Characterization of genomic subtypes of Ewing sarcoma based on copy number alterations

Ewing sarcoma is the second most common bone cancer in children, with an incidence of 3 cases per million people in the United States. The 5-year survival for patients with primary tumors is approximately 70%, while 5-year survival for patients with metastasis at the time of diagnosis is less than 30%. These numbers have not improved over the past 3 decades, suggesting a need for superior therapies and a better understanding of the molecular pathogenesis of this disease. Ewing sarcoma is characterized by an EWS-ETS translocation, which functions as an aberrant and oncogenic transcription factor. Like other translocation-driven cancers, point mutations are infrequent in Ewing sarcoma. The more frequent and recurring genetic aberration in Ewing sarcoma, aside from the EWS-ETS translocation, is copy number alteration (CNA). This dissertation characterizes two genomic subtypes of Ewing sarcoma based on CNAs: simple (containing CNAs in <1% of the length of the genome), and complex (containing CNAs in ≥1% of the length of the genome). Our data show that the complex subtype trends toward worse outcome than the simple subtype. We further explore specific genes and pathways that are altered by CNAs in the complex subtype of Ewing sarcoma, including the TP53 pathway, CEBPB, and HOTAIR. Although occurring in only 5% of Ewing sarcoma cases, TP53 mutations are tightly associated with the complex subtype. We test the functional consequences of TP53 mutations in Ewing sarcoma cell lines and observe that TP53 mutant cell lines are iv resistant to chemotherapeutic drugs and radiation. We also explore alternate mechanisms of p53 pathway suppression through CNAs of p53 pathway members CDKN2A, MDM2, and MDM4. Copy number gain of CEBPB occurs in 15% of Ewing sarcoma. We demonstrate how CEBPB acts as an oncogene by promoting attachment-independent cell growth, resistance to chemotherapeutic drugs, and regulation of the cancer stem cell marker ALDH1A1. HOTAIR is a long non-coding RNA that is upregulated and unmethylated in the genomically complex Ewing sarcoma subtype. We demonstrate that Ewing sarcoma cells deficient in HOTAIR have increased cell growth, viability, and migration, and we propose future studies to understand the exact mechanisms through which HOTAIR is involved in these processes.

CHARACTERIZATION OF GENOMIC SUBTYPES OF EWING SARCOMA BASED ON COPY NUMBER ALTERATIONS by Jamie Diane Gardiner A dissertation submitted to the faculty of The University of Utah in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Oncological Sciences The University of Utah December 2018 Copyright © Jamie Diane Gardiner 2018 All Rights Reserved The University of Utah Graduate School STATEMENT OF DISSERTATION APPROVAL The dissertation of Jamie Diane Gardiner has been approved by the following supervisory committee members: Joshua David Schiffman , Chair 10/22/2018 Date Approved Kevin Bruce Jones , Member 10/22/2018 Date Approved Katherine Elena Varley , Member 10/22/2018 Date Approved Christopher T. Gregg , Member 10/22/2018 Date Approved Trudy Oliver , Member Date Approved and by the Department/College/School of Bradley Cairns , Chair/Dean of Oncological Sciences and by David B. Kieda, Dean of The Graduate School. ABSTRACT Ewing sarcoma is the second most common bone cancer in children, with an incidence of 3 cases per million people in the United States. The 5-year survival for patients with primary tumors is approximately 70%, while 5-year survival for patients with metastasis at the time of diagnosis is less than 30%. These numbers have not improved over the past 3 decades, suggesting a need for superior therapies and a better understanding of the molecular pathogenesis of this disease. Ewing sarcoma is characterized by an EWS-ETS translocation, which functions as an aberrant and oncogenic transcription factor. Like other translocation-driven cancers, point mutations are infrequent in Ewing sarcoma. The more frequent and recurring genetic aberration in Ewing sarcoma, aside from the EWS-ETS translocation, is copy number alteration (CNA). This dissertation characterizes two genomic subtypes of Ewing sarcoma based on CNAs: simple (containing CNAs in <1% of the length of the genome), and complex (containing CNAs in ≥1% of the length of the genome). Our data show that the complex subtype trends toward worse outcome than the simple subtype. We further explore specific genes and pathways that are altered by CNAs in the complex subtype of Ewing sarcoma, including the TP53 pathway, CEBPB, and HOTAIR. Although occurring in only 5% of Ewing sarcoma cases, TP53 mutations are tightly associated with the complex subtype. We test the functional consequences of TP53 mutations in Ewing sarcoma cell lines and observe that TP53 mutant cell lines are resistant to chemotherapeutic drugs and radiation. We also explore alternate mechanisms of p53 pathway suppression through CNAs of p53 pathway members CDKN2A, MDM2, and MDM4. Copy number gain of CEBPB occurs in 15% of Ewing sarcoma. We demonstrate how CEBPB acts as an oncogene by promoting attachment-independent cell growth, resistance to chemotherapeutic drugs, and regulation of the cancer stem cell marker ALDH1A1. HOTAIR is a long non-coding RNA that is upregulated and unmethylated in the genomically complex Ewing sarcoma subtype. We demonstrate that Ewing sarcoma cells deficient in HOTAIR have increased cell growth, viability, and migration, and we propose future studies to understand the exact mechanisms through which HOTAIR is involved in these processes. iv TABLE OF CONTENTS ABSTRACT ....................................................................................................................... iii LIST OF FIGURES .......................................................................................................... vii LIST OF TABLES ............................................................................................................. ix ACKNOWLEDGEMENTS ................................................................................................ x Chapters 1. INTRODUCTION .......................................................................................................... 1 Ewing sarcoma presentation and incidence ............................................................... 1 Treatment ................................................................................................................... 2 EWS-ETS translocations............................................................................................. 3 Cell of origin .............................................................................................................. 4 Genomic landscape of Ewing sarcoma ...................................................................... 4 Copy number alterations in Ewing sarcoma .............................................................. 5 TP53 ........................................................................................................................... 7 CEBPB ....................................................................................................................... 8 HOTAIR ..................................................................................................................... 9 Dissertation goals ..................................................................................................... 10 References ................................................................................................................ 12 2. TP53 STATUS ASSOCIATES WITH GENOMIC COMPLEXITY, DNA DAMAGE RESPONSE, AND CLINICAL OUTCOME IN EWING SARCOMA ........................... 26 Abstract .................................................................................................................... 26 Introduction .............................................................................................................. 27 Results ...................................................................................................................... 28 Discussion ................................................................................................................ 36 Materials and methods ............................................................................................. 40 Acknowledgements .................................................................................................. 44 References ................................................................................................................ 44 3. C/EBPβ-1 PROMOTES TRANSFORMATION AND CHEMORESISTANCE IN EWING SARCOMA CELLS ........................................................................................... 78 Abstract .................................................................................................................... 79 Introduction .............................................................................................................. 79 Results ...................................................................................................................... 80 Discussion ................................................................................................................ 87 Materials and methods ............................................................................................. 88 Acknowledgements .................................................................................................. 90 References ................................................................................................................ 91 Supplementary materials .......................................................................................... 93 4. CONCLUSIONS AND FUTURE DIRECTIONS ....................................................... 97 Genomic subtypes of Ewing sarcoma ...................................................................... 98 TP53 in Ewing sarcoma ........................................................................................... 99 CEBPB in Ewing sarcoma ..................................................................................... 101 HOTAIR in Ewing sarcoma ................................................................................... 102 Other genes ............................................................................................................ 104 References .............................................................................................................. 105 APPENDIX: THE LONG NON-CODING RNA HOTAIR IN EWING SARCOMA ... 110 Abstract .................................................................................................................. 110 Introduction ............................................................................................................ 111 Materials and methods ........................................................................................... 113 Results .................................................................................................................... 116 Discussion and future directions ............................................................................ 118 References .............................................................................................................. 120 vi LIST OF FIGURES Figures 1.1. Schematic of EWSR1 and FLI1 gene domains. ......................................................... 22 1.2. The p53 pathway. ....................................................................................................... 23 1.3. C/EBPβ isoforms ....................................................................................................... 24 1.4. Schematic of HOTAIR interaction with chromatin remodelers ................................. 25 2.1. Simple and complex subtypes of Ewing sarcoma...................................................... 48 2.2. TP53 mutated Ewing sarcoma cell lines are resistant to chemotherapies ................. 49 2.3. TP53 mutated Ewing sarcoma cell lines are resistant to IR....................................... 50 2.4. P53 pathway CNAs in Ewing sarcoma ...................................................................... 52 S2.1. Profile of the simple and complex Ewing sarcoma subtypes .................................. 53 S2.2. Copy number profile of Ewing sarcoma cell lines .................................................. 54 S2.3. TP53 mutated Ewing sarcoma cell lines are resistant to chemotherapies ............... 55 S2.4. TP53 mutated Ewing sarcoma cell lines are resistant to IR .................................... 56 S2.5. Relative TP53 mRNA expression in Ewing sarcoma samples with 2 or 1 copies of TP53 .................................................................................................................................. 57 3.1. C/EBPβ expression in Ewing sarcoma ...................................................................... 81 3.2. C/EBPβ expression in Ewing sarcoma cells is regulated by EWS-FLI1 ................... 83 3.3. C/EBPβ isoform expression does not affect cell proliferation or viability in 2D culture ............................................................................................................................... 84 3.4. C/EBPβ-1 promotes attachment-independent cellular transformation ...................... 85 3.5. ALDH1A1 is a target of C/EBPβ................................................................................ 86 3.6. ALDH activity is influenced by C/EBPβ expression ................................................. 87 3.7. C/EBPβ overexpression leads to chemoresistance in Ewing sarcoma cells .............. 88 S3.1. CEBPB expression decreases with EWS-FLI1 knockdown .................................... 93 S3.2. Depletion of C/EBPβ by shRNA ............................................................................. 93 S3.3. Microarray individual gene mRNA expression ....................................................... 94 S3.4. ALDH activity is regulated by C/EBPβ expression ................................................ 95 S3.5. C/EBPβ overexpression leads to chemoresistance in Ewing sarcoma cells ............ 95 A.1. HOTAIR expression in Ewing sarcoma .................................................................. 124 A.2. CRISPR knockout of HOTAIR in Ewing sarcoma cells ......................................... 125 A.3. HOTAIR deficiency affects Ewing sarcoma cell growth ........................................ 126 A.4. Chemosensitivity of HOTAIR-deficient Ewing sarcoma cells. ............................... 127 viii LIST OF TABLES Tables 2.1 Factors associated with the simple and complex Ewing sarcoma subtypes ............... 58 2.2 Univariate and multivariate survival analysis ............................................................. 59 2.3 Relationship between simple/complex subtypes and TP53/STAG2 mutation status .. 60 2.4 Univariate and multivariate survival analysis of p53 pathway CNAs ........................ 61 S2.1 Summary of Ewing sarcoma patients included in the discovery cohort ................... 62 S2.2 Summary of Ewing sarcoma public data included in the validation cohort ............. 68 S2.3 Influence of metastases at diagnosis and the simple/complex subtypes on event-free and overall survival ........................................................................................................... 73 S2.4 Relationship between TP53 CNA and TP53 mutation status in Ewing sarcoma ..... 74 S2.5 TP53 pathway copy number and mutation status in Ewing sarcoma cell lines ........ 75 S2.6 IC50 of Ewing sarcoma cell lines treated with doxorubicin, etoposide, or Nutlin-3a 76 S3.1 Most differentially expressed genes between C/EBPβ-1 knockdown and overexpression .................................................................................................................. 96 ACKNOWLEDGEMENTS I would like to thank my dissertation committee: Josh Schiffman, Kevin Jones, Chris Gregg, Trudy Oliver, and K-T Varley. Each has contributed valuable insight and advice throughout my graduate years. As my mentor, Josh has inspired me to continually focus on the patients, which has motivated me to work hard and efficiently. I would also like to thank members of the Schiffman lab, past and present, for encouragement, optimism, and friendship over the years. I especially want to thank my family for their continual support, perspective, and love. I could not have accomplished this without the lifetime example of hard work and persistence from my parents. Thank you. CHAPTER 1 INTRODUCTION Ewing sarcoma presentation and incidence Ewing sarcoma is the second most common bone cancer in children, with an incidence of approximately 3 cases per million people in the United States [1]. It was first discovered in 1921 by James Ewing, who described the tumor as a “diffuse endothelioma of bone” [2]. Ewing sarcoma is morphologically characterized as a small, round, blue-cell tumor and can arise in any bones in the body, but is more commonly found in the larger bones, such as the pelvis, femur, and ribs. Ewing sarcoma has a slight male bias (male to female ratio of 1.6:1), and arises more frequently in Caucasians than Hispanics or African Americans, suggesting that a genetic component exists that is not entirely understood [3, 4]. The 5-year survival for patients with primary Ewing sarcoma is approximately 70%. However, if there are metastases present at the time of diagnosis -- which happens in about 25% of cases -- that survival drops to <30% [1, 5]. These numbers have remained unchanged for nearly 3 decades, suggesting the need for improved therapies for these tumors and a better understanding of their pathogenesis. Although the presence of metastases at the time of diagnosis remains the leading negative prognostic factor for Ewing sarcoma, other negative prognostic factors include age (>14 years), and tumor size 2 (>8cm) [5]. Treatment Currently, nearly all treatment regimens for patients with Ewing sarcoma are roughly the same. Therapy includes a combination chemotherapy cycle of DNA damaging agents and includes vincristine, doxorubicin, ifosfamide, etoposide, actinomycin-D, and cyclophosphamide, followed by local control with surgery and/or radiation therapy when needed [6]. These aggressive treatments can have devastating long-term effects on children and young adults, such as infertility, anemia, heart problems, and physical, social, and emotional stresses. New targeted therapies for Ewing sarcoma have undergone clinical trials over the years, including IGF1R inhibitors, PD-1 inhibitors, and PARP inhibitors [7-9]. However, none of these therapies have significantly improved patient outcome, suggesting a need to better understand the molecular mechanisms involved in the pathogenesis of Ewing sarcoma, so novel targets and therapies can be developed. If all cases of Ewing sarcoma could be subtyped into more specific categories based on their genetic, morphological, or clinical features, we could devise more precise therapies for each individual patient. This has been done in other cancer types, such as breast cancer, lung cancer, and medulloblastoma [10-12]. Through the identification of specific molecular anomalies in these tumors, novel targeted therapies have been developed and treatments have improved. Genomic subtyping is likewise necessary for Ewing sarcoma. 3 EWS-ETS translocations Ewing sarcoma is characterized by translocations between the Ewing sarcoma breakpoint region 1 (EWSR1) gene on chromosome 22 with a member of the E26 transformation-specific (ETS) family of transcription factors, most commonly Friend leukemia integration 1 (FLI1) (85% of cases) on chromosome 11 [13]. This creates a fused gene that encodes the DNA-binding domain of FLI1 with the transcriptional activation domain of EWSR1, creating a powerful and aberrant transcription factor (Figure 1.1) [14]. In a minority of cases, EWSR1 fuses with other members of the ETS family of transcription factors, including ERG (10%), ETV4 (<5%), and ETV1 (<5%) [13]. EWS-ETS fusion proteins bind GGAA microsatellite regions to regulate transcription of target genes [15]. Microsatellites are short sequence repeats found in thousands of locations throughout the genome and were previously thought to be “junk DNA.” The DNA-binding domain of ETS proteins binds DNA at these GGAA microsatellite regions, and binding affinity is dependent on the length of these repeats [15, 16]. EWS-ETS can bind GGAA microsatellites in promoter regions (close to the target gene) or in enhancer regions (distant to the target gene) [17]. EWS-ETS fusions have the ability to activate or repress expression of its targets, and this ability is largely affected by the length of the GGAA repeats [18-20]. Furthermore, there are racial disparities found in the length of GGAA microsatellite repeats, which may account for the racial disparities found in the incidence of Ewing sarcoma [21]. 