 expression levels (Figures 12A and 12B). Although there are four RSPOs, RSP02 has been shown to have the highest affinity for both LGR4 and LGRS (Carmon eta!., 2011). In addition, RSP02 is highly expressed in developing bone (Nam eta!., 2007;Hankenson eta!., 2010). Therefore, we next exposed ES cells to RSP02 either alone or in combination with Wnt3a and measured ~-catenin nuclear localization. The presence of RSP02 alone had no effect on sub-cellular localization of ~-catenin (Figures 13A and 13B). In contrast, the combination of Wnt3a conditioned media plus RSP02 resulted in potentiation of Wnt/~-catenin signaling in CHLA25 cells as demonstrated by further up regulation and nuclear translocation of ~-catenin (Figures 13A and 13B). We also measured the expression of the Wnt/~-catenin target gene AXIN2 in these four conditions. In keeping with increased nuclear localization of ~-catenin, we observed induction of AXIN2 expression following Wnt3a stimulation and, in CHLA25 cells, potentiation of this up-regulation in the presence of RSP02 (Figure 13A). Likewise, TCF reporter activity increased in a stepwise fashion when cells were exposed first to Wnt3a alone and then both Wnt3a and RSP02 together (Figure 13B). Notably, potentiation ofWnt/~ catenin signaling by RSP02 was only statistically significant in CHLA25 cells, the cell 48 line with the highest level of LGR5. To determine if LGRS was responsible for mediating this robust RSP02-dependent potentiation ofWnt/~-catenin signaling we evaluated the consequences of LGR5 knockdown on this signaling axis in CHLA25 cells. Significantly, in the context of Wnt3a and RSP02 ligands, knockdown of LGR5 resulted in reduced potentiation of Wnt/~-catenin transcriptional activity (Figure 13C) Interestingly, we also observed down-regulation of TCF reporter activity in Wnt3a-only treated cells (Figure 13C) suggesting that low-level endogenous production of RSPO by CHLA25 cells might also contribute to potentiation of the Wnt/~-catenin axis in a Wnt- rich microenvironment. 49 Figure 11. No correlation exists between LGR5 expression and Wnt/P-catenin activity in ES cells in standard culture. (A) Canonical Wnt/~-catenin target gene expression was measured by qRT -PCR in ES cell lines. There was no apparent correlation between LGR5 expression and Wnt/~-catenin target genes AXIN2 and CD44. (B) No consistent or significant change occurs in the expression of a panel ofWnt/~ catenin target genes following LGR5 knockdown. The dotted line represents a ratio of 1 between shLGR5#1 and shNS cells. (C) LGR5 knockdown does not measurably decrease Wnt/~-catenin transcriptional activity. Luciferase activity was determined using TCF reporter-transduced cell lines as described in Section "Material and methods." All data shown is from three independent experiments and error bars are SEM. (A) 12 (B)- s (f) z g? ~ 4 ~ .. 6 (Q 3 '00 (9 <n..J @ ..c 2 c."' x'Q L.LJ 2 1 _§_ LGRS 12 2 ~ <!' 1 c 0 -~ ~ ~ CHLA25 A673 TC7 1 AXIN2 CHLA25 A673 0 CHLA25 • A673 • TC71 TC71 (C) o. t Q) ~B ]; E 0. o.9 "'u LJ_Q) (:!~o E 0 E. 12 CD44 2 ~ ~ 1 c 0 -~ ~ st w CHLA25 A673 TC71 D shNS • shLGR5#1 0 .. 0( J..L..L-- 293A CHLA25 50 Figure 12. Exogenous Wnt3a induces nuclear localization of P-catenin and is potentiated by RSP02. (A) CHLA25 cells grown in the presence ofWnt3a conditioned medium (CM) induced nuclear localization of ~-catenin and this was robustly potentiated by the addition of RSP02. RSP02 alone had no effect on ~-catenin nuclear localization. (B) The percentage of CHLA25, A673, and TC71 cells with nuclear ~-catenin all increased with Wnt3a CM, but CHLA25 showed the greatest increase in nuclear localization with the addition of RSP02. Data from three independent experiments and error bars are SEM. (A) CHLA25 (B) 2 0 D L-CeiiCM D L-Cell CM + RSP02 a; ';' • Wnt3aCM • Wnt3a CM + RSP02 ...J N o._ 0 (fJ cr •p<.05 + CHLA25 A673 TC71 2 0 a; (.) _j 2 0 "' c::l c s N 0 o._ (fJ cr + 2 0 "' c::l c s 51 Figure 13. RSP02 potentiates Wnt/P-catenin signaling in LGR5-high ES cells. (A) Exposure ofES cell lines to Wnt3a conditioned medium (CM) increased AXIN2 expression, but expression was only potentiated by RSP02 in the LGR5-high cell line, CHLA25. (B) TCF reporter activity was induced in ES cells exposed to Wnt3a CM, but was only potentiated in CHLA25 cells by the addition ofRSP02. (C) TCF reporter activity was measured in control and LGR5 knockdown CHLA25 cells following exposure to Wnt3a with/without RSP02. Reporter activity in the presence ofRSP02 was significantly diminished in LGR5 knockdown cells. All data shown is from three independent experiments and error bars are SEM. (A) 1 -~ QJ() > ., - Q) <'~Y <-' ~:§ ~ (C) "1 200 -"'~ 150 ~ ~ 100 ~ ~ 50 u.s I-~ 1 CHLA25 A673 TC71 CHLA25 j Ill shNS shLGR5#1 D L-CeiiCM 0 L-Cell CM + RSP02 • Wnt3aCM • Wnt3a CM + RSP02 .. p<.005 ' p<.05 (B) sool ~ 600 ::2 ~~ 400 ,.,., "" ::'.--" 200 ()£ I-! 'p<.05 CHLA25 A673 TC71 52 LGR5 mediates Wnt/PCP signaling in ES cells Previous work has demonstrated that LGRS and/or the RSPO family of ligands mediate Wnt/PCP signaling in mammalian and non-mammalian cells (Glinka et a!., 2011;0hkawara eta!., 2011). We hypothesized LGRS could increase Wnt/PCP signaling in ES cells and investigated this possibility using cells grown under both standard culture conditions and in the presence of exogenous ligands. We used Western blot analysis to probe for phospho-JNK expression, a mediator ofWnt/PCP signaling, to assess basal levels of activation. We saw no difference in phospho-JNK levels in A673, CHLA25, and TC71 cell lines (Figure 14A). There are thought to be multiple ways Wnt/PCP signals can be transduced so we looked at a panel of important pathway mediators as part of a Wnt/PCP gene signature originally described for adenocarcinoma of the lung (Hu eta!., 2011 ). Again, we found that LGR5 knockdown did not change or consistently affect the expression of these mediators under basal conditions (Figure 14B). Although both the phospho-JNK and the Wnt/PCP gene signature results showed no change in Wnt/PCP signaling with LGR5 knockdown, these assays are measurements of components in the pathway and not measurements of downstream transcriptional activity. To directly measure Wnt/PCP-induced transcriptional activity, we focused on the gene CHOP. The promoter for this gene is part of an ATF2luciferase reporter construct that has been used for measuring Wnt/PCP signaling in Xenopus embryos (Ohkawara et a!., 2011;0hkawara and Niehrs, 2011). We reasoned that expression of this gene could be used in a similar fashion to measure Wnt/PCP signaling in ES cells. For cell lines grown under standard culture conditions, CHLA25 cells were the only ones to show a change in CHOP expression with LGR5 knockdown. These cells showed a decrease in a CHOP 53 levels, whereas the expression was unchanged in A673 and TC71 cells (Figure 14C). We then examined the effect on CHOP in CHLA25 cells grown in the presence of Wnt5a and/or RSP02. Both the non-silencing control cells and LGR5 knockdown cells showed a nearly 2-fold increase in CHOP levels when cultured with Wnt5a alone or in combination with RSP02 (Figure 14D). In the presence ofRSP02 alone, however, the non-silencing control cells showed an increase in CHOP expression, whereas LGR5 knockdown cells showed a trend towards little or no CHOP induction (Figure 14D). This finding indicates that Wnt/PCP may be dependent on LGR5 in CHLA25 cells grown under both basal conditions and in the presence of RSP02 alone. 54 Figure 14. The role of LGR5 in Wnt/PCP signaling in ES. (A) Western blot analysis shows there was no change in phospho-JNK (P-JNK) levels in ES cell lines with LGR5 knockdown grown under standard culture conditions. (B) There was no consistent or significant change in the expression of defined Wnt/PCP signature genes (Hu et al., 2011) with LGR5 knockdown. The dotted line represents a ratio of 1 between shLGR5#1 and shNS cells. (C) Only CHLA25 cells demonstrated a decrease in CHOP expression with LGR5 knockdown. (D) CHOP expression was measured in control and LGR5 knockdown CHLA25 cells following exposure to Wnt5a with/without RSP02. There was a trend towards reduced CHOP induction by RSP02 alone with decreased LGR5. All data shown is from two to three independent experiments and error bars are SEM. (A) uv shNS shLGR5#1 P-JNK Total JNK GAPDH 293 (D) + + + + + + + ~ .......... -.........-.,., .-.r-·~.4~ ... ~··- -J, ~ ..... . :I' - < .. '-> .. il' .., .. A673 CHLA25 D CH LA25 A673 (C) rn 2. • TC71 CHLA25 z .<::: ..@ 1 . ~~ ..9!o:: ~ s 1. iJ~ g 0. ~ 0. TC71 D L-ceiiCM D L-cell CM + RSP02 Wnt5aCM • Wnt5a CM + RSP02 shNS shLGR5#1 55 LGR5 does not promote ES cell proliferation in vitro LGR5 functions as a growth-promoting oncogene in several adult malignancies including gastrointestinal tumors and gliomas (McClanahan et a!., 2006; Tanese et a!., 2008;Nakata eta!., 2013). In addition, LGR5 was recently shown to promote proliferation of neuroblastoma (Balamuth eta!., 2010), another highly aggressive pediatric tumor of neural crest origin (Jiang eta!., 2011). In order to determine if LGR5 promotes ES cell proliferation, we generated ES cell lines with altered LGR5 expression using stable RNA interference-mediated knockdown or ectopic over-expression (Figures !SA and ISB). As shown, although reduced proliferation of the CHLA25 cell line was observed with one LGR5-targeted knockdown sequence (shLGRS#l), this phenotype was not observed in A673 cells (Figure ISC). In addition, a secondLGR5-targeted shRNA sequence (shLGR5#3) did not inhibit proliferation in any of the cell lines (Figure ISC). Forced over-expression of LGR5 in the LGR5- low cell line A673 also had no impact on proliferation (Figure lSD). Cellular viability was also measured by trypan blue staining and no change was observed in any of the cell lines following LGR5 knockdown or over expression. Therefore, in contrast to neuroblastoma cells (Balamuth eta!., 2010), LGR5 does not promote the proliferation of ES cells in standard in vitro culture conditions. Wnt signals produced by the tumor microenvironment promote growth of cancer or cancer cells (Clevers and Nusse, 2012) and we reasoned that proliferation downstream of W nt/~-catenin activation would only be apparent in ES cells exposed to the necessary ligands. Therefore, we repeated proliferation studies in non-transduced, parent cells in the presence of Wnt and RSPO, reasoning that changes in proliferation downstream of Wnt/~-catenin activation would only be apparent in cells that were exposed to the 56 necessary ligands. Interestingly, no change in proliferation was observed in either A673 or CHLA25 cells with the addition ofWNT3A and/or RSP02 (Figure 15E). Studies are ongoing to evaluate the phenotypic consequences of LGRS-mediated potentiation of Wnt/~-catenin signaling but these data support prior observations that this pathway does not promote ES cell proliferation (Endo eta!., 2008). 57 Figure 15. LGR5 does not promote ES cell proliferation. (A) qRT -PCR confirmation of LGR5 knockdown in ES cell lines transduced with LGR5-targeted shRNAs (shNS = non-silencing control). (B) Confirmation of LGR5 over-expression in A673 cells transduced with LGR5 over-expression construct (Empty, empty vector control). (C) ES cell expansion was measured by MTS assay and was not affected by LGR5 knockdown. (D) LGR5 over-expression did affect cell growth in A673 cells. (E) Parental cell growth was not affected by the addition ofWNT3A and/or RSP02 to standard culture conditions. All data shown is from two independent experiments and error bars are SEM. (A) CHLA25 A673 (B) '""] A673 iil ~ a. ~ E w .9 1. .9 200 ~ ~ a; a; > > ~ ~ "' "' C>: C>: C!l C!l ..J ..J ;:;."' ~" ~"' q:- ,;:? or:-"' 0 0 ""'v ""'v ~ ~ ;:;."' ~" ~"' """' q:- ,;:? or:-"' <v<$' 0 0 ""'v ""'v ~ ~ (C) 8 CHLA25 -- shNS A673 (D) A673 -- shNS --Empty 6 -a- shLGR5#1 -a- shLGR5#1 -B- LGRS -+- shLGR5#3 4 •p<.OS 0 0 4 Days Days Days (E) CHLA25 A673 --PBS --PBS -a- RSP02 -a- RSP02 6 -+- WNT3A 6 -+- WNT3A -e- WNT3A + RSP02 -e- WNT3A + RSP02 4 4 Days Days 58 LGR5 does not promote chemoresistance in ES cells in standard in vitro culture We previously showed LGR5 was increased in a primary drug resistance case of ES and that higher levels are associated with increased resistance to doxorubicin in ES cell lines (Figure 8). In order to understand the biologic basis of these findings, we explored whether LGR5 directly contributes to chemoresistance using in vitro drug sensitivity assays for ES cells grown under standard culture conditions. We utilized the isogenic cell lines CHLA9 and CHLAl 0 which have low and high LGR5 levels (Figure SA) and ICSO values for doxorubicin (Figure 8B), respectively. We first over-expressed LGR5 in CHLA9 cells and found no effect on the ICSO for doxorubicin compared to Empty vector control cells (Figure 16B). Similarly, we stably knocked down LGR5 in CHLAlO cells and only observed a slight decrease in the ICSO for doxorubicin (Figure 16D). We confirmed these results using CHLA25 cells withLGR5 knockdown exposed to several different chemotherapeutic agents. The ICSO values for treatment with doxorubicin, etoposide, and vincristine were slightly reduced or did not change with decreasedLGR5 (Figure 16E). These observations lead us to conclude thatLGR5 does not promote chemoresistance under standard culture conditions. 59 Figure 16. LGRS does not promote chemoresistance. (A) qRT-PCR analysis demonstrating LGR5 over-expression in stably transduced CHLA9 cells. (B) CHLA9 cells with increased LGR5 expression showed no change in sensitivity to doxorubicin compared to Empty vector control cells. (C) Confirmation of LGR5 knockdown in stably transduced CHLAlO cells. (D) CHLAlO cells with decreased LGR5 expression grown in the presence of increasing doses of doxorubicin showed minimal change in the IC50 compared to non-silencing control cells. (E) IC50 values for CHLA25 non-silencing control and LGR5 knockdown cells grown in the presence of several chemotherapeutic agents. The IC50 values showed little or no change based on decreased LGR5 expression. All data is from two to three independent experiments and error bars are SEM. (A) CHLA9 (B) 1.5 • Empty (IC50 = .23 nM) ~ D LGR5 (IC50 = .22 nM) c. E u w 2 1.0 .9 ?;>(I) ·- Q) :=:.)::> ~ .Oc (I) :::J Qi >_g 0.5 > ~ ..9:? l!) 0:: (!) ....J 0.0 Empty LGR5 -10 -8 -6 Log [Doxorubicin (nM)] (C) CHLA10 (D) 1.5 • shNS (IC50 = 8.03 nM) U) z D shLGR5#1 (IC50 = 5.61 nM) .r:. CfJ u .9 2 1.0 ?;>(I) ~ = ~ :oc Qi (I) :::J > >.9 ..9:? 0 ~ 0.5 l!) 0:: (!) ....J 0 0.0 shNS shLGR5#1 -10 -8 -6 Log [Doxorubicin (nM)] (E) CHLA25 Drug shNS shLGR5#1 Doxorubicin 2.61 nM 2.58 nM Etoposide 0.21 1-1M .15 1-1M Vincristine .62 nM .40nM 60 LGR5 promotes anchorage-independent growth in ES cells Anchorage-independent growth is considered a hallmark of the transformed cellular phenotype and ES cells have been shown to grow in spheroid aggregates with limited extracellular attachments in tumor sections (Lawlor eta!., 2002). These factors indicate that anchorage-independent growth is key pathologic feature of ES cells and we assessed the role of LGR5 in mediating this phenotype. We found the colony-forming ability of both CHLA25 and A673 cell lines was significantly impaired with LGR5 knockdown. (Figures 17 A and 17B). The decrease in colony formation, however, differed between cell lines. Both LGR5-targeted shRNA constructs decreased colony formation in CHLA25 cells, whereas only one construct (shLGR5#1) decreased colony formation in A673 cells (Figures 17 A and 17B). This finding may be due to the greater level of LGR5 in CHLA25 cells (Figure 3B) and greater dependence on its expression. 61 Figure 17. LGR5 promotes anchorage-independent growth. (A) Representative images of CHLA25 and A673 macroscopic colonies demonstrate that stable LGR5 knockdown impairs colony formation in soft agar. (B) Colonies were quantified after cells were grown in soft agar for an average of two weeks. Data is from two to four independent experiments and error bars are SEM. (A) (B) c: 0 0 u CHLA25 A673 shNS CHLA25 # shLGR5#1 Q) ~ .'!! c: 0 0 u shLGR5#3 A673 *p<.05 #p=.05 ***p<5x1 o- 5 shNS shLGR5#1 shLGR5#3 shNS shLGR5#1 shLGR5#3 62 LGR5 promotes chemotaxis in high expressing ES cells Tumor metastasis is a complex process involving the release of cells from the primary tumor site, invasion through the basement membrane, intravasion and extravasion, and adaptation to microenvironmental conditions encountered at the ectopic site. Part of this process can involve the directed movement of cells along a chemoattractant gradient (also known as chemotaxis) facilitating localization of tumor cells to the metastatic site. A previous report demonstrated that the LGRS homologue, LGR6, promotes this process in HeLa cells (Gong eta!., 2012) and we wanted to determine if LGRS itself could promote chemotaxis in ES. We measured chemotaxis using the xCelligence System and monitored ES cells migrating towards serum in real time. In particular, we measured chemotaxis in CHLA25 and A673 cells with stable LGR5 knockdown. We found thatLGR5 knockdown significantly reduced CHLA25 migration towards serum, but not in A673 cells (Figures 18A and 18B). We attempted to replicate these findings using two additional ES cell lines with differing levels of LGR5, but both cell lines showed limited migratory potential and did not provide a suitable comparison. The observed pro-chemotactic effect of LGR5 is thus far limited to LGR5- high cell lines. 63 Figure 18. LGR5 promotes chemotaxis in LGR5-high cells. (A) Migration towards serum was monitored in realtime over a 12-hour period for CHLA25 (top) and A673 (bottom) with the xCELLigence System. CHLA25 cells (LGR5-high cells) transduced with two different shRNA-LGR5 targeting constructs showed decreased migration compared to non-silencing control cells. In contrast, LGR5 knockdown did not affect migration in A673 cells (LGR5-1ow cells). (B) Cell index values were plotted for the 0, 6, and 12-hour time points for CHLA25 (left) and A673 (right) cells. Data is from two to four independent experiments and error bars are SEM. (A) CHLA25 .s - shNS ~ 0.5 - shLGRS#l - shLGR5#3 Time (in Hour) A673 - shNS .s ~0 .25 - shLGRS#l - shLGR5#3 Time (in Hour) (B) CHLA25 A673 D shNS 0 • shLGR5#1 0 X • shLGR5#3 X Q) Q) ""0 ""0 .!: :§ 0 Qi Qi (.) (.) 0 *p<.05 0 0 6 12 0 6 12 Hours Hours 64 Discussion In this study, for the first time, we have identified the existence of the LGRS-Wnt signaling axis in ES. Although LGRS was identified some years ago as a Wnt-responsive gene, its role as a potentiating receptor upstream ofWnt signaling was only recently discovered (Carmon eta!., 2011;de Lau eta!., 20ll;Glinka eta!., 2011;Gong eta!., 2012). LGRS and its related receptor family members LGR4 and LGR6 have now all been established as a receptors for the RSPO-family of ligands, which act to potentiate Wnt/B-catenin signaling by complexing with Frizzled/LRP receptors (Carmon eta!., 2011;de Lau eta!., 20ll;Glinka eta!., 2011;Gong eta!., 2012). Additionally, LGR4 and LGRS have been shown to mediate Wnt/PCP signaling (Glinka eta!., 2011). There is now substantial evidence that hijacking ofWnt signaling may be important for the initiation and maintenance of ES, although the precise contribution of B-catenin dependent and independent pathways remains to be elucidated (Uren et a!., 2004;Miyagawa eta!., 2009;Navarro eta!., 2010;Vijayakumar eta!., 2011;Hauer eta!., 2013). Significantly, Oren eta!. measured expression of the canonical Wnt ligands WNTJ, WNT2, and WNT3 and found them to be undetectable in ES. In addition, only low levels of B-catenin nuclear localization were detected in cell lines and tumor samples (Uren eta!., 2004). Likewise, Navarro eta!. reported low levels of basal TCF transcriptional activity in ES cell lines (Navarro eta!., 2010). Studies ofDKK1 and DKK2, known modulators ofWnt signaling, have demonstrated that while the former is repressed in ES (Navarro eta!., 2010) the latter is induced by EWS-FLI1 and promotes cellular invasion and metastasis (Miyagawa eta!., 2009;Hauer eta!., 2013). A recent report also showed that the non-canonical Wnt, Wnt5A, induces ES cell migration and 65 that the Wnt inhibitor SFRPS is silenced in ES cells by DNA promoter methylation (Jin eta!., 20 12). Thus, although there is still much to be understood about the role of Wnt signaling in ES pathogenesis, the data thus far support a more dominant role for Wnt deregulation in altering cell morphology, motility and invasion rather than classical B catenin-mediated cellular proliferation. Consistent with these prior reports we found that, in standard culture conditions, basal levels ofWnt/B-catenin signaling are low. In addition, despite the fact that Wnt3a and RSP02 were able to robustly activate and potentiate canonical Wnt signaling, we did not observe changes in cellular proliferation. Even the addition of RSP02 to the culture media of LGR5-high CHLA25 cells had no impact on cell proliferation. Likewise, LGR5 knockdown had no discernible impact on the expression ofWnt/B-catenin target gene expression nor TCF -reporter activity under basal conditions. However, when growth medium was supplemented with Wnt3a, B-catenin nuclear localization and TCF transcriptional activity were robustly activated. In ES cells that express LGR5 this signaling was further potentiated by the addition of RSP02 confirming a functional role for this receptor in cells that are exposed to LGRS ligand. Our phenotypic assays also revealed that LGR5 does not directly contribute to chemoresistance in the absence of exogenous ligands. We found that altering LGR5 levels in CHLA9, CHLlO, and CHLA25 cells did significantly change the resistance of these cells to various chemotherapeutic agents. Our previous results indicated that basal LGR5 levels are associated with higher resistance to doxorubicin in ES cell lines and that levels are increased upon doxorubicin treatment. This result is in agreement with a previous report demonstrating that LGR5 levels are increased in gastric carcinoma biopsy 66 samples after chemotherapeutic treatment (Bauer eta!., 2012). We conclude thatLGR5 acts as a marker of chemoresistance in ES, but does not directly promote this phenotype under standard culture conditions. The designation of LGR5 as a marker rather than an effector of chemoresistance may reflect other upstream signaling pathways that are directly involved in mediating chemoresistance and lead to an up-regulation of LGR5. In contrast to the effects of LGR5 on proliferation and chemoresistance, we observed that LGR5 promoted anchorage-independent growth and chemotaxis in ES cell lines under standard culture conditions. Furthermore, RSP02 induced expression of CHOP, a putative Wnt/PCP target gene in CHLA25 cells. There was a trend towards decreased induction, however, when LGR5 was knocked down. Expression was also decreased in CHLA25 cells under standard culture conditions, indicating some level of autocrine Wnt/PCP signaling in these cells. Together our results demonstrate that in the presence of appropriate ligands, Wnt/B-catenin and Wnt/PCP signaling can be activated in ES cells and this activation is maximal in cells that express high levels of LGR5. The phenotypic effects of this increased Wnt activity remain unclear but our studies thus far suggest that proliferation is unaffected, whereas LGR5 promotes anchorage-independent growth and chemotaxis. We believe that both cell autonomous as well microenvironmental factors combine to determine the functional consequences of enhanced Wnt signaling in LGRS+ ES cells. Emerging evidence from other model systems supports the possibility that LGRS function might be highly contextually dependent. Mouse studies have shown that Lgr5 is widely expressed during embryogenesis but expression becomes limited to discrete stem cell populations post-natally, in particular stem cells of hair follicles and the 67 gastrointestinal tract (Barker eta!., 2007;Barker and Clevers, 2010). In normal stem cells secretion of RSPO from neighboring cells supports Lgr5-dependent self-renewal and proliferation by activating canonical Wnt signaling (Barker and Clevers, 2010). Likewise, the initiation and proliferation of intestinal adenomas and carcinomas is driven by Lgr5+ stem cells (Barker eta!., 2009;Schepers eta!., 2012). However, data from both colorectal carcinoma cell lines as well as other tumor types complicate the scenario. Consistent with our own data, Walker eta!. found that LGRS did not affect proliferation of colorectal cancer cell lines but instead negatively regulated Wnt-signaling, colony formation, and migration (Walker eta!., 2011). Thus, in this cellular context LGRS functioned more as a tumor suppressor than tumor promoter. In addition, promoter hypermethylation and loss of function mutations in LGR5 and LGR6 have been discovered in some colorectal tumor samples, again suggesting that these genes could act as tumor suppressors in some contexts (Sjoblom eta!., 2006;Chan eta!., 2008;de Sousa eta!., 2011). These conflicting reports demonstrate that LGR5 has differing roles during development and in different cellular contexts. Thus, its contribution to cancer maintenance and progression is likely to be determined by both the genetic and epigenetic state of the affected cell as well as the surrounding microenvironment. In summary, we have shown that LGR5 is expressed by ES, in particular by putative cancer stem cells and, in the context of a Wnt and RSPO-rich microenvironment, LGRS functions to potentiate canonical Wnt/B-catenin signaling. Given the profound complexity ofWnt signaling and its dependence on both cell autonomous as well as micro environmental cues it is now essential that functional studies of LGRS and the 68 Wnt/B-catenin axis in ES be performed in model systems that faithfully recapitulate the in vivo tumor microenvironment. 69 ChapterS Basic science and clinical implications ofLGR5 expression in ES This thesis represents the culmination of an effort to identify new molecular prognostic biomarkers in ES. To achieve this aim, we used a genome-wide microarray screening approach to assess gene expression changes in a primary drug resistant case before and after treatment. This approach allowed us to control for genetic differences between patients and to identify gene expression changes that were specific to tumor progression. We identified the stem cell marker LGR5 as a gene of interest and proceeded to investigate its role in the origin and development of ES. We found LGR5 is highly expressed in neural crest stem cell-derived populations, a putative cell of origin. It is also up-regulated in response to extracellular stressors such as low serum and chemotherapeutic agents and is expressed by discrete subpopulations of ES cells, including those enriched for CSC-like properties. Finally, we found thatLGR5 contributes to the malignant phenotype and enhances both canonical Wnt/B-catenin and non-canonical Wnt/PCP signal transduction in ES cells. These findings raise new questions about the involvement Wnt signaling in the pathogenesis of ES and highlight the prospect of using LGR5 expression as the basis of future prognostic and therapeutic app li cations. Our results suggest a role for Wnt signaling in the initiation and maintenance of the malignant phenotype in ES. In particular, we found that LGRS promotes anchorage independent growth and chemotaxis and that LGR5 knockdown decreases activation of Wnt/B-catenin and Wnt/PCP signaling. Both Wnt/B-catenin and Wnt/PCP pathways have been shown to contribute to the malignant phenotype in various epithelial cancers (Katoh, 70 2005;Clevers and Nusse, 2012) and we propose these also pathways promote aggressive features in ES cells downstream ofLGRS stimulation. The Wnt and RSPO-ligands that activate these pathways may be produced by tumor cells or secreted by non-malignant cells in the tumor microenvironment. Based on the low level of endogenous Wnt/B catenin signaling in ES cells, it is likely that the activating ligands are provided by the tumor microenvironment. In contrast, ES cells demonstrated high levels of endogenous Wnt/PCP signaling and the activating ligands may be secreted in an autocrine fashion. Wnt signaling is highly contextual and the interplay between canonical and non canonical Wnt signals may determine what phenotypic properties ES cells adopt. For example, samples derived from colorectal cancer patients show a higher level ofWnt/B catenin target gene expression in pre-malignant lesions but then a down-regulation in progressive disease (de Sousa et a!., 2011 ). Likewise, blockade of TCF4, a transcription factor important in activating Wnt/B-catenin target gene expression, leads to a decrease in metastatic spread of colorectal cancer cells (V arnat et a!., 201 0). Canonical and non canonical Wnt signaling can also be active at different times and mediate a phenotypic switch in colorectal cancer (Bordonaro eta!., 2011). This model would fit with the phenotypic data we observed for ES cell lines grown under basal conditions or in the presence of exogenous ligands. The decreases in anchorage-independent growth and chemotaxis with LGR5 knockdown, both of which are important in tumor spread, were observed under basal conditions and these findings could reflect a state where ES cells are demonstrating a predominantly Wnt/PCP signaling pattern. Wnt/B-catenin signaling, on the other hand, was only stimulated in the presence of exogenous ligands and any 71 phenotypic effects this pathway has may only be seen in a system that accurately models the tumor microenvironment. The putative neural crest origin of ES and our finding that neural crest-derived stem cell populations express LGR5 indicate that Wnt/B-catenin signaling may be important in the early development of ES. Previous work demonstrated the importance of Wnt/B-catenin at all stages of neural crest development, including during induction and delamination when NCSC first arise and migrate to their final destinations (Schmidt and Patel, 2005). Given the high level of LGR5 expression in NCSC, exposure to RSPO would likely potentiate Wnt/B-catenin signaling and maintain the undifferentiated, epigenetic state of these cells. The epigenetic state ofNCSC is thought to be critical for tolerance of EWS-FLI1 (von Levetzow eta!., 2011) and thus, Wnt/B-catenin signaling may facilitate malignant transformation. Once NCSC become transformed, however, Wnt/B-catenin signaling may become attenuated as evident by the absence of endogenous signaling in both ES tumors and cell lines (Uren eta!., 2004). Navarro eta!. has demonstrated that EWS-FLI1 binds LEF1 and inhibits Wnt/B-catenin transcriptional activity (Navarro eta!., 2010). The blockade ofWnt/B-catenin signaling by EWS-FLI1 may be overcome in an RSPO-rich microenvironment when peak signaling activity is achieved. We therefore propose a model in which Wnt/B-catenin contributes to tumor initiation and then becomes repressed after NCSC are transformed by EWS-FLI1 (Figure 19). This pathway may become active again when ES cells are exposed to RSPO and promote NCSC-associated features. 72 Figure 19. Proposed trechanism ofWntfll-cateninsignaling dwing initiation and progt-ession ofES tw\lllrs. Neural crest stem cells (NCSC) are a putative cell of origin for ES. NC SC have high levels ofWntl~·catenin signaling during development and signal activation may facilitate EWS·FLII·driven malignant transfonnation. Once transformed, EWS·FLII represses endogenous Wntl~·catenin signaling. Repression may be relieved in ES cells, however, in a Wntand RSPO·rich microenvironment that could lead to as of yet undefined phenotypic effects. EW&FU1 NCSC Wntl~·catenin•~• ES cells Wntl~·catenin"" ES cells Wntl~·catenin 1 ~ 1 r-...., ,..., Phenotypic c:;;: r effects??? 73 In addition to the biologic implications of this work, our results suggest that LGRS may have prognostic and therapeutic applications for ES. We demonstrated that LGR5 is a potential prognostic biomarker and future work in the lab will be directed at firmly establishing this finding. We will need to validate our earlier results using an independent cohort of ES biopsy samples and then show LGR5 is predictive of outcome in a prospective clinical trial. This work could lead to biopsy or possibly less invasive, circulating tumor cell-based assays for assessing patient risk at diagnosis and assigning patients to treatment protocols that reflect the likelihood of response. The high expression of LGR5 in a subset of tumor samples also makes it promising target for the development of new biologic-based therapies. Several academic and biotechnology groups have developed monoclonal antibodies that specifically target LGRS (Kemper et a!., 2012;Kobayashi eta!., 2012). A group from Genentech, Inc. has further refined this technology by covalently attaching chemotherapeutic agents to anti-LGRS monoclonal antibodies (i.e. antibody drug conjugates or ADC) (Mao, 201 0) They found good in vivo efficacy using several animal models and a single dose of the anti-LGRS ADC. Small molecule inhibitors of Wnt signaling may be another way of targeting LGRS+ cells. Drugs targeting both canonical Wnt and non-canonical Wnt signal transduction proteins are in various stages development (Anastas and Moon, 2013) and could eventually be used for clinical trials in ES patients. The combination of better risk-stratification and biologically-targeted therapies based on LGRS expression in ES patients offers the prospect of more efficacious treatments that increase survival with less toxic long-term side effects. 74 Tumor samples and cell lines Chapter6 Materials and methods Primary tumor RNA was obtained from Children's Hospital Los Angeles (CHLA) and Children's Oncology Group (COG) tumor biorepositories. All primary tissue samples were coded according to an anonymous numbering scheme and acquired in accordance with approval from the CHLA Committee for Clinical Investigation. Primary tumor expression data were kindly provided by the COG (Lawlor E, unpublished data). ES cell lines were kindly provided by Dr. Timothy Triche (CHLA, Los Angeles, CA), Dr. Heinrich Kovar (CCRI, St. Anna Kinderkrebsforschung, Vienna, Austria) and the COG cell bank ( cogcell.org) and identities confirmed by short tandem repeat profiling courtesy of Dr. Pat Reynolds (Texas Tech University, Lubbock, TX). AllES cell lines were maintained in RPMI with L-glutamine and 10% FBS on tissue culture treated polystyrene plates, except for CHLA25 and STA-ET-8.2 cells. These cell lines were grown on plates coated with fibronectin. Human bone marrow mesenchymal stem cell lines (MSC) (obtained from Dr. Darwin Prockop, Tulane University, New Orleans, LA), HI and H9 human embryonic stem cell (ESC) lines (Wicell, Madison, WI) and ESC-derived neural crest cells were cultured and differentiated using standard protocols as previously described by our lab (Jiang et al., 2009;von Levetzow et al., 20ll). MRCS fibroblasts were obtained from American Type Culture Collection and maintained in DMEM with L glutamine and 10% FBS. 75 Microarray analysis Samples were prepared in the CHLA genome core using the GeneChip whole transcript sense target labeling assay manual (Affymetrix; Santa Clara, CA) and hybridization to HuEx 1.0 ST microarrays was carried out following the manufacturer's instructions. Cell intensity file data were quantile-normalized and summarized using the iterative probe logarithmic intensity error estimation (iterPLIER) within the Affymetrix expression console software. Differentially expressed genes were defined as genes with a 2-fold or greater difference in expression between drug-resistant and drug-sensitive tumors. Network Analysis of the 130 differentially expressed genes was performed using Ingenutiy Pathway Analysis software. Quantitative real-time reverse transcription PCR eDNA was generated from RNA (iScript; Bio-Rad, Hercules, CA) and qRT-PCR was performed with validated TaqMan primers (LGR5, CJTED2, CYCLINDI, MYC, GAP DH, and ISS; Life Technologies, Grand Island, NY) or primers designed using qPrimerDepot (primerdepot.nci.nih.gov) (Table 2) (Cui eta!., 2007). Assays were performed in triplicate using the Applied Biosystems 7900HT Fast realtime PCR system or Roche Light Cycler 480 and average Ct values were normalized relative to GAPDH or ISS expression in the same sample. 76 Table 2. SYBR primer sequences Gene Fotward primer sequence Reverse primer sequence ISS GCAATTATTCCCCATGAACGA GGCCTCACTAAACCATCCAAT AXIN2 AAGTGCAAACTTTCGCCAAC ACAGGATCGCTCCTCTTGAA CD44 GACAAGTTTTGGTGGCACG CACGTGGAATACACCTGCAA CDC25A CCAGCCCCAAAGAGTCAAC AAGGTCCCTTGGGTCATTGT CELSR2 GACAACAACCGGCCTCTG TCTCATCGGTGATGATGGTC CHOP CCAAAATCAGAGCTGGAACC CCATCTCTGCAGTTGGATCA DVL3 ACTTTAAGGGCGTTTTGCAG AAGCATGGTAGCTTGGCATT EPHB2 CTCTACTGTAACGGGGACGG CCTTGAAAGTCCCAGATGGA FZD7 CGCCTCTGTTCGTCTACCTC GTCGTGTTTCATGATGGTGC GAPDH AAGGTGAAGGTCGGAGTCAA AATGAAGGGGTCATTGATGG HIG2 CTTCTGCGCTGGTGCTTAGT GCAGAGAAACAGAGCTGCCT ID2 GACAGCAAAGCACTGTGTGG TCAGCACTTAAAAGATTCCGTG LEFJ TGGATCTCTTTCTCCACCCA CACTGTAAGTGATGAGGGGG RYK CTCTACCTGAGCGAGGACGA CAGACTAAAGGATAGAGCGTAGTGA TROY CCCTCCTCCTCCTTACGAAC CCAGAGCGCTGCAGATAAC VANGL2 ACCGCTCTAAGAGTCGAGATG GTTACTACTGTCGTCGTTTCCC WNT5A TCGACTATGGCTACCGCTTTG CACTCTCGTAGGAGCCCTTG 77 Immuno-detection assays For immunocytochemistry, cells were grown on fibronectin-coated chamber slides (Thermo Fisher Scientific, Waltham, MA), fixed in 4% paraformaldehyde and permeabilized with 0.1% Triton X-100. After rinsing in PBS the slides were incubated for one hour in blocking solution (1% BSA and 5% donkey serum in PBS). Blocked slides were incubated with primary ~-catenin antibody (1 :300; BD Biosciences, San Jose, CA) overnight at 4°C, washed in PBS, incubated with Alexa Fluor 488-conjugated, donkey-anti mouse secondary antibody (1:500; Life Technologies) for one hour, and then visualized with a Leica DMI6000 B Fluorescence Microscope. Nuclei were counterstained with 4', 6-diamidino-2- phenylindole (DAPI). Cytoplasmic and nuclear~ catenin staining cells were manually counted in a minimum of three high-power fields. For Western blot analysis, whole celllysates were separated by SDS-PAGE followed by transfer onto a PVDF membrane. The membrane was probed with the following primary antibodies: JNK (1:1000; Cell Signaling; Danvers, MA), phosho-JNK (1:1000; Cell Signaling), and GAPDH (1:5000; Cell Signaling). Next, the membrane was incubated with an anti-rabbit HRP-conjugated antibody (1:5000; Thermo Fischer Scientific) and immersed with ECL and visualized on X-ray films (Kodak; Rochester, NY). Stress condition assays All cells were plated at a density of 4.0 x 10 5 cells/well and cultured in medium consisting ofRPMI and L-glutamine with or without 10% FBS. To mimic nutrient starvation, cells were grown without FBS. Cells grown under low attachment conditions 78 were seeded onto ultra low-attachment plates (Thermo Fisher Scientific). To simulate hypoxia, cells were grown in a Biospherix Hypoxia System chamber at 1% oxygen. RNA was harvested after 48 hours of stress exposure for qRT-PCR analysis. Drug sensitivity assays Cell growth and viability following drug treatment was assessed using the Cell Titer 96 AQueous One Solution Cell Proliferation Assay (Promega, Madison, WI). Cells were plated at a density of2.0 x 10 4 cells/well in 8-well replicated in 96 flat bottomed plates, allowed to attach overnight, and then chemotherapeutic agents were added at increasing concentrations. Survival of cells was assessed 72-96 hrs post treatment with Doxorubicin (0-1 11M), Etoposide (0 - 30 11M), or Vincristine (0 - 100 nM) (Calbiochem, San Diego, CA). The normalized cell index was calculated relative to the absorbance of untreated cells minus background. RNA was also collected from these cells from these cells for qRT-PCR analysis. Cell sorting and flow cytometry For CD133 sorts, cultured STA-ET-8.2 cells were trypsinized and resuspended in FeR Blocking Reagent (Miltenyi Biotec, Auburn, CA) with 0.5% bovine serum albumin (Sigma, StLouis, MO) in phosphate-buffered saline (PBS). Mouse anti-human CD133/2- PE (Miltenyi) monoclonal antibody was then added (1: 11 dilution) and incubated for 15 minutes at 4°C in the dark. After two washes, cells were sorted on a MoFlo Astrios instrument (Beckman Coulter) at the University of Michigan Flow Cytometry Core. Positive and negative gates were determined using IgG stained and unstained controls. 