4 Cell of origin Though James Ewing first theorized an endothelial lineage because of its morphology, the cell of origin for Ewing sarcoma is still debated today [2]. Neural crest cells and mesenchymal stem cells are the most popular putative cells of origin. The neural crest lineage is supported by genomic expression signatures that are similar between Ewing sarcoma cells and neural crest stem cells, but not mesenchymal stem cells [22]; upregulation of neural crest development genes by EWS-FLI1 expression [23]; and differentiation of Ewing sarcoma cells into neural cells upon treatment with differentiation-inducing agents [24].On the other hand, the mesenchymal lineage is supported by expression signatures of Ewing sarcoma cells with EWS-FLI1 knockdown that are similar to that of mesenchymal stem cells [25], and expression of EWS-FLI1 in murine mesenchymal stem cells leads to transformation and tumorigenesis in vivo, with similar cellular morphology and expression signatures to Ewing sarcoma [26, 27]. Studies are still underway to characterize the cell of origin of this disease. Genomic landscape of Ewing sarcoma As is common with most translocation-driven pediatric cancers, Ewing sarcoma has a relatively low burden of point mutations compared to adult cancers. STAG2 and TP53 are the two most frequently mutated genes in Ewing sarcoma, with mutations arising in 15-20% and 5-10% of cases, respectively [28-30]. Furthermore, STAG2 and TP53 mutations co-associate in more aggressive Ewing sarcoma tumors [30]. Aside from STAG2 and TP53, the next most frequently mutated genes include epigenetic regulators such as EZH2, ZMYM3, and PRDM9, with a frequency of less than 5% each [30]. 5 Structural variants are common in Ewing sarcoma, including the EWS-FLI1 fusion and many recurrent copy number alterations (CNAs). Focal deletions of cyclindependent kinase inhibitor 2A (CDKN2A) on chromosome 9 occur in 10-20% of Ewing sarcoma. Copy number gains are often found in whole chromosome arms including 1q, 2, 5, 8, 12, and 20. Copy number loss is often observed in chromosomes 16q and chromosome 10. Copy number alterations in Ewing sarcoma A common characteristic of cancer is chromosome instability, or recurrent gains and losses of chromosomes throughout cell divisions [31]. Different cancer types are often hallmarked by specific copy number signatures. Through the use of comparative genomic hybridization (CGH) microarray and genotype SNP technology, these CNAs can be detected [32, 33]. Newer molecular inversion probe (MIP) technology offers similar results in tissues with fragmented DNA, such as frozen-fixed, paraffin-embedded (FFPE) samples [34]. This is an extraordinary advantage for Ewing sarcoma. Because of the low incidence of Ewing sarcoma in the population, it is often necessary to collect archived FFPE samples to obtain a sufficient sample size for studies. Compared to many adult cancers, Ewing sarcoma is considered relatively stable in terms of genomic integrity. These tumors do, however, frequently acquire copy number gains and losses throughout their genome. The most frequent CNA in Ewing sarcoma is gain of chromosome 8, followed by gain of chromosome 1q, gain of chromosome 12, loss of chromosome 16q, and loss of chromosome 10. The frequency and clinical impact that these CNAs have varied depending on the 6 study and number of samples, though common trends exist. Worse outcome is predicted in tumors with more CNAs than those with fewer CNAs [35-37]. Chromosome 8 trisomy is the most frequent CNA in Ewing sarcoma, found in 35-50% of cases [36, 38, 39]. Chromosome 8 trisomy trends towards worse outcome, though this does not always reach statistical significance [6, 38-40]. Gain of chromosome 1q occurs in approximately 1520% of cases and has been shown to have worse outcome [36, 39, 41]. Copy number loss of chromosome 16q occurs in about 10% of Ewing sarcoma, and is associated with metastatic disease at diagnosis and worse outcome [36, 38, 39]. Additionally, chromosome 1q gain and 16q loss often co-associate. Gain of chromosome 12 occurs in 15-30% of cases and is associated with worse outcome [36, 38-40]. Trisomy 8 and trisomy 12 often co-associate in Ewing sarcoma, but can also arise independently [38, 40, 42]. Although these CNAs in Ewing sarcoma have been observed by many independent studies, little is known about the molecular consequences of these changes. It is still not understood why certain regions of the genome are consistently gained or lost, and what gene or genes within those regions affect tumor growth and clinical outcome. Since CNAs are more common in Ewing sarcoma than mutations, one can hypothesize that CNAs are a greater contributor to pathogenesis. Other contributors may include epigenetic regulators, but epigenetics is also not well understood in the context of Ewing sarcoma. 7 TP53 Tumor protein 53 (TP53), named after the estimated weight of 53kDa of the p53 protein it encodes, is a tumor suppressor gene nicknamed “the guardian of the genome” [43-45]. P53 is a highly conserved transcription factor that binds DNA in tetramers to modulate expression of a myriad of signaling pathways, from DNA damage repair to cell death [46]. Although it is networked into several signaling pathways, p53 has four major functions in the cell: 1- it helps repair damaged DNA; 2- it arrests cell growth if extensive DNA repair is required; 3- it can direct a cell to apoptosis; and 4- it is involved in cell senescence (Figure 1.2) [47, 48]. Because of its powerful tumor suppressor capabilities, p53 is inactivated -- either directly through mutation or indirectly through aberrations of other members of the p53 pathway -- in a majority of cancers. TP53 is mutated in more than 50% of human cancers, but the frequency of TP53 mutations varies depending on the cancer type. TP53 mutations arise in 38-50% of ovarian, lung, esophageal, and colorectal cancers, but are found in less than 20% of leukemias, sarcomas, and cervical cancers [49-51]. For cancers with a low frequency of TP53 mutations, the TP53 pathway is often abrogated by other means, such as TP53 or cyclin-dependent kinase 2A (CDKN2A) deletions, and mouse double minute 2 (MDM2) or mouse double minute 4 (MDM4) amplifications [49, 50]. CDKN2A is a tumor suppressor that encodes two overlapping transcripts: P16-INK4A and P14ARF. P16-INK4a leads to cell cycle arrest while p14ARF binds to and inhibits MDM2 from shuttling p53 out of the nucleus for degradation [52, 53]. MDM2 is an E3-ubiquitin ligase that targets and shuttles p53 out of the nucleus for proteasomal degradation, thus negatively regulating p53 [54, 55]. MDM4 is also a 8 negative regulator of p53 and has structural homology to MDM2; however, MDM4 inhibits p53 by interacting with its transcriptional activation domain (see Figure 1.2) [56, 57]. MDM2 amplification occurs in less than 10% of cancers overall, but arises most frequently (20%) in soft tissue sarcomas and bone cancers [58, 59]. MDM4 is amplified in 10% of all cancers, but is less explored than MDM2 [60]. The frequency of CDKN2A alterations are also highly variable depending on the cancer type. The primary mechanisms through which CDKN2A is suppressed, however, include copy number deletions and promoter methylation [61]. CEBPB CCAAT/enhancer binding proteins (C/EBPs) are a family of basic-leucine zipper (bZIP) transcription factors composed of six members: C/EBPα, C/EBPβ, C/EBPγ, C/EBPδ, C/EBPε, and C/EBPζ. C/EBPs recognize and bind to CCAAT box motifs in gene promoters [62, 63]. The C-terminal end is highly conserved among members of this family and contains the bZIP domain, which is essential for binding DNA and forming homodimers and heterodimers with other members of the C/EBP family or other bZIP transcription factors [64, 65]. The N-terminus contains transcriptional activation domains for regulating expression of genes involved with cell proliferation, growth, metabolism, inflammation, liver function, adipocyte differentiation, and neural development [66-71]. The second member of this family to be discovered is CEBPB, located on chromosome 20, which encodes C/EBPβ. CEBPB is most highly expressed in the liver, but is also expressed in the intestine, lung, adipose tissue, kidney, bone, and spleen, among others [67, 72-74]. CEBPB encodes 3 protein isoforms from a single mRNA by 9 leaky ribosome slipping to alternate start codons. The isoforms produced are LAP* (liver activating protein, 38kDa), LAP (35kDa), and LIP (liver inhibitor protein, 20kDa) [7577]. LAP* and LAP contains transcriptional activation domains while LIP is lacking one (Figure 1.3). LIP thus functions primarily as a transcription repressor by forming nonfunctional heterodimers with other C/EBP transcription factors. C/EBPβ also functions in osteoblast differentiation and bone development, and is expressed in many cancers [78, 79]. It is thought that the ratio of isoform expression (LAP*/LAP/LIP) affects expression of downstream target genes and must be tightly regulated for proper cellular function [75]. HOTAIR As next-generation sequence technologies improve, more is understood about the human genome and its many functions. It is estimated that up to 98% of the human genome contains non-protein-coding genes [80]. These genes were originally thought to have no function, but researchers are continuously discovering the roles of non-coding RNAs in human biology. Humans have four clusters of sequentially regulated homeobox, or HOX, genes: HOXA-HOXD. HOX genes are important for proper development of embryos along their head-tail axis and are involved in establishing limb identity. HOX antisense intergenic RNA, or HOTAIR, is a long non-coding RNA on chromosome 12 near the HOXC gene cluster [81]. Although located within the HOXC regulatory region, HOTAIR does not regulate HOXC expression, but acts in trans and affects expression of the HOXD locus on chromosome 2 [82]. HOTAIR functions as a molecular scaffold between the polycomb 10 repressive complex 2 (PRC2) on its 5’ end and the lysine specific demethylase 1 (LSD1)/RE1-silencing transcription factor (REST) complex on its 3’ end [81, 83]. This couples the histone H3 lysine 27 methylation ability of PRC2 with the lysine 4 demethylation ability of LSD1, repressing the expression of target genes, including tumor suppressors (Figure 1.4) [83]. HOTAIR expression is associated with poor prognosis and plays a role in the development of several cancers, including breast, ovarian, colorectal, lung, and pancreatic cancers [84-86]. Higher HOTAIR expression is found in metastatic tumors compared to primary tumors in breast cancer, so is thought to contribute to metastatic spread [87]. The field of non-coding RNAs is still relatively new, so much remains unknown about how these molecules function. Dissertation goals Ewing sarcoma is an aggressive disease that affects children and young adults. Although outcome for patients with primary disease is 75%, outcome for patients with metastatic disease is dismal; and these outcomes have not improved over the past 3 decades [1, 5]. A better understanding of the molecular pathogenesis of this disease is required in order to develop superior treatments. The treatment protocol for Ewing sarcoma is roughly the same for all patients. These young children are treated with toxic chemotherapy and radiation that can have negative long-term consequences. If subtypes of Ewing sarcoma can be sufficiently characterized, then perhaps the treatment regimens can be tailored to personally meet the patients’ needs based on their tumors’ subtype. Perhaps not every child with Ewing sarcoma requires the same toxic doses of drugs. 11 Perhaps novel targeted therapies could be added to the treatment regimen for the more aggressive subtypes. The overarching goal of this dissertation is to characterize molecular subtypes of Ewing sarcoma and identify potential new therapeutic targets. The majority of Ewing sarcoma cases is characterized by the t(11;22) translocation encoding the oncogenic transcription factor EWS-FLI1 [13]. Aside from this translocation, the genomic landscape of Ewing sarcoma is relatively quiet, especially compared to other adult cancers. Point mutations are rarely seen in Ewing sarcoma, with the most common occurring in less than 20% of cases [28-30]. However, whole chromosome, partial chromosome, and focal copy number gains and losses frequently recur in these tumors. We utilized these copy number changes to characterize genomic subtypes of Ewing sarcoma, identify potential new therapeutic targets, and understand molecular mechanisms of Ewing sarcoma development. The goal of this dissertation is to answer these questions: 1- Are there clinically relevant genomic subtypes of Ewing sarcoma defined by CNAs? Through whole-genome copy number analysis, we characterize two distinct subtypes of Ewing sarcoma: simple and complex. We define the simple subtype as those tumors with less than 1% of the length of the genome changed by CNAs, whereas the complex subtype is defined as those tumors with 1% or greater of the length of the genome changed by CNAs. We couple these subtypes with outcome data and further characterize them with clinical factor information. This work is summarized in Chapter 2. 2- How is the TP53 pathway impacted by CNAs in the complex subtype? TP53 mutations in Ewing sarcoma are infrequent but are often associated with poor prognosis [28-30, 88-91]. In Chapter 2, we explore how CNAs are capable of abrogating the TP53 12 pathway in Ewing sarcoma. We also explore the functional consequences of TP53 mutations in Ewing sarcoma cell lines, and their responses to chemotherapies and radiation. 3- What is the consequence of CEBPB amplification in Ewing sarcoma? In a previous study by our lab, the region of chromosome 20 containing CEBPB was amplified by CNAs in 15% of Ewing sarcoma [92]. In Chapter 3 of this dissertation, we explore how CEBPB is regulated by EWS-FLI1, how it promotes chemoresistance and attachment-independent cellular transformation, and how it regulates downstream factors that could lead to clinically actionable therapeutic targets. 4- What is the function of the long non-coding RNA HOTAIR in Ewing sarcoma? In addition to whole-genome copy number analysis, we were able to perform microarray gene expression analysis and DNA methylation analysis on a subset of our Ewing sarcoma samples. Through these methods, we identified HOTAIR as a gene that was upregulated and whose promoter region was unmethylated in the genomically unstable Ewing sarcoma subtype compared to the stable subtype. We performed CRISPR knockout to study the function of HOTAIR in Ewing sarcoma cell lines. These data are outlined in the Appendix. References 1 Esiashvili N, Goodman M, Marcus RB, Jr. Changes in incidence and survival of Ewing sarcoma patients over the past 3 decades: Surveillance epidemiology and end results data. Journal of Pediatric Hematology/Oncology 2008; 30: 425-430. 2 Ewing J. Diffuse endothelioma of bone. Proceedings of the New York Pathological Society 1921; 21: 17-24. 13 3 Burt M, Karpeh M, Ukoha O, Bains MS, Martini N, McCormack PM et al. Medical tumors of the chest wall. Solitary plasmacytoma and Ewing's sarcoma. The Journal of Thoracic and Cardiovascular Surgery 1993; 105: 89-96. 4 Jawad MU, Cheung MC, Min ES, Schneiderbauer MM, Koniaris LG, Scully SP. Ewing sarcoma demonstrates racial disparities in incidence-related and sexrelated differences in outcome: An analysis of 1631 cases from the seer database, 1973-2005. Cancer 2009; 115: 3526-3536. 5 Gaspar N, Hawkins DS, Dirksen U, Lewis IJ, Ferrari S, Le Deley MC et al. Ewing sarcoma: Current management and future approaches through collaboration. Journal of Clinical Oncology 2015; 33: 3036-3046. 6 Ozaki T. Diagnosis and treatment of Ewing sarcoma of the bone: A review article. Journal of Orthopaedic Science 2015; 20: 250-263. 7 Choy E, Butrynski JE, Harmon DC, Morgan JA, George S, Wagner AJ et al. Phase II study of olaparib in patients with refractory Ewing sarcoma following failure of standard chemotherapy. BMC Cancer 2014; 14: 813. 8 Tawbi HA-H, Burgess MA, Crowley J, Van Tine BA, Hu J, Schuetze S et al. Safety and efficacy of PD-1 blockade using pembrolizumab in patients with advanced soft tissue (STS) and bone sarcomas (BS): Results of SARC028—a multicenter phase II study. Journal of Clinical Oncology 2016; 34: 11006-11006. 9 van Maldegem AM, Bovee JV, Peterse EF, Hogendoorn PC, Gelderblom H. Ewing sarcoma: The clinical relevance of the insulin-like growth factor 1 and the poly-adp-ribose-polymerase pathway. European Journal of Cancer 2016; 53: 171-180. 10 Curtis C, Shah SP, Chin SF, Turashvili G, Rueda OM, Dunning MJ et al. The genomic and transcriptomic architecture of 2,000 breast tumours reveals novel subgroups. Nature 2012; 486: 346-352. 11 Taylor MD, Northcott PA, Korshunov A, Remke M, Cho Y-J, Clifford SC et al. Molecular subgroups of medulloblastoma: The current consensus. Acta Neuropathologica 2012; 123: 465-472. 12 West L, Vidwans SJ, Campbell NP, Shrager J, Simon GR, Bueno R et al. A novel classification of lung cancer into molecular subtypes. PloS One 2012; 7: e31906. 13 Delattre O, Zucman J, Plougastel B, Desmaze C, Melot T, Peter M et al. Gene fusion with an ets DNA-binding domain caused by chromosome translocation in human tumours. Nature 1992; 359: 162-165. 14 14 May WA, Gishizky ML, Lessnick SL, Lunsford LB, Lewis BC, Delattre O et al. Ewing sarcoma 11;22 translocation produces a chimeric transcription factor that requires the DNA-binding domain encoded by FLI1 for transformation. Proceedings of the National Academy of Sciences of the United States of America 1993; 90: 5752-5756. 15 Gangwal K, Sankar S, Hollenhorst PC, Kinsey M, Haroldsen SC, Shah AA et al. Microsatellites as EWS/FLI response elements in Ewing's sarcoma. Proceedings of the National Academy of Sciences of the United States of America 2008; 105: 10149-10154. 16 Seth A, Watson DK. ETS transcription factors and their emerging roles in human cancer. European Journal of Cancer 2005; 41: 2462-2478. 17 Johnson KM, Taslim C, Saund RS, Lessnick SL. Identification of two types of GGAA-microsatellites and their roles in EWS/FLI binding and gene regulation in Ewing sarcoma. PloS One 2017; 12: e0186275. 18 Guillon N, Tirode F, Boeva V, Zynovyev A, Barillot E, Delattre O. The oncogenic EWS-FLI1 protein binds in vivo GGAA microsatellite sequences with potential transcriptional activation function. PloS One 2009; 4: e4932. 19 Owen LA, Kowalewski AA, Lessnick SL. EWS/FLI mediates transcriptional repression via NKX2.2 during oncogenic transformation in Ewing's sarcoma. PloS One 2008; 3: e1965. 20 Sankar S, Theisen ER, Bearss J, Mulvihill T, Hoffman LM, Sorna V et al. Reversible LSD1 inhibition interferes with global EWS/ETS transcriptional activity and impedes Ewing sarcoma tumor growth. Clinical Cancer Research 2014; 20: 4584-4597. 21 Beck R, Monument MJ, Watkins WS, Smith R, Boucher KM, Schiffman JD et al. EWS/FLI-responsive GGAA microsatellites exhibit polymorphic differences between European and African populations. Cancer Genetics 2012; 205: 304-312. 22 von Levetzow C, Jiang X, Gwye Y, von Levetzow G, Hung L, Cooper A et al. Modeling initiation of Ewing sarcoma in human neural crest cells. PloS One 2011; 6: e19305. 23 Hu-Lieskovan S, Zhang J, Wu L, Shimada H, Schofield DE, Triche TJ. EWSFLI1 fusion protein up-regulates critical genes in neural crest development and is responsible for the observed phenotype of Ewing's family of tumors. Cancer Research 2005; 65: 4633-4644. 15 24 Cavazzana AO, Miser JS, Jefferson J, Triche TJ. Experimental evidence for a neural origin of Ewing's sarcoma of bone. The American Journal of Pathology 1987; 127: 507-518. 25 Tirode F, Laud-Duval K, Prieur A, Delorme B, Charbord P, Delattre O. Mesenchymal stem cell features of Ewing tumors. Cancer Cell 2007; 11: 421429. 26 Castillero-Trejo Y, Eliazer S, Xiang L, Richardson JA, Ilaria RL, Jr. Expression of the EWS/FLI-1 oncogene in murine primary bone-derived cells results in EWS/FLI-1-dependent, Ewing sarcoma-like tumors. Cancer Research 2005; 65: 8698-8705. 27 Riggi N, Cironi L, Provero P, Suva ML, Kaloulis K, Garcia-Echeverria C et al. Development of Ewing's sarcoma from primary bone marrow-derived mesenchymal progenitor cells. Cancer Research 2005; 65: 11459-11468. 28 Brohl AS, Solomon DA, Chang W, Wang J, Song Y, Sindiri S et al. The genomic landscape of the Ewing sarcoma family of tumors reveals recurrent STAG2 mutation. PLoS Genetics 2014; 10: e1004475. 29 Crompton BD, Stewart C, Taylor-Weiner A, Alexe G, Kurek KC, Calicchio ML et al. The genomic landscape of pediatric Ewing sarcoma. Cancer Discovery 2014; 4: 1326-1341. 30 Tirode F, Surdez D, Ma X, Parker M, Le Deley MC, Bahrami A et al. Genomic landscape of Ewing sarcoma defines an aggressive subtype with co-association of STAG2 and TP53 mutations. Cancer Discovery 2014; 4: 1342-1353. 31 Burkard ME, Weaver BA. Tuning chromosomal instability to optimize tumor fitness. Cancer Discovery 2017; 7: 134-136. 32 Chee M, Yang R, Hubbell E, Berno A, Huang XC, Stern D et al. Accessing genetic information with high-density DNA arrays. Science (New York, NY) 1996; 274: 610-614. 33 Pinkel D, Segraves R, Sudar D, Clark S, Poole I, Kowbel D et al. High resolution analysis of DNA copy number variation using comparative genomic hybridization to microarrays. Nature Genetics 1998; 20: 207-211. 34 Absalan F, Ronaghi M. Molecular inversion probe assay. Methods in Molecular Biology 2007; 396: 315-330. 35 Ferreira BI, Alonso J, Carrillo J, Acquadro F, Largo C, Suela J et al. Array cgh and gene-expression profiling reveals distinct genomic instability patterns 16 associated with DNA repair and cell-cycle checkpoint pathways in Ewing's sarcoma. Oncogene 2008; 27: 2084-2090. 36 Ozaki T, Paulussen M, Poremba C, Brinkschmidt C, Rerin J, Ahrens S et al. Genetic imbalances revealed by comparative genomic hybridization in Ewing tumors. Genes, Chromosomes & Cancer 2001; 32: 164-171. 37 Savola S, Klami A, Tripathi A, Niini T, Serra M, Picci P et al. Combined use of expression and CGH arrays pinpoints novel candidate genes in Ewing sarcoma family of tumors. BMC Cancer 2009; 9: 17. 38 Hattinger CM, Pötschger U, Tarkkanen M, Squire J, Zielenska M, KiuruKuhlefelt S et al. Prognostic impact of chromosomal aberrations in Ewing tumours. British Journal of Cancer 2002; 86: 1763-1769. 39 Tarkkanen M, Kiuru-Kuhlefelt S, Blomqvist C, Armengol G, Bohling T, Ekfors T et al. Clinical correlations of genetic changes by comparative genomic hybridization in Ewing sarcoma and related tumors. Cancer Genetics and Cytogenetics 1999; 114: 35-41. 40 Maurici D, Perez-Atayde A, Grier HE, Baldini N, Serra M, Fletcher JA. Frequency and implications of chromosome 8 and 12 gains in Ewing sarcoma. Cancer Genetics and Cytogenetics 1998; 100: 106-110. 41 Roberts P, Burchill SA, Brownhill S, Cullinane CJ, Johnston C, Griffiths MJ et al. Ploidy and karyotype complexity are powerful prognostic indicators in the Ewing's sarcoma family of tumors: A study by the united kingdom cancer cytogenetics and the children's cancer and leukaemia group. Genes, Chromosomes & Cancer 2008; 47: 207-220. 42 Armengol G, Tarkkanen M, Virolainen M, Forus A, Valle J, BÃhling T et al. Recurrent gains of 1q, 8 and 12 in the Ewing family of tumours by comparative genomic hybridization. British Journal of Cancer 1997; 75: 1403-1409. 43 Lane DP. Cancer. P53, guardian of the genome. Nature 1992; 358: 15-16. 44 Lane DP, Crawford LV. T antigen is bound to a host protein in SV40-transformed cells. Nature 1979; 278: 261-263. 45 Linzer DI, Levine AJ. Characterization of a 54k dalton cellular SV40 tumor antigen present in SV40-transformed cells and uninfected embryonal carcinoma cells. Cell 1979; 17: 43-52. 46 Friedman PN, Chen X, Bargonetti J, Prives C. The p53 protein is an unusually shaped tetramer that binds directly to DNA. Proceedings of the National Academy of Sciences of the United States of America 1993; 90: 3319-3323. 17 47 Bieging KT, Mello SS, Attardi LD. Unravelling mechanisms of p53-mediated tumour suppression. Nature Reviews Cancer 2014; 14: 359-370. 48 Zilfou JT, Lowe SW. Tumor suppressive functions of p53. Cold Spring Harbor Perspectives in Biology 2009; 1: a001883. 49 Joerger AC, Fersht AR. The p53 pathway: Origins, inactivation in cancer, and emerging therapeutic approaches. Annual Review of Biochemistry 2016; 85: 375404. 50 Leroy B, Anderson M, Soussi T. TP53 mutations in human cancer: Database reassessment and prospects for the next decade. Human Mutation 2014; 35: 672-688. 51 Olivier M, Hollstein M, Hainaut P. TP53 mutations in human cancers: Origins, consequences, and clinical use. Cold Spring Harbor Perspectives in Biology 2010; 2: a001008. 52 Pomerantz J, Schreiber-Agus N, Liegeois NJ, Silverman A, Alland L, Chin L et al. The INK4a tumor suppressor gene product, p19ARF, interacts with MDM2 and neutralizes MDM2's inhibition of p53. Cell 1998; 92: 713-723. 53 Stott FJ, Bates S, James MC, McConnell BB, Starborg M, Brookes S et al. The alternative product from the human CDKN2A locus, p14(ARF), participates in a regulatory feedback loop with p53 and MDM2. The EMBO Journal 1998; 17: 5001-5014. 54 Haupt Y, Maya R, Kazaz A, Oren M. MDM2 promotes the rapid degradation of p53. Nature 1997; 387: 296-299. 55 Kubbutat MH, Jones SN, Vousden KH. Regulation of p53 stability by MDM2. Nature 1997; 387: 299-303. 56 Danovi D, Meulmeester E, Pasini D, Migliorini D, Capra M, Frenk R et al. Amplification of MDMX (or MDM4) directly contributes to tumor formation by inhibiting p53 tumor suppressor activity. Molecular and Cellular Biology 2004; 24: 5835-5843. 57 Shvarts A, Bazuine M, Dekker P, Ramos YF, Steegenga WT, Merckx G et al. Isolation and identification of the human homolog of a new p53-binding protein, MDMX. Genomics 1997; 43: 34-42. 58 Momand J, Jung D, Wilczynski S, Niland J. The MDM2 gene amplification database. Nucleic Acids Research 1998; 26: 3453-3459. 18 59 Oliner JD, Kinzler KW, Meltzer PS, George DL, Vogelstein B. Amplification of a gene encoding a p53-associated protein in human sarcomas. Nature 1992; 358: 80-83. 60 Toledo F, Wahl GM. Regulating the p53 pathway: In vitro hypotheses, in vivo veritas. Nature Reviews Cancer 2006; 6: 909-923. 61 Zhao R, Choi BY, Lee MH, Bode AM, Dong Z. Implications of genetic and epigenetic alterations of CDKN2A (p16(INK4A)) in cancer. EBioMedicine 2016; 8: 30-39. 62 Johnson PF, Landschulz WH, Graves BJ, McKnight SL. Identification of a rat liver nuclear protein that binds to the enhancer core element of three animal viruses. Genes & Development 1987; 1: 133-146. 63 Landschulz WH, Johnson PF, Adashi EY, Graves BJ, McKnight SL. Isolation of a recombinant copy of the gene encoding C/EBP. Genes & Development 1988; 2: 786-800. 64 Landschulz WH, Johnson PF, McKnight SL. The leucine zipper: A hypothetical structure common to a new class of DNA binding proteins. Science 1988; 240: 1759-1764. 65 Landschulz WH, Johnson PF, McKnight SL. The DNA binding domain of the rat liver nuclear protein C/EBP is bipartite. Science 1989; 243: 1681-1688. 66 Calella AM, Nerlov C, Lopez RG, Sciarretta C, von Bohlen und Halbach O, Bereshchenko O et al. Neurotrophin/TRK receptor signaling mediates C/EBPalpha, -beta and neurod recruitment to immediate-early gene promoters in neuronal cells and requires C/EBpPs to induce immediate-early gene transcription. Neural Development 2007; 2: 4. 67 Cao Z, Umek RM, McKnight SL. Regulated expression of three C/EBP isoforms during adipose conversion of 3T3-l1 cells. Genes & Development 1991; 5: 15381552. 68 Ménard C, Hein P, Paquin A, Savelson A, Yang XM, Lederfein D et al. An essential role for a mek-C/EBP pathway during growth factor-regulated cortical neurogenesis. Neuron 2002; 36: 597-610. 69 Tomizawa M, Garfield S, Factor V, Xanthopoulos KG. Hepatocytes deficient in CCAAT/enhancer binding protein α (C/EBPα) exhibit both hepatocyte and biliary epithelial cell character. Biochemical and Biophysical Research Communications 1998; 249: 1-5. 19 70 Wang GL, Salisbury E, Shi X, Timchenko L, Medrano EE, Timchenko NA. Hdac1 promotes liver proliferation in young mice via interactions with C/EBPbeta. The Journal of Biological Chemistry 2008; 283: 26179-26187. 71 Zhu S, Oh HS, Shim M, Sterneck E, Johnson PF, Smart RC. C/EBPbeta modulates the early events of keratinocyte differentiation involving growth arrest and keratin 1 and keratin 10 expression. Molecular and Cellular Biology 1999; 19: 7181-7190. 72 Chang CJ, Chen TT, Lei HY, Chen DS, Lee SC. Molecular cloning of a transcription factor, AGP/EBP, that belongs to members of the C/EBP family. Molecular and Cellular Biology 1990; 10: 6642-6653. 73 Descombes P, Chojkier M, Lichtsteiner S, Falvey E, Schibler U. Lap, a novel member of the C/EBP gene family, encodes a liver-enriched transcriptional activator protein. Genes & Development 1990; 4: 1541-1551. 74 Kousteni S, Kockar FT, Sweeney GE, Ramji DP. Characterisation and developmental regulation of the xenopus laevis CCAAT-enhancer binding protein beta gene. Mechanisms of Development 1998; 77: 143-148. 75 Calkhoven CF, Muller C, Leutz A. Translational control of C/EBPalpha and C/EBPbeta isoform expression. Genes & Development 2000; 14: 1920-1932. 76 Descombes P, Schibler U. A liver-enriched transcriptional activator protein, LAP, and a transcriptional inhibitory protein, LIP, are translated from the same mRNA. Cell 1991; 67: 569-579. 77 Xiong W, Hsieh CC, Kurtz AJ, Rabek JP, Papaconstantinou J. Regulation of CCAAT/enhancer-binding protein-beta isoform synthesis by alternative translational initiation at multiple AUG start sites. Nucleic Acids Research 2001; 29: 3087-3098. 78 Gomis RR, Alarcón C, Nadal C, Van Poznak C, Massagué J. C/EBPβ at the core of the TGFβ cytostatic response and its evasion in metastatic breast cancer cells. Cancer Cell 2006; 10: 203-214. 79 Lamb J, Ramaswamy S, Ford HL, Contreras B, Martinez RV, Kittrell FS et al. A mechanism of cyclin d1 action encoded in the patterns of gene expression in human cancer. Cell 2003; 114: 323-334. 80 Harrow J, Frankish A, Gonzalez JM, Tapanari E, Diekhans M, Kokocinski F et al. Gencode: The reference human genome annotation for the encode project. Genome Research 2012; 22: 1760-1774. 20 81 Rinn JL, Kertesz M, Wang JK, Squazzo SL, Xu X, Brugmann SA et al. Functional demarcation of active and silent chromatin domains in human HOX loci by noncoding RNAs. Cell 2007; 129: 1311-1323. 82 Zhang J, Zhang P, Wang L, Piao HL, Ma L. Long non-coding RNA HOTAIR in carcinogenesis and metastasis. Acta Biochimica et Biophysica Sinica 2014; 46: 15. 83 Tsai MC, Manor O, Wan Y, Mosammaparast N, Wang JK, Lan F et al. Long noncoding RNA as modular scaffold of histone modification complexes. Science 2010; 329: 689-693. 84 Kogo R, Shimamura T, Mimori K, Kawahara K, Imoto S, Sudo T et al. Long noncoding RNA HOTAIR regulates polycomb-dependent chromatin modification and is associated with poor prognosis in colorectal cancers. Cancer Research 2011; 71: 6320-6326. 85 Liu XH, Liu ZL, Sun M, Liu J, Wang ZX, De W. The long non-coding RNA HOTAIR indicates a poor prognosis and promotes metastasis in non-small cell lung cancer. BMC Cancer 2013; 13: 464. 86 Xue X, Yang YA, Zhang A, Fong KW, Kim J, Song B et al. LncRNA HOTAIR enhances er signaling and confers tamoxifen resistance in breast cancer. Oncogene 2016; 35: 2746-2755. 87 Gupta RA, Shah N, Wang KC, Kim J, Horlings HM, Wong DJ et al. Long noncoding RNA HOTAIR reprograms chromatin state to promote cancer metastasis. Nature 2010; 464: 1071-1076. 88 Abudu A, Mangham DC, Reynolds GM, Pynsent PB, Tillman RM, Carter SR et al. Overexpression of p53 protein in primary Ewing's sarcoma of bone: Relationship to tumour stage, response and prognosis. British Journal of Cancer 1999; 79: 1185-1189. 89 de Alava E, Antonescu CR, Panizo A, Leung D, Meyers PA, Huvos AG et al. Prognostic impact of p53 status in Ewing sarcoma. Cancer 2000; 89: 783-792. 90 Huang HY, Illei PB, Zhao Z, Mazumdar M, Huvos AG, Healey JH et al. Ewing sarcomas with p53 mutation or p16/p14ARF homozygous deletion: A highly lethal subset associated with poor chemoresponse. Journal of Clinical Oncology 2005; 23: 548-558. 91 Lerman DM, Monument MJ, McIlvaine E, Liu XQ, Huang D, Monovich L et al. Tumoral TP53 and/or CDKN2A alterations are not reliable prognostic biomarkers in patients with localized Ewing sarcoma: A report from the Children's Oncology Group. Pediatric Blood & Cancer 2015; 62: 759-765. 21 92 Jahromi MS, Putnam AR, Druzgal C, Wright J, Spraker-Perlman H, Kinsey M et al. Molecular inversion probe analysis detects novel copy number alterations in Ewing sarcoma. Cancer Genetics 2012; 205: 391-404. 22 Figure 1.1 Schematic of EWSR1 and FLI1 gene domains. The breakpoint regions are shown where the two genes combine to form the EWS-FLI1 fusion gene. TAD= transcriptional activation domain; RBD= RNA binding domain; DBD= DNA binding domain. 23 Figure 1.2 The p53 pathway. Basic schematic showing major players of the p53 pathway. CDKN2A encodes proteins from two overlapping transcripts: p16INK4a, which leads to cell cycle arrest, and p14ARF, which inhibits MDM2. MDM2 and MDM4 are both p53 inhibitors, while p53 activation leads to cell cycle arrest, DNA repair, apoptosis, or cell senescence. 24 Figure 1.3 C/EBPβ isoforms. Schematic showing the 3 C/EBPβ isoforms (LAP*, LAP, and LIP) and their protein domain. LAP* and LAP contain transcriptional activation domains (TAD) while LIP does not. All 3 isoforms include the basic leucine zipper (bZIP) domain. The molecular mass of each protein isoform is indicated. 25 Figure 1.4 Schematic of HOTAIR interaction with chromatin remodelers. The lncRNA HOTAIR binds to PRC2 on its 5’ end and LSD1 on its 3’ end. The protein complexes are then localized to target genes, where lysine residues of histone 3 can be trimethylated by PRC2 or de-methylated by LSD1 to suppress transcription of target genes, as indicated by the red X on the arrow. CHAPTER 2 TP53 STATUS ASSOCIATES WITH GENOMIC COMPLEXITY, DNA DAMAGE RESPONSE, AND CLINICAL OUTCOME IN EWING SARCOMA Abstract Ewing sarcoma is a primary bone tumor in children and young adults characterized by EWSR1-ETS gene fusions. Somatic mutations are uncommon in Ewing sarcoma, but a well-defined set of copy number alterations (CNAs) occur frequently. In this study, we evaluate copy number in a large cohort of 500 primary, metastatic, and relapsed tumors (N=474 individual patients) to define genomic subtypes of Ewing sarcoma based on CNAs and their association with TP53 mutations. The “complex subtype” contains focal and chromosomal arm CNAs while the “simple subtype” lacks any detectable CNAs. We strengthen previous associations between distinct and recurring CNAs with patient survival and report a striking association between the complex subtype and the presence of TP53 alterations. We explore the functional consequences of TP53 mutations in Ewing sarcoma cell lines and demonstrate that those with TP53 mutations tend to be more resistant to chemotherapies and irradiation than cell lines with wild-type TP53. Finally, we investigate CNAs in the p53 pathway from tumors in our patient cohort and their contribution to p53 pathway suppression. This study broadens our understanding of TP53 and its relationship to Ewing sarcoma genomics, and emphasizes 27 the need for molecular-based risk stratification for future clinical trials. Introduction Ewing sarcoma (ES) is the second most common bone cancer in children, with an incidence of 3 cases per million people [1, 2]. The 5-year overall survival for patients with localized ES approaches 75%, but survival for patients with metastasis at diagnosis remains poor [1]. Few successful new treatments for ES have been introduced over the past two decades, emphasizing the need to better understand the molecular differences between treatment responders and nonresponders. Such knowledge may improve outcomes by offering more targeted treatments based on molecular alterations in tumors. ES is defined most commonly by a translocation between EWSR1 and a member of the ETS family of transcription factors, usually FLI1 [3]. This translocation combines the transcription regulating domain of EWS with the DNA binding domain of FLI1, encoding the oncogenic transcription factor EWS-FLI1 [3, 4]. EWSR1-ETS translocations drive ES development and have been described to induce a p53-dependent growth arrest in primary human fibroblasts [5]. ES has a relatively low burden of point mutations [6]; however, STAG2 and TP53 were recently confirmed as the most commonly mutated genes in ES, at a rate of 15-20% and 5-10%, respectively [7-9]. The most common genetic aberrations in ES aside from the characteristic fusions are copy number alterations (CNAs), including gain of chromosomes 8, 12, and 1q, and loss of chromosomes 10, 16q, and focal deletion of CDKN2A [10-13]. Although these CNAs have been reported for many years in the literature [14], little is known about how or why they develop in ES. One recent study demonstrates that EWSR1-ETS fusion formation 28 occurs early in the development of ES, often through complex genomic rearrangements involving multiple genes and chromosomes [15]. Despite these known CNAs, most consider ES to be relatively “quiet” in terms of its genomic alterations [8]. Currently, ES is not risk-stratified based on molecular genomics, and clinical “subtypes” do not exist based on molecular genomics. In this study, we explore global CNAs in a large number of tumors from patients with ES and characterize two distinct subtypes of ES defined by CNAs (genomic simple vs. complex). We identify a strong association between TP53 mutations, clinical outcome, and the complex subtype. We further explore the functional consequences of TP53 mutations in ES cell lines, and characterize CNAs and mutational alterations in the p53 pathway in ES. These findings support further investigation of TP53, copy number, clinical outcome, and genomic risk classification in future clinical trials of ES. Results Copy number alterations define two subtypes of Ewing sarcoma For our discovery cohort, 179 clinically diagnosed ES tissues from 157 individual patients were obtained for analysis (primary tumor [N=124], metastasis [N=30], relapse [N=16], unknown [N=9]); this included fresh and formalin-fixed paraffin-embedded (FFPE) samples obtained from Intermountain Primary Children’s Hospital in Utah (N=84), University of Michigan (N=29), University of California San Francisco (N=19), Seattle Children’s Hospital (N=15), Nagoya University Hospital in Japan (N=10), St. Luke’s Health System Idaho (N=8), University of Texas Southwestern (N=7), and The Children’s Hospital at Westmead in Australia (N=7) (Figure S2.1A, Table 2.1). Genome- 29 wide CNAs were determined by molecular inversion probe technology using OncoScan microarrays (Affymetrix/Thermo Fisher Scientific, Waltham, MA) [16, 17] and analyzed by Nexus Copy Number (BioDiscovery, El Segundo, CA). In addition, we obtained copy number data on 88 samples from the Children’s Oncology Group (COG, primary tumors [N=85], metastasis [N=3]) generated from the Human SNP Array 6.0 (Affymetrix/Thermo Fischer Scientific, Waltham, MA). Results were combined for the primary tumors for the non-COG (N=126) and COG samples (N=85) to comprise a total of 209 primary tumors from unique individuals in the discovery cohort. For our validation cohort, we obtained copy number calls from whole-genome and whole-exome sequencing data from 4 publicly available datasets (N=233 samples from 229 unique individuals, including N=107 primary tumors, N=48 metastatic or relapsed tumors, and N=78 tumors without clinical data) (Figure S2.1A, Table 2.2) [7-9, 15]. The discovery and validation cohorts together comprise 500 ES samples (N=316 primary tumors, N=97 metastasis/relapse, N=87 unknown) from 474 unique individuals (Figure 2.1A). Two genomic subtypes of ES were classified from the discovery and validation cohorts for the purposes of our analysis: (1) the simple subtype, defined as tumors with less than 1% of the entire genome affected by CNAs, and (2) the complex subtype, defined as those with greater than or equal to 1% of the length of the genome changed by CNAs. Of the 209 primary tumors evaluated by microarray in the discovery cohort, 75 (36%) were simple and 134 (64%) were complex (Figure S2.1B). The average percent of the genome changed by CNAs was 0.22% for the simple subtype and 11.93% for the complex subtype. The average percent of the genome changed by CNAs was 7.71% 30 across all samples (Table S2.1). Copy number analysis was likewise performed on 8 ES cell lines, revealing similar CNAs to those of patient samples (Figure S2.2). However, as is common in cultured cell lines, genomic instability is frequent, and all of the ES cell lines are defined as genomically complex. Consistent with previous reports in ES [14], the primary tumors of our discovery cohort (N=209) included frequent copy number gains in chromosome 8 (N=83 [40%]), 12 (N=39 [18%]), and 1q (N=31 [15%]), and copy number loss in chromosome 10 (N=18 [9%]), and 16q (N=28 [13%]) (Figure 2.1B, Figure S2.1C). Of these CNAs, gain of chromosome 8 co-occurs with gain of chromosome 12 more frequently than would be expected by chance alone (Fisher’s exact test; P<0.0001); likewise, gain of chromosome 12 co-associates with loss of chromosome 16q (P=0.0016); and gain of chromosome 1q co-associates with loss of chromosome 16q (P<0.0001) (Table S2.1). Our validation cohort (N=233) showed nearly identical findings with frequent copy number gains in chromosome 8 (N=98 [42%]), 12 (N=47 [20%]), and 1q (N=40 [17%]), and copy number loss in chromosome 10 (N=19 [8%]), and 16q (N=36 [15%]) (Figure 2.1B). Similarly, in the validation cohort, gain of chromosome 8 co-associates with gain of 12 (P<0.0001), gain of chromosome 12 co-associates with loss of 16q (P=0.0012), and gain of chromosome 1q co-associates with loss of 16q (P<0.0001). Complex subtype of Ewing sarcoma associates with older age and metastasis To determine if particular clinical features associate with the simple and complex subtypes, we combined the CNA data with available patient data for the primary tumors in the discovery cohort. Patients ≥ 10 years old at the time of diagnosis were more likely to 31 have tumors in the complex subtype (Fisher’s exact test; P=0.0198). The complex subtype was more likely to be found in patients with metastatic tumors at the time of diagnosis (P=0.0047) (Table 2.1). There was no significant association between the simple or complex subtypes for sex, tissue type (bone/soft tissue), or tumor size (Table 2.1). Complex subtype of Ewing sarcoma associates with clinical outcome Clinical outcome data were available for a subset of patients in the discovery cohort (N=155 unique patients, N=125 primary at diagnosis, N=17 metastatic at diagnosis, N=13 primary posttreatment). Similar to previous reports on smaller cohorts [10, 11, 13], 5-year event-free (EFS) and overall (OS) survival in our cohort trended towards worse outcome in the complex subtype compared to the simple subtype (Figure 2.1C). Kaplan-Meier analysis also demonstrated a nonsignificant trend towards worse EFS and OS in complex ES with chromosome 16q loss or chromosome 1q gain compared to simple ES (Figure 2.1D). To determine whether inferior outcome of the complex subtype is dependent on the presence of metastasis, the main clinical adverse prognostic factor for ES [1], univariate and multivariate outcome analysis was performed using the variables of metastasis at diagnosis and complex subtype. The hazard ratio (HR) for EFS in complex ES is 1.37 (95% CI 0.82-2.30, P=0.225) when evaluated as a single factor, but when controlled for the presence of metastasis at diagnosis, the HR for EFS in complex ES decreases to 1.05 (95% CI 0.52-2.13, P=0.899) (Table S2.3). Similarly, the HR for OS in complex ES is 1.75 (95% CI 0.90-3.38, P=0.097) as an independent variable, but when evaluated in combination with metastasis at diagnosis, it lowers slightly to 1.61 (95% CI 32 0.65-4.01, P=0.308). This suggests that some additional feature of metastasis drives EFS and OS more than the presence of genomic complexity alone. (Table S2.3). To eliminate the observed contribution of metastasis in our outcome analysis, we performed univariate survival analysis in patients without known metastatic disease at the time of diagnosis (N=71) to identify associations between recurring CNAs and clinical outcome. Chromosome 1q gain (95% CI 0.15-1.10, P=0.077) and 16q loss (95% CI 0.060.56, P=0.003) were the strongest negative predictors of OS (Table 2.2). Outcome was not significantly worse for patients with gain of chromosome 8 or 12, or loss of chromosome 10 (Table 2.2). Subtype (simple/complex), tissue type (bone/soft tissue), and sex had no significant impact on OS. When evaluated as continuous variables, age at diagnosis and tumor size had no significant impact on OS. The percent of genome changed by CNAs as a continuous variable predicted OS (95% CI 1.01-1.15, P=0.027) (Table 2.2). In the multivariate Cox analysis, chromosome 1q gain no longer predicted worse OS outcome (95% CI 0.22-2.49, P=0.623), but the total percent of genome changed by CNAs as a continuous variable (95% CI 0.98-1.13, P=0.174) and chromosome 16q loss (95% CI 0.07-1.30, P=0.110) trended towards worse OS (Table 2.2). TP53 mutations associate with complex Ewing sarcoma Given the role of TP53 in DNA repair and genome integrity, and its previously described mutation status in ES, we sequenced TP53 from samples with available tissue in our discovery cohort (N=41) to determine if TP53 mutations might be related to the complex subtype. Five percent (N=2/41) of samples had pathogenic or likely pathogenic 33 variants in TP53 according to ClinVar significance calling, consistent with the previously reported rate of 5-10% of TP53 mutations in ES [18-21]. Although limited, we observed that both samples with TP53 mutations were found exclusively in the complex subtype of ES. To expand this finding, we turned to our validation cohort with sequencing data [7-9, 15]. We combined these additional 230 samples with our own into a larger, single cohort (N=271), and discovered a significant association between TP53 mutations and the complex subtype: 100% of TP53 altered ES samples (N=18) fell exclusively within the complex category (P=0.0047) (Table 2.3, Table S2.2). Additionally, TP53 copy number loss was seen only in the complex ES subtype, regardless of TP53 mutation status (Table S2.4). Although TP53 and STAG2 mutations co-associate [7-9], we found no association between STAG2 mutations and complex subtypes (P=0.5021) (Table 2.3, Table S2.2). Previous reports have shown that ES samples with STAG2 loss regardless of TP53 status had significantly more somatic CNAs than those with wild-type STAG2. Conversely, samples with STAG2 loss and wild-type TP53 show no significant difference in somatic CNAs compared to those with normal STAG2 [8]; our expanded cohort also supports this finding. TP53 mutated Ewing sarcoma cell lines are resistant to p53 activating drugs To explore the functional consequences of mutated TP53 in the context of ES, TP53 status was verified by Sanger sequencing in 8 different Ewing sarcoma cell lines. All cell lines contained ≥1% CNAs, and therefore fall within the complex subtype for ES. A673, CHLA10, ES1, and ES8 cells had mutant TP53, and CHLA9, TC252, TC32, and 34 CHLA258 had wild-type TP53 (Table S2.5). To confirm that ES cells respond as predicted to p53 activation, each cell line was treated with Nutlin-3a, an MDM2 antagonist. Following Nutlin-3a treatment, increased protein expression of p21 (a direct transcriptional target of p53) was observed in TP53 wild-type ES cell lines (with the exception of CHLA258), while no p21 increase was observed in TP53-mutant ES cell lines (Figure 2.2A). The response of these cells to DNA damage-inducing chemotherapeutic drugs also was measured. Viability of cells with wild-type TP53 was more sensitive (lower IC50) to doxorubicin and etoposide than mutant TP53 cells, with the exception of wild-type CHLA258, which behaved like a mutant TP53 cell line in its sensitivity response (Figure 2.2B, Table S2.6). When wild-type TP53 was knocked out of TC32 and TC252 cells by CRISPR, sensitivity to doxorubicin, etoposide, and Nutlin-3a significantly decreased similar to the mutant TP53 cell lines (Figure 2.2C, Figure S2.3AB). Furthermore, when wild-type TP53 expression was rescued following knockout, sensitivity to all three drugs was restored (Figure 2.2D, Figure S2.3C-D). TP53 mutated Ewing sarcoma cell lines are more resistant to γ-irradiation To determine the response of wild-type and mutant p53 to γ-irradiation (IR) in ES, we subjected the same ES cell lines to 5 Gray (Gy) IR and measured apoptosis by Annexin-V signal over time. TP53 wild-type cells were more sensitive to IR and underwent apoptosis more abundantly and rapidly than TP53 mutant ES cells (Figure 2.3A; Mann-Whitney test, TC252 P=0.0006, CHLA9 P<0.0001, CHLA10 P=0.0679; Figure S2.4A). Furthermore, TP53 wild-type cells increase p21 protein expression within 2 hours of IR treatment, whereas TP53 mutant cells do not (Figure 2.3B, Figure S2.4A). 35 Using a lentiviral-transduced doxycycline-inducible expression plasmid, wild-type TP53 expression was forced in a mutant background, and cells became more sensitive to IR and underwent apoptosis more rapidly compared to cells with the mutant background alone (Figure 2.3C; P<0.0001). When TP53 was knocked out of ES cells, they became more resistant to IR (Figure 2.3D; P<0.0001). P53 is also involved in DNA damage repair, and it was recently reported that ES cells are defective in DNA repair compared to osteosarcoma cells [22]. To determine if DNA-damage repair in ES relates to TP53 status, wild-type TP53 and mutant TP53 ES cells were treated with 5Gy radiation. DNA double-strand breaks were measured by quantifying γH2AX foci by immunofluorescence in wild-type or mutant TP53 ES cells. By 24 hours post-IR, cells with wild-type TP53 had fewer γH2AX foci (P=0.0201, Figure 2.3E) compared to TP53 mutant cells, indicating wild-type TP53 cells repair more DNA following IR-induced DNA damage. With knockout of wild-type TP53, cells repair DNA slower than their wild-type controls at 8 or 24 hours postradiation (Figure 2.3F). TP53 pathway altered by copy number in Ewing sarcoma patient samples Although TP53 is mutated in only 5-10% of ES, we hypothesized that TP53 and other key members of the TP53 pathway might be altered by CNAs as an alternate mechanism to p53 pathway suppression. To explore this further, we evaluated the presence of CNAs and their clinical impact on 4 important members of the p53 pathway: TP53, CDKN2A, MDM2, and MDM4 (Figure 2.4A). Loss of TP53 (N=6) and CDKN2A (N=14) each resulted individually in worse 5-year EFS (P<0.0001 and P=0.0065, respectively) and OS (P<0.0001 and P=0.0005, respectively) compared to samples of the 36 simple subtype (Figure 2.4B-C). Additionally, combined samples with TP53 loss and/or CDKN2A loss had worse EFS (P=0.0010) and OS (P=0.0003) than samples in the simple category. Samples with gain of MDM2 (N=34) or MDM4 (N=34) did not have any significant difference in outcome. Overall survival of patients without metastasis was evaluated by univariate Cox proportional hazards modeling, and TP53 loss (95% CI 0.004-0.44, P=0.007) and CDKN2A loss (95% CI 0.04-0.33, P=0.00007) were both significant negative predictors of outcome (Table 2.4). In a multivariate analysis including the most significant clinical and genetic predictors of outcome, loss of CDKN2A remained the strongest negative predictor of outcome in ES (95% CI 0.03-0.29, P=0.00006). We evaluated co-association of CNAs in these p53 pathway genes, and determined that CDKN2A deletions are mutually exclusive to MDM4 gains and, to a lesser extent, MDM2 gains (Figure 2.4A). Additionally, gains of MDM4 were largely mutually exclusive to gains of MDM2. Collectively, p53 pathway CNAs were found in 33% (N=70/211) of ES primary tumors from unique individuals in the discovery cohort. Additionally, five-year OS is significantly worse in patients with CDKN2A loss and/or MDM4 gain (P=0.0417) (Figure 2.4D), and CDKN2A loss and/or MDM2 gain (P=0.0387) (Figure 2.4E). Discussion TP53 mutations are found in 5-10% of ES cases, in contrast to most adult cancers in which TP53 is mutated in 38-50% of tumors [23]. Previous studies conflict as to whether TP53 mutations significantly affect outcome in patients with ES; however, a 37 trend toward worse outcome has been consistently described [9, 18, 20, 24, 25]. It is interesting that other aggressive childhood cancers like osteosarcoma or glioblastoma contain TP53 mutations, yet TP53 mutations are rarely described in ES – a fusion-driven but equally aggressive tumor. Here, we explore alternate methods of p53 pathway disruption, specifically by CNAs. In four representative p53 pathway genes, we found over a third of ES tumors contain often mutually exclusive CNAs. TP53 copy number loss corresponds to decreased TP53 mRNA expression, supporting CNAs as an alternate method of p53 pathway suppression (Figure S2.5). If additional upstream and downstream targets in the p53 pathway were interrogated for alterations, it is possible that an even larger percentage of ES contains CNAs with disruption of the p53 pathway. ES is characterized by EWSR1-ETS translocations [26, 27]. An alternative EWSETS-driven p53 suppression could help to explain the abundance of wild-type TP53 in ES – and perhaps lead to far greater p53 pathway misregulation than previously recognized in this disease. It has been suggested that EWS-FLI1 suppresses p53 in ES, either directly [28] or indirectly through NOTCH-signaling [29]. However, our data suggest that TP53 suppression by EWSR1-ETS fusions does not completely render the p53 pathway nonfunctional, as the ES cell lines with wild-type TP53 show normal response to p53activating drugs [30]. Alternatively, a yet-to-be-described mechanism could be in place in ES with wild-type TP53 that somehow overcomes the suppression by EWSR1-ETS. Further exploration of a potential compensatory mechanism for p53 suppression will help explain the phenomenon of normal p53 activity in the context of EWSR1-ETS. Previous studies have shown varying results to forced p53 expression in ES cells, resulting in different rates of apoptosis and cell cycle arrest [31]. The reason for these 38 different response rates remains elusive, although CNA analysis of genes within the p53 pathway in ES could be informative and help to explain these differences. It is possible that p53 pathways are regulated epigenetically, by promoter or histone methylation, in the TP53 wild-type cases of ES without CNAs or known mutations. Indeed, Tirode et al. reported that histone methyltransferases and chromatin remodelers were the next most commonly mutated genes in ES after STAG2 and TP53 [9]. This area of subtle but perhaps universal genomic involvement of the p53 pathway still remains relatively unexplored in the context of ES, and warrants further investigation. We showed that CHLA258 cell line with wild-type TP53 consistently shows a lack of response to chemotherapies and IR very similar to mutant TP53 cell lines, further supporting the theory that alternative mechanisms may be involved in suppressing the p53 pathway in ES in the presence of wild-type p53. Interestingly, CHLA258 cells repair DNA damage at rates similar to other TP53 wild-type cell lines but lacked normal apoptotic function. Since p53 is integral to many cellular pathways, it is possible that some p53 pathways may be altered in CHLA258 cells while others remain intact. Based on microarray data, CDKN2A is homozygous deleted in CHLA258 cell lines, but TP53, MDM2, and MDM4 remain diploid (Table S2.5). Perhaps CHLA258 cells are more heavily dependent on CDKN2A than the other cell lines in response to chemotherapies, accounting for its TP53 mutant-like response. Identifying the exact mechanism of selective p53 loss of function in CHLA258 will require further investigation, which may identify mechanisms at play in patients with wild-type TP53 and poor response to ES treatment. In addition to investigating CNAs in the p53 pathway in ES, we explored CNAs 39 at a global level and characterized two subtypes of ES: simple and complex. The complex subtype trends towards worse outcome, consistent with other reports [9-11, 13, 32], though this relationship may be influenced by the presence of metastatic tumors at diagnosis, which are associated with the complex subtype and remain a negative prognostic factor in ES. However, we still found worse clinical outcome in patients presenting with the complex subtype of ES without metastasis at diagnosis. We also describe a striking association between TP53 mutations and the complex ES subtype. A recent study exploring chromosome rearrangements in ES also found enrichment of TP53 mutations in tumors with complexly rearranged genomes, further supporting our findings [15]. Although CNAs have been described for many years in ES, they are still not used to risk-stratify patients on current clinical trials. With further confirmation, the simple versus complex genomic subtypes in ES may be a useful way to divide treatments for future clinical investigation. In our study, we describe two subtypes of ES defined through genome-wide CNAs, simple and complex. The increased CNAs found in the complex subtype were associated with worse patient outcome. With future understanding of why and how these frequent and recurring CNAs occur, we may identify targeted therapies to prevent complex tumors from progressing. Extending our analysis to CNAs in the p53 pathway highlights the importance of integrating different types of genomic alterations, as the majority of ES tumors have wild-type TP53, but over a third of tumors also include CNAs within p53 pathway genes. Our study also shows through functional analysis of the CHLA258 cell line that ES can express wild-type TP53 but still functionally behave as if lacking p53 function; this lack of DNA damage response in the setting of wild-type TP53 40 may be reflected in patients with ES who are treatment-resistant. In summary, we report in ES an association between TP53 mutations and 1) genomic complexity, 2) chemotherapy and radiation resistance in vitro, and 3) clinical outcome. Genomic complexity, including alterations in the TP53 pathway, should be considered when designing future clinical trials for ES. Materials and methods Tumor collection IRB approval for the Pediatric Biobank (University of Utah IRB #00049682), Ewing sarcoma study (University of Utah IRB #00062545), and Cancer Genetics Study (University of Utah IRB #00041211) were obtained. De-identified FFPE scrolls and H&E slides from clinically diagnosed ES from Seattle Children’s Hospital, Primary Children’s Hospital in Utah, UTSW, Nagoya University Hospital, University of Michigan, St. Luke’s Health System Idaho, and UCSF were obtained and tumor content was determined by a pathologist. Samples with >60% tumor content were used. DNA/RNA was simultaneously extracted using the RecoverAllTM Multi-Sample RNA/DNA Isolation Workflow (Invitrogen). Fresh frozen samples were embedded in OCT, cut onto slides, and stained with H&E. Tumor content was marked by a pathologist where 2mm tumor punches were taken, and DNA/RNA was extracted using QIAGEN’s AllPrep DNA/RNA Mini Kit. 41 Copy number analysis DNA from patient ES tumors underwent copy number analysis with either the MIP Copy Number assay, the OncoScan FFPE Express 2.0, or the OncoScan CNV FFPE assay (Affymetrix) [33]. Data visualization and copy number analysis was performed with Nexus Copy Number 9.0 (BioDiscovery). Heatmap was created using probe median values (ratio of copy number to 2 in log2 space). Tumors from discovery cohort were assigned to genomically simple/complex subtypes based on the % Genome Changed in Nexus Copy Number 9.0 (<1% for simple and ≥1% for complex) with a significance cutoff of P<0.05 for CNA calls. For the validation cohort, BAM files from non-tumor controls were imported into the Multiscale BAM Reference Builder from Nexus. The BAM files from tumor samples were loaded into Nexus against the compiled pool reference to generate log ratios and β-allele frequencies. Gains and losses spanning greater than 120 probes were considered CNAs, and simple/complex assignments were made based on % Genome Changed. Simple/complex subtypes were estimated by copy number calls from Tirode, et al. and Anderson, et al. [9, 15]. Cell culture A673 cells were from ATCC. CHLA9, CHLA10, CHLA258, and TC32 cells were from Children’s Oncology Group Cell Culture and Xenograft Repository. ES1 and ES8 cells were from St. Jude Childhood Solid Tumor Network. All experiments were performed within 3 months of cell line resuscitation, and cell line authentication was performed by providers by STR profiling. 42 P53 knockdown and overexpression P53 knockout cells were generated by transfection with 1ug of p53 CRISPR/Cas9 KO plasmid (Santa Cruz Biotechnology) using TransIT-LT1 Transfection Reagent (Mirus). Two days later, cells were treated with 5uM Nutlin-3a for 10 days to enrich for p53 knockout cells and confirmed by Western blot. To overexpress p53, cells were transduced with lentivirus containing pLVX-TetOne empty vector or pLVX-TetOne-p53DYK (Clontech) and were selected with puromycin (2ug/ml). P53 expression was induced with 150-200ng/ml doxycycline 16 hours prior to treatments. Western blot Antibodies included p53 DO-1-HRP (Santa Cruz Biotechnology sc-126), anti-p21 (BD Bioscience #556430), anti-GAPDH mouse (Sigma-Aldrich G8795), and HRP-antimouse IgG secondary (GE Healthcare NA931V). Western Lightning Ultra (PerkinElmer NEL112001EA) was used as substrate. Images were taken using the BioRad ChemiDoc Gel Imager. Drug response assay Doxorubicin, etoposide, or Nutlin-3a were added at the indicated concentrations, and cell viability was measured using CellTiter-Glo (Promega) 48-72 hours later. Apoptosis assay P53 expression was induced, IncuCyte Annexin-V green reagent (Essen Bioscience) was added, and plates received either no treatment or irradiation. Plates were 43 imaged every 2h with 10x objective on phase and green channels in the IncuCyte Zoom (Essen Bioscience). Confluency and apoptosis (indicated by green object count) was measured over time using IncuCyte ZOOM software. Graphs are representative of 2 experimental replicates, each including at least 3 technical replicates. γ-H2AX DNA damage repair assay Cells were seeded onto fibronectin-treated coverslips and p53 expression was induced. Cells were treated with either 0Gy or 5Gy, and fixed in 4% PFA at 0 hours (no treatment), 5 minutes, 2 hours, 8 hours, or 24 hours post-irradiation. γ-H2AX was detected using Anti-phospho-histone-H2A.X primary (Millipore, JBW301) and Alexafluor-488 goat anti-mouse secondary (Life technologies, A11029). Coverslips were mounted using Prolong Diamond Antifade Mountant with DAPI (Life Technologies). Coverslip were imaged on an Olympus BX61 epifluorescent microscope. γ-H2AX loci were enumerated using GenASIs capture and analysis system software, version 8.0.0.42469 (Applied Spectal Imaging Ltd, Yokneam 20692 Israel). A minimum of 100 nuclei per treatment in 2 experimental replicates was analyzed. Statistical analysis Significance of survival curves was determined using the Log-rank (Mantel-Cox) test. Association of the simple/complex subtypes with categorical factors was determined by Chi-square and two-tailed Fisher’s exact test. IC50 was recorded and differences between curves were calculated using student t-test or one-way ANOVA. Mann-Whitney test was used to test differences in apoptotic response. Student t-test was used for each 44 time point during DNA repair. The relationship between outcome and other variables was tested using the Cox proportional hazards model for univariate analysis on samples that did not have metastasis at the time of diagnosis, and a multivariate model was fitted for variables with P<0.1 at the univariate level. Acknowledgements We thank Hyundai Hope on Wheels (Meeker), Kneader Bakery & Café Hope Campaign, Five For The Fight (Qualtrics), St. Baldrick’s Foundation, Damon Runyon Cancer Research Foundation, SARC, the Huntsman Cancer Institute’s Sarcoma DOT, Seattle Children’s Foundation, from Kat’s Crew Guild through the Sarcoma Research Fund, RO1CA161780 and R21CA187516 for their funding. Thanks to the University of Utah Mutation Generation and Detection Core. Sequencing was performed at the DNA Sequencing Core Facility, University of Utah. References 1 Esiashvili N, Goodman M, Marcus RB, Jr. Changes in incidence and survival of Ewing sarcoma patients over the past 3 decades: Surveillance epidemiology and end results data. Journal of Pediatric Hematology/Oncology 2008; 30: 425-430. 2 Jawad MU, Cheung MC, Min ES, Schneiderbauer MM, Koniaris LG, Scully SP. Ewing sarcoma demonstrates racial disparities in incidence-related and sexrelated differences in outcome: An analysis of 1631 cases from the seer database, 1973-2005. Cancer 2009; 115: 3526-3536. 3 Delattre O, Zucman J, Plougastel B, Desmaze C, Melot T, Peter M et al. Gene fusion with an ETS DNA-binding domain caused by chromosome translocation in human tumours. Nature 1992; 359: 162-165. 4 May WA, Gishizky ML, Lessnick SL, Lunsford LB, Lewis BC, Delattre O et al. Ewing sarcoma 11;22 translocation produces a chimeric transcription factor that requires the DNA-binding domain encoded by FLI1 for transformation. 45 Proceedings of the National Academy of Sciences of the United States of America 1993; 90: 5752-5756. 5 Lessnick SL, Dacwag CS, Golub TR. The Ewing's sarcoma oncoprotein EWS/FLI induces a p53-dependent growth arrest in primary human fibroblasts. Cancer Cell 2002; 1: 393-401. 6 Lawrence MS, Stojanov P, Polak P, Kryukov GV, Cibulskis K, Sivachenko A et al. Mutational heterogeneity in cancer and the search for new cancer-associated genes. Nature 2013; 499: 214-218. 7 Brohl AS, Solomon DA, Chang W, Wang J, Song Y, Sindiri S et al. The genomic landscape of the Ewing sarcoma family of tumors reveals recurrent STAG2 mutation. PLoS Genetics 2014; 10: e1004475. 8 Crompton BD, Stewart C, Taylor-Weiner A, Alexe G, Kurek KC, Calicchio ML et al. The genomic landscape of pediatric Ewing sarcoma. Cancer Discovery 2014; 4: 1326-1341. 9 Tirode F, Surdez D, Ma X, Parker M, Le Deley MC, Bahrami A et al. Genomic landscape of Ewing sarcoma defines an aggressive subtype with co-association of STAG2 and TP53 mutations. Cancer Discovery 2014; 4: 1342-1353. 10 Ferreira BI, Alonso J, Carrillo J, Acquadro F, Largo C, Suela J et al. Array CGH and gene-expression profiling reveals distinct genomic instability patterns associated with DNA repair and cell-cycle checkpoint pathways in Ewing's sarcoma. Oncogene 2008; 27: 2084-2090. 11 Jahromi MS, Putnam AR, Druzgal C, Wright J, Spraker-Perlman H, Kinsey M et al. Molecular inversion probe analysis detects novel copy number alterations in Ewing sarcoma. Cancer Genetics 2012; 205: 391-404. 12 Roberts P, Burchill SA, Brownhill S, Cullinane CJ, Johnston C, Griffiths MJ et al. Ploidy and karyotype complexity are powerful prognostic indicators in the Ewing's sarcoma family of tumors: A study by the united kingdom cancer cytogenetics and the children's cancer and leukaemia group. Genes, Chromosomes & Cancer 2008; 47: 207-220. 13 Savola S, Klami A, Tripathi A, Niini T, Serra M, Picci P et al. Combined use of expression and CGH arrays pinpoints novel candidate genes in Ewing sarcoma family of tumors. BMC Cancer 2009; 9: 17. 14 Jahromi MS, Jones KB, Schiffman JD. Copy number alterations and methylation in Ewing's sarcoma. Sarcoma 2011; 2011: 362173. 46 15 Anderson ND, de Borja R, Young MD, Fuligni F, Rosic A, Roberts ND et al. Rearrangement bursts generate canonical gene fusions in bone and soft tissue tumors. Science 2018; 361. 16 Singh RR, Mehrotra M, Chen H, Almohammedsalim AA, Sahin A, Bosamra A et al. Comprehensive screening of gene copy number aberrations in formalin-fixed, paraffin-embedded solid tumors using molecular inversion probe-based singlenucleotide polymorphism array. The Journal of Molecular Diagnostics 2016; 18: 676-687. 17 Wang Y, Cottman M, Schiffman JD. Molecular inversion probes: A novel microarray technology and its application in cancer research. Cancer Genetics 2012; 205: 341-355. 18 Huang HY, Illei PB, Zhao Z, Mazumdar M, Huvos AG, Healey JH et al. Ewing sarcomas with p53 mutation or p16/p14ARF homozygous deletion: A highly lethal subset associated with poor chemoresponse. Journal of Clinical Oncology 2005; 23: 548-558. 19 Kovar H, Auinger A, Jug G, Aryee D, Zoubek A, Salzer-Kuntschik M et al. Narrow spectrum of infrequent p53 mutations and absence of MDM2 amplification in Ewing tumours. Oncogene 1993; 8: 2683-2690. 20 Lerman DM, Monument MJ, McIlvaine E, Liu XQ, Huang D, Monovich L et al. Tumoral TP53 and/or CDKN2A alterations are not reliable prognostic biomarkers in patients with localized Ewing sarcoma: A report from the Children's Oncology Group. Pediatric Blood & Cancer 2015; 62: 759-765. 21 Neilsen PM, Pishas KI, Callen DF, Thomas DM. Targeting the p53 pathway in Ewing sarcoma. Sarcoma 2011; 2011: 746939. 22 Stewart E, Goshorn R, Bradley C, Griffiths LM, Benavente C, Twarog NR et al. Targeting the DNA repair pathway in Ewing sarcoma. Cell Reports 2014; 9: 829841. 23 Olivier M, Hollstein M, Hainaut P. TP53 mutations in human cancers: Origins, consequences, and clinical use. Cold Spring Harbor Perspectives in Biology 2010; 2: a001008. 24 Abudu A, Mangham DC, Reynolds GM, Pynsent PB, Tillman RM, Carter SR et al. Overexpression of p53 protein in primary Ewing's sarcoma of bone: Relationship to tumour stage, response and prognosis. British Journal of Cancer 1999; 79: 1185-1189. 25 de Alava E, Antonescu CR, Panizo A, Leung D, Meyers PA, Huvos AG et al. Prognostic impact of p53 status in Ewing sarcoma. Cancer 2000; 89: 783-792. 47 26 Grunewald TGP, Cidre-Aranaz F, Surdez D, Tomazou EM, de Alava E, Kovar H et al. Ewing sarcoma. Nature Reviews Disease Primers 2018; 4: 5. 27 Toomey EC, Schiffman JD, Lessnick SL. Recent advances in the molecular pathogenesis of Ewing's sarcoma. Oncogene 2010; 29: 4504-4516. 28 Li Y, Tanaka K, Fan X, Nakatani F, Li X, Nakamura T et al. Inhibition of the transcriptional function of p53 by EWS-FLI1 chimeric protein in Ewing family tumors. Cancer Letters 2010; 294: 57-65. 29 Ban J, Bennani-Baiti IM, Kauer M, Schaefer K-L, Poremba C, Jug G et al. EWSFLI1 suppresses notch-activated p53 in Ewing’s sarcoma. Cancer Research 2008; 68: 7100-7109. 30 Stolte B, Iniguez AB, Dharia NV, Robichaud AL, Conway AS, Morgan AM et al. Genome-scale CRISPR-cas9 screen identifies druggable dependencies in TP53 wild-type Ewing sarcoma. The Journal of Experimental Medicine 2018; 215: 2137-2155. 31 Kovar H, Pospisilova S, Jug G, Printz D, Gadner H. Response of Ewing tumor cells to forced and activated p53 expression. Oncogene 2003; 22: 3193-3204. 32 Ozaki T, Paulussen M, Poremba C, Brinkschmidt C, Rerin J, Ahrens S et al. Genetic imbalances revealed by comparative genomic hybridization in Ewing tumors. Genes, Chromosomes & Cancer 2001; 32: 164-171. 33 Absalan F, Ronaghi M. Molecular inversion probe assay. Methods in Molecular Biology 2007; 396: 315-330. 48 Figure 2.1 Simple and complex subtypes of Ewing sarcoma. A- Heatmap of copy number profile across all samples from unique patients (N=474). Chromosomal coordinates are on x-axis and samples are plotted along the y-axis. Color in heatmap represents log2 ratio of integer copy number to 2. B- Frequency of common CNAs in all primary tumors in both the discovery and validation cohorts. Red represents copy number gain and blue represents copy number loss. C- 5-year event-free and overall survival for Ewing sarcoma patients with simple or complex tumors. B- 5-year event-free and overall survival for Ewing sarcoma patients with loss of chromosome 16q or gain of chromosome 1q compared to patients with simple tumors. 49 Figure 2.2 TP53 mutated Ewing sarcoma cell lines are resistant to chemotherapies. AWestern blot of Ewing sarcoma cell lines with and without Nutlin-3a treatment to show p53 activation as measured by increased expression of p21, a direct target of p53. BViability of Ewing sarcoma cell lines after 72h of doxorubicin, etoposide, or Nutlin-3a treatment. TP53 mutant cell lines are in gray while TP53 wild-type cell lines are in red. CHLA258 is dashed, because of its TP53 wild-type genotype and TP53 mutant phenotype. C- Viability of TC252 cells with wild-type p53 (NT) or with p53 knockout (KO) after 52h of doxorubicin, etoposide, or Nutlin-3a treatment. D- Viability of TC252 cells with wildtype p53 (Parental + Empty), p53 knockout (KO + Empty), p53 knockout followed by rescue (KO + p53), or p53 overexpression (Parental + p53) after 55h of doxorubicin, etoposide, or Nutlin-3a treatment. 50 Figure 2.3 TP53 mutated Ewing sarcoma cell lines are resistant to IR. A- Apoptotic response of TC252 (P=0.0006), CHLA9 (P<0.0001), and CHLA10 (P=0.0679) cell lines as measured by Annexin-V positive cells in response to 5Gy IR over time. B- Western blot showing p53 and p21 expression with no treatment (NT) or at 2, 8, and 24 hours following 5Gy IR. C- Annexin-V positive CHLA10 cells with (TP53) or without (Empty) TP53 expression induced and treated (5Gy) or without (NT) IR (5Gy Empty vs. 5Gy p53 P<0.0001). D- Annexin-V positive TC252 and TC32 cells with wild-type (Parental) or knocked out (TP53 KO) TP53 treated with (5Gy) or without (NT) IR (Parental + 5Gy vs. TP53 KO + 5Gy P<0.0001). E- Percent of cells with greater than 20 γH2AX foci as a measure of DNA double strand breaks following treatment with 5Gy IR (TP53 mut vs. TP53 WT, P=0.0201 at 24h). Representative images of γH2AX foci in CHLA9 (wild-type TP53) and CHLA10 (mutant TP53) cells over time following IR. F- Percent of cells with greater than 20 γH2AX foci in TC252 wild-type (Parental) and TP53 knockout (TP53 KO) (P=0.0448 at 8h) and TC32 wild-type (Parental) and TP53 knockout (TP53 KO) (P=0.0016 at 24h) cells following 5Gy IR treatment. 51 52 Figure 2.4 P53 pathway CNAs in Ewing sarcoma. A- Profile of CDKN2A, MDM4, MDM2, and TP53 CNAs in Ewing sarcoma tumors (N=211). Homozygous loss represents two copy deletion, loss represents single copy deletion, amplifications represent >1 copy number gain, and gains represent 1 copy number gain. CDKN2A loss is mutually exclusive with MDM4 gain and MDM2 gain. 5-year overall survival for Ewing sarcoma patients with BTP53 copy number loss, C- CDKN2A copy number loss, D- MDM4 gain and/or CDKN2A loss, E- MDM2 gain and/or CDKN2A loss. 53 Figure S2.1 Profile of the simple and complex Ewing sarcoma subtypes. A- Summary of the Ewing sarcoma samples in the discovery and validation cohorts. B- Frequency distribution of Ewing sarcoma primary tumors (N=209) by the percent of the genome length altered by CNAs. 36% of samples have <1% of the length of the genome altered by CNAs and are classified in the simple subtype, whereas 64% have ≥1% of the genome altered by CNAs and are classified as complex. C- Profile of common CNAs in Ewing sarcoma primary tumors (N=209), including alterations in chromosomes 8, 12, 1q, 16q, and 10. 54 Figure S2.2 Copy number profile of Ewing sarcoma cell lines. Heatmap of copy number profile across 8 Ewing sarcoma cell lines. Chromosomal coordinates are on x-axis and samples are plotted along the y-axis. Color in heatmap represents log2 ratio of integer copy number to 2. 55 Figure S2.3 TP53 mutated Ewing sarcoma cell lines are resistant to chemotherapies. AWestern blot verifying CRISPR knockout of TP53 in TC252 and TC32 cells as measured by a lack of p53 protein expression. GAPDH was included as a loading control. B- Viability of TC32 cells with wild-type TP53 (NT) or with TP53 knockout (TP53 KO) after 52h of doxorubicin, etoposide, or Nutlin-3a treatment. C- Western blot showing knockout of TP53 in TC252 and TC32 cells, followed by expression of p53 protein from a doxycycline (doxy) inducible vector. D- Viability of TC32 cells with wild-type TP53 (Parental + Empty), TP53 knockout (TP53 KO + Empty), TP53 knockout followed by rescue (TP53 KO + p53), or TP53 overexpression (Parental + p53) after 55h of doxorubicin, etoposide, or Nutlin-3a treatment. 56 Figure S2.4 TP53 mutated Ewing sarcoma cell lines are resistant to IR. A- Apoptotic response of CHLA258 (P<0.0001), and A673 (P=0.8026) cell lines as measured by Annexin-V positive cells in response to 5Gy IR over time. B- Western blot showing p53 and p21 expression with no treatment (NT) or at 2, 8, and 24 hours following 5Gy IR. 57 Figure S2.5 Relative TP53 mRNA expression in Ewing sarcoma samples with 2 or 1 copies of TP53. 58 Table 2.1 Factors associated with the simple and complex Ewing sarcoma subtypes Simple Complex P-val 72 (69%) 42 (69%) P=1.0000 73 (72%) 14 (50%) P=0.0198 41 (68%) 29 (64%) P=0.6820 19 (95%) 30 (75%) P=0.0852 11 (92%) 38 (59%) P=0.0047 Sex Male 33 (31%) Female 19 (31%) Age at Diagnosis ≥10 years 29 (28%) <10 years 14 (50%) Tissue Type Bone 19 (32%) Soft Tissue 16 (36%) Tumor Size ≥10 cm 1 (5%) <10 cm 10 (25%) Metastasis at Diagnosis Yes 1 (8%) No 26 (41%) 59 Table 2.2 Univariate and multivariate survival analysis Overall Survival Univariate Variable Overall Survival Multivariate HR 95% CI P-val HR 95% CI P-val % of Genome Changed 1.07 1.01-1.15 0.027 1.05 0.98-1.13 0.174 Complex vs. Simple 1.64 0.66-4.10 0.288 - - - Age at DX 1.01 0.97-1.05 0.591 - - - Tumor size 0.91 0.65-1.28 0.605 - - - Tissue type (soft vs. bone) 0.90 0.29-2.81 0.860 - - - Gender (female vs. male) 0.66 0.28-1.54 0.335 - - - Chr8 (gain vs. diploid) 0.96 0.41-2.24 0.924 - - - Chr10 (diploid vs. loss) 1.04 0.24-4.59 0.959 - - - Chr12 (diploid vs. gain) 0.86 0.24-3.09 0.817 - - - Chr1q (diploid vs. gain) 0.40 0.15-1.10 0.077 0.73 0.22-2.49 0.623 Chr16q (diploid vs. loss) 0.18 0.06-0.56 0.003 0.31 0.07-1.30 0.110 60 Table 2.3 Relationship between simple/complex subtypes and TP53/STAG2 mutation status Simple Complex P-val WT Mut 71 (28%) 0 (0%) 182 (72%) 18 (100%) P=0.0047 WT Mut 53 (27%) 9 (32%) 149 (73%) 19 (68%) P=0.5021 TP53 status STAG2 status 61 Table 2.4 Univariate and multivariate survival analysis of p53 pathway CNAs Overall Survival Univariate Variable Overall Survival Multivariate HR 95% CI P-val HR 95% CI P-val TP53 (diploid vs. loss) 0.05 0.004-0.44 0.007 NA NA NA CDKN2A (diploid vs. loss) 0.11 0.04-0.33 0.00007 0.09 0.03-0.29 0.00006 MDM2 (diploid vs. gain) 0.37 0.12-1.11 0.075 2.11 0.37-12.13 0.404 MDM4 (diploid vs. gain) 0.67 0.20-2.28 0.520 - - - 62 Table S2.1 Summary of Ewing sarcoma patients included in the discovery cohort % Genome Simple/ TP53 chr8 chr10 chr12 chr1q chr16q TP53 CDKN2A MDM2 MDM4 Changed Complex (seq) Sample ID Institution EWS_01_PT Utah 0.0907 simple EWS_02_PT Utah 3.9395 complex gain gain n/a gain gain n/a n/a EWS_10_PT Utah 18.4025 complex gain EWS_11_ML Utah 7.7437 complex n/a EWS_11_RX Utah 2.2069 complex n/a EWS_13_PT Utah 11.1987 complex gain EWS_14_PT Utah 0.7164 simple EWS_15_PT Utah 6.6717 complex gain gain n/a n/a gain n/a EWS_16_PT Utah 4.6981 complex gain EWS_16_ML Utah 1.0479 complex n/a EWS_17_PT Utah 10.7037 complex gain EWS_18_PT Utah 4.9084 complex EWS_19_PT Utah 36.7719 complex gain EWS_20_PT Utah 7.9101 complex n/a n/a n/a gain gain gain gain loss gain gain n/a gain n/a gain n/a EWS_22_PT Utah 0.1888 simple EWS_22_RX Utah 0.1589 simple EWS_23_ML Utah 12.6017 EWS_24_PT Utah 0.285 EWS_24_RX Utah 0.1866 simple EWS_25_PT Utah 3.479 complex gain gain n/a EWS_26_PT Utah 3.5359 complex gain gain n/a EWS_27_PT Utah 5.2125 complex gain EWS_28_PT Utah 19.3091 complex gain EWS_29_PT Utah 5.4635 complex gain EWS_30_PT Utah 32.8073 complex EWS_31_PT Utah 14.5071 complex gain EWS_32_PT Utah 25.1937 complex gain EWS_34_ML Utah 6.0253 complex EWS_35_PT Utah 0.0486 simple n/a EWS_36_PT Utah 0.089 simple n/a n/a complex gain loss loss n/a simple n/a n/a n/a gain n/a n/a loss gain loss gain n/a gain n/a n/a gain loss gain n/a EWS_37_PT Utah 1.9735 complex EWS_39_ML Utah 11.6964 complex n/a EWS_40_PT Utah 1.5281 complex EWS_41_ML Utah 15.3614 complex gain EWS_42_PT Utah 10.148 complex gain EWS_44_PT Utah 0.2048 simple n/a EWS_44_ML Utah 0.0944 simple n/a loss gain gain n/a n/a gain gain n/a loss n/a EWS_45_PT Utah 9.0721 complex gain loss EWS_45_ML Utah 13.7683 complex gain loss loss EWS_46_PT Utah 15.545 complex gain EWS_48_PT Utah 0.1442 simple EWS_49_PT Utah 0.0071 simple n/a EWS_49_ML Utah 0.019 simple n/a EWS_50_PT1 Utah 4.8505 complex gain loss gain n/a EWS_50_PT2 Utah 4.8971 complex gain loss gain n/a gain n/a n/a loss gain n/a loss n/a 63 Table S2.1 continued Sample ID Institution EWS_51_PT EWS_51_ML EWS_54_PT EWS_56_PT EWS_57_PT EWS_58_PT EWS_59_PT EWS_60_PT EWS_61_PT EWS_62_PT EWS_83_PT EWS_84_RX EWS_86_ML EWS_87_ML EWS_88_PT EWS_89_PT EWS_90_PT EWS_91_PT EWS_92_PT EWS_94_RX EWS_95_PT EWS_96_ML EWS_97_RX EWS_98_RX EWS_99_PT EWS_100_RX EWS_102_PT EWS_103_PT EWS_104_PT EWS_105_PT EWS_106_PT EWS_107_PT EWS_108_PT EWS_109_PT EWS_110_PT1 EWS_110_PT2 EWS_111_PT EWS_112_PT EWS_113_PT EWS_114_PT EWS_115_PT EWS_116_ML EWS_116_PT EWS_117_PT1 EWS_117_PT2 EWS_118_PT EWS_121_PT EWS_122_PT EWS_123_PT EWS_125_ML EWS_126_PT Utah Utah Utah Utah Utah Utah Utah Utah Utah Utah Utah UCSF UCSF UCSF UCSF UCSF UCSF UCSF UCSF UCSF UCSF UCSF UCSF UCSF UCSF UCSF UCSF UCSF UCSF UCSF Utah Utah Utah Utah Utah Utah Utah Utah Utah Utah Utah Utah Utah Utah Utah Utah Utah Utah Utah Utah Utah % Genome Simple/ TP53 chr8 chr10 chr12 chr1q chr16q TP53 CDKN2A MDM2 MDM4 Changed Complex (seq) 15.8364 0.116 9.2936 0.1739 10.7718 0.0691 26.7781 37.0155 0.7708 0.2103 4.819 0.0047 6.7563 5.6389 4.226 0 3.2301 0.0689 0 0.0184 3.1225 1.0319 0.002 1.4086 10.0964 9.2966 2.2511 10.9349 0.005 5.9792 5.4297 4.6907 25.5574 0.0014 15.2442 17.766 12.6865 20.263 0.0567 14.8437 0.0923 0.7202 0.0124 22.1255 14.4372 21.413 16.8772 20.3711 1.2243 19.027 10.7857 complex simple complex simple complex simple complex complex simple simple complex simple complex complex complex simple complex simple simple simple complex complex simple complex complex complex complex complex simple complex complex complex complex simple complex complex complex complex simple complex simple simple simple complex complex complex complex complex complex complex complex gain gain gain loss gain gain gain gain gain loss gain gain gain gain loss loss gain gain gain gain gain gain gain gain gain loss gain gain loss gain gain gain loss gain loss gain gain gain loss gain loss loss loss loss loss gain gain gain loss gain gain gain loss gain gain loss loss loss loss loss gain gain gain loss loss gain loss gain gain gain loss gain gain loss gain gain gain gain loss loss loss gain gain gain gain loss loss loss gain gain gain gain gain loss gain loss gain gain gain n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a wt n/a n/a wt n/a wt n/a n/a wt n/a wt wt n/a n/a wt n/a n/a wt wt n/a n/a n/a n/a wt n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a mut wt 64 Table S2.1 continued Sample ID Institution EWS_127_ML EWS_128_PT EWS_135_PT EWS_137_PT EWS_138_PT EWS_144_RX EWS_145_PT EWS_146_RX EWS_147_ML EWS_148_PT EWS_149_PT EWS_149_ML EWS_152_PT EWS_153_PT EWS_154_RX EWS_155_PT EWS_157_ML1 EWS_157_ML2 EWS_160_ML EWS_161_PT EWS_162_ML EWS_163_RX EWS_164_PT EWS_166_RX EWS_167_ML EWS_168_PT EWS_169_RX EWS_171_RX EWS_172_PT EWS_173_PT EWS_174_PT EWS_176_RX EWS_178_PT EWS_179_ML EWS_180_PT EWS_464_PT EWS_465_ML1 EWS_465_PT EWS_466_ML EWS_466_PT EWS_467_ML2 EWS_467_ML1 EWS_467_PT EWS_667_ML EWS_839_1 EWS_839_3 EWS_840 EWS_841 EWS_842_1 EWS_842_2 EWS_844_1 Utah Utah Utah Utah Utah Michigan Michigan Michigan Michigan Michigan Michigan Michigan Michigan Michigan Michigan Michigan Michigan Michigan Michigan Michigan Michigan Michigan Michigan Michigan Michigan Michigan Michigan Michigan Michigan Michigan Michigan Michigan Michigan Michigan Utah St. Luke's St. Luke's St. Luke's St. Luke's St. Luke's St. Luke's St. Luke's St. Luke's Utah Seattle Seattle UTSW UTSW UTSW UTSW UTSW % Genome Simple/ TP53 chr8 chr10 chr12 chr1q chr16q TP53 CDKN2A MDM2 MDM4 Changed Complex (seq) 10.1045 1.9562 11.5778 0.1389 11.3625 1.2316 0.124 0 3.8252 6.8278 0.0105 0.019 0.0012 2.2595 0 19.4339 2.6598 3.828 4.6435 1.9138 0.5238 2.5353 2.1651 12.7056 8.781 0.0355 0 4.8042 4.0515 6.2724 0.0381 0 0.1202 0.3889 5.1115 15.8676 2.3825 15.5622 14.6657 4.6786 8.5102 7.243 11.6987 12.972 6.6001 0.1016 0.1834 0.0377 0.2646 7.5032 9.0523 complex complex complex simple complex complex simple simple complex complex simple simple simple complex simple complex complex complex complex complex simple complex complex complex complex simple simple complex complex complex simple simple simple simple complex complex complex complex complex complex complex complex complex complex complex simple simple simple simple complex complex gain gain gain loss loss gain gain gain gain gain loss gain gain gain gain gain gain gain gain gain loss loss gain gain loss gain gain gain gain gain gain gain gain gain gain gain gain gain gain gain gain loss gain gain gain gain gain gain gain gain loss gain wt n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a wt mut n/a wt wt n/a n/a wt n/a n/a wt wt wt wt wt 65 Table S2.1 continued Sample ID Institution EWS_844_1 EWS_844_2 EWS_846 EWS_865_PT EWS_866_PT EWS_867_PT EWS_868_PT EWS_870_PT EWS_871_PT EWS_872_PT EWS_873_PT EWS_875_PT EWS_876_PT EWS_878_PT EWS_881_PT1 EWS_882_PT EWS_884_PT EWS_888_PT EWS_891_PT EWS_892_PT EWS_893_PT EWS_895_ML EWS_896_PT EWS_881_PT2 EWS_902_PT EWS_903_PT EWS_394_PT EWS_397_PT EWS_401_PT EWS_415_PT EWS_440_PT EWS_447_PT EWS_451_PT L308 L309 L310 L311 L312 L313 L314 L315 L386 L387 L388 L389 L390 L391 L392 L393 L394 L395 UTSW UTSW UTSW Japan Japan Japan Japan Japan Japan Japan Japan Japan Japan Seattle Seattle Seattle Seattle Seattle Seattle Seattle Seattle Seattle Seattle Seattle Seattle Seattle Westmead Westmead Westmead Westmead Westmead Westmead Westmead COG COG COG COG COG COG COG COG COG COG COG COG COG COG COG COG COG COG % Genome Simple/ TP53 chr8 chr10 chr12 chr1q chr16q TP53 CDKN2A MDM2 MDM4 Changed Complex (seq) 9.0523 0.1633 0.5144 0.3607 0.4271 0.206 3.9746 13.2559 14.745 3.513 12.1056 25.0709 9.9572 11.1722 8.1124 10.8279 2.5502 11.698 38.2102 6.9914 9.0373 10.5002 1.7245 8.3999 10.8016 14.0987 18.2913 12.1779 13.487 4.7167 10.4287 13.7193 0.6625 11.9155 4.7313 23.2735 0.1438 14.3706 4.0721 0.2174 6.1282 1.3462 1.572 68.212 0.2562 4.6089 0.2052 39.0738 15.2787 0.2798 0.2638 complex simple simple simple simple simple complex complex complex complex complex complex complex complex complex complex complex complex complex complex complex complex complex complex complex complex complex complex complex complex complex complex simple complex complex complex simple complex complex simple complex complex complex complex simple complex simple complex complex simple simple gain gain gain gain gain gain loss loss gain gain loss gain gain gain gain loss loss gain gain gain gain gain gain gain gain gain loss loss gain gain gain loss gain gain gain gain gain gain gain gain gain loss gain loss loss gain gain gain gain loss gain gain gain gain gain gain gain gain gain gain gain gain loss gain gain gain loss loss loss gain gain gain gain gain gain gain loss gain gain gain gain wt wt wt n/a n/a n/a n/a n/a n/a wt wt wt n/a wt wt wt wt wt wt wt wt wt wt wt wt wt n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a 66 Table S2.1 continued Sample ID Institution % Genome Changed L674 L676 L677 L678 L805 L806 L807 L808 L809 L810 L811 L813 L814 L815 L816 L819 M942 M943 M944 M945 M946 M947 M949 M951 M952 M953 M954 M955 M956 N830 N831 N832 N833 N834 N836 N837 N838 N846 N848 N850 N854 N855 N856 O434 O435 O436 O437 O438 O439 O440 O441 COG COG COG COG COG COG COG COG COG COG COG COG COG COG COG COG COG COG COG COG COG COG COG COG COG COG COG COG COG COG COG COG COG COG COG COG COG COG COG COG COG COG COG COG COG COG COG COG COG COG COG 0.2565 0.1808 21.2778 0.2437 0.932 0.2625 0.214 28.1995 0.2415 0.1928 2.8595 0.3224 0.2894 0.1992 3.7347 2.1595 13.7265 15.8223 4.4127 4.6445 0.3616 20.0556 19.8211 0.2543 0.1479 3.0793 0.3036 4.7359 8.9468 0.2047 0.2237 0.5326 2.1219 24.5789 0.1683 0.1544 4.9309 11.8744 2.3565 0.2605 0.1174 0.173 1.8991 0.3452 0.1907 0.1086 0.13 0.1778 9.2156 14.1883 0.0857 Simple/ TP53 chr8 chr10 chr12 chr1q chr16q TP53 CDKN2A MDM2 MDM4 Complex (seq) simple simple complex simple simple simple simple complex simple simple complex simple simple simple complex complex complex complex complex complex simple complex complex simple simple complex simple complex complex simple simple simple complex complex simple simple complex complex complex simple simple simple complex simple simple simple simple simple complex complex simple gain gain gain gain loss loss gain gain gain gain gain gain gain gain gain gain gain gain loss gain gain gain loss loss gain gain gain gain loss gain gain loss gain gain gain loss gain loss loss loss gain gain gain gain gain gain gain gain loss gain gain gain loss gain gain gain gain loss gain n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a 67 Table S2.1 continued Sample ID Institution % Genome Changed O442 O443 O444 O445 O456 O457 O458 O459 O460 P097 P100 P101 P102 P103 P105 P106 P107 P108 Q164 COG COG COG COG COG COG COG COG COG COG COG COG COG COG COG COG COG COG COG 17.7425 0.1634 0.2531 34.1668 29.3852 0.1785 5.2431 7.3577 13.2625 0.7434 0 0.2295 6.4745 21.8404 20.7089 7.5782 0.8456 2.8422 19.5966 Simple/ TP53 chr8 chr10 chr12 chr1q chr16q TP53 CDKN2A MDM2 MDM4 Complex (seq) complex simple simple complex complex simple complex complex complex simple simple simple complex complex complex complex simple complex complex loss gain gain gain loss loss loss gain gain gain gain gain gain gain gain gain gain gain loss gain gain gain gain gain loss loss gain loss gain gain n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a 68 Table S2.2 Summary of Ewing sarcoma public data included in the validation cohort Sample ID SJEWS001301 SJEWS001302 SJEWS001303 SJEWS001304 SJEWS001305 SJEWS001306 SJEWS001307 SJEWS001308 SJEWS001309 SJEWS001311 SJEWS001312 SJEWS001313 SJEWS001314 SJEWS001316 SJEWS001317 SJEWS001318 SJEWS001319 SJEWS001320 IC009 IC015 IC024 IC034 IC044 IC046 IC049 IC053 IC054 IC057 IC058 IC066 IC067 IC076 IC077 IC080 IC082 IC086 IC092 IC093 IC096 IC105 IC106 IC111 IC112 IC114 IC116 IC121 IC128 IC130 IC147 IC149 IC151 Dataset Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tumor type TP53 status PT PT PT PT PT PT PT PT PT PT PT PT PT ML ML ML PT ML ML PT PT PT PT PT PT ML ML PT PT PT PT PT NA ML PT PT ML ML PT ML ML PT ML PT ML ML PT PT ML PT PT wt wt wt wt wt wt wt wt wt wt wt wt wt wt wt wt wt wt wt wt wt mut wt wt wt mut wt wt wt wt wt wt wt wt wt mut mut wt wt wt wt wt wt wt wt mut wt wt wt wt wt STAG2 status wt wt mut wt wt wt wt wt wt wt mut wt wt wt wt wt wt wt wt wt wt wt wt mut wt wt wt wt wt wt wt wt wt mut mut mut mut wt wt wt wt wt wt wt wt mut wt mut wt wt wt Simple/Complex complex simple complex simple simple complex complex simple complex complex simple simple complex complex complex complex complex simple simple complex simple complex complex simple simple complex complex complex complex complex complex complex simple complex simple complex complex simple complex complex complex simple complex simple complex complex complex complex complex complex complex TP53 CN loss gain gain gain 69 Table S2.2 continued Sample ID IC158 IC165 IC168 IC174 IC193 IC196 IC197 IC198 IC224 IC242 IC248 IC254 IC262 IC263 IC264 IC267 IC268 IC270 IC272 IC273 IC274 IC275 IC277 IC278 IC279 IC280 IC282 IC283 IC284 IC286 IC288 IC294 IC295 IC296 IC297 IC299 IC300 IC301 IC302 IC303 IC305 IC306 IC309 IC311 IC315 IC316 IC318 IC319 IC323 IC324 IC325 Dataset Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tirode Tumor type TP53 status PT PT PT PT ML ML ML ML PT PT PT PT PT NA ML ML PT PT ML NA ML PT NA PT PT NA PT PT PT ML ML PT PT ML ML ML ML PT PT PT ML PT ML ML NA NA PT NA PT PT ML wt wt wt wt wt wt wt wt wt wt wt wt wt wt wt wt wt wt wt wt wt wt wt wt mut mut wt wt wt wt wt wt wt wt wt wt wt wt wt wt wt wt wt wt na na wt na wt wt wt STAG2 status mut wt wt wt wt wt wt wt wt mut mut wt wt wt wt wt wt wt wt wt wt wt wt wt wt mut wt wt wt wt wt wt wt mut wt wt mut wt wt wt mut wt wt wt na na wt na wt wt wt Simple/Complex simple complex simple complex simple complex complex complex complex simple simple complex complex complex complex complex simple simple complex complex complex complex complex complex complex complex simple simple complex simple complex simple simple complex complex complex complex simple complex simple complex complex complex complex simple complex complex simple complex complex complex TP53 CN loss gain 70 Table S2.2 continued Sample ID IC340 IC343 IC831 IC929 IC973 EWS2006 EWS2008 EWS2009 EWS2012 EWS2017 EWS2020 1996-01-P034_T 1996-07-P075_T 1997-03-P152_T 1997-06-P036_T 1997-08-P462_T 1997-10-P083_T 1997-11-P048_T 1997-12-P412_T 1997-12-P616_T 1998-01-P131_T 1998-02-P170_T 1998-05-P103_T 1998-08-P054_T 1998-12-P279_T 1999-01-P459_T 1999-12-P2194_T 2000-01-P1216_T 2000-06-P2007_T 2000-07-P1079_T 2000-08-P1004_T 2000-08-P1177_T 2001-08-P8005_T 2001-10-P1058_T 2001-10-P4110_T 2001-10-P8045_T 2001-11-P8004_T 2001-12-P4072_T 2001-12-P8001_T 2002-03-P1004_T 2002-05-P8014_T 2002-05-P8068_T 2002-07-P8041_T 2003-01-P8014_T CHEWS001-1 CHEWS002-1 CHEWS003-1 CHEWS004-1 CHEWS005-1 CHEWS006-1 CHEWS007-1 Dataset Tirode Tirode Tirode Tirode Tirode Brohl Brohl Brohl Brohl Brohl Brohl Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Tumor type NA PT PT ML PT PT PT PT NA PT ML/RX NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA TP53 status wt wt wt wt mut wt wt wt wt wt wt mut wt wt wt wt wt wt wt wt wt wt wt mut wt wt wt wt wt wt wt wt wt wt wt wt wt mut wt wt wt wt wt wt wt wt wt wt wt wt wt STAG2 status wt wt wt wt wt wt wt wt wt mut mut wt wt wt wt wt wt wt wt wt wt wt wt wt wt wt wt mut wt wt wt wt wt wt wt wt wt wt wt wt wt wt wt wt wt mut wt wt wt wt wt Simple/Complex simple simple simple complex complex complex complex complex complex complex complex complex simple complex complex complex complex complex complex complex complex simple simple complex complex complex complex complex complex simple complex complex simple complex simple complex complex complex complex complex complex complex complex complex simple complex simple complex complex complex simple TP53 CN loss loss loss 71 Table S2.2 continued Sample ID CHEWS008-1 CHEWS009-1 CHEWS010-1 CHEWS011-1 CHEWS012-1 CHEWS013-1 CHEWS014-1 CHEWS015-1 CHEWS016-1 CHEWS017-1 CHEWS018-1 CHEWS019-1 CHEWS020-1 CHEWS021-1 CHEWS023-1 CHEWS024-1 CHEWS025-1 CHEWS026-1 CHEWS027-1 CHEWS028-1 CHEWS029-1 CHEWS030-1 CHEWS031-1 CHEWS032-1 CHEWS033-1 CHEWS034-1 CHEWS035-1 CHEWS036-1 SJDES00028-1 SJDES00029-1 SJDES001-1 SJDES002-1 SJDES003-1 SJDES004-1 SJDES005-1 SJDES006-1 SJDES007-1 SJDES008-1 SJDES009-1 SJDES010-1 SJDES011-1 SJDES012-1 SJDES013-1 SJDES014-1 SJDES015-1 SJDES016-1 SJDES017-1 SJDES018-1 SJDES019-1 SJDES020-1 SJDES021-1 Dataset Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Crompton Tumor type TP53 status NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA PT PT ML ML/RX ML/RX PT PT PT RX PT PT PT ML PT PT RX PT PT PT RX PT PT PT wt wt wt wt wt wt wt wt wt wt wt wt wt wt wt wt wt wt wt wt wt wt wt wt wt wt wt wt wt wt wt wt wt wt wt wt mut wt wt wt wt wt wt wt wt wt wt mut wt wt wt STAG2 status wt wt wt wt wt wt wt wt wt wt mut wt wt wt wt wt wt wt wt wt wt mut wt wt wt wt wt wt wt wt wt wt wt wt wt mut mut wt wt wt wt wt wt wt wt wt wt mut wt wt wt Simple/Complex complex complex simple simple simple complex complex complex complex complex simple complex simple complex simple complex complex complex simple complex simple complex complex simple complex complex simple simple simple complex complex complex complex simple simple simple complex simple complex complex complex complex complex simple complex complex complex complex complex complex complex TP53 CN gain gain loss gain 72 Table S2.2 continued Sample ID SJDES022-1 SJDES023-1 SJDES025-1 SJDES026-1 SJDES027-1 2213 2226 2234 2440 274194 2925 4004 4021 4022 4094 4117 4120 4143 4197 4226 4311 4434 4459 4460 4461 4462 4464 4465 4466 Dataset Crompton Crompton Crompton Crompton Crompton Anderson Anderson Anderson Anderson Anderson Anderson Anderson Anderson Anderson Anderson Anderson Anderson Anderson Anderson Anderson Anderson Anderson Anderson Anderson Anderson Anderson Anderson Anderson Anderson Tumor type TP53 status ML/RX ML ML ML/RX ML/RX PT PT PT PT PT PT PT PT PT PT PT PT PT PT PT PT PT PT PT PT PT PT PT PT wt mut wt wt wt wt wt wt wt wt wt wt wt wt mut wt wt wt wt wt wt wt wt wt wt wt wt mut wt STAG2 status wt mut wt wt wt wt wt wt wt wt wt wt wt wt wt wt wt wt wt mut wt wt wt wt wt wt wt mut wt Simple/Complex complex complex complex complex complex complex complex complex complex complex complex complex simple complex complex complex simple simple complex simple simple complex complex complex complex complex complex complex complex TP53 CN loss gain loss loss loss loss loss 73 Table S2.3 Influence of metastases at diagnosis and the simple/complex subtypes on event-free and overall survival Event Free Survival Event Free Survival Overall Survival Overall Survival Univariate Multivariate Univariate Multivariate Variable HR 95% CI P-val HR 95% CI P-val HR 95% CI P-val HR 95% CI P-val Mets at DX (yes) 2.68 1.146.29 0.024 2.63 1.066.50 0.037 2.82 1.037.70 0.043 2.34 0.826.69 0.112 Complex 1.37 0.822.30 0.225 1.05 0.522.13 0.899 1.75 0.903.38 0.097 1.61 0.654.01 0.308 74 Table S2.4 Relationship between TP53 CNA and TP53 mutation status in Ewing sarcoma TP53 Alteration Mut WT P-val TP53 Loss 7 (50%) 7 (50%) P=0.0001 TP53 No Loss 11 (4%) 246 (96%) 75 Table S2.5 TP53 pathway copy number and mutation status in Ewing sarcoma cell lines Cell line TP53 status TP53 CNA CDKN2A MDM2 MDM4 A673 c.354_355insCA (p.Ala119fs*5) Het loss Homo del Diploid Diploid CHLA9 WT Diploid Homo del Diploid 1 copy gain CHLA10 c.783-1G>T Het loss Diploid Het loss (partial) 1 copy gain CHLA258 WT Diploid Homo del Diploid Diploid ES1 c.743G>A (p.Arg248Gln) Diploid Homo del Diploid Diploid ES8 c.404G>T (p.Cys135Phe) Het loss Homo del 1 copy gain 1 copy gain TC32 WT Diploid Homo del 1 copy gain 2 copy gain TC252 WT Diploid Homo del Het loss (partial) 2 copy gain 76 Table S2.6 IC50 of Ewing sarcoma cell lines treated with doxorubicin, etoposide, or Nutlin-3a Cell line Doxorubicin IC50 (uM) Etoposide IC50 (uM) Nutlin-3a IC50 (uM) A673 0.5066 1.319 >10 CHLA258 0.6186 1.743 >10 CHLA9 0.005413 0.08748 2.125 CHLA10 0.07529 0.3777 >10 TC252 0.005782 0.04365 0.4559 TC32 0.01108 0.0573 0.7228 ES1 0.04352 0.3693 >10 ES8 0.08079 0.2133 >10 Cell line Doxorubicin IC50 (uM) Etoposide IC50 (uM) Nutlin-3a IC50 (uM) TC252 Parental 0.03427 0.04661 2.862 TC252 KO 0.171 0.1599 >10 Unpaired t test P-val <0.0001 <0.0001 Cell line Doxorubicin IC50 (uM) Etoposide IC50 (uM) Nutlin-3a IC50 (uM) TC252 Parental + Empty 0.02153 0.04251 0.7708 TC252 Parental + p53 0.06956 0.02685 1.47 TC252 KO + Empty 0.1045 0.2437 >10 TC252 KO + p53 0.03956 0.03317 2.983 Parental + Empty vs. Parental + p53 0.0013 n.s. Parental + Empty vs. KO + Empty <0.0001 <0.0001 Parental + Empty vs. KO + p53 n.s. n.s. KO + Empty vs. KO + p53 <0.0001 <0.0001 Cell line Doxorubicin IC50 (uM) Etoposide IC50 (uM) Nutlin-3a IC50 (uM) TC32 Parental 0.03185 0.07784 0.7763 TC32 KO 0.1906 0.297 >10 Unpaired t test P-val <0.0001 <0.0001 One-way ANOVA P-val 77 Table S2.6 continued Cell line Doxorubicin IC50 (uM) Etoposide IC50 (uM) Nutlin-3a IC50 (uM) TC32 parental, Empty 0.02999 0.1168 2.252 TC32 parental, p53 0.04437 0.4454 1.002 TC32 KO, Empty 0.2749 0.708 >10 TC32 KO, p53 0.1968 0.7626 1.369 Parental + Empty vs. Parental + p53 n.s. n.s. Parental + Empty vs. KO + Empty <0.0001 0.0018 Parental + Empty vs. KO + p53 <0.0001 0.0006 KO + Empty vs. KO + p53 0.0036 n.s. One-way ANOVA P-val CHAPTER 3 C/EBPβ-1 PROMOTES TRANSFORMATION AND CHEMORESISTANCE IN EWING SARCOMA CELLS This work is reprinted with the permission of Oncotarget through the Creative Commons Attribution 3.0 License. The manuscript was originally published in Oncotarget on January 27, 2017. 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 CHAPTER 4 CONCLUSIONS AND FUTURE DIRECTIONS Ewing sarcoma was first described in 1921 by James Ewing, who described the tumor as a “diffuse endothelioma of bone” [1]. Since then, researchers have learned more about the molecular, genetic, and morphological characteristics of this disease. The most unique feature of this bone and soft tissue cancer is the presence of an EWS-ETS translocation, most commonly EWS-FLI1 [2]. This encodes a powerful and oncogenic transcription factor that alters expression of many genes in the cell. EWS-FLI1 binds to GGAA microsatellite repeats in the promoter regions of cells, where it can then activate or suppress transcription [3-6]. Since the majority of Ewing sarcoma has an EWS-ETS translocation, it is likely that the fusion protein is a primary driver in the development of this cancer. This is further supported by the fact that Ewing sarcoma has few recurring somatic point mutations, an observation common to translocation-driven pediatric cancers. Aside from STAG2 and TP53 mutations occurring in 20% and 10% of cases, respectively, there are few other genes that are frequently mutated in Ewing sarcoma [79]. The genetic alteration that arises more frequently than point mutations, however, is copy number alterations (CNAs). Trisomy 8 occurs in up to 50% of Ewing sarcoma. Other common CNAs include amplifications of chromosomes 1q, 2, 5, 8, 12, and 20, and 98 copy number loss of chromosomes 10 and 16q. Focal deletion of chromosome 9p21, the region containing CDKN2A, is deleted in 10-20% of Ewing sarcoma. Focal CNAs have also been reported in regions containing CEBPB and SMARCB1, among others (refer to Chapter 3) [10, 11]. Although the frequency of these CNAs has been well characterized, few have more deeply explored the functional consequences of these genomic changes. This dissertation is the first to describe different subtypes of Ewing sarcoma based on CNAs: those containing CNAs (complex) and those without (simple). It also highlights specific genes associated with the complex subtype and examines their molecular functions in the context of Ewing sarcoma. Genomic subtypes of Ewing sarcoma Our lab and others have shown frequent and recurring CNAs in Ewing sarcoma [12-14]. We classified Ewing sarcoma tumors into two subtypes based on CNAs: those containing CNAs in greater than or equal to 1% of the length of the genome we defined as genomically complex; those containing CNAs in less than 1% of the length of the genome we defined as genomically simple (see Chapter 2). The complex subtype was associated with worse patient outcome, consistent with observations from previous studies that show inferior outcome with increasing CNAs [12, 14, 15]. We characterized only these two subtypes through copy number, but there are likely more subtypes that can be defined with additional profiling. It would be most valuable to include somatic, copy number, gene expression, and methylation profiles of the same tissue simultaneously for a comprehensive profile of this disease. This can be challenging as Ewing sarcoma samples are often acquired through tissue punches at the 99 time of diagnosis, and thus genetic material is limited due to the size of the tissue. Tumors are often surgically removed, but this usually occurs after treatment, when much of the tumor is necrotic and is not viable for genomic profiling platforms. Archived Ewing sarcoma samples are also challenging, as the genetic material from formalin-fixed paraffin-embedded (FFPE) tissues are degraded compared to that of fresh tissue. These hurdles can be overcome, however, especially as DNA/RNA extraction and genomic profiling technologies improve. The somatic and structural landscape of Ewing sarcoma has been well characterized. Future genomic studies of Ewing sarcoma should include gene expression and methylation profiling, as these are less understood. Since point mutations are rare, there are likely powerful epigenetic regulators at play in the progression of Ewing sarcoma. DNA and histone methylation can suppress expression of genes without altering the gene sequence, so will not be detected on whole-genome or whole-exome platforms that have been used in the past [7-9]. The ultimate goal of characterizing and subtyping Ewing sarcoma by copy number and other genomic signatures is to improve diagnosis and provide superior personalized treatments. With a comprehensive understanding of the CNAs, mutations, gene expression, and methylation signatures involved in Ewing sarcoma subtype development, we can provide more personal therapies tailored to each specific subtype. TP53 in Ewing sarcoma TP53 is mutated in only 5-10% of Ewing sarcoma cases, a stark contrast from the 30-50% that arise in certain adult cancers [7-9, 16]. Many previous studies have explored 100 the impact TP53 mutations in Ewing sarcoma have on outcome. Some demonstrate that TP53 mutations result in significantly worse outcome than those with wild-type TP53, while others claim there is no statistically significant difference [9, 17-20]. In all of these studies, however, tumors with TP53 mutations trend toward worse clinical outcome, suggesting that mutant TP53 results in more aggressive tumors or in treatment resistance. Indeed, studies have shown increased frequency of TP53 mutations in treated Ewing sarcoma samples compared to those acquired at the time of diagnosis (pretreatment) [8]. This suggests that TP53 is more involved in treatment resistance and relapse than in the initiation of Ewing sarcoma, but the mechanism has not been described. Few studies, aside from our own, have explored CNAs of TP53 in Ewing sarcoma. Heterozygous copy number loss occurs in 5% of Ewing sarcoma and results in decreased TP53 expression and significantly worse 5-year event-free and overall survival (refer to Chapter 2). This further emphasizes the importance of functional p53 in Ewing sarcoma, and suggests that a variety of mechanisms exist to suppress p53 function. Additional mechanisms of p53 suppression include CNAs of p53 pathway genes, such as CDKN2A, MDM2, and MDM4. CDKN2A deletion has been well characterized in Ewing sarcoma and occurs in 10-20% of cases [19-21]. CDKN2A deletion is typically mutually exclusive with TP53 mutations and copy number loss, suggesting that either mechanism alone is sufficient to compromise this pathway (Chapter 2). MDM2 and MDM4 CNAs are less characterized in the context of Ewing sarcoma. We demonstrate that MDM2 and MDM4 are each amplified in approximately 15% of Ewing sarcoma, and these amplifications are also generally mutually exclusive to each other, CDKN2A deletions, and TP53 loss. This further supports the hypothesis that multiple mechanisms 101 are utilized to abrogate the p53 pathway in Ewing sarcoma. Future studies should investigate CNAs of additional players of this pathway, such as E2F1, CDKN1A, and ATM. CEBPB in Ewing sarcoma From one of our earliest Ewing sarcoma cohorts, we identified a region of copy number gain on chromosome 20 that contains CEBPB. This gain occurred in 15% of the samples of that cohort and correlated with worse patient outcome [11]. In support of this, CEBPB gain appears at a frequency of 18% in all our cohorts combined (N=179, data not shown). Chapter 3 shows how CEBPB is indirectly regulated by EWS-FLI1, and how CEBPB-1, the largest of the 3 CEBPB isoforms, functions as an oncogene in Ewing sarcoma cells by promoting 3D colony formation and resistance to chemotherapies [10]. Furthermore, ALDH1A1 is identified as a critical target of CEBPB-1. ALDH1A1 is a marker of cancer stem cells and had previously been shown to increase colony formation and chemoresistance, so we hypothesized that CEBPB functions through ALDH1A1 in Ewing sarcoma cells [22]. Cancer stem cells are not well characterized in Ewing sarcoma, however, though they have been identified [23]. Cancer stem cells are unique in their ability to both selfrenew and to differentiate into different cell lineages. ALDH1A1 is a stem cell marker in various cancers [24-26]. It is possible that ALDH1A1 is involved in cancer stem cell development or maintenance in Ewing sarcoma, possibly mediated through CEBPB. An in-depth analysis of these genes is required to better understand their relationship. Regardless of the involvement of CEBPB and ALDH1A1, it would be valuable to 102 investigate the establishment of cancer stem cell populations in Ewing sarcoma. Cancer stem cells are refractory to treatments, including chemotherapies, the primary treatment modality in Ewing sarcoma [27]. Cancer stem cells may partially explain the adverse outcome in some patients, as cancer stem cells will evade chemotherapy-induced cell death and will re-populate the tumor at relapse. A better understanding of cancer stem cells in Ewing sarcoma may lead to new therapies that can specifically target that population and reduce relapse. HOTAIR in Ewing sarcoma HOX antisense intergenic RNA, or HOTAIR, is a long non-coding RNA on chromosome 12q13.13 near the HOXC gene cluster [28]. HOTAIR, though not proteincoding, interacts with the chromatin remodeling complexes PRC2 (including EZH2) and LSD1/REST and localizes them to target genes where they suppress transcription [28, 29]. High HOTAIR expression has been implicated in many cancers, and is associated with metastasis, more aggressive disease, and increased cell growth and proliferation [3034]. The appendix of this dissertation is one of the first studies to describe the role of HOTAIR in Ewing sarcoma. We utilized CRISPR technology to delete the HOTAIR gene from two Ewing sarcoma cell lines. To determine how HOTAIR affects cell growth, proliferation, migration, and colony formation, we assayed multiple single-cell knockout clones alongside the untreated wild-type parental cell lines. HOTAIR knockout increased cell growth, proliferation, migration, and colony formation (refer to the Appendix), which is the opposite of what we expected based on HOTAIR function in other cancer types. 103 HOTAIR appears to function uniquely in Ewing sarcoma compared to other cancer types, but the exact mechanisms involved are not known. It is possible that HOTAIR activity is altered by the EWS-ETS translocation, since these translocations are also unique to Ewing sarcoma. RNA immunoprecipitation experiments are needed to determine whether HOTAIR interacts with PRC2 and LSD1/REST in Ewing sarcoma as it does in other cancer types. It is possible that HOTAIR binds to other chromatin remodelers or transcription factors as well, which would explain its unique function in these cells. Another experiment that will reveal HOTAIR function in Ewing sarcoma is RNA sequencing. By comparing the gene expression signatures of cells with HOTAIR to those without, we can determine which pathways are altered by HOTAIR. LSD1 and EZH2 inhibitors could also be included as treatments in the RNA sequencing experiment. This will help determine whether inhibition of these chromatin remodelers is negatively or positively affected by HOTAIR. LSD1 inhibitors are currently in clinical trials for the treatment of Ewing sarcoma (ClinicalTrials.gov identifier NCT03600649). If high HOTAIR expression affects the sensitivity of cells to LSD1 inhibition, it would be crucial to know the status of HOTAIR expression in patients prior to administering treatment. To truly understand the relationship of HOTAIR with LSD1, EZH2, and other chromatin remodeling complexes in the context of Ewing sarcoma, future studies are required. Little is known about the regulation of HOTAIR expression in Ewing sarcoma. Our data suggest that its promoter region is de-methylated in some samples of the complex subtype compared to the simple subtype, suggesting DNA methylation as one method of expression regulation (refer to the Appendix). Other studies have shown 104 HOTAIR regulation is mediated by micro-RNAs [35, 36]. To my knowledge, these mechanisms have not been studied in Ewing sarcoma and thus warrant further investigation. Other genes A preliminary study in our lab compared gene expression to CNAs in 37 Ewing sarcoma samples from Michigan and the Children’s Oncology Group. We identified paternally expressed gene 10 (PEG10) as being the most differentially expressed gene between samples with trisomy of chromosome 8 compared to samples with diploid 8 (unpublished data). PEG10 was overexpressed in the trisomy 8 samples compared to the diploid 8 samples. Interestingly, PEG10 is located on chromosome 7, so the increase in expression cannot be explained by the trisomy alone. One proposed mechanism of PEG10 upregulation in trisomy 8 Ewing sarcoma is through MYC regulation, because MYC is located on chromosome 8. PEG10 has been shown by others to be a transcriptional target of MYC in other cancer types [37]. We explored PEG10 expression with MYC suppression and overexpression in Ewing sarcoma cells and saw no corresponding change in PEG10 levels, suggesting that MYC does not regulate PEG10 expression. Additionally, MYC expression did not increase in our trisomy 8 samples compared to diploid 8 samples, suggesting that copy number gain of chromosome 8 is not a primary mechanism to increase MYC expression. This is contrary to other cancers, such as medulloblastoma and others, where MYC amplification corresponds to increased MYC mRNA levels [38, 39]. It is possible that a single copy gains of MYC, which occurs with trisomy 8 in Ewing sarcoma, is not enough to cause a 105 detectable change in MYC expression levels. An alternative hypothesis is that MYC expression is already high in Ewing sarcoma because it is a transcriptional target of EWSFLI1, minimizing the effect of a single copy gain of the gene [40]. However, according to the Broad Institute’s Cancer Cell Encyclopedia, Ewing sarcoma cell lines do not have exceptionally high MYC levels compared to other cancer types [41]. Collectively, this suggests MYC’s role in the development and progression of Ewing sarcoma is minimal compared to other drivers, such as EWS-FLI1. PEG10 is an imprinted gene. It is expressed only from the paternal allele, while the maternal allele is methylated and silenced [42]. Our group began exploring the consequence of CNAs of imprinted genes in Ewing sarcoma. We wanted to determine whether specific maternal or paternal alleles had copy number gain or loss in order to favor expression or silencing of imprinted genes. For example, if the paternal allele of PEG10 was preferentially lost over the maternal allele, we would expect to see no PEG10 mRNA expression because the only copy present would be silenced by imprinting. However, this study was limited by the number of samples for which we had both copy number and gene expression data. It would be valuable for future studies to increase the sample size on both platforms so that significant patterns can be detected. If there is evidence of preferential CNAs of imprinted alleles, it further emphasizes the need to understand the epigenomic landscape of Ewing sarcoma, both through DNA methylation and imprinting, and histone modifications. References 1 Ewing J. Diffuse endothelioma of bone. Proceedings of the New York Pathological Society 1921; 21: 17-24. 106 2 Delattre O, Zucman J, Plougastel B, Desmaze C, Melot T, Peter M et al. Gene fusion with an ETS DNA-binding domain caused by chromosome translocation in human tumours. Nature 1992; 359: 162-165. 3 Gangwal K, Sankar S, Hollenhorst PC, Kinsey M, Haroldsen SC, Shah AA et al. Microsatellites as EWS/FLI response elements in Ewing's sarcoma. Proceedings of the National Academy of Sciences of the United States of America 2008; 105: 10149-10154. 4 Guillon N, Tirode F, Boeva V, Zynovyev A, Barillot E, Delattre O. The oncogenic EWS-FLI1 protein binds in vivo GGAA microsatellite sequences with potential transcriptional activation function. PloS One 2009; 4: e4932. 5 Owen LA, Kowalewski AA, Lessnick SL. EWS/FLI mediates transcriptional repression via NKX2.2 during oncogenic transformation in Ewing's sarcoma. PloS One 2008; 3: e1965. 6 Sankar S, Theisen ER, Bearss J, Mulvihill T, Hoffman LM, Sorna V et al. Reversible LSD1 inhibition interferes with global EWS/ETS transcriptional activity and impedes Ewing sarcoma tumor growth. Clinical Cancer Research 2014; 20: 4584-4597. 7 Brohl AS, Solomon DA, Chang W, Wang J, Song Y, Sindiri S et al. The genomic landscape of the Ewing sarcoma family of tumors reveals recurrent STAG2 mutation. PLoS Genetics 2014; 10: e1004475. 8 Crompton BD, Stewart C, Taylor-Weiner A, Alexe G, Kurek KC, Calicchio ML et al. The genomic landscape of pediatric Ewing sarcoma. Cancer Discovery 2014; 4: 1326-1341. 9 Tirode F, Surdez D, Ma X, Parker M, Le Deley MC, Bahrami A et al. Genomic landscape of Ewing sarcoma defines an aggressive subtype with co-association of STAG2 and TP53 mutations. Cancer Discovery 2014; 4: 1342-1353. 10 Gardiner JD, Abegglen LM, Huang X, Carter BE, Schackmann EA, Stucki M et al. C/EBPbeta-1 promotes transformation and chemoresistance in Ewing sarcoma cells. Oncotarget 2017. 11 Jahromi MS, Putnam AR, Druzgal C, Wright J, Spraker-Perlman H, Kinsey M et al. Molecular inversion probe analysis detects novel copy number alterations in Ewing sarcoma. Cancer Genetics 2012; 205: 391-404. 12 Ferreira BI, Alonso J, Carrillo J, Acquadro F, Largo C, Suela J et al. Array CGH and gene-expression profiling reveals distinct genomic instability patterns associated with DNA repair and cell-cycle checkpoint pathways in Ewing's sarcoma. Oncogene 2008; 27: 2084-2090. 107 13 Jahromi MS, Jones KB, Schiffman JD. Copy number alterations and methylation in Ewing's sarcoma. Sarcoma 2011; 2011: 362173. 14 Savola S, Klami A, Tripathi A, Niini T, Serra M, Picci P et al. Combined use of expression and CGH arrays pinpoints novel candidate genes in Ewing sarcoma family of tumors. BMC Cancer 2009; 9: 17. 15 Ozaki T, Paulussen M, Poremba C, Brinkschmidt C, Rerin J, Ahrens S et al. Genetic imbalances revealed by comparative genomic hybridization in Ewing tumors. Genes, Chromosomes & Cancer 2001; 32: 164-171. 16 Olivier M, Hollstein M, Hainaut P. TP53 mutations in human cancers: Origins, consequences, and clinical use. Cold Spring Harbor Perspectives in Biology 2010; 2: a001008. 17 Abudu A, Mangham DC, Reynolds GM, Pynsent PB, Tillman RM, Carter SR et al. Overexpression of p53 protein in primary Ewing's sarcoma of bone: Relationship to tumour stage, response and prognosis. British Journal of Cancer 1999; 79: 1185-1189. 18 de Alava E, Antonescu CR, Panizo A, Leung D, Meyers PA, Huvos AG et al. Prognostic impact of p53 status in Ewing sarcoma. Cancer 2000; 89: 783-792. 19 Huang HY, Illei PB, Zhao Z, Mazumdar M, Huvos AG, Healey JH et al. Ewing sarcomas with p53 mutation or p16/p14ARF homozygous deletion: A highly lethal subset associated with poor chemoresponse. Journal of Clinical Oncology 2005; 23: 548-558. 20 Lerman DM, Monument MJ, McIlvaine E, Liu XQ, Huang D, Monovich L et al. Tumoral TP53 and/or CDKN2A alterations are not reliable prognostic biomarkers in patients with localized Ewing sarcoma: A report from the Children's Oncology Group. Pediatric Blood & Cancer 2015; 62: 759-765. 21 Lopez-Guerrero JA, Pellin A, Noguera R, Carda C, Llombart-Bosch A. Molecular analysis of the 9p21 locus and p53 genes in Ewing family tumors. Laboratory Investigation 2001; 81: 803-814. 22 Awad O, Yustein JT, Shah P, Gul N, Katuri V, O'Neill A et al. High ALDH activity identifies chemotherapy-resistant Ewing's sarcoma stem cells that retain sensitivity to EWS-FLI1 inhibition. PloS One 2010; 5: e13943. 23 Suva ML, Riggi N, Stehle JC, Baumer K, Tercier S, Joseph JM et al. Identification of cancer stem cells in Ewing's sarcoma. Cancer Research 2009; 69: 1776-1781. 108 24 Ginestier C, Hur MH, Charafe-Jauffret E, Monville F, Dutcher J, Brown M et al. ALDH1 is a marker of normal and malignant human mammary stem cells and a predictor of poor clinical outcome. Cell Stem Cell 2007; 1: 555-567. 25 Jiang F, Qiu Q, Khanna A, Todd NW, Deepak J, Xing L et al. Aldehyde dehydrogenase 1 is a tumor stem cell-associated marker in lung cancer. Molecular Cancer Research : MCR 2009; 7: 330-338. 26 Li T, Su Y, Mei Y, Leng Q, Leng B, Liu Z et al. ALDH1A1 is a marker for malignant prostate stem cells and predictor of prostate cancer patients' outcome. Laboratory Investigation 2010; 90: 234-244. 27 Abdullah LN, Chow EK-H. Mechanisms of chemoresistance in cancer stem cells. Clinical and Translational Medicine 2013; 2: 3-3. 28 Rinn JL, Kertesz M, Wang JK, Squazzo SL, Xu X, Brugmann SA et al. Functional demarcation of active and silent chromatin domains in human HOX loci by noncoding RNAs. Cell 2007; 129: 1311-1323. 29 Tsai MC, Manor O, Wan Y, Mosammaparast N, Wang JK, Lan F et al. Long noncoding RNA as modular scaffold of histone modification complexes. Science 2010; 329: 689-693. 30 Chen FJ, Sun M, Li SQ, Wu QQ, Ji L, Liu ZL et al. Upregulation of the long noncoding RNA HOTAIR promotes esophageal squamous cell carcinoma metastasis and poor prognosis. Molecular Carcinogenesis 2013; 52: 908-915. 31 Gupta RA, Shah N, Wang KC, Kim J, Horlings HM, Wong DJ et al. Long noncoding RNA HOTAIR reprograms chromatin state to promote cancer metastasis. Nature 2010; 464: 1071-1076. 32 Kogo R, Shimamura T, Mimori K, Kawahara K, Imoto S, Sudo T et al. Long noncoding RNA HOTAIR regulates polycomb-dependent chromatin modification and is associated with poor prognosis in colorectal cancers. Cancer Research 2011; 71: 6320-6326. 33 Liu XH, Liu ZL, Sun M, Liu J, Wang ZX, De W. The long non-coding RNA HOTAIR indicates a poor prognosis and promotes metastasis in non-small cell lung cancer. BMC Cancer 2013; 13: 464. 34 Sorensen KP, Thomassen M, Tan Q, Bak M, Cold S, Burton M et al. Long noncoding RNA HOTAIR is an independent prognostic marker of metastasis in estrogen receptor-positive primary breast cancer. Breast Cancer Research and Treatment 2013; 142: 529-536. 109 35 Chiyomaru T, Fukuhara S, Saini S, Majid S, Deng G, Shahryari V et al. Long non-coding RNA HOTAIR is targeted and regulated by miR-141 in human cancer cells. The Journal of Biological Chemistry 2014; 289: 12550-12565. 36 Dasgupta P, Kulkarni P, Majid S, Shahryari V, Hashimoto Y, Bhat NS et al. MicroRNA-203 inhibits long noncoding RNA HOTAIR and regulates tumorigenesis through epithelial-to-mesenchymal transition pathway in renal cell carcinoma. Molecular Cancer Therapeutics 2018; 17: 1061-1069. 37 Li CM, Margolin AA, Salas M, Memeo L, Mansukhani M, Hibshoosh H et al. PEG10 is a c-MYC target gene in cancer cells. Cancer Research 2006; 66: 665672. 38 Kalkat M, De Melo J, Hickman KA, Lourenco C, Redel C, Resetca D et al. MYC deregulation in primary human cancers. Genes 2017; 8: 151. 39 Robinson G, Parker M, Kranenburg TA, Lu C, Chen X, Ding L et al. Novel mutations target distinct subgroups of medulloblastoma. Nature 2012; 488: 43-48. 40 Dauphinot L, De Oliveira C, Melot T, Sevenet N, Thomas V, Weissman BE et al. Analysis of the expression of cell cycle regulators in Ewing cell lines: EWS-FLI-1 modulates p57KIP2and c-Myc expression. Oncogene 2001; 20: 3258-3265. 41 Barretina J, Caponigro G, Stransky N, Venkatesan K, Margolin AA, Kim S et al. The Cancer Cell Line Encyclopedia enables predictive modelling of anticancer drug sensitivity. Nature 2012; 483: 603-607. 42 Ono R, Kobayashi S, Wagatsuma H, Aisaka K, Kohda T, Kaneko-Ishino T et al. A retrotransposon-derived gene, PEG10, is a novel imprinted gene located on human chromosome 7q21. Genomics 2001; 73: 232-237. APPENDIX THE LONG NON-CODING RNA HOTAIR IN EWING SARCOMA Abstract HOTAIR is a long non-coding RNA that is involved in growth, invasion, and metastasis in many cancers. However, until now, the role of HOTAIR has not been explored in Ewing sarcoma, an aggressive bone cancer found in children and young adults. HOTAIR is highly expressed in Ewing sarcoma compared to other cancer types. HOTAIR is especially high in the genomically complex Ewing sarcoma subtype, where its promoter region is demethylated compared to the genomically simple Ewing sarcoma subtype. We employed CRISPR knockout technology to explore the consequences of HOTAIR deletion on Ewing sarcoma growth. Knockout of HOTAIR led to increased cell viability, proliferation, colony formation, and migration. HOTAIR knockout also increased sensitivity of Ewing sarcoma cells to chemotherapeutic agents. We further discuss potential mechanisms through which HOTAIR may be regulated in Ewing sarcoma, and provide future experiments that can explore HOTAIR activity further. 111 Introduction As next-generation sequencing technologies improve, more is understood about the human genome and its many functions. It is estimated that up to 98% of the human genome contains non-protein-coding genes [1, 2]. These genes were originally thought to have no function, but researchers are continuously discovering the roles of non-coding RNAs in human biology. Long non-coding RNAs (lncRNAs) are typically greater than 200 nucleotides in length and are found throughout the genome. Their roles include transcriptional and posttranscriptional regulation within the cell [3, 4]. Many lncRNAs affect gene expression by recruiting and localizing chromatin remodelers to target sites [5]. LncRNAs are also implicated in several diseases, including cancer [6]. Overexpression of some lncRNAs increases tumor growth, proliferation, and metastasis [7, 8]. Humans have four clusters of sequentially regulated homeobox, or HOX, genes: HOXA-HOXD. HOX genes are important for proper development of embryos along their head-tail axis and are involved in establishing limb identity [9]. HOX antisense intergenic RNA, or HOTAIR, is a long non-coding RNA on chromosome 12q13.13 near the HOXC gene cluster [10]. Although located within the HOXC regulatory region, HOTAIR does not regulate HOXC expression, but acts in trans and affects expression of the HOXD locus on chromosome 2 [11]. HOTAIR functions as a molecular scaffold between the polycomb repressive complex 2 (PRC2) on its 5’ end and the lysine specific demethylase 1 (LSD1)/RE1-silencing transcription factor (REST) complex on its 3’ end [10, 12]. This couples the histone H3 lysine 27 methylation ability of PRC2 with the lysine 4 demethylation ability of LSD1, repressing the expression of target genes, including tumor 112 suppressors [12, 13]. LSD1 inhibitors are currently used in preclinical and clinical trials for the treatment of several cancers, including leukemia, non-small cell lung cancer, and Ewing sarcoma [14-17]. Inhibition of enhancer of zeste homologue 2 (EZH2), the catalytic subunit of PRC2, is also in progress for preclinical and clinical trials [18]. However, the effect of these inhibitors on HOTAIR activity and the effect of HOTAIR expression on inhibitor function is currently unexplored. HOTAIR expression is associated with poor prognosis and plays a role in the development of several cancers, including breast, ovarian, colorectal, lung, and pancreatic cancers [19-21]. Higher HOTAIR expression is found in metastatic tumors compared to primary tumors in breast cancer, so is thought to contribute to metastatic spread [13]. HOTAIR plays a role in initiating epithelial to mesenchymal transition and maintaining cancer stem cells [22]. Furthermore, HOTAIR is involved in cell proliferation, invasion, metastasis, and chemoresistance, and suppression of HOTAIR expression results in the opposite effect [20, 23-25]. The field of non-coding RNAs is still relatively new, so much can still be learned about how these molecules function. This study explores the role of HOTAIR in the context of Ewing sarcoma, which has previously remained unexplored. Ewing sarcoma is the second most common bone cancer in children in young adults, and outcome is poor [26]. We relate HOTAIR expression and promoter methylation to the genomically complex Ewing sarcoma subtype, defined in more detail in Chapter 2. We explore the functional consequences of HOTAIR knockout on cell growth, transformation, and chemosensitivity in Ewing sarcoma cells. 113 Materials and methods Tumor collection and DNA/RNA extraction Deidentified FFPE samples from clinically diagnosed Ewing sarcoma tumors were cut into 5um scrolls as well as cut onto slides. 1-2 slides were stained with H&E and tumor content was determined by a pathologist. Samples with >60% tumor content were deparaffinized in Hemo-De and washed in ethanol. DNA and RNA were simultaneously extracted using the RecoverAllTM Multi-Sample RNA/DNA Isolation Workflow (Invitrogen). DNA was used for copy number and methylation analysis. RNA was used for gene expression microarray analysis. Genomic copy number analysis DNA extracted from patient Ewing sarcoma FFPE tumors underwent copy number analysis with either the MIP Copy Number assay, the OncoScan FFPE Express 2.0, or the OncoScan CNV FFPE assay (Affymetrix) [27]. Data visualization and copy number analysis was performed with Nexus Copy Number 9.0 (BioDiscovery). Samples were classified based on the percentage of the length of the genome changed by CNA (less than 1% for simple; greater than 1% for complex). DNA methylation array DNA extracted from patient Ewing sarcoma FFPE tumors underwent bisulfite conversion using the EZ DNA Methylation Kit (Zymo Research), followed by DNA methylation analysis with the Infinium Human Methylation 450k Beadchip (Illumina, San Diego, CA). Raw data underwent quality control and normalization using ChAMP 114 [28] and differentially methylated regions were identified and visualized using ChAMP, GenomeStudio (Illumina), and Partek Genomics Suite (Partek). Gene expression microarray RNA extracted from patient Ewing sarcoma FFPE tumors underwent gene expression analysis with the Human Transcriptome Array 2.0 (Affymetrix). Raw Affymetrix CEL files were processed by Affymetrix Expression Console as described previously [29]. Cell culture A673 cells were from American Type Culture Collection and ES8 cells were from St. Jude Childhood Solid Tumor Network. All experiments were performed within 3 months of cell line resuscitation, and cell line authentication was performed by providers by STR profiling. A673 and ES8 cells were cultured in DMEM supplemented with 10% FBS, 1% glutamax, and 1% sodium pyruvate. CRISPR knockout A673 and ES8 cell lines were transfected with CRISPR/Cas9 reagents using the TransIT-LT1 Transfection protocol (Mirus). Following selection with puromycin and blasticidin, single cells were plated and allowed to grow out. Single cell clones were screened for successful HOTAIR knockout by PCR. HOTAIR knockout was further validated by qPCR using TaqMan Gene Expression Assays (Thermo Fisher Scientific, HOTAIR variant 1= Hs03296630_m1; HOTAIR variant 3= Hs03296661_mH) normalized 115 to GAPDH (Thermo Fisher Scientific, 4310884E-1605067). Cell viability, proliferation, and colony formation assays A673 and ES8 cells that are either HOTAIR wild-type (+/+), heterozygous (+/-), or knockout (-/-) were plated in 96-well plates at 2500 cells/well in triplicate. Viability of cells was measured 96 hours after plating using Cell Titer Glo (Promega). Proliferation was measured as previously described [30]. Briefly, 500,000 cells were plated in 10-cm dishes and grown in normal conditions for 3 days. Cells were counted and 500,000 cells were re-plated. This process repeated every 3 days for 15 days. For the colony formation assay, 625 cells were plated per well as described previously [29]. Colonies were counted and imaged after 14 days. Scratch assay 600,000 cells/well were plated in 24-well plates. After cells adhered to the surface, a scratch was made using a pipet tip. Pictures of the scratch were taken every 12 hours for 36 hours. Area of the scratch that was filled in was measured using ImageJ. Drug response assay 2500 cells/well were plated in 96-well plates and allowed to adhere. Doxorubicin, etoposide, and HCI-2509 were added at the indicated concentrations, and cell viability was measured using Cell Titer Glo (Promega) 48-72 hours later. 116 Results HOTAIR is highly expressed in Ewing sarcoma To determine whether HOTAIR was expressed in Ewing sarcoma cells, we interrogated the Broad Institute’s Cancer Cell Line Encyclopedia (CCLE) for expression of HOTAIR [31]. Out of the 37 different cancer types included, Ewing sarcoma cell lines had the highest HOTAIR expression on average (Figure A.1A). For a subset of Ewing sarcoma patient samples, we evaluated gene expression by using the Human Transcriptome Array 2.0 (N=17) and whole-genome DNA methylation by using the Infinium Human Methylation 450k Beadchip (N=24). When comparing gene expression levels between the simple and complex genomic subtypes of Ewing sarcoma, we identified HOTAIR as a gene that is significantly upregulated in the complex subtype compared to the simple subtype (P=0.0106; Figure A.1B). Furthermore, the HOTAIR promoter was significantly demethylated in the genomically complex Ewing sarcoma subtype compared to the genomically simple subtype (P=0.0009; Figure A.1C). Design of HOTAIR knockout in Ewing sarcoma cells by CRISPR To evaluate the function of HOTAIR in Ewing sarcoma cells, we used clustered regularly interspaced short palindromic repeats and CRISPR associated protein 9 (CRISPR/Cas9) technology to remove the HOTAIR gene. Guide RNA was designed to target the intronic region prior to exon 2 and the 3’ UTR following exon 7 (Figure A.2A). This removes the majority of the HOTAIR gene, with the exception of exon 1, since exon 1 overlaps with the HOXC11 gene. Following transfection and antibiotic selection for the CRISPR reagents, Ewing sarcoma cell lines A673 and ES8 were plated as single cell 117 clones. PCR around the CRISPR cut site was used on the single cell clones to screen for successful HOTAIR knockout (data not shown), followed by qPCR for 2 HOTAIR transcripts (variant 1 and variant 3) as further validation of knockout (Figure A.2B). The parental cell line was included as a HOTAIR wild-type (+/+) control, as was ES8 clone 21, which underwent the transfection process but maintained wild-type HOTAIR. For both A673 and ES8, 3 heterozygous clones (+/-) and 3 homozygous deletion clones (-/-) were used in functional assays (Figure A.2B). HOTAIR deficiency promotes Ewing sarcoma cell growth and migration To evaluate the functional consequences of HOTAIR loss on Ewing sarcoma cells, viability was measured after 96 hours of growth under normal cell culture conditions. In general, clones with one or both alleles of HOTAIR missing had increased viability compared to controls (Figure A.3A). Cells were plated in soft agar to evaluate the role of HOTAIR in attachmentindependent cell growth. After 14 days of growth, A673 clones that were heterozygous or homozygous null for HOTAIR had more colonies than the wild-type control (Figure A.3B). The ES8 clones were slightly more variable in their colony formation, but also trended towards increased colony growth with HOTAIR deficiency (Figure A.3B). The population doubling time of each clone was evaluated to compare proliferation rates. In general, A673 and ES8 clones with one or both alleles of HOTAIR deleted had increased proliferation rates compared to those with both HOTAIR alleles intact (Figure A.3C). Finally, a scratch assay was used to measure cell motility in each of these clones. 118 The A673 clones with +/- or -/- HOTAIR filled the area of the scratch days before the wild-type parental cell line (Figure A.3D). However, the same was not true of the ES8 clones, where the HOTAIR -/- clones were the slowest at migrating into the scratch area (Figure A.3D). Collectively, however, the data indicate that Ewing sarcoma cells with HOTAIR knockout have increased viability, increased attachment-independent colonyformation, increased proliferation rates, and increased cell motility. HOTAIR deficiency promotes Ewing sarcoma chemotherapy sensitivity To determine the response of HOTAIR-deficient Ewing sarcoma cells to chemotherapeutic reagents, we treated the A673 and ES8 clones with doxorubicin and etoposide. The clones with one or both HOTAIR alleles deleted consistently showed increased sensitivity to both doxorubicin (Figure A.4A) and etoposide (Figure A.4B). The clones were also treated with HCI-2509, a reversible LSD1 inhibitor, to determine whether HOTAIR’s interaction with LSD1 affects the sensitivity of these cells to LSD1 inhibition. There was no difference in sensitivity to LSD1 in any of the HOTAIR knockout clones tested (Figure A.4C). Discussion and future directions HOTAIR has been implicated as an oncogene in several types of cancer, including breast, colorectal, ovarian, lung, and pancreatic cancer, among others. As reviewed by Hajjari and Salavaty, HOTAIR expression is associated with poor prognosis, metastasis, proliferation, invasion, and tumor aggression [32]. However, to our knowledge, HOTAIR has not been studied in Ewing sarcoma. This is the first study to evaluate HOTAIR 119 function in Ewing sarcoma cell lines using CRISPR knockout technology. Interestingly, HOTAIR appears to function differently in Ewing sarcoma cells from other cancer types. When HOTAIR was deleted from Ewing sarcoma cells, they became more viable, more proliferative, and more migratory, contrary to what is expected based on its function in other cancer types. Ewing sarcoma cells with HOTAIR knockout are also more sensitive to doxorubicin and etoposide, which aligns more with what is expected from other cancers. However, this may be an artifact of increased cell proliferation, as doxorubicin and etoposide primarily affect dividing cells. Future studies are required to understand the different HOTAIR functions in Ewing sarcoma compared to other cancers. One method to determine the downstream effects of HOTAIR, including its impact on chromatin remodelers such as LSD1 and PRC2, is to perform RNA sequencing on Ewing sarcoma cell lines with or without HOTAIR that have been treated with or without LSD1 (or EZH2) inhibitors. Doing this will help answer the following questions: (1) Which pathways in Ewing sarcoma cells are regulated by HOTAIR? (2) Is HOTAIR dependent upon proper LSD1 (or EZH2) activity to regulate those pathways? (3) If HOTAIR function is independent of LSD1 (or EZH2) inhibition, what downstream genes or pathways are regulated by LSD1 (or EZH2)? The results of this experiment will give us a global understanding of the regulatory effects of HOTAIR, and its chromatin remodeling partners, in Ewing sarcoma. Ewing sarcoma is characterized by a translocation between EWSR1 and a member of the ETS family of transcription factors, most commonly FLI1, creating the oncogenic transcription factor EWS-FLI1 [33, 34]. Since HOTAIR seems to behave uniquely in Ewing sarcoma, it is possible that EWS-FLI1 regulates its expression or 120 downstream activity either directly or through some intermediate regulator. This relationship has not been explored and warrants further investigation. Another potential regulator of HOTAIR in Ewing sarcoma is through promoter or CpG island methylation and/or histone remodeling [35]. Our data show that the HOTAIR promoter is de-methylated in the complex Ewing sarcoma subtype compared to the simple subtype, corresponding to an increase in expression levels. Understanding the upstream regulators of HOTAIR can help explain its increased levels in Ewing sarcoma compared to other cancer types, as demonstrated by the CCLE. Collectively, this information will help us understand the unique HOTAIR biology in the context of Ewing sarcoma. References 1 Harrow J, Frankish A, Gonzalez JM, Tapanari E, Diekhans M, Kokocinski F et al. GENCODE: The reference human genome annotation for The ENCODE Project. Genome Research 2012; 22: 1760-1774. 2 Nie L, Wu HJ, Hsu JM, Chang SS, Labaff AM, Li CW et al. Long non-coding RNAs: Versatile master regulators of gene expression and crucial players in cancer. American Journal of Translational Research 2012; 4: 127-150. 3 Kurokawa R. Long noncoding RNA as a regulator for transcription. Progress in Molecular and Subcellular Biology 2011; 51: 29-41. 4 Mercer TR, Dinger ME, Mattick JS. Long non-coding RNAs: Insights into functions. Nature Reviews Genetics 2009; 10: 155-159. 5 Mercer TR, Mattick JS. Structure and function of long noncoding RNAs in epigenetic regulation. Nature Structural & Molecular Biology 2013; 20: 300-307. 6 Wapinski O, Chang HY. Long noncoding RNAs and human disease. Trends in Cell Biology 2011; 21: 354-361. 121 7 Du Z, Fei T, Verhaak RG, Su Z, Zhang Y, Brown M et al. Integrative genomic analyses reveal clinically relevant long noncoding RNAs in human cancer. Nature Structural & Molecular Biology 2013; 20: 908-913. 8 Qiu MT, Hu JW, Yin R, Xu L. Long noncoding RNA: An emerging paradigm of cancer research. Tumour Biology 2013; 34: 613-620. 9 Mallo M, Alonso CR. The regulation of Hox gene expression during animal development. Development 2013; 140: 3951-3963. 10 Rinn JL, Kertesz M, Wang JK, Squazzo SL, Xu X, Brugmann SA et al. Functional demarcation of active and silent chromatin domains in human HOX loci by noncoding RNAs. Cell 2007; 129: 1311-1323. 11 Zhang J, Zhang P, Wang L, Piao HL, Ma L. Long non-coding RNA HOTAIR in carcinogenesis and metastasis. Acta Biochimica et Biophysica Sinica 2014; 46: 15. 12 Tsai MC, Manor O, Wan Y, Mosammaparast N, Wang JK, Lan F et al. Long noncoding RNA as modular scaffold of histone modification complexes. Science 2010; 329: 689-693. 13 Gupta RA, Shah N, Wang KC, Kim J, Horlings HM, Wong DJ et al. Long noncoding RNA HOTAIR reprograms chromatin state to promote cancer metastasis. Nature 2010; 464: 1071-1076. 14 Harris WJ, Huang X, Lynch JT, Spencer GJ, Hitchin JR, Li Y et al. The histone demethylase KDM1A sustains the oncogenic potential of MLL-AF9 leukemia stem cells. Cancer Cell 2012; 21: 473-487. 15 Maiques-Diaz A, Somervaille TCP. LSD1: Biologic roles and therapeutic targeting. Epigenomics 2016; 8: 1103-1116. 16 Mohammad HP, Smitheman KN, Kamat CD, Soong D, Federowicz KE, Van Aller GS et al. A DNA hypomethylation signature predicts antitumor activity of LSD1 inhibitors in SCLC. Cancer Cell 2015; 28: 57-69. 17 Sankar S, Theisen ER, Bearss J, Mulvihill T, Hoffman LM, Sorna V et al. Reversible LSD1 inhibition interferes with global EWS/ETS transcriptional activity and impedes Ewing sarcoma tumor growth. Clinical Cancer Research 2014; 20: 4584-4597. 18 Kim