79 TC71 and MHH-ES cells were sorted on the basis of aldehyde dehydrogenase activity using the Aldefluor® assay (Stem Cell Technologies, Vancouver, BC). Cells were analyzed and sorted as previously described (A wad et al., 2010). Stably transfected CHLA25 cells were sorted on the basis of LGR5-GFP reporter expression (GeneCopoeia; Rockville, MD). Cells were first selected in puromycin (lflg/ml) and later sorted into GFP!ow and GFPhhigh populations. Positive and negative gates were set on the basis of transfected and non-transfected parent cells, respectively. For flow cytometry experiments, CHLA25 and A673 cells were washed in FeR Blocking Reagent with 0. 5% bovine serum albumin in PBS after which they were stained with an anti-LGR5 antibody ( 4Dll; BD Biosciences) at a concentration of 1 11g/ml for 20 min at 4° C. Cell were then incubated with an Alexa Fluor® 594 goat anti-rat IgG (Life Technologies), diluted 1:500, for 20 min at 4° C. Labeled cells were then analyzed using an Accuri C6 Flow Cytometer. Positive and negative gates were determined using anti LGR5 and isotype rat IgG2b antibodies, respectively. In situ hybridization (ISH) ISH for LGR5 mRNA was performed using the RNAscope kits (Advanced Cell Diagnostics; Hayward, CA) for formalin-fixed, paraffin-embedded tissue and formalin fixed cultured adherent cells according to the manufacturer's instructions for detection of HRP conjugated probes. Briefly, 5 11m thick TMA sections or adherent cells were treated with heat and protease digestion followed by hybridization with target probes to LGR5, the housekeeping gene Peptidyl-prolyl cis-trans isomerase B (PPIB) as a positive control, 80 or the bacterial gene DapB as a negative control. Probe-specific hybridization signals were detected with DAB. Generation of stable knockdown and over-expression cell lines For knockdown studies, cell lines were transduced with pLK0.1-puro lentiviral short hairpin RNAs vectors targetingLGR5 expression (shLGR5# 1- TAAGTGCCAGAACTGCTATGG and shLGR5#3 - TTGTCCAAATTGCAGTAGAGC; Thermo Fisher Scientific). Control cells were transduced with a non-silencing small hairpin RNA vector containing an inert sequence (shNS- CAACAAGATGAAGAGCACCAA). Cells were selected in puromycin (1-2 11g/ml) for at least 48 hours before use in subsequent experiments. For over-expression studies, cells were transduced with one of two expression constructs. A673 cells were transduced with a lentiviral pCL6 expression vector into which the full length open reading frame for LGR5 eDNA (NP _003658; Thermo Scientific, Clone ID BC09624) was cloned. The pCL6 vector is a modified version of the pCL1 vector including an internal ribosome entry site and woodchuck hepatitis virus post-transcription regulatory element up- and downstream, respectively, ofthe EGFP reporter gene (Feldhahn et al., 2007;von Levetzow et al., 2011). Control cells were transduced with empty vector that lacked the LGR5 sequence (Empty). Cells were selected by F ACS on the basis of GFP expression. CHLA9 cells were transduced with the pLentiLox RSV-puro vector (University of Michigan Vector Core) containing the mature human form of LGR5 (amino acids 21-907) fused at theN-terminus with sequence encoding a Myc tag. Control cells were transduced with empty vector (Empty). Cells were selected in puromycin as 81 described above. Lentiviral supernatant for each construct was generated from plasmid transfected 293FT packaging cells as previously described (von Levetzow eta!, 2011). Wnt3a and Wnt5a conditioned medium Conditioned medium (CM) was collected from control mouse fibroblasts (L-cells) or L-cells that have been engineered to over-express Wnt3a (L-Wnt3a-cells) or Wnt5a (L-Wnt5a-cells) obtained from the American Type Culture Collection (A TCC CRL-2648, CRL-2647, and CRL-2814, respectively). The final dilution ofL-cell CM, Wnt3a CM, and Wnt5a CM was I :2 in RPMI containing 5% FBS. Recombinant human RSP02 (R&D Systems, Minneapolis, MN) was added to CM at a concentration of 20 ng/mL. TCF reporter assay To assess the transcriptional activity of the Wnt/~-catenin axis we used previously characterized TCF-promoter luciferase-reporter constructs (Fuerer and Nusse, 2010). The 7xTcf-FFluc II SV40-mCherry (p7TFC) and 7xTcf-FFluc II SV40-PuroR (p7TFP) plasmids were purchased from Addgene (plasmids 24307 and 24308, respectively). Cells were sorted for mCherry expression (p7TFC) by F ACS or cultured with puromycin (p7TFP) to select for cells that were successfully transduced. To assess the effects of exogenous Wnt3a ligand onES cells, p7TFC and p7TFP-transduced cells were plated at a density of 4.0xl0 4 cells/well on 96-well plates and allowed to attach overnight. The cells were cultured for another 24 hours in standard RPMI with I 0% FBS or treated with L- cell CM or Wnt3a CM. Luciferase measurements were carried out using the Luciferase Assay System (Promega) according to the manufacturer's protocol and read on a 82 Molecular Devices LMAX plate reader. Measurements were normalized to mCherry mean fluorescent intensity for cells transduced with the p7TFC vector, otherwise TCF reporter activity was expressed relative to the L-ee!! CM control. Cell proliferation assays Cell growth was assessed using the Cell Titer 96 AQueous One Solution Cell Proliferation Assay (Promega). Cells were plated at a density of 5 x 10 3 cells/well in 8- well replicates on 96-well flat-bottomed plates, allowed to attach overnight, and growth was assessed on Days 0, 2 and 4 post-attachment. For cells grown in the presence of exogenous Wnt ligands, WNT3A (R&D Systems) and/or RSP02 (R&D Systems) were supplemented on Days 0 and 3 at 500 ng/mL and 20 ng/mL, respectively. The normalized cell index was calculated relative to the Day 0 absorbance reading minus background. Soft agar assays For the study of anchorage-independent growth, transduced cells were plated as single cell suspensions in 0.35% noble agar (Life Technologies) in Iscove's medium supplemented with 20% fetal bovine serum and 2 mmol/L of L-glutarnine. The cellular layer was placed on top of a feeder layer consisting of 0. 7% noble agar in Is cove's medium supplemented with 10% fetal bovine serum and 2 mmol/L of L-glutamine. cells were seeded at a density of 5 x 10 3 to 8 x 10 3 cells/well, respectively, in 3-well replicates on 6-well plates and allowed to form macroscopic colonies between 12 and 16 days. Colonies were then stained with .005% crystal violet for 24 hours and then photographed and manually counted. 83 xCELLligence chemotaxis assays Migration was measured using the xCelligence System (Roche) for real-time cell analysis. These assays were performed using CIM-plates modeled after a Boyden chamber apparatus. Media containing I 0% FBS was added to the lower chambers of a CIM-plate pre-coated with fibronectin. The upper chamber was then assembled on top of the lower chamber and serum free media was added to the wells. After I hour of equilibration with media, a cell suspension (8.0xl0 4 cells/well in serum free media) was added to 4-well replicates. Cell migration to the lower chamber was monitored for 12 hours and expressed as cell index. Statistical analysis Unless otherwise indicated data from all experiments are expressed as mean± SEM from a minimum of 3 independent experiments. The data was analyzed using GraphPad Prism software by a Student's t test and a p value of <0.05 was considered significant. 84 Bibliography Aguilera, C., Nakagawa, K., Sancho, R., Chakraborty, A., Hendrich, B., and Behrens, A. (2011). c-Jun N-terminal phosphorylation antagonises recruitment of the Mbd3/NuRD repressor complex. Nature 469, 231-235. Anastas, J.N., and Moon, R.T. (2013). WNT signalling pathways as therapeutic targets in cancer. Nat Rev Cancer 13, 11-26. Atkinson, A.J., Colburn, W.A., Degruttola, V.G., Demets, D.L., Downing, G.J., Hoth, D.F., Oates, J.A., Peck, C.C., Schooley, R.T., Spilker, B.A., Woodcock, J., and Zeger, S.L. (2001). Biomarkers and surrogate endpoints: preferred definitions and conceptual framework. Clin Pharmacal Ther 69, 89-95. A wad, 0., Yustein, J.T., Shah, P., Gul, N., Katuri, V., O'neill, A., Kong, Y., Brown, M.L., Toretsky, J.A., and Loeb, D.M. (2010). High ALDH activity identifies chemotherapy-resistant Ewing's sarcoma stem cells that retain sensitivity to EWS FLil inhibition. PLoS One 5, e13943. Balamuth, N.J., and Womer, R.B. (2010). Ewing's sarcoma. The lancet oncology 11, 184- 192. Balamuth, N.J., Wood, A., Wang, Q., Jagannathan, J., Mayes, P., Zhang, Z., Chen, Z., Rappaport, E., Courtright, J., Pawel, B., Weber, B., Wooster, R., Sekyere, E.O., Marshall, G.M., and Maris, J.M. (2010). Serial transcriptome analysis and cross species integration identifies centromere-associated protein E as a novel neuroblastoma target. Cancer Res 70, 2749-2758. Barker, N., and Clevers, H. (2010). Leucine-rich repeat-containing G-protein-coupled receptors as markers of adult stem cells. Gastroenterology 138, 1681-1696. Barker, N., Huch, M., Kujala, P., Van De Wetering, M., Snippert, H.J., VanEs, J.H., Sato, T., Stange, D.E., Begthel, H., Van Den Born, M., Danenberg, E., Van Den Brink, S., Korving, J., Abo, A., Peters, P.J., Wright, N., Poulsom, R., and Clevers, H. (2010). Lgr5(+ve) stem cells drive self-renewal in the stomach and build long lived gastric units in vitro. Cell Stem Cell6, 25-36. Barker, N., Ridgway, R.A., VanEs, J.H., Van De Wetering, M., Begthel, H., Van Den Born, M., Danenberg, E., Clarke, A.R., Sansom, O.J., and Clevers, H. (2009). Crypt stem cells as the cells-of-origin of intestinal cancer. Nature 457, 608-611. Barker, N., Rookmaaker, M.B., Kujala, P., Ng, A., Leushacke, M., Snippert, H., Van De Wetering, M., Tan, S., VanEs, J.H., Huch, M., Poulsom, R., Verhaar, M.C., Peters, P.J., and Clevers, H. (2012). Lgr5(+ve) Stem/Progenitor Cells Contribute to Nephron Formation during Kidney Development. Cell reports 2, 540-552. 85 Barker, N., VanEs, J.H., Kuipers, J., Kujala, P., Van Den Born, M., Cozijnsen, M., Haegebarth, A., Korving, J., Begthel, H., Peters, P.J., and Clevers, H. (2007). Identification of stem cells in small intestine and colon by marker gene Lgr5. Nature 449, 1003-1007. Barretina, J., Taylor, B.S., Banerji, S., Ramos, A.H., Lagos-Quintana, M., Decarolis, P.L., Shah, K., Socci, N.D., Weir, B.A., Ho, A., Chiang, D.Y., Reva, B., Mermel, C. H., Getz, G., Antipin, Y., Beroukhim, R., Major, J.E., Hatton, C., Nicoletti, R., Hanna, M., Sharpe, T., Fennell, T.J., Cibulskis, K., Onofrio, R.C., Saito, T., Shukla, N., Lau, C., Nelander, S., Silver, S.J., Sougnez, C., Viale, A., Winckler, W., Maki, R.G., Garraway, L.A., Lash, A., Greulich, H., Root, D. E., Sellers, W.R., Schwartz, G.K., Antonescu, C.R., Lander, E.S., Vannus, H.E., Ladanyi, M., Sander, C., Meyerson, M., and Singer, S. (2010). Subtype-specific genomic alterations define new targets for soft-tissue sarcoma therapy. Nat Genet 42, 715- 721. Bauer, L., Langer, R., Becker, K., Hapfelmeier, A., Ott, K., Novotny, A., Hofler, H., and Keller, G. (2012). Expression profiling of stem cell-related genes in neoadjuvant treated gastric cancer: a NOTCH2, GSK3B and beta-catenin gene signature predicts survival. PLoS One 7, e44566. Becker, L., Huang, Q., and Mashimo, H. (2010). Lgr5, an intestinal stem cell marker, is abnormally expressed in Barrett's esophagus and esophageal adenocarcinoma. Dis Esophagus 23, 168-174. Becker, M.W., and Jordan, C.T. (2011). Leukemia stem cells in 2010: current understanding and future directions. Blood Rev 25, 75-81. Ben-Porath, I., Thomson, M.W., Carey, V.J., Ge, R., Bell, G.W., Regev, A., and Weinberg, R.A. (2008). An embryonic stem cell-like gene expression signature in poorly differentiated aggressive human tumors. Nat Genet 40, 499-507. Bordonaro, M., Tewari, S., Cicco, C.E., Atamna, W., and Lazarova, D.L. (2011). A switch from canonical to noncanonical Wnt signaling mediates drug resistance in colon cancer cells. P LoS One 6, e27308. Carmon, K.S., Gong, X., Lin, Q., Thomas, A., and Liu, Q. (2011). R-spondins function as ligands of the orphan receptors LGR4 and LGR5 to regulate Wnt/beta-catenin signaling. Proc Natl Acad Sci US A 108, 11452-11457. Chaffer, C.L., and Weinberg, R.A. (2011). A perspective on cancer cell metastasis. Science 331, 1559-1564. Chan, T.A., Glockner, S., Yi, J.M., Chen, W., Van Neste, L., Cope, L., Herman, J.G., Velculescu, V., Schuebel, K.E., Ahuja, N., and Baylin, S.B. (2008). Convergence 86 of mutation and epigenetic alterations identifies common genes in cancer that predict for poor prognosis. PLoS M ed 5, e 114. Clevers, H., and Nusse, R. (2012). Wnt/beta-catenin signaling and disease. Cell149, ll92-1205. Cui, W., Taub, D.D., and Gardner, K. (2007). qPrimerDepot: a primer database for quantitative real time PCR. Nucleic Acids Res 35, D805-809. De Alava, E., Kawai, A., Healey, J.H., Fligman, I., Meyers, P.A., Huvos, A. G., Gerald, W.L., Jhanwar, S.C., Argani, P., Antonescu, C.R., Pardo-Mindan, F.J., Ginsberg, J., Womer, R., Lawlor, E.R., Wunder, J., Andrulis, I., Sorensen, P.H., Barr, F.G., and Ladanyi, M. (1998). EWS-FLI1 fusion transcript structure is an independent determinant of prognosis in Ewing's sarcoma. J Clin Onco/16, 1248-1255. De Lau, W., Barker, N., Low, T.Y., Koo, B.K., Li, V.S., Teunissen, H., Kujala, P., Haegebarth, A., Peters, P.J., Van De Wetering, M., Stange, D.E., VanEs, J.E., Guardavaccaro, D., Schasfoort, R.B., Mohri, Y., Nishimori, K., Mohammed, S., Heck, A.J., and Clevers, H. (2011). Lgr5 homologues associate with Wnt receptors and mediate R-spondin signalling. Nature 476, 293-297. De Lau, W.B., Snel, B., and Clevers, H. C. (2012). The R-spondin protein family. Genome Bio/13, 242. De Sousa, E.M.F., Colak, S., Buikhuisen, J., Koster, J., Cameron, K., DeJong, J.H., Tuynman, J.B., Prasetyanti, P.R., Fessler, E., Van Den Bergh, S.P., Rodermond, H., Dekker, E., VanDer Loos, C.M., Pals, S.T., Van De Vijver, M.J., Versteeg, R., Riche!, D.J., Vermeulen, L., and Medema, J.P. (2011). Methylation of cancer stem-cell-associated Wnt target genes predicts poor prognosis in colorectal cancer patients. Cell Stem Cell9, 476-485. Delattre, 0., Zucman, J., Plougastel, B., Desmaze, C., Melot, T., Peter, M., Kovar, H., Joubert, I., DeJong, P., Rouleau, G., and Et AI. (1992). Gene fusion with an ETS DNA-binding domain caused by chromosome translocation in human tumours. Nature 359, 162-165. Deneen, B., and Denny, C.T. (2001). Loss ofp16 pathways stabilizes EWS/FLI1 expression and complements EWS/FLI1 mediated transformation. Oncogene 20, 6731-6741. Douglas, D., Hsu, J.H., Hung, L., Cooper, A., Abdueva, D., Van Doorninck, J., Peng, G., Shimada, H., Triche, T.J., and Lawlor, E.R. (2008). BMI-1 promotes ewing sarcoma tumorigenicity independent of CDKN2A repression. Cancer Res 68, 6507-6515. 87 Endo, Y., Beauchamp, E., Woods, D., Taylor, W.G., Toretsky, J.A., Uren, A., and Rubin, J.S. (2008). Wnt-3a and Dickkopf-1 stimulate neurite outgrowth in Ewing tumor cells via a Frizzled3- and c-Jun N-terminal kinase-dependent mechanism. Molecular and cellular biology 28, 2368-2379. Eppert, K., Takenaka, K., Lechman, E.R., Waldron, L., Nilsson, B., Van Galen, P., Metzeler, K.H., Poeppl, A., Ling, V., Beyene, J., Canty, A.J., Danska, J.S., Bohlander, S.K., Buske, C., Minden, M.D., Golub, T.R., Jurisica, I., Ebert, B.L., and Dick, I.E. (20 11 ). Stem cell gene expression programs influence clinical outcome in human leukemia. Nature medicine 17, 1086-1093. Feldhahn, N., Henke, N., Melchior, K., Duy, C., Soh, B.N., Klein, F., Von Levetzow, G., Giebel, B., Li, A., Hofmann, W.K., Jumaa, H., and Muschen, M. (2007). Activation-induced cytidine deaminase acts as a mutator in BCR-ABLl transformed acute lymphoblastic leukemia cells. J Exp Med 204, 1157-1166. Friedman, M.S., Oyserman, S.M., and Hankenson, K.D. (2009). Wnt11 promotes osteoblast maturation and mineralization through R-spondin 2. J Biol Chem 284, 14117-14125. Fuerer, C., and Nusse, R. (2010). Lentiviral vectors to probe and manipulate the Wnt signaling pathway. PLoS One 5, e9370. Gentles, A.J., Plevritis, S.K., Majeti, R., and Alizadeh, A.A. (2010). Association of a leukemic stem cell gene expression signature with clinical outcomes in acute myeloid leukemia. JAMA : the journal of the American Medical Association 304, 2706-2715. Ginsberg, J.P., Goodman, P., Leisenring, W., Ness, K.K., Meyers, P.A., Wolden, S.L., Smith, S.M., Stovall, M., Hammond, S., Robison, L.L., and Oeffinger, K.C. (2010). Long-term survivors of childhood Ewing sarcoma: report from the childhood cancer survivor study. J Natl Cancer Inst 102, 1272-1283. Glinka, A., Dolde, C., Kirsch, N., Huang, Y.L., Kazanskaya, 0., Ingelfinger, D., Boutros, M., Cruciat, C.M., and Niehrs, C. (2011). LGR4 and LGR5 are R-spondin receptors mediating Wnt/beta-catenin and Wnt/PCP signalling. EMBO Rep 12, 1055-1061. Gong, X., Carmon, K.S., Lin, Q., Thomas, A., Yi, J., and Liu, Q. (2012). LGR6 is a high affinity receptor ofR-spondins and potentially functions as a tumor suppressor. PLoS One 7, e37137. Hanahan, D., and Weinberg, R.A. (2011). Hallmarks of cancer: the next generation. Cell 144, 646-674. 88 Hankenson, K.D., Sweetwyne, M.T., Shitaye, H., and Posey, K.L. (2010). Thrombospondins and novel TSR-containing proteins, R-spondins, regnlate bone formation and remodeling. Curr Osteoporos Rep 8, 68-76. Hauer, K., Calzada-Wack, J., Steiger, K., Grunewald, T.G., Baumhoer, D., Plehm, S., Buch, T., DaCosta, O.P., Esposito, I., Burdach, S., and Richter, G.H. (2013). DKK2 Mediates Osteolysis, Invasiveness, and Metastatic Spread in Ewing Sarcoma. Cancer research 73, 967-977. Hu, Y., Galkin, A.V., Wu, C., Reddy, V., and Su, A.I. (2011). CAFET algorithm reveals Wnt/PCP signature in lung squamous cell carcinoma. PLoS One 6, e25807. Huang, H.Y., Illei, P.B., Zhao, Z., Mazumdar, M., Huvos, A. G., Healey, J.H., Wexler, L.H., Gorlick, R., Meyers, P., and Ladanyi, M. (2005). Ewing sarcomas with p53 mutation or p16/p14ARF homozygous deletion: a highly lethal subset associated with poor chemoresponse. J Clin Oncol23, 548-558. Jaks, V., Barker, N., Kasper, M., VanEs, J.H., Snippert, H.J., Clevers, H., and Toftgiird, R. (2008). Lgr5 marks cycling, yet long-lived, hair follicle stem cells. Nature genetics 40, 1291-1299. Jiang, M., Stanke, J., and Lahti, J.M. (2011). The connections between neural crest development and neuroblastoma. Curr Top Dev Biol94, 77-127. Jiang, X., Gwye, Y., Mckeown, S.J., Bronner-Fraser, M., Lutzko, C., and Lawlor, E.R. (2009). Isolation and characterization of neural crest stem cells derived from in vitro-differentiated human embryonic stem cells. Stem Cells Dev 18, 1059-1070. Jiang, X., Gwye, Y., Russell, D., Cao, C., Douglas, D., Hung, L., Kovar, H., Triche, T.J., and Lawlor, E.R. (2010). CD133 expression in chemo-resistant Ewing sarcoma cells. BMC Cancer 10, 116. Jin, Z., Zhao, C., Han, X., and Han, Y. (2012). Wnt5a promotes ewing sarcoma cell migration through upregulating CXCR4 expression. BMC Cancer 12, 480. Kanwar, S.S., Yu, Y., Nautiyal, J., Patel, B.B., and Majumdar, A.P. (2010). The Wnt/beta-catenin pathway regulates growth and maintenance of colonospheres. Mol Cancer 9, 212. Katoh, M. (2005). WNT/PCP signaling pathway and human cancer (review). Oneal Rep 14, 1583-1588. Kawauchi, D., Robinson, G., Uziel, T., Gibson, P., Rehg, J., Gao, C., Finkelstein, D., Qu, C., Pounds, S., Ellison, D.W., Gilbertson, R.J., and Roussel, M.F. (2012). A mouse model of the most aggressive subgroup of human medulloblastoma. Cancer cell21, 168-180. 89 Kemper, K., Prasetyanti, P.R., De Lau, W., Rodermond, H., Clevers, H., and Medema, J.P. (2012). Monoclonal Antibodies Against Lgr5 Identify Human Colorectal Cancer Stem Cells. Stem Cells. Khan, J., Wei, J.S., Ringner, M., Saal, L.H., Ladanyi, M., Westermann, F., Berthold, F., Schwab, M., Antonescu, C.R., Peterson, C., and Meltzer, P.S. (2001). Classification and diagnostic prediction of cancers using gene expression profiling and artificial neural networks. Nat Med 7, 673-679. Kim, S.H., Kim, K.H., Yoo, B.C., and Ku, J.L. (2012). Induction ofLGR5 by H202 treatment is associated with cell proliferation via the JNK signaling pathway in colon cancer cells. Int J Onco/41, 1744-1750. Kobayashi, S., Yamada-Okabe, H., Suzuki, M., Natori, 0., Kato, A., Matsubara, K., Jau Chen, Y., Yamazaki, M., Funahashi, S., Yoshida, K., Hashimoto, E., Watanabe, Y., Mutoh, H., Ashihara, M., Kato, C., Watanabe, T., Yoshikubo, T., Tamaoki, N., Ochiya, T., Kuroda, M., Levine, A.J., and Yamazaki, T. (2012). LGR5- Positive Colon Cancer Stem Cells Interconvert with Drug-Resistant LGR5- Negative cells and are Capable of Tumor Reconstitution. Stem Cells 30,2631- 2644. Lawlor, E.R., Scheel, C., Irving, J., and Sorensen, P.H. (2002). Anchorage-independent multi-cellular spheroids as an in vitro model of growth signaling in Ewing tumors. Oncogene 21,307-318. Le Deley, M.C., Delattre, 0., Schaefer, K.L., Burchill, S.A., Koehler, G., Hogendoorn, P.C., Lion, T., Poremba, C., Marandet, J., Ballet, S., Pierron, G., Brownhill, S.C., Nesslbock, M., Ranft, A., Dirksen, U., Oberlin, 0., Lewis, I.J., Craft, A.W., Jurgens, H., and Kovar, H. (2010). Impact ofEWS-ETS fusion type on disease progression in Ewing's sarcoma/peripheral primitive neuroectodermal tumor: prospective results from the cooperative Euro-E.W.I.N.G. 99 trial. J Clin Oneal 28, 1982-1988. Lessnick, S.L., Dacwag, C.S., and Golub, T.R. (2002). The Ewing's sarcoma oncoprotein EWS/FLI induces a p53-dependent growth arrest in primary human fibroblasts. Cancer Celli, 393-401. Longley, D.B., and Johnston, P.G. (2005). Molecular mechanisms of drug resistance. J Patho/205, 275-292. Luo, W., Gangwal, K., Sankar, S., Boucher, K.M., Thomas, D., and Lessnick, S.L. (2009). GSTM4 is a microsatellite-containing EWS/FLI target involved in Ewing's sarcoma oncogenesis and therapeutic resistance. Oncogene 28,4126- 4132. 90 Mao, W., Tien, J., Goldenberg, D., Stawicki, S., Vij, R., Xia, X. Hongo, J.H., Ross, S., Kozak, K., Firestein, R., Wu, Y., Spencer, S., and Polakis, P. (2010). "Early study on LGR5/GPR49 molecule as a potential colon cancer stem cell target for the antibody conjugated drug treatment", in: Proceedings of the 101 st Annual Meeting of the American Association for Cancer Research;. (Washington, D.C.: AACR). Mcclanahan, T., Koseoglu, S., Smith, K., Grein, J., Gustafson, E., Black, S., Kirschmeier, P., and Samatar, A.A. (2006). Identification of overexpression of orphan G protein-coupled receptor GPR49 in human colon and ovarian primary tumors. Cancer Biol Ther 5, 419-426. Meltzer, P.S. (2007). Is Ewing's sarcoma a stem cell tumor? Cell stem cell!, 13-15. Merlos-Suarez, A., Barriga, F.M., Jung, P., Iglesias, M., Cespedes, M.V., Rossell, D., Sevillano, M., Hernando-Momblona, X., Da Silva-Diz, V., Munoz, P., Clevers, H., Sancho, E., Mangues, R., and Batlle, E. (2011). The intestinal stem cell signature identifies colorectal cancer stem cells and predicts disease relapse. Cell Stem CellS, 511-524. Miyagawa, Y., Okita, H., Itagaki, M., Toyoda, M., Katagiri, Y.U., Fujimoto, J., Hata, J., Umezawa, A., and Kiyokawa, N. (2009). EWS/ETS regulates the expression of the Dickkopffamily in Ewing family tumor cells. PLoS One 4, e4634. Nagoshi, N., Shibata, S., Kubota, Y., Nakamura, M., Nagai, Y., Satoh, E., Morikawa, S., Okada, Y., Mabuchi, Y., Katoh, H., Okada, S., Fukuda, K., Suda, T., Matsuzaki, Y., Toyama, Y., and Okano, H. (2008). Ontogeny and multipotency of neural crest-derived stem cells in mouse bone marrow, dorsal root ganglia, and whisker pad. Cell Stem Cell2, 392-403. Nakata, S., Campos, B., Bageritz, J., Lorenzo Bermejo, J., Becker, N., Engel, F., Acker, T., Momma, S., Herold-Mende, C., Lichter, P., Radlwimmer, B., and Goidts, V. (2013). LGR5 is a Marker of Poor Prognosis in Glioblastoma and is Required for Survival of Brain Cancer Stem-Like Cells. Brain Pathol. Nam, J.S., Turcotte, T.J., and Yoon, J.K. (2007). Dynamic expression ofR-spondin family genes in mouse development. Gene Expr Patterns 7, 306-312. Navarro, D., Agra, N., Pestana, A., Alonso, J., and Gonzalez-Sancho, J.M. (2010). The EWS/FLII oncogenic protein inhibits expression of the Wnt inhibitor DICKKOPF-1 gene and antagonizes beta-catenin/TCF-mediated transcription. Carcinogenesis 31, 394-401. Ohali, A., Avigad, S., Zaizov, R., Ophir, R., Horn-Saban, S., Cohen, I.J., Meller, I., Kollender, Y., Issakov, J., and Yaniv, I. (2004). Prediction of high risk Ewing's sarcoma by gene expression profiling. Oncogene 23, 8997-9006. 91 Ohkawara, B., Glinka, A., and Niehrs, C. (2011). Rspo3 binds syndecan 4 and induces Wnt/PCP signaling via clathrin-mediated endocytosis to promote morphogenesis. Dev Cell 20, 303-314. Ohkawara, B., and Niehrs, C. (2011 ). An ATF2-based luciferase reporter to monitor non canonical Wnt signaling in Xenopus embryos. Dev Dyn 240, 188-194. Pece, S., Tosoni, D., Confalonieri, S., Mazzarol, G., Vecchi, M., Ronzoni, S., Bernard, L., Viale, G., Pelicci, P.G., and DiFiore, P.P. (2010). Biological and molecular heterogeneity of breast cancers correlates with their cancer stem cell content. Cell 140, 62-73. Plaks, V., Brenot, A., Lawson, D.A., Linnemann, J.R., Van Kappel, E.C., Wong, K.C., De Sauvage, F., Klein, O.D., and Werb, Z. (2013). Lgr5-expressing cells are sufficient and necessary for postnatal mammary gland organogenesis. Cell Rep 3, 70-78. Richter, G.H., Plehm, S., Fasan, A., Rossler, S., Unland, R., Bennani-Baiti, I.M., Hotfilder, M., Lowe!, D., Von Luettichau, I., Mossbrugger, I., Quintanilla Martinez, L., Kovar, H., Staege, M.S., Muller-Tidow, C., and Burdach, S. (2009). EZH2 is a mediator ofEWS/FLI1 driven tumor growth and metastasis blocking endothelial and neuro-ectodermal differentiation. Proc NatlAcad Sci USA 106, 5324-5329. Riggi, N., Suva, M.L., Suva, D., Cironi, L., Provero, P., Tercier, S., Joseph, J.M., Stehle, J.C., Baumer, K., Kindler, V., and Starnenkovic, I. (2008). EWS-FLI-1 expression triggers a Ewing's sarcoma initiation program in primary human mesenchymal stem cells. Cancer Res 68, 2176-2185. Sato, T., Vries, R.G., Snippert, H.J., Van De Wetering, M., Barker, N., Stange, D.E., Van Es, J.H., Abo, A., Kujala, P., Peters, P.J., and Clevers, H. (2009). Single Lgr5 stem cells build crypt-villus structures in vitro without a mesenchymal niche. Nature 459, 262-265. Schaefer, K.L., Eisenacher, M., Braun, Y., Brachwitz, K., Wai, D.H., Dirksen, U., Lanvers-Karninsky, C., Juergens, H., Herrero, D., Stegmaier, S., Koscielniak, E., Eggert, A., Nathrath, M., Gosheger, G., Schneider, D.T., Bury, C., Diallo Danebrock, R., Ottaviano, L., Gabbert, H.E., and Poremba, C. (2008). Microarray analysis of Ewing's sarcoma family of tumours reveals characteristic gene expression signatures associated with metastasis and resistance to chemotherapy. Eur J Cancer 44, 699-709. Schepers, A.G., Snippert, H.J., Stange, D.E., Van Den Born, M., VanEs, J.H., Van De Wetering, M., and Clevers, H. (2012). Lineage tracing reveals Lgr5+ stem cell activity in mouse intestinal adenomas. Science 337, 730-735. 92 Schmidt, C., and Patel, K. (2005). Wnts and the neural crest. A nat Embryo! (Berl) 209, 349-355. Scotlandi, K., Remondini, D., Castellani, G., Manara, M.C., Nardi, F., Cantiani, L., Francesconi, M., Mercuri, M., Caccuri, A.M., Serra, M., Knuutila, S., and Picci, P. (2009). Overcoming resistance to conventional drugs in Ewing sarcoma and identification of molecular predictors of outcome. J Clin Onco/27, 2209-2216. Simon, E., Petke, D., Boger, C., Behrens, H.M., Warneke, V., Ebert, M., and Rocken, C. (2012). The spatial distribution ofLGR5+ cells correlates with gastric cancer progression. P LoS One 7, e35486. Sjoblom, T., Jones, S., Wood, L.D., Parsons, D.W., Lin, J., Barber, T.D., Mandelker, D., Leary, R.J., Ptak, J., Silliman, N., Szabo, S., Buckbaults, P., Farrell, C., Meeh, P., Markowitz, S.D., Willis, J., Dawson, D., Willson, J.K., Gazdar, A.F., Hartigan, J., Wu, L., Liu, C., Parmigiani, G., Park, B. H., Bachman, K.E., Papadopoulos, N., Vogelstein, B., Kinzler, K.W., and Velculescu, V.E. (2006). The consensus coding sequences of human breast and colorectal cancers. Science 314,268-274. Staege, M.S., Hutter, C., Neumann, I., Foja, S., Hattenhorst, U.E., Hansen, G., Afar, D., and Burdach, S.E. (2004). DNA microarrays reveal relationship of Ewing family tumors to both endothelial and fetal neural crest-derived cells and define novel targets. Cancer Res 64, 8213-8221. Suva, M.L., Riggi, N., Stehle, J.C., Baumer, K., Tercier, S., Joseph, J.M., Suva, D., Clement, V., Provero, P., Cironi, L., Osterheld, M.C., Guillou, L., and Stamenkovic, I. (2009). Identification of cancer stem cells in Ewing's sarcoma. Cancer Res 69, 1776-1781. Takahashi, H., Ishii, H., Nishida, N., Takemasa, I., Mizushima, T., Ikeda, M., Yokobori, T., Mimori, K., Yamamoto, H., Sekimoto, M., Doki, Y., and Mori, M. (2011). Significance of Lgr5( +ve) cancer stem cells in the colon and rectum. Ann Surg On col 18, 1166-117 4. Takashima, Y., Era, T., Nakao, K., Kondo, S., Kasuga, M., Smith, A. G., and Nishikawa, S. (2007). Neuroepithelial cells supply an initial transient wave of MSC differentiation. Cell129, 1377-1388. Tanese, K., Fukuma, M., Yamada, T., Mori, T., Yoshikawa, T., Watanabe, W., Ishiko, A., Amagai, M., Nishikawa, T., and Sakamoto, M. (2008). G-protein-coupled receptor GPR49 is up-regulated in basal cell carcinoma and promotes cell proliferation and tumor formation. Am J Patho/173, 835-843. Tirode, F., Laud-Duval, K., Prieur, A., Delorme, B., Charbord, P., and Delattre, 0. (2007). Mesenchymal stem cell features of Ewing tumors. Cancer Cell11, 421- 429. 93 Uchida, H., Yamazaki, K., Fukuma, M., Yamada, T., Hayashida, T., Hasegawa, H., Kitajima, M., Kitagawa, Y., and Sakamoto, M. (2010). Overexpression of leucine-rich repeat-containing G protein-coupled receptor 5 in colorectal cancer. Cancer Sci 101, 1731-1737. Uren, A., Wolf, V., Sun, Y.F., Azari, A., Rubin, J.S., and Toretsky, J.A. (2004). Wnt/Frizzled signaling in Ewing sarcoma. Pediatr Blood Cancer 43, 243-249. Valladares-Ayerbes, M., Blanco-Calvo, M., Reboredo, M., Lorenzo-Patino, M.J., Iglesias-Diaz, P., Haz, M., Diaz-Prado, S., Medina, V., Santamarina, I., Pertega, S., Figueroa, A., and Anton-Aparicio, L.M. (2012). Evaluation of the Adenocarcinoma-Associated Gene AGR2 and the Intestinal Stem Cell Marker LGR5 as Biomarkers in Colorectal Cancer. Int J Mol Sci 13, 4367-4387. VanDer Flier, L.G., Sabates-Bellver, J., Oving, I., Haegebarth, A., De Palo, M., Anti, M., Van Gijn, M.E., Suijkerbuijk, S., Van De Wetering, M., Marra, G., and Clevers, H. (2007). The Intestinal Wnt/TCF Signature. Gastroenterology 132, 628-632. Van Doorninck, J.A., Ji, L., Schaub, B., Shimada, H., Wing, M.R., Krailo, M.D., Lessnick, S.L., Marina, N., Triche, T.J., Sposto, R., Womer, R.B., and Lawlor, E.R. (2010). Current treatment protocols have eliminated the prognostic advantage of type 1 fusions in Ewing sarcoma: a report from the Children's Oncology Group. J Clin Oncol28, 1989-1994. Van Maldegem, A.M., Hogendoorn, P.C., and Hassan, A.B. (2012). The clinical use of biomarkers as prognostic factors in Ewing sarcoma. Clin Sarcoma Res 2, 7. Varnat, F., Siegl-Cachedenier, I., Malerba, M., Gervaz, P., and Ruiz I Altaba, A. (2010). Loss ofWNT-TCF addiction and enhancement ofHH-GLil signalling define the metastatic transition of human colon carcinomas. EMBO Mol Med 2, 440-457. Vijayakumar, S., Liu, G., Rus, LA., Yao, S., Chen, Y., Akiri, G., Grnmolato, L., and Aaronson, S.A. (2011). High-frequency canonical Wnt activation in multiple sarcoma subtypes drives proliferation through a TCF/beta-catenin target gene, CDC25A. Cancer celll9, 601-612. Visvader, J.E., and Lindeman, G.J. (2012). Cancer stem cells: current status and evolving complexities. Cell Stem CelllO, 717-728. Von Levetzow, C., Jiang, X., Gwye, Y., Von Levetzow, G., Hung, L., Cooper, A., Hsu, J.H., and Lawlor, E.R. (2011). Modeling initiation of Ewing sarcoma in human neural crest cells. PLoS One 6, el9305. 94 Walker, F., Zhang, H. H., Odorizzi, A., and Burgess, A.W. (2011). LGR5 is a negative regulator oftumourigenicity, antagonizes Wnt signalling and regulates cell adhesion in colorectal cancer cell lines. PLoS One 6, e22733. Wislet-Gendebien, S., Laudet, E., Neirinckx, V., Alix, P., Leprince, P., Glejzer, A., Poulet, C., Hennuy, B., Sommer, L., Shakhova, 0., and Rogister, B. (2012). Mesenchymal stem cells and neural crest stem cells from adult bone marrow: characterization of their surprising similarities and differences. Cell Mol Life Sci 69, 2593-2608. Wu, X.-S., Xi, H.-Q., and Chen, L. (2012). Lgr5 is a potential marker of colorectal carcinoma stem cells that correlates with patient survival. World journal of surgical oncology 10, 244. Zoubek, A., Dockhorn-Dworniczak, B., Delattre, 0., Christiansen, H., Niggli, F., Gatterer-Menz, I., Smith, T.L., Jurgens, H., Gadner, H., and Kovar, H. (1996). Does expression of different EWS chimeric transcripts define clinically distinct risk groups of Ewing tumor patients? J Clin Oncol14, 1245-1251. 95 ORIGINAL RESEARCH ARTICLE publ1shed 15 Apnl 2013 do 10 3389/fonc 2013 00081 LGR5 is expressed by Ewing sarcoma and potentiates Wnt/~-catenin signaling Christopher A. Scannef11. 2 , Elisabeth A. Pedersen 2 , Jack T. Mosher 2 , Melanie Anne Krook 2 , LaurenA. Nicholls 2 , BreelynA. Wilky 3 , David M. Loeb 4 and Elizabeth R. Lawlor 2 ,H 1 Keck School of Medicine, University Southern California, Los Angeles, CA, USA 2 Department of Pediatrics, University of Michigan, Ann Arbor, Ml, USA 3 Department of Medical Oncology, Sidney Kimmel Comprehensive Cancer Center, Johns Hopkins University, Baltimore, MD, USA 4 Division of Pediatric Oncology, Sidney Kimmel Comprehensive Cancer Center, Johns Hopkins University, Baltimore, MD, USA 5 Department of Pathology, University of Michigan, Ann A rbor, M l, USA Edited by: Stephen Lessnick, University of Utah, USA Reviewed by: Michael Engel, University of Utah School of Medicine, US A Frederic Barr, National Cancer Institute, USA *Correspondence: Elizabeth R. Lawlor, North Campus Research Complex, B520 Room 1352 , 1600 Huron Parkway, Ann Arbor, M l 481 09, US A e--mail. elawlor@med.umich.edu INTRODUCTION Ewing sarcoma (ES) is an aggressive bone and soft tissue tumor of putative stem cell ori gin that predominantly occurs in children and young adults. Although most patients with localized ES can be cured with intensive therapy, the clinical course is variable and up to one third of patients relapse following initial remission. Unfortunately, little is yet known about the biologic features that distinguish low-risk from high-risk disease or the mech anisms of ES disease progression. Recent reports have suggested that putative cancer stem cells exist in ES and may contribute to an aggressive phenotype. The cell surface receptor leucine-rich repeat-containing G-protein coupled receptor 5 (LGR5) is a somatic stem cell marker that functions as an oncogene in several human cancers, most notably colorectal carcinoma. LGR5 is a receptor for the R-spondin (RSPO) family of ligands and RSPO-mediated activation of LGR5 potentiates Wnt/1)-catenin signaling, contributing to stem cell proliferation and self-renewal. Given its presumed stem cell origin, we investi gated whether LGR5 contributes to ES pathogenesis. We found that LGR5 is expressed byES and that its expression is relatively increased in cells and tumors that display a more aggressive phenotype. In particular, LGR5 expression w as increased in putative cancer stem cells. We also found that neural crest-derived stem cells express LGR5, raising the possibility that expression of LGR5 may be a feature of ES cells of origin. LGR5-high ES cells showed nuclear localization of I)-eaten in and robust activation ofTCF reporter activity when exposed to Wnt ligand and this was potentiated by RSPO. However, modulation of LGR5 or exposure to RSPO had no impact on proliferation confirming that Wnt/1)-catenin signaling in ES cells does not recapitulate signaling in epithelial cells. Together these stud ies show that the RSPO-LGR5-Wnt-J)-catenin axis is present and active in ES and may contribute to tumor pathogenesis. Keywords: LGR5, Ewing sarcoma, stem cell, R-spondin, Wnt, ~-catenin Ewing sarcoma (ES) is a malignant tumor of the bone and soft tissue that can present at any age but predominantly occurs in adolescents and young adults. These tumors are genetically defined by recurrent chromosomal translocations that result in the creation of novel fusion oncogenes, most commonly EWS FLII (Balamuth and Womer, 2010). Clinically, these tumors often have an aggressive course with one quarter of patients present ing with gross metastatic disease at the time of diagnosis. In addition, nearly one third of patients will relapse after an initial clinical remission and patients who have metastatic or relapsed ES have 5-year event free survival rates of only 10-20% (Bala muth and Womer, 2010). Unfortunately, there are no clinical or pathologic criteria apart from metastases that can reliably pre dict whether a newly diagnosed patient with ES is likely to be cured or to relapse. Histologically, tumors are characterized by an undifferentiated small round blue cell phenotype with features of primitive neuroectodermal cells. Although predominantly a bone and connective tissue tumor, clinically ES can present in multiple organs and tissue types throughout the body, suggest ing a relatively undifferentiated and potentially highly migratory cell of origin (Meltzer, 2007). Indeed, current evidence supports the hypothesis that ES arise from either mesenchymal stem cells (MSC) or neural crest stem cells (NCSC) or their early progeni tors (Staege et al., 2004; Tirode et al., 2007; Riggi et al., 2008; von Levetzow et al., 2011 ). Importantly, poorly differentiated tumors in other classes of human malignancy often express stem cell associated markers and an undifferentiated phenotype combined with high -level expression of stem cell genes is associated with worse clinical outcomes (Phillips et al., 2006; Ben-Porath et al., 2008; Spike et al., 2012). The stem cell phenotype and aggressive nature of ES raise the question of whether stem cell markers could be useful in understanding the origin and pathogenesis of this enigmatic disease. www.frontiersin.org April 2013 fVolume 3 1Article81 f1 96 Scannell et al Leucine-rich repeat-containing G-protein coupled receptor 5 (LGRS) is a seven transmembrane spanning receptor that has recently been identified as a somatic stem cell marker that plays key functional roles in both normal development and cancer. Mouse studies have demonstrated that Lgr5 is widely expressed during embryonic development but expression in postnatal tissues is lim ited to discrete stem cell populations (Barker et al., 2007). Such stem cells can be found in the small and large intestine, stom ach, hair follicles, and kidney (Barker et al., 2007,2010, 2012; Jaks et al., 2008). The self-renewal and proliferation of normal murine intestinal stem cells (ISC) is dependent on LgrS (Barker et al., 2007). Significantly, Lgrs+ ISC have also recently been identified as both the cells of origin for murine intestinal tumors and tumor maintaining cancer stem cells in established adenomas (Barker et al., 2009; Schepers et al., 2012). Studies of human cancer cell lines have now confirmed that LGR5 promotes the growth and/or survival of colorectal and basal cell carcinoma (McClanahan et al., 2006; Tanese et al., 2008), glioblastoma (Nakata et al., 2013), and neuroblastoma (Balamuth et al., 201 0). Thus, LGRS has now been shown to contribute functionally to normal and malignant biol ogy in tissues of both epithelial and neural origin. To determine if high levels of LGR5 are associated with a more aggressive clinical course, retrospective studies of archived tumors were undertaken and showed diminished survival in gastrointestinal carcinoma and glioblastoma patients whose tumors expressed high levels of LGR5 (Becker et al., 2010; Wu et al., 2012; Nakata et al., 2013). In addi tion, in the case of colorectal carcinoma, LGRS is expressed by a subpopulation of cells with stem cell-like properties (i.e., cancer stem cells or CSC) (Kemper et al., 2012; Kobayashi et al., 2012). The LGRS + colorectal CSC have increased clonogenic and tumori genic potential compared to bulk tumor cells and lose expression ofLGRS upon in vi tro differentiation (Kemper et al., 20 12). Gene expression profiling experiments have also demonstrated that an LgrS-stem cell gene signature predicts disease relapse in colorectal cancer patients (Merlos-Suarez et al., 2011 ). Thus, there is com pelling evidence in both human and murine intestinal tumors that LGRS+ stem cells contribute to cancer initiation and progression and that high LGR5 expression is associated with worse clinical outcomes. The mechanism by which LGRS promotes stem cell self-renewal and proliferation has only recently been elucidated. It is now known that, along with its closely related homologs LGR4 and LGR6, LGRS functions to potentiate canonical Wnt/1)-catenin sig naling (Carmon et al., 2011; de Lau et al., 2011 ; Glinka et al., 2011 ; Gong et al., 2012). This potentiation ofWnt signaling is achieved when LGRS is bound by its ligand R-spondin (RSPO). RSPOs are a recently described family of secreted proteins that function as Wnt agonists and play pivotal roles as regulators of normal embryonic development and stem cell proliferation (Glinka et al., 2011 ). When RSPO binds to LGRS the receptor associates with the Frizzled/LRP complex to increase the activation of ~-catenin and downstream TCF reporter activity as well as non-canonical Wnt signaling (Glinka et al., 2011). The potentiation of Wnt/~-catenin signaling is now believed to mediate the self-renewal and prolif eration of LGRs+ stem cells, both normal and malignant, and provides a mechanistic link between LGRS, Wnt signaling, cancer stem cells, and cancer progression (Clevers and Nusse, 2012). Frontiers in Oncology 1 Pediatric Oncology LGR5 in Ew ing sarcoma Given the presumed stem cell origin of ES, the presence of RSP02 in developing bone (Friedman et al., 2009) and prior evi dence that Wnt signaling is abnormal in these tumors (Uren et al., 2004; Navarro et al., 201 0; Vijayakumar et al., 20 11 ) we hypoth esized that LGR5 and its downstream impact on Wnt signaling might contribute toES pathogenesis. In the current study we have investigated this hypothesis using studies of ES primary tumors and cell lines and normal neuro-MSC, candidate cells ofES origin. Our findings indicate that LGR5 is indeed expressed by some pop ulations ofES cells and that these cells upregulate Wnt/~-catenin signaling when exposed to RSPO. Together our studies suggest that LGRS-mediated potentiation ofWnt signaling may be an impor tant contributor to ES initiation and maintenance, especially in RSPO-rich microenvironments like developing bone. MATERIALS AND METHODS TUMOR SAMPLES AND CELL LINES Primary tumor RNA was obtained from Children's Hospital Los Angeles (CHLA) and Children's Oncology Group (COG) tumor biorepositories. All primary tissue samples were coded according to an anonymous numbering scheme and acquired in accor dance with approval from the CHLA Committee for Clinical Investigation. Primary tumor expression data were kindly pro vided by the COG (Lawlor, unpublished data). ES cell lines were kindly provided by Dr. Timothy Triche ( CHLA, Los Angeles, CA, USA), Dr. Heinrich Kovar ( CCRI, St. Anna Kinderkrebsforschung, Vienna, Austria), and the COG cell bank (cogcell.org) and identi ties confirmed by short tandem repeat profiling courtesy of Dr. Pat Reynolds (Texas Tech University, Lubbock, TX, USA). All ES cell lines were maintained in RPMI with L -glutamine and 1 Oo/o FBS on tissue culture treated polystyrene plates, except for CHLA25 and STA-ET -8.2 cells. These cell lines were grown on plates coated with fibronectin. Human bone marrow mesenchymal stem cell lines (MSC; obtained from Dr. Darwin Prockop, Tulane University, New Orleans, LA, USA), H1 and H9 human embryonic stem cell (ESC) lines (Wicell, Madison, WI, USA) and ESC-derived neural crest cells were cultured and differentiated using standard protocols as previously described by our lab (Jiang et al., 2009; von Levetzow et al., 201 1). MRCS fibroblasts were obtained from American Type Culture Collection and maintained in DMEM with L-glutamine and 10% FBS. QUANTITATIVE REAL-TIME REVERSE TRANSCRIPTION PCR eDNA was generated from RNA (iScript; Bio-Rad, Hercules, CA, USA) and qRT-PCR was performed with validated TaqMan primers (LGR5, CITED2, CYCLINDI, MYC, GAPDH, and ISS; Life Technologies, Grand Island, NY, USA) or primers designed using qPrimerDepot (primerdepot.nci.nih.gov) (Table 1) (Cui et al., 2007). Assays were performed in triplicate using the Applied Biosystems 7900HT Fast real-time PCR system or Roche Light Cycler 480 and average Ct values were normalized relative to GAPDH or ISS expression in the same sample. IMMUNOCYTOCHEMISTRY Cells were grown on fibronectin-coated chamber slides (Thermo Fisher Scientific, Waltham, MA, USA), fixed in 4% paraformalde hyde, and permeabilized with 0.1% Triton X-100. After rinsing April 2013 1 Volume 3 1 Article 81 12 97 Scannell et al Table 1 1 Prime sequences for RT-PCR. Gene Forward primer sequence Reverse primer sequence AXIN2 AAGTGCAAACTTTCGCCAAC ACAGGATCGCTCCTCTTGAA CD44 GACAAGTTTTGGTGGCACG CACGTGGAATACACCTGCAA CDC25A CCAGCCCCAAAGAGTCAAC AAGGTCCCTTGGGTCATTGT EPHB2 CTCTACTGTAACGGGGACGG CCTTGAAAGTCCCAGATGGA HIG2 CTTCTGCGCTGGTGCTTAGT GCAGAGAAACAGAGCTGCCT 102 GACAGCAAAGCACTGTGTGG TCAGCACTTAAAAGATTCCGTG LEFJ TGGATCTCTTTCTCCACCCA CACTGTAAGTGATGAGGGGG TROY CCCTCCTCCTCCTTACGAAC CCAGAGCGCTGCAGATAAC GAPDH AAGGTGAAGGTCGGAGTCAA AATGAAGGGGTCATTGATGG 185 GCAATTATTCCCCATGAACGA GGCCTCACTAAACCATCCAAT in phosphate-buffered saline (PBS) the slides were incubated for 1 h in blocking solution (1 o/o BSA and 5% donkey serum in PBS). Blocked slides were incubated with primary 1)-catenin anti body (1 :300; BD Biosciences, San Jose, CA, USA) overnight at 4°C, washed in PBS, incubated with Alexa Fluor 488-conjugated, donkey-anti mouse secondary antibody ( 1 :500; Life Technologies) for 1 h, and then visualized with a Leica DMI6000 B Fluorescence Microscope. Nuclei were counterstained with 4 1 ,6-diamidino-2- phenylindole (DAPI). Cytoplasmic and nuclear 1)-catenin staining cells were manually counted in a minimum of three high -power fields. CELL SORTING FOR SUBPOPULATION STUDIES For CD133 sorts, cultured STA-ET-8.2 cells were trypsinized and resuspended in FeR Blocking Reagent (Miltenyi Biotec, Auburn, CA, USA) with O.So/o bovine serum albumin (Sigma, St Louis,MO, USA) in PBS. Mouse anti-human CD133/2-PE (Miltenyi) mono clonal antibody was then added ( 1:11 dilution) and incubated for 15 min at 4°C in the dark. After two washes, cells were sorted on a MoFlo Astrios instrument (Beckman Coulter) at the University of Michigan Flow Cytometry Core. Positive and negative gates were determined using IgG stained and unstained controls. TC71 and MHH -ES cells were sorted on the basis of aldehyde dehydrogenase activity using the Aldefluor® assay (Stem Cell Technologies, Van couver, BC, USA). Cells were analyzed and sorted as previously described (:\wad c! 10). GENERATION OF STABLE KNOCKDOWN AND OVER-EXPRESSION CELL LINES For knockdown studies, cell lines were transduced with pLK0.1-puro lentiviral short hairpin RNAs vectors targeting LGR5 expression (shLGRS#l-TAAGTGCCAGAACTGCTATGG and shLGR5#3-TTGTCCAAATTGCAGTAGAGC; Thermo Fisher Scientific). Control cells were transduced with a non-silencing small hairpin RNA vector containing an inert sequence (shNS-CAACAAGATGAAGAGCACCAA). Cells were selected in puromycin (1-2~-Lg/mL) for at least 48 h before use in subsequent experiments. For over-expression studies, cells were transduced with a lentiviral pCL6 expression vector into which the full length open reading frame for LGR5 eDNA (NP _003658; Thermo Scien tific, Clone ID BC09624) was cloned. The pCL6 vector is a mod ified version of the pCLl vector including an internal ribosome www.frontiersin.org LGR5 1n Ewmg sarcoma entry site and woodchuck hepatitis virus post -transcription regulatory element up- and dovmstream, respectively, of the EGFP reporter gene (!'eJ,Jh:lhn d aL, 2007; von ! e! aL, II). Control cells were transduced with empty vector that lacked the LGR5 sequence (Empty). Cells were selected by FACS on the basis of GFP expression. Lentiviral supernatant for each construct was generated from plasmid-transfected 293FT packaging cells as previously described (von ! ci 201] ). TCF REPORTER ASSAY To assess the transcriptional activity of the Wnt/1)-catenin axis we used previously characterized TCF-promoter luciferase-reporter constructs (Fth:rer [() ). The 7xTcf-FFluc//SV 40- mCherry (p7TFC) and 7xTcf-FFluc//SV40-PuroR (p7TFP) plas mids were purchased from Addgene (plasmids 24,307 and 24,308, respectively). Cells were sorted for mCherry expression (p7TFC) by FACS or rultured with puromycin (p7TFP) to select for cells that were successfully transduced. To assess the effects of exoge nous Wnt ligand on ES cells we collected conditioned media ( CM) from control mouse fibroblasts (L-cells) or L-cells that have been engineered to over-express Wnt3a (L-Wnt3a-cells) obtained from the American Type Culture Collection (ATCC CRL-2648 and CRL-2647, respectively). The final dilution of L-cell CM or Wnt3a CM was 1:2 in RPMI containing 5% FBS. Recom binant human RSP02 (R&D Systems, Minneapolis, MN, USA) was added to L-cell CM or Wnt3a CM at a concentration of 20 ng/mL. p7TFC and p7TFP-transduced cells were plated at a density of 4.0 x 10 4 cells/well on 96-well plates and allowed to attach overnight. The cells were cultured for another 24 h in standard RPMI with 1 Oo/o FBS or treated with L-cell CM or Wnt3a CM. Luciferase measurements were carried out using the Luciferase Assay System (Promega) according to the manufac turer's protocol and read on a Molecular Devices LMAX plate reader. Measurements were normalized to mCherry mean flu orescent intensity for cells transduced with the p7TFC vector, otherwise TCF reporter activity was expressed relative to the L-cell CM control. CELL PROLIFERATION ASSAYS Cell growth was assessed using the Cell Titer 96AQueous One Solu tion Cell Proliferation Assay (Pro mega, Madison, WI, USA). Cells were plated at a density of 5 x 10 3 cells/well in 8-well replicates on 96-well flat-bottomed plates, allowed to attach overnight, and growth was assessed on Days 0, 2, and 4 post -attachment. For cells grown in the presence of exogenous Wnt ligands, WNT3A (R&D Systems), and/or RSP02 (R&D Systems) were supplemented on Days 0 and 3 at 500 and 20 ng/mL, respectively. The normalized cell index was calculated relative to the Day 0 absorbance reading minus background. STATISTICAL ANALYSIS Unless otherwise indicated data from all experiments are expressed as mean± SEM from a minimum of three independent exper iments. The data was analyzed using GraphPad Prism software by a Student's t-test and a p value of <0.05 was considered significant. April 2013 1 Volume 3 1 Art1cle 81 13 98 Scannell et al. RESULTS LGR5 IS EXPRESSED BY EWING SARCOMA AND NEURAL CREST-DERIVED STEM CELLS Ewing sarcoma are usually highly undifferentiated tumors that dis play features of both NCSC and MSC stem cells. Thus, although their precise cellular ontogeny remains to be fully elucidated, ES are presumed to arise from malignant transformation of nor mal NCSC and/or MSC. We therefore sought to determine if LGR5 is expressed by ES and the stem cells from which they are thought to arise. To address this we analyzed the expression of LGR5 in a cohort of 49 primary ES tumor samples and 15 ES cell lines. Unfortunately, the specificity of commercially available antibodies for studies of human LGRS remains inadequate for immuno-histochemical studies so we limited our evaluation of LGR5 expression to quantitative RT-PCR analysis. As shown, LGR5 was widely detectable in both ES tumors and cell lines, however, the level of expression was highly variable (Figures lA,B). Given its designation as a stem cell marker in epithelial tis sues we next assessed whether LGR5 might be expressed by NCSC and/or MSC. To address this we evaluated adult human bone marrow-derived MSC as well as human ESC-derived NCSC A I" 0 E 2 1 0 .. LGR5 1 evel (%rei to GAPDH) FIGURE 1 I Leucine-rich repeat-containing G-protein coupled receptor 5 is expressed by Ewing sarcomas. qRT -PCR analys1 s revealed low to very high-level expression of LGR5 in (A) 49 primary Ewing sarcoma (ES) and (B) A~ ; : l ~ 5: • ~ (3 5 • :Q ~ • (!) 0 --- ~ 0.10 •• 0.05 •• ..... i 0.151 • 0 .00-'-~fl-'"--t----,..L_;---r-.,..L_r---~ TI - t:' ~ S2 ~ S2 ::::. "' Ql :§ ~ (/) UJ (.) (/) UJ .s::; (.) (/) (.) z .s::; (.) (/) :2 u z .s::; FIGURE 2 I Leucine-rich repeat-containing G-protein coupled receptor 5 is expressed by neural crest stem cells (NCSC). (A) qRT -PCR analysis of human embryonic stem cells (hESC), hESC-derived NCSC (hNCSC), hNCSC-derived mesenchymal stem cells (hNC-MSCI, bone marrow-derived mesenchymal stem cells lhBM-M SCi. and human lung embryo fibroblasts Frontiers in Oncology I Pediatnc Oncology LGR5 in Ew ing sarcoma and their MSC progeny, neural crest-derived MSC (NC-MSC) as previously described by our group (von Levetzow et al., 2011 ). Significantly, our analyses revealed that undifferentiated NCSC express relatively high levels of LGR5 whereas the transcript is undetectable in adult bone marrow-derived MSC (Figure 2A). Consistent with this observation, LGRS expression declined when NCSC were differentiated toward an MSC identity (NC-MSC) (Figure 2A). Interestingly, the level of LGR5 expression in these neural crest-derived stem cells was comparable to that of ES cell lines (Figure 2A). To determine if the observed differential expres sion of LGR5 by neural crest cells was merely an artifact of the human ESC culture system we next interrogated publicly avail able gene expression data that compared murine bone marrow stem cells of mesodermal origin to bone marrow stem cells of neural crest origin (GEO accession 30,419) (Wislet-Gendebien et al., 2012). Interestingly, consistent with our human cell stud ies, murine bone marrow-derived stem cells of neural crest origin expressed higher levels of Lgr5 than MSC of mesodermal origin (Figure 2B). Together these studies show that LGR5 is expressed, to varying degrees, by established ES tumors and cell lines. Moreover, they 15 ES cell lines. For primary tumors levels of expression relative to GAPDH were <0.001 % in 1 tumor, 0.001- 0.01% in 7 tumors, 0.01- 0.1 % in 17 tumors, 0.1- 1 % in 12 tumors, 1 - 10% in 9 tumors, and 10- 100% in 3 tumors. B 8 ~ -~ 6 .$ .!: ~ 4 0> 'iii 'g, 2 ..,J • • • •• • 0~----~--------~~--- Mesodermal (3) Neural crest (3) Bone marrow stromal cells (MRC5) showed that undifferentiated hNCSC express the highest levels of LGR5. (B) Murine bone marrow stromal cells of neural crest origin express h1gher levels of Lgr5 than cells of mesodermal origin (from publically available microarray data GEO accession GSE3041 9) M/islet-Gendebien et al., 2012). N = 3 ± SEM . Lines represent arithmetic means of replicate samples. April 2013 1 Volume 3 1 Artr cle 81 14 99 Scannell et al. demonstrate that while LGR5 is not highly expressed by MSC of mesodermal origin, it is expressed by neural crest-derived stem cells, in particular undifferentiated NCSC. LGR5 EXPRESSION IS ELEVATED IN CLINICALLY AGGRESSIVE ES AND IN PUTATIVE ES CANCER STEM CELLS Over-expression of LGR5 has been linked to clinically aggressive disease in colorectal carcinoma as well as several other epithelial malignancies. Therefore, we sought to determine if high expres sion of LGR5 might also be associated with aggressive disease in ES. To address this we first compared LGR5 expression in the CHLA9 and CHLAl 0 cell lines, which were derived from the same patient prior to (CHLA9) and after (CHLAlO) chemotherapy. It is note worthy that the CHLA10 cell line was derived from a metastatic focus and therefore represents a progressive state of disease. As shown, LGR5 levels were found to be 10-fold higher in CHLA10 than CHLA9 cells (Figure 3A). To evaluate whether clinically aggressive ES express higher levels of LGR5 in vivo we interro gated whole genome expression data that were generated from primary human tumor samples (ERL unpublished data, kindly provided by the COG). LGR5 expression levels from diagnostic biopsies were compared in 4 patients who succumbed to rapidly progressive disease (survival 5-13 months) and 10 patients who remained disease free for at least 48 months. LGR5 expression was found to be extremely high in three of four patients with aggressive disease and mean expression was significantly higher than in long term event free survivors (Figure 3B). Interestingly, two of the three patients with the highest expression of LGR5 had primary drug-resistant tumors. A _C) Q)<{ >...J ~I "'u f§~ -J~ CHLA9 CHLA10 c 4 STA-ET-8.2 - ci> 3 "'"' ~0 "'U f§.a 2 -~~ 0 CD 133- CD1 33+ FIGURE 31 Leucine-rich repeat-containing G-protein coupled receptor 5 is increased in aggressive disease and putative cancer stem cells. (A) Metastatic tumor-derived CHLA 10 cells express higher levels of LGR5 than CH LA9 cells, w hich were derived from the primary tumor at diagnosis. (B) LGR5 expression was found by microarray analysis to be increased 1 n tumors from 4 patients w 1 th rapidly progressive and fatal pr1mary ES (DOD-dead of disease) compared to 10 pat1 ents with at least 48 months disease free survival (L TS-Iong term www.frontiersin.org LGR5 in Ew ing sarcoma It has been previously reported that LGR5 expression is enriched in cancer cells with stem cell-like properties (Merlos Suarez et al., 2011; Kemper et al., 2012; Kobayashi et al., 2012; Nakata et al., 2013). These cells have the unique ability to self renew and differentiate compared to bulk tumor cells and can contribute to metastasis and drug resistance (Visvader and Lin deman, 2012). Although putative esc were identified in a small cohort of primary tumors (Suva et al., 2009) it has been chal lenging to isolate putative CSC from established ES cell lines. Nevertheless, CD133 surface expression (Jiang et al., 2010) and high-level of ALDH activity (Awad et al., 2010) have been used to successfully enrich for CSC populations in STA-ET-8.2 and MHH-ES and TC71 cells, respectively. Therefore, we used these previously reported ES cell lines and CSC-enrichment assays to determine whether LGR5 expression is up regulated in putative CSC. qRT-PCR analysis of CD133-sorted STA-ET-8.2 popula tions consistently demonstrated increased expression of LGR5 in the CSC-enriched CD1 33+ fraction (Figure 3C). Similarly, levels of LGR5 were reproducibly higher in MHH-ES cells that dis played high ALDH activity than MHH-ES cells with low ALDH activity (Figure 3D). In contrast, data from TC71 were incon sistent. Although increased expression of LGR5 was detected in ALDHhigh CSC populations in one experiment this finding was not reproducible in two other independent experiments (Figure 3D). In summary, these findings suggest that LGR5 is highly expressed by at least some populations of CSC-like ES cells. In addition, they provide preliminary evidence in support of the hypothesis that over-expression of LGR5 in primary tumors may be associated with an aggressive drug-resistant clinical phenotype. B 10 ?;- ••• ·u; 8 c • J!l .!::: 6 ro • •• c "' p; .001 ·u; 4 "' =:· 0::: ~ 2 .... 0 DOD LTS Patient Outcome D 4 • j" 3 a;~ > I ~0 • "' -' 2 O:::<t: 3~ •• g_1 • • 0 MHH-ES TC71 survivors) (C) LGR5 levels are higher in CD 133+ compared to CD133- STA-El'8.2 cells. Data from two independent sort1ng experiments are shown and expression on they-axis is of CD133+ cells relative to t he corresponding CD133- cells. (D) LGR5 expression IS increased in ALDH""" compared to ALDH 1 '""' MHH-ES and TC71 cells. Data from three independent sorts are shown and expression on the y-axis is that of ALDH" 9 " relative to the corresponding ALDH'~., cells. The horizontal line in (B- D) all represent arithmet iC mean values. April 201 3 I Volume 3 1 Article 81 I 5 100 Scannell et al Studies are ongoing to determine if this association can be validated in a larger cohort of tumors. R-SPONDIN POTENTIATES Wnt/~-CATENlN SIGNALING IN AN LGR5-DEPENDENT MANNER Given its role as a potentiator of Wnt/f -\-catenin signaling, we hypothesized that ES cells with high levels of LGR5 expression would demonstrate relatively high Wnt/fl-catenin transcriptional activity. To address this we first evaluated basal levels of the Wnt/[-1-catenin axis in ES cells that were grown in standard tis sue culture conditions. To begin we measured the expression of known Wnt/['1-catenin target genes in three different ES cell lines (A673, CHLA25, and TC71) and discovered that although the lev els of gene expression varied significantly among the three cell lines, there was no correlation with LGR5 levels (Figure 4A). In support of this, we observed no significant or reproducible change in the expression of Wnt/f-\-catenin target genes following LGR5 knockdown (Figure 48). In particular, loss of LGR5 did not result in down-regulation ofW nt targets suggesting that, in the context of standard tissue culture (i.e., in the absence of exogenous RSPO), LGRS has little impact on Wnt/~-catenin transcriptional activ ity (Figure 4B). Next, we used TCF-promoter luciferase-reporter assays to directly measure the level of TCF transcriptional activ ity in CHLA25 ES cells before and after LGR5 knockdown. These assays confirmed that LGR5 knockdown had only minimal impact on basal Wnt/~-catenin transcriptional activity in standard culture conditions (Figure 4C). Thus, in standard culture conditions ES cells show only minimal canonical Wnt activity and this activity is not impacted by LGRS. Next, we investigated whether the Wnt/f\-catenin axis is intact in ES cells and capable of activation in the presence of exoge nous ligands. In particular, we hypothesized that cells with high levels of endogenous LGR5 expression might be susceptible A ~ LGRS ~ " " e; e; c " i ~ ~ ·~ • ~ 3 .n FIGURE 41 No correlation exists between LGR5 expression and Wnt/~-catenin activity in ES cells in standard culture. (A) Canonical WntJti-catenin target gene expression was measured by qRT -PCR in ES ce ll lines. There was no apparent correlation between LGR5 expression and W nt/rl-catenin target genes AXIN2 and CD44. (B) No consist ent or significant change occurs in the expression of a panel ofWnt/1'1-catenin target genes Frontiers in Oncology I Pediatric Oncology LGR5 in Ew ing sarcoma to RSPO-mediated potentiation of Wnt/~-catenin signaling. To address this we first evaluated sub-cellular localization of f'l catenin in ES cells before and after exposure to exogenous Wnt3a. Consistent with a previous report (Uren et al., 2004), ES cells demonstrated little evidence of ~-catenin in their nuclei under basal conditions (Figure SA). In contrast, exposure of ES cells to Wnt3a CM resulted in robust nuclear translocation of ~-catenin (Figures SA,B). Indeed, Wnt3a induced nuclear translocation of ~-catenin in all three lines irrespective of their endogenous LGR5 expression levels (Figure SB). Although there are four RSPOs, RSP02 has been shown to have the highest af finity for both LGR4 and LGRS (Carmon et al., 2011 ). In addition, RSP02 is highly expressed in developing bone (Nam eta!., 2007; Hankenson et a!., 2010). Therefore, we next exposed ES cells to RSP02 either alone or in combination with Wnt3a and measured ~-catenin nuclear localization. The presence of RSP02 alone had no effect on sub cellular localization of 1 3-catenin (Figures SA,B). In contrast, the combination of Wnt3a CM plus RSP02 resulted in potentia tion ofWntm-catenin signaling in CHLA25 cells as demonstrated by further up-regulation and nuclear translocation of ~-catenin (Figures SA,B). We also measured the expression of the Wnt/f 'l catenin target gene AXIN2 in these four conditions. In keeping with increased nuclear localization of ~-catenin, we observed induction of AXIN2 expression following W nt3a stimulation and, in CHLA25 cells, potentiation of this up-regulation in the presence of RSP02 (Figure 6A). Likewise, TCF reporter activity increased in a stepwise fashion when cells were exposed first to Wnt3a alone and then both Wnt3a and RSP02 together (Figure 6B). Notably, potentiation of WntfB-catenin signaling by RSP02 was only sta tistically significant in CHLA25 cells, the cell line with the highest level of LGR5. To determine if LGRS was responsible for mediat ing this robust RSP0 2-dependent potentiation of Wnt/~-catenin signaling we evaluated the consequences of LGR5 knockdown on AXIN2 CD44 c D shNS • shLGR5#1 293A CHLA25 following LGR5 knockdow n. The dotted line represents a ratio of 1 between shLGR5#1 and shNS ce lls. (C) LGR5 knockdow n does not measurably decrease Wnt/1·1-catenln transcriptional activ1 ty. Luc1 ferase activity was determined using TCF reporter-transduced cell lines as described in Section " Material and Methods." A ll data shown is from three independent experiments and error bars are SEM . April 2013 1 Volume 3 1 Article 81 16 101 Scannell et al. A CHLA25 B n C] l-Ceii CM Cl L-Cell CM + RSP02 . Wnt3a CM • Wnt3a CM + RSP02 LGR5 in Ew ing sarcoma FIGURE 51 ExogenousWnt3a induces nuclear localization of p-catenin effect on ~-{;aten1n nuclear localization. (B) The percentage of CHLA25, A673, and is potentiated by RSP02. (A) CHLA25 cells grown in the presence of and TC71 cells w ith nuclear ~-catenin all increased with W nt3a CM, but Wnt3a conditioned medium (CM) rnduced nuclear localizatron of ~-catenin and CHLA25 showed the greatest increase in nuclear localrzation w ith the addrtion thrs was robustly potentiated by the addition of RSP02. RSP02 alone had no of RSP0 2. Data from three independent experiments and error bars are SEM. A D L·Ceii CM D L·Cell CM + RSP02 • Wnt3aCM • Wnt3a CM + RSP02 ""p<.005 "p<.05 c CHLA25 2 5 0 1 200 ~~ 150 ~~ 100 ~ -l 50 us I- ! 1 J FIGURE 61 RSP02 potentiates Wnt/P-catenin signaling in LGR5-high ES cells. (A) Exposure of ES cell lines to W nt3a conditioned medium ICMI rncreased AXIN2 expression, but expression was only potentiated by RSP02 in the LGR5-high cell line, CHLA25. (B) TCF reporter activity was induced in ES cells exposed to Wnt3a CM, but was only potentiated in CHLA25 cells by this signaling axis in CHLA25 cells. Significantly, in the context of Wnt3a and RSP02 ligands, knockdown of LGR5 resulted in reduced potentiation of WntfB-catenin transcriptional activity (Figure 6C). Interestingly, we also observed down-regulation of TCF reporter activity in Wnt3a-only treated cells (Figure 6C) sug gesting that low level endogenous production of RSPO by CHLA25 cells might also contribute to potentiation of the Wnt/~-catenin axis in a Wnt-rich microenvironment. www.frontiersin.org • *p<.OS the addition of RSP02. (C) TCF reporter activity was measured in control and LGR5 knockdown CHLA25 cells following exposure toWnt3a with/w ithout RSP02. Reporter activrty in the presence of RSP02 w as significantly diminished in LGR5 knockdown cells. All data shown is from three independent experiments and error bars are SEM. LGR5 DOES NOT PROMOTE ES CELL PROLIFERATION IN VITRO Leucine-rich repeat-containing G-protein coupled receptor 5 functions as a growth-promoting oncogene in several adult malig nancies including gastrointestinal tumors and gliomas (McClana han et al., 2006; Tanese et al., 2008; Nakata et al., 2013). In addition, LGR5 was recently shown to promote proliferation of neuroblastoma (Balamuth et al., 2010), another highly aggressive pediatric ttunor of neu ral crest origin (Jiang et al., 20 11 ). In order April 2013 1 Volume 3 1 Artr cle 81 17 102 Scannell et al. to determine if LGR5 promotes ES cell proliferation we gener ated ES cell lines with altered levels of expression of LGR5 as a consequence of stable RNA-interference-mediated knockdown or ectopic over-expression (Figures 7 A,B). As shown, although reduced proliferation of the CHLA25 cell line was observed with one IGR5-targeted knockdown sequence (shLGR5#1), this phe notype was not observed in A673 cells (Figure ?C). In addition, a second IGR5-targeted shRNA sequence (shLGR5#3) did not inhibit proliferation in any of the cell lines (Figure 7C). Forced over-expression of LGR5 in the IGR5-low cell line A673 also had no impact on proliferation (Figure 7D). Cellular viability was also measured by trypan blue staining and no change was observed in any of the cell lines following LGR5 knockdown or over-expression (data not shown). Therefore, in contrast to neuroblastoma cells (Balamuth et al., 2010), LGR5 does not promote the proliferation of ES cells in standard in vitro culture conditions. Next, we repeated proliferation studies in non-transduced, par ent cells in the presence ofWNT and RSPO, reasoning that changes in proliferation downstream of Wnt/B-catenin activation would only be apparent in cells that were exposed to the necessary ligands. Interestingly, no change in proliferation was observed in either A673 or CHLA25 cells with the addition ofWNT3A and/ or RSP02 (Figure 7E). Studies are ongoing to evaluate the phenotypic con sequences of LGR5-mediated potentiation of Wnt/B-catenin sig naling but these data support prior observations that this pathway does not promote ES cell prolif eration (Endo et al., 2008). CHLA25 iii z ,; ! 1 } ~ ~ 0: s s c CHLA25 . ..,.. shNS . ......, shNS LGR5 in Ew ing sarcoma In summary, these studies demonstrate that the WNT/fl catenin signaling axis is present and intact in ES cells but is not activated in standard cell culture conditions. Addition ofWnt lig and to ES cell culture media can activate the canonical pathway and, in CHLA25 cells that express high levels of LGR5, activation is potentiated by the addition of RSP02. The generalizability of this observation to other LGRS over-expressing ES cells and pri mary tumors now requires fmther investigation to fully elucidate the contribution of the LGRS-WNT /fl-catenin signaling axis toES pathogenesis. DISCUSSION In this study we have, for the first time, shown that LGR5 is expressed by ES cells, both in the context of primary tumors and in cell culture, but that expression levels vary significantly among different ES cell populations. In particular, we found that levels of LGR5 were highest in putative ES cancer stem cells as well as primary tumor biopsies obtained from patients with rapidly pro gressive and drug-resistant disease. The wide range and pattern of LGR5 expression in ES tumor samples and cell lines is consis tent with patterns of LGR5 expression that have been discovered in other human cancers. Specifically, recent studies of colorectal carcinoma support the hypothesis that LGRS is a marker of can cer stem cells in these tumors (Kemper et al., 2012; Kobayashi et al., 2012). In addition, studies of colorectal and gastric car cinoma, glioblastoma, and esophageal adenocarcinoma have all A673 B '1 A673 "' ~ w 200 .. ! ~ l--- § o"'- 'v<' !<'' 'v (j A673 D ~ 6 .,. shLGR51#1 ~ 6 ..... shlGR5#1 1i 1l l ] ~ l 2 !! •p<. OS Days E CHLA25 -e- PBS -- PBS . 4- RSP02 . ... RSP02 ~ ~ .!: 6 - WNT3A .!: 6 - 'MIIT3A 1l 1- VYf'.lT3A + RSP02 1l Days FIGURE 71 Leucine·rich repeat-containing G·protein coupled receptor 5 does not promote ES cell proliferation. (A) qRT -PCR confirmation of LGR5 knockdown in ES cell lines transduced w it h LGR5-targeted shRNAs lshNS = non-silencing cont rol). (B) Confirmation of LGR5 over-expression in A673 cells transduced w ith LGR5 over -expression construct (Empty, empty Frontiers in Oncology I Pediatric Oncology Days Days A673 Days vector control). (C) ES cell expansion was measured by MTS assay and was not affected by LGR5 knockdown. (D) LGR5 ove r-expression did affect cell growth in A673 cells. (E) Parental cell growth was not affected by the addition of WNT3A and/or RSP02 to standard culture conditions All data shown IS from two independent experiments and error bars are SEM. April 201 3 I Volume 3 I Article 81 18 103 Scannell et al demonstrated heterogeneity of LGR5 expression and shown that high IGR5levels are associated with worse outcomes (Becker et aL, 0; Simon el 2; et :1L, 2; ,;; I:;)_ Like- wise, several studies have found a connection between high LGR5 expression and chemoresistance (Bau;.;r d al ., 20 12) and metastasis (t IO;TJ.kah:lshict:1l.,201 l;Va!larl:lrcs aL, 2; \Vn ei aL, 2012) in gastrointestinal malignancies. Interest ingly, recent studies in mouse models of neuroblastoma (BaL1 1nu1 h aL, 0) and medulloblastoma (K;rh·;mchi <:i 2) also discovered increased expression of Lgr5 in the most aggressive tumors. Thus, there is now substantial evidence in tumors of both epithelial and neural origin to implicate LGRS as a marker of an aggressive cellular phenotype. Data to support a potential role for LGRS in tumors of mesenchymal origin is also now beginning to emerge.A recent report by !\ol l) described a novel splice variant of LGR5 in the context of soft tissue sarcoma and reported that low level expression of this variant transcript (which lacks exon 5) was associated with worse overall and event free survival. In addition, the LGR5 locus was recently identified to be among the most frequently amplified loci in a genome-wide study of soft tissue sarcomas (copy number in top 1 o/o of nearly 19,000 genes) suggesting that up-regulation of LGR5 may contribute to sarcoma pathogenesis (B:HTei tna ct aL, 0). Our current study adds ES to the list of tumors where LGR5 may serve as a marker of aggres sive disease. We are now actively pursuing this hypothesis in large cohorts of primary tumor samples as well as model systems. Although LGRS was identified some years ago as a Wnt responsive gene, its role as a potentiating receptor upstream of Wnt signaling was only recently discovered (C:umon cl aL, II; d,; Lm ct aL, 2011; CHnl,_a ct 201 l; ct aL, 2012). LGRS and its related receptor family members LGR4 and LGR6 have now all been established as a receptors for the RSPO-family of ligands, which act to potentiate Wnt/1)-catenin signaling by complexing with Frizzled!LRP receptors (Cannon ct aL, II; d,; L:m aL, l; c;Hnka ei II; ei 12). There is now substan- tial evidence that hijacking ofWnt signaling may be important for the initiation and maintenance of ES, although the precise con tribution of 1)-catenin-dependent and - independent pathways remains to be elucidated (Urcn ci aL, 2004; aL, 0; 2011; J-l:_mcr ct aL, 20 U). Significantly, l aL (20(H) measured expression ofthe canon- ical Wnt ligands WNTl, WNT2, and WNT3 and found them to be undetectable in ES. In addition, only low levels of 1)-catenin nuclear localization were detected in cell lines and tumor sam ples. Likewise, et aL 0) reported low levels of basal TCF transcriptional activity in ES cell lines. Studies of DKK1 and DKK2, known modulators ofWnt signaling, have demonstrated that while the former is repressed in ES et l 0) the latter is induced by EWS-FLI1 and promotes cellular inva sion and metastasis (:'vl 20()'); ! Cl aL, l 3). In addition, a recent report showed that the non-canonical Wnt, WNTSA, induces ES cell migration and that the Wnt inhibitor SFRPS is silenced in ES cells by DNA promoter methylation (J fn ct al., 2). Thus, although there is still much to be understood about the role ofWnt signaling in ES pathogenesis, the data thus far support a more dominant role for Wnt deregulation in alter ing cell morphology, motility, and invasion rather than classical www.frontiersin.org LGR5 1n Ewmg sarcoma 1)-catenin-mediated cellular proliferation. Consistent with these prior reports we found that, in standard culture conditions, basal levels of Wnt/1)-catenin signaling are low. In addition, despite the fact that Wnt3a and RSP02 were able to robustly activate and potentiate canonical Wnt signaling, respectively, we did not observe changes in cellular proliferation. Even the addition of RSP02 to the culture media of LGR5-high CHLA25 cells had no impact on cell proliferation. Likewise, LGR5 knockdown had no discernible impact on the expression ofWnt/1)-catenin target gene expression nor TCF reporter activity under basal conditions. However, when growth medium was supplemented with Wnt3a, 1)-catenin nuclear localization, and TCF transcriptional activity were robustly activated. In ES cells that express LGR5 this signal ing was further potentiated by the addition of RSP02 confirming a functional role for this receptor in cells that are exposed to LGRS ligand. Thus, our results together demonstrate that, in the presence of appropriate ligands, Wnt/1)-catenin signaling can be stimulated in ES cells and this stimulation is maximal in cells that express high levels of LGR5. The phenotypic effects of this increased Wnt activity remain nnclear but our studies thus far suggest that pro liferation is nnaffected and that both cell autonomous as well microenvironmental factors combine to determine the functional consequences of enhanced Wnt signaling and nuclear 1)-catenin localization in LGRs+ ES cells. Emerging evidence from other model systems supports the pos sibility that LGRS function might be highly contextually depen dent. Mouse studies have shown that Lgr5 is widely expressed during embryogenesis but only becomes limited to discrete stem cell populations post-natally, in particular stem cells of hair folli cles and the gastrointestinal tract (Barker 2007; B:nkcr :mJ '"''"'"'- uu l 0). In normal stem cells secretion ofRSPO from neigh boring cells supports Igr5-dependent self-renewal and prolifera tion by activating canonical Wnt signaling (B:_lrkcr ~Hhl I 0 ). Likewise, the initiation and proliferation of intestinal ade nomas and carcinomas is driven by Lgrs+ stem cells ( Bn rker el al., 12). However, data from both colorectal carcinoma cell lines as well as other tumor types complicate the scenario. Consistent with our ovm data, et l) found that LGRS did not affect proliferation of colorectal cancer cell lines but instead negatively regulated Wnt signaling, colony formation, and migration. Thus, in this cellular context LGRS functioned more as a tumor suppressor than tumor promoter. In addition, promoter hypermethylation and loss of function mutations in LGR5 and LGR6 have been discovered in some colorectal tumor samples, again suggesting that these genes could act as tumor sup pressors in some contexts (Sjoblom cl 2006; aL, 2000; Sousa ct aL, II). These conflicting reports demonstrate that LGR5 has differing roles during development and in different cel lular contexts. Thus, its contribution to cancer maintenance and progression is likely to be determined by both the genetic and epigenetic state of the affected cell as well as the surrounding microenvironment. The mechanism of transcriptional up-regulation of LGR5 in ES cell populations is unknown. Importantly, however, LGR5 is not induced by EWS-FLI1 and unpublished data from our ovm lab as well a previously published report (see d :1L, 0 supplementary data) suggest that LGR5 is, in fact, repressed by Apr112013IVolume31Artlcle81l9 104 Scannell et al EWS-FLI 1. Thus, we reasoned that expression of LGR5 in ES cells may instead be a reflection of their putative stem cell origins. Interestingly, among different neuro-mesenchymal stem cell pop ulations we discovered that undifferentiated NCSC expressed the highest levels of LGR5. Expression was still detectable, albeit at lower levels, inN CSC that had undergone epithelial-mesenchymal transition to an MSC-state (von Levetzow et al., 2011 ). In con trast, bone marrow-derived MSC did not express detectable levels of LGR5. Intriguingly, it is now established that rare MSC in the bone marrow are derived from the neural crest (Takashima et al., 2007; Nagoshi et al., 2008) and gene expression profiling data from mouse MSC (GEO accession GSE30419) showed increased expres sion of Lgr5 in neural crest -compared to mesoderm-derived MSC populations (Wislet-Gendebien et al., 2012). Based on these stud ies we now speculate that at least some ES might arise from neural crest -derived LGRs+ stem cells in the bone marrow. Further, we speculate that the RSPO-rich microenvironment of developing bone may contribute to malignanttransformation of these LGRs+ cells in the event that they acquire an EWS-ETS fusion. Ongo ing studies in our laboratory are now specifically addressing these intriguing hypotheses. In summary, we have shown that LGR5 is expressed by ES, in particular by putative cancer stem cells and, in the context of a Wnt REFERENCES A wad, 0., Yustein, J. T., Shah, P, Gul, N., Katuri, V, O'Neill, A., et al. (2010). High ALDH activity identifies chemotherapy-resist ant Ewing's sarcoma stem cells that retain sensitivity to EWS-FU1 inhibition. PI.oS ONE 5:e13943. doi:l 0. 13 71/journal. pone.0013943 Balam uth, N. J., and Womer, R. B. (2010). Ewing's sarcoma. lAncet Oncol. 11,184-192. Balamuth, N. J., Wood, A. , Wang, Q., Jagannathan, J., Mayes, P, Zhang, Z., et al. (2010). Serial transcrip tome analysis and cross-species integration identifies centromere associated protein E as a novel neu roblastoma target Canccr Res. 70, 2749-2758. Barker, N. , and Clevers, H. (2010). Leucine-rich repeat-containing G. protein-coupled receptors as mark ers of adult stem cells. Gastroenterol ogy 138, 1681- 1696. Barker, N., Huch, M., Kujala, P , Van De Wetering, M., Snippert, H. J., Van Es, J. H., et a!. (2010). Lgr5( + ve) stem cells drive self-renewal in the stom ach and build long-lived gastric units in vitro. Cell Stem Cell 6, 25-36. Barker, N., Ridgway, R. A., Van Es, J. H., Van De Wetering, M., Begthel, H., Van Den Born, M., et al. (2009). Crypt stem cells as the cells-of origin of intestinal cancer. Nature 457, 608- 611. Barker, N., Rookmaaker, M. B., Kujala, P , Ng, A., Leushacke, M., Snippert, H., et al. (2012). Lgr5(+ve) stem/progenitor cells contribute to nephro n formatio n during kidney development Cell Rep. 2, 540- 552. Barker, N., Van Es, J. H., Kuipers, L Kujala, P, Van D en Born, M., Cozi jnsen, M., eta!. (2007). Identification of stem cells in small intestine and colon by marker gene Lgr5. Nature 449, 1003- 1007. Barretina, J., Taylor, B. S., Banerji, S., Ramos, A. H., Lagos-Quintana, M., Decarolis, P L., et al. (2010). S ubtype-specific genomic alter ations define new targets for soft-tissue sarcoma therapy. Na t. Genet. 42, 715-721. Bauer, L., Langer, R., Becker, K., Hapfelmeier, A., Ott, K., Novotny, A., et al. (2012). Expression pro filing of stem cell-related gen es in neoadjuvant-treated gast ric can cer: a NOTCH2, GSK3B and beta-catenin gene signature pre dicts survival. PI.oS ONE 7:e44566. doi:l 0.13 71/journal.pone.0044566 Becker, L.,Huang, Q., and Mashimo, H. (2010). Lgr5, an intestinal stem cell marker, is abnormally expressed in Barrett's eso phagus and esophageal adenocarcinoma. Dis. Esophagus 2 3, 168- 174. Ben-Porath, I., Thomson, M. W., Carey, V J., Ge, R., Bell, G. W., Regev, A., et al. (2008). An embryonic stem cell-like gene expression signa ture in poorly differentiated aggres sive human tumors. Nat. Genet. 40, 499-507. Frontiers in Oncology 1 Ped1atr1 c Oncology LGR5 in Ew ing sarcoma and RSPO-rich microenvironment, LGRS fimctions to potentiate canonical Wnt/~-catenin signaling. In addition, we have discov ered that LGR5 is expressed by neural crest -derived stem cells, putative cells of ES origin, demonstrating that LGRS may also be a marker of some non-epithelial stem cells. Given the pro found complexity ofWnt signaling and its dependence on both cell autonomous as well as micro environmental cues it is now essen tial that functional studies of LGRS and the Wnt/~-catenin axis in ES be performed in model systems that faithfully recapitulate the in vivo tumor microenvironment. ACKNOWLEDGMENTS The authors wish to thank Drs. Diana Abdueva, Cornelia von Levetzow, Gregor von Levetzow, as well as Nikhil S hyam, and Matthew Thayer for their t echnical support and members of the Lawlor lab for helpful discussion. We are grateful to Drs. Tim othy Triche, Heinrich Kovar, Darwin Prockop for cell lines, the CHLA and Children's Oncology Group Biorepositories for ES tumor specimens, and Dr. Pat Reynolds for STR profiling. Fund ing for this work was provided by the following research grants: 1R01-CA134604, 1U01 CA114757,SU2C-AACR-IRG1309,Nancy Newton Loeb Pediatric Cancer Research Award (to EliZ'l.beth R. Lawlor) and 1R01CA138212-01 (to David M. Loeb). Carmon, K. S., Gong, X., Lin, Q., Thomas, A. , and Liu, Q. (2011). R spondins functio n as ligands of the orphan receptors LGR4 and LGR5 to regulate Wnt/beta-catenin sign al ing. Proc. Natl. Acad. Sci. U.SA. 108, 11452-11457. Chan, T. A., Glockner, S., Yi, J. M., Chen, W , Van Neste, L, Cope, L., et al. (2008). Con vergence of mutation and epige netic alterations identifies common genes in can cer that predict for poor prognosis. PLoS Med. S:e114. doi:l 0.1 371/journal. pmed.0050114 Clevers, H., and Nusse, R. (2012). Wnt/beta-catenin signaling and d is ease. Cel/ 149, 11 92- 1205. Cui, W, Taub, D. D., and Gardner, K. (2007). qPrimerDepot: a primer database for quantitative real time PCR. Nucleic Acids Res. 35, DSOS- 809. de L.au, W , Barker, N., Low, T. Y, Koo, B. K., Li, V S ., Teunissen, H. , et al. (2011 ). Lgr5 homologues associate with Wnt recepto rs and m ediate R-spondin signalling. Nature 4 76, 293-297. de Sousa, E. M. F., Colak, S., Buikhuisen, J., Koster, J., Cameron, K., De Jong, J. H., et al. (2011). Methylation of cancer-stem -cell associated Wnt target gen es pre dicts poor prognosis in colorectal cancer patients. Cell SIEm Cell 9, 476-485. Endo, Y, Beauchamp, E., Woods, D., Taylor, W G., Toretsky, J. A., Uren, A., et al. (2008). Wnt- 3a and Dickkopf-1 stimulate neu rite outgrowth in Ewing tumor cells via a Frizzled3- and c Jun N-terminal kinase-dependent mechanism Mol. CelL Bioi. 2 8, 2368-2379. Feldhahn, N., Henke, N., Melchior, K., D uy, C., Soh, B. N. , Klein, F., et al. (2007). Activation induced cytidine deaminase acts as a mutator in BCR-ABLl transformed acute lymphoblastic leukemia cells. f. Exp. Med. 204, 1157- 1166. Friedman, M. S., Oyserman, S. M., and Hankenson, K. D. (2009). Wnt11 promotes osteoblast matu ration and mineralization through R-spondin 2. ]. BioL Chern. 284, 14117- 14125. Fuerer, C., and Nu sse, R. (2010). Lentiviral vectors to probe and manipulate the Wnt signaling pathway. P I.oS ONE S:e9370. doi:l 0.13 71/journal. pone.0009370 Glinka,A.,Dolde, C., Kirsch, N.,Huang, Y L, Kazanskaya, 0 ., Ingelfinger, D. , et al. (2011). LGR4 and LGR5 are R-spondin receptors mediating Wnt/beta-catenin and Wnt/PCP sig nalling. EMBO Rep. 1 2, 1055-1061. Gong, X., Carmon, K. S., Lin, Q., Thomas, A., Y i, J., and Liu, Q. (2012). LGR6 is a high affin ity receptor of R-spondins and potentially functions as a tumor suppressor. PI.oS ONE 7:e37137. doi:10.1 371/journal.pone.0037137 Apni 2013 1Volu me 3 1 Art icle 81 110 105 Scannell et al Hankenson, K. D., Sweetwyne, M. T., Shitaye, H., and Posey, K. L. (2010). Thrombospondins and novel TSR-containing proteins, R spondins, regulate bone formation and remodeling. Curr. Osteoporos. Rep. 8, 68-76. Hauer, K., Calzada-Wack, J., Steiger, K., Grunewald, T. G., Baumhoer, D., Plehm, S., eta!. (2013). DKK2 medi ates osteolysis, invasiveness, and metastatic spread in Ewing sarcoma. Camer Res. 73, 967-977. Jaks, V, Barker, N., Kasper, M., VanEs, J. H., Snippert, H. J., Clevers, H., eta!. (2008). Lgr5 marks cycling, yet long lived, hair follicle stem cells. Nat. Genet. 40,1291-1299. Jiang, M., Stanke, J., and Lahti, J. M. (2011). The connections between neural crest development and neu roblastoma. Curr. Top. Dev. Bioi. 94, 77-127. Jiang, X., Gwye, Y, McKeown, S. J., Bronner-Fraser, M., Lutzko, C., and Lawlor, E. R. (2009). Isolation and characterization of neural crest stem cells derived from in vitro differentiated human embryonic stem cells. Stem Cells Dev. 18, 1059-1070. Jiang, X., Gwye, Y, Russell, D., Cao, C., Douglas, D., Hung, L., et a!. (2010). CD133 expressioninchemo resistant Ewing sarcoma cells. BMC Camer 10:116. doi:10.1186/1471- 2407-10-116 Jin, Z., Zhao, C., Han, X., and Han, Y (2012). Wnt5a promotes ewing sar coma cell migration through upreg ulating CXCR4 expression. BMC Camer 12:480. doi:10.1186/1471- 2407-12-480 Kawauchi, D., Robinson, G., Uziel, T., Gibson, P, Rehg, J., Gao, C., et a!. (2012). A mouse model of the most aggressive subgroup of human medulloblastoma. Cancer Cell 21, 168--180. Kemper, K., Prasetyanti, P. R., De Lau, W., Rodermond, H., Clevers, H., and Medema, J. P (2012). Mono clonal antibodies against Lgr5 iden tify human colorectal cancer stem cells. Stem Cells 30,2378--2386. Kobayashi, S., Yamada-Okabe, H., Suzuki, M., Natori, 0., Kato, A., Matsubara, K., eta!. (2012). LGR5- positive colon cancer stem cells interconvert with drug-resistant LGR5-negative cells and are capable of tumor reconstitution. Stem Cells 30,2631-2644. McClanahan, T., Koseoglu, S., Smith, K., Grein, J., Gustafson, E., Black, S., et a!. (2006). Identification of overexpression of orphan G protein coupled receptor GPR49 in human www.frontiersin.org colon and ovarian primary tumors. Camer Bioi. Ther. 5, 419-426. Meltzer,P. S. (2007). Is Ewing's sarcoma a stem cell tumor? Cell Stem Cell 1, 13-15. Merlos-Suarez, A., Barriga, F. M., Jung, P., Iglesias, M., Cespedes, M. V, Rossell, D., eta!. (2011 ). The intesti nal stem cell signature identifies col- o rectal cancer stem cells and pre dicts disease relapse. Cell Stem Cell 8, 511-524. Miyagawa, Y, Okita, H., Itagaki, M., Toyoda, M., Katagiri, Y. U., Fuji moto, J., et a!. (2009). EWS/ETS regulates the expression of the Dickkopf family in Ewing family tumor cells. PLoS ONE 4:e4634. d oi: 1 0.13 71/ journal. pone. 0004634 Nagoshi, N., Shibata, S., Kubota, Y, Nakamura, M., Nagai, Y., Satoh, E., eta!. (2008). Ontogeny and multi potency of neural crest-derived stem cells in mouse bone marrow, dorsal root ganglia, and whisker pad. Cell Stem Cell2, 392-403. Nakata, S., Campos, B., Bageritz, J., Lorenzo Bermejo, J., Becker, N., Engel, F., et a!. (2013). LGR5 is a marker of poor prognosis in glioblastoma and is required for sur vival of brain cancer stem-like cells. Brain Pathol. 23,60-72. Nam, J. S., Turcotte, T. J., and Yoon, J. K. (2007). Dynamic expression of R-spondin family genes in mouse development Gene Expr. Patterns 7, 306--312. Navarro, D., Agra, N., Pestana, A., Alonso, J., and Gonzalez-Sancho, J. M. (2010). TheEWS/FLil oncogenic protein inhibits expression of the Wnt inhibitor DICKKOPF-1 gene and antagonizes beta-catenin!TCF mediated transcription. Carcinogen esis 31, 394-401. Phillips, H. S., Kharbanda, S., Chen, R., Forrest, W. F., Soriano, R. H., Wu, T. D., et a!. (2006). Molecular subclasses ofhigh-gradeglioma pre dict prognosis, delineate a pattern of disease progression, and resemble stages in neurogenesis. Cancer Cell9, 157-173. Riggi, N., Suva, M. L., Suva, D., Cironi, L., Provero, P., Tercier, S., et a!. (2008). EWS-FLI-1 expression trig gers a Ewing's sarcoma initiation program in primary human mes enchymal stem cells. Camer Res. 68, 2176-2185. Rot, S., Taubert, H., Bache, M., Gre ither, T., Wurl, P., Eckert, A. W., et a!. (2011). A novel splice variant of the stem cell marker LGR5/GPR49 is correlated with the risk of tumor-related death in soft-tissue sarcoma patients. BMC Camer 11:429. doi:10.1186/1471- 2407-11-429 Schepers, A. G., Snippert, H. J., Stange, D. E., VanDen Born, M., VanEs, J. H., Van De Wetering, M., eta!. (2012). Lineage tracing reveals Lgr5+ stem cell activity in mouse intestinal ade nomas. Science 337,730-735. Simon, E., Petke, D., Boger, C., Behrens, H. M., Warneke, V, Ebert, M., et a!. (2012). The spatial dis tribution of LGR5+ cells cor relates with gastric cancer pro gression. PLoS ONE 7:e35486. d oi: 1 0.13 71/j ournal. pone. 003 5486 Sjoblom, T., Jones, S., Wood, L. D., Parsons, D. W., Lin, J., Barber, T. D., et a!. (2006). The consensus coding sequences of human breast and colorectal cancers. Science 314, 268--274. Spike, B. T., Engle, D. D.,Lin, J. C., Che ung, S. K., La, J., and Wahl, G. M. (2012). A mammary stem cell pop ulation identified and characterized in late embryogenesis reveals simi larities to human breast cancer. Cell Stem Cell10, 183-197. Staege, M. S., Hutter, C., Neumann, 1., Foja, S., Hattenhorst, U. E., Hansen, G., et a!. (2004). DNA microarrays reveal relationship of Ewing fam ily tumors to both endothelial and fetal neural crest-derived cells and define novel targets. Camer Res. 64, 8213-8221. Suva, M L., Riggi, N., Stehle, J. C., Baumer, K., Tercier, S., Joseph, J. M., eta!. (2009). Identification of can cer stem cells in Ewing's sarcoma. Camer Res. 69,1776-1781. Takahashi, H., Ishii, H., Nishida, N., Takemasa, 1., Mizushima, T., Ikeda, M., et a!. (2011). Significance of Lgr5(+ve) cancer stem cells in the colon and rectum. Ann. Surg. Omol. 18,1166-1174. Takashima, Y,Era, T.,Nakao,K.,Kondo, S., Kasuga, M., Smith, A. G., et a!. (2007). Neuroepithelial cells supply an initial transient wave of MSC differentiation. Cell129, 1377-1388. Tanese, K., Fukuma, M., Yamada, T., Mori, T., Yoshikawa, T., Watanabe, W., eta!. (2008). G-protein-coupled receptor GPR49 is up-regulated in basal cell carcinoma and promotes cell proliferation and tumor forma tion. Am. f. Patho/.173, 835-843. Tirade, F., Laud-Duval, K., Prieur, A., Delorme, B., Charbord, P, and Delattre, 0. (2007). Mesenchymal stem cell features of Ewing tumors. Camer Cell11, 421-429. Uchida, H., Yamazaki, K., Fukuma, M., Yamada, T., Hayashida, T.,Hasegawa, H., et a!. (2010). Overexpression of leucine-rich repeat-containing G LGR5 1n Ewmg sarcoma protein-coupled receptor 5 in col orectal cancer. Camer Sci. 101, 1731-1737. Uren, A., Wolf, V., Sun, Y. F., Azari, A., Rubin, J. S., and Toretsky, J. A. (2004). Wnt/Frizzled signaling in Ewing sarcoma. Pediatr. Blood Can cer 43,243-249. Valladares-Ayerbes, M., Blanco-Calvo, M., Reboredo, M., Lorenzo-Patina, M. J., Iglesias-Diaz, P., Haz, M., et a!. (2012). Evaluation of the adenocarcinoma-associated gene AGR2 and the intestinal stem cell marker LGR5 as biomarkers in colorectal cancer. Int. f. MoL Sci. 13, 4367-4387. Vijayakumar, S., Liu, G., Rus, I. A., Yao, S., Chen, Y., Akiri, G., et a!. (2011). High-frequency canoni cal Wnt activation in multiple sar coma subtypes drives proliferation through a TCF/beta-catenin tar get gene, CDC25A. Camer Cell 19, 601-612. Visvader, J. E., and Lindeman, G. J. (2012). Cancer stem cells: current status and evolving complexities. Cell Stem Cell10, 717-728. von Levetzow, C., Jiang, X., Gwye, Y., Von Levetzow, G., Hung, L., Cooper, A., eta!. (2011). Modeling initiation of Ewing sarcoma in human neural crest cells. PLoS ONE 6:e19305. d oi: 1 0.13 71/jo urnal. pone. 00 19305 Walker, F., Zhang, H. H., Odor izzi, A., and Burgess, A. W. (2011). LGR5 is a negative reg ulator of tumourigenicity, antag onizes Wnt signalling and regu lates cell adhesion in colorectal can cer cell lines. PLoS ONE 6:e22733. d oi: 1 0.13 71/jo urnal. pone. 0022 733 Wislet-Gendebien, S., Laudet, E., Neir inckx, V., Alix, P., Leprince, P, Gle jzer, A., eta!. (2012). Mesenchymal stem cells and neural crest stem cells fromadultbone marrow: characteri zation oftheir surprising similarities and differences. Cell. Mol. Lifo Sci. 69,2593-2608. Wu, X.-S., Xi, H. Q., and Chen, L. (2012). Lgr5 is a potential marker of colorectal carcinoma stem cells that correlates with patient survival. World]. Surg. Oncol. 10,244. Conflict of Interest Statement: The authors declare that the research was conducted in the absence of any com mercial or financial relationships that could be construed as a potential con flict of interest Received: 25 February 2013; paper pend ing published: 19 March 2013; accepted: 28March2013; published online: 15April 2013. April 2013 1 Volume 3 1 Art1cle 81 111 106 Scannell et al Citation: Scannell CA, Pedersen .£4, Mosher JT, Krook .MA, Nicholls LA, Wilky BA, Loeb DM and Lawlor ER (2013) LGRS is expressed by Ewing sarcoma and potentiates Wnt/(3 -catenin signaling. Front. Oncol. 3:81. doi: 10.3389/fonc.2013.00081 This article was submitted to Frontiers in Pediatric Oncology, a specialty of Fron tiers in Oncology. Frontiers in Oncology 1 Ped1atr1c Oncology Copyright © 2013 Scannell, Pedersen, Mosher, Krook, Nicholls, Wtlky, Loeb and Lawlor. This is an open-access arti cle distributed under the terms of the Creative Commons Attribution License, LGR5 1n Ewmg sarcoma which permits use, distribution and reproduction in other forums, provided the original authors and source are cred ited and subject to any copyright notices concerning any third-party graphics etc. Apr112013IVolume 31Artlcle 81 112 107

Abstract (if available)

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Harnessing the power of stem cell self-renewal pathways in cancer: dissecting the role of BMI-1 in Ewing’s sarcoma initiation and maintenance

PDF

Molecular role of EZH2 overexpression in Colorectal Cancer progression

PDF

Identification and characterization of adult stem cells in the oral cavity

PDF

Identification of CBP/FOXM1 as a molecular target in triple negative breast cancer

PDF

Differential role of two coactivators, CCAR1 and CARM1, for dysregulated beta-catenin activity in colorectal cancer cell growth and gene expression

PDF

β-catenin couples self-renewal, induction and epithelial morphogenesis and patterning at the outset of mammalian nephrogenesis

PDF

The role of fibroblast growth factor signaling on postnatal hepatic progenitor cell expansion

PDF

Wnt/β-catenin/p300 induced transcription is critical for the differentiation and maintenance of Paneth cells

PDF

The cancer stem-like phenotype: therapeutics, phenotypic plasticity and mechanistic studies

PDF

Genomic risk factors associated with Ewing Sarcoma susceptibility

PDF

Steatosis induces Wnt/β-catenin pathway to stimulate proliferation of hepatic tumor initiating cells and promote liver cancer development

PDF

Exploration of the roles of cancer stem cells and survivin in the pathogenesis and progression of prostate cancer

PDF

Immunomodulatory and regenerative potential of amniotic fluid stem cells as a treatment strategy for pulmonary fibrosis

PDF

Role of cancer-associated fibroblast secreted annexin A1 in generation and maintenance of prostate cancer stem cells

PDF

Functional role of a Ewing-sarcoma-specific vlncRNA in tumor growth and progression

PDF

The noncanonical role of telomerase in prostate cancer cells: exploring a non-telomeric signaling role for telomerase protein (TERT) in a cancer cell line

PDF

The role of adipocytes in acute lymphoblastic leukemia cell migration and survival against daunorubicin

PDF

PTEN deletion induced tumor initiating cells: Strategies to accelerate the disease progression of liver cancer

PDF

Bone marrow derived mesenchymal stem cells promote survival and drug resistance in tumor cells

PDF

Identification and characterization of immune-escape mechanisms in solid tumors

Asset Metadata

Creator Scannell, Christopher (author)

Core Title The role of LGR5 in the pathogenesis of Ewing sarcoma: a marker of aggressive disease and a contributor to the malignant phenotype

School Keck School of Medicine

Degree Doctor of Philosophy

Degree Program Systems Biology and Disease

Publication Date 07/29/2014

Defense Date 06/12/2013

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag Ewing sarcoma,LGR5,OAI-PMH Harvest,R-spondin,stem cell,Wnt,β-catenin

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Taylor, Clive R. (committee chair), Asgharzadeh, Shahab (committee member), Epstein, Alan L. (committee member), Goldkorn, Amir (committee member)

Creator Email scannell@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c3-304380

Unique identifier UC11293726

Identifier etd-ScannellCh-1872.pdf (filename),usctheses-c3-304380 (legacy record id)

Legacy Identifier etd-ScannellCh-1872.pdf

Dmrecord 304380

Document Type Dissertation

Format application/pdf (imt)

Rights Scannell, Christopher

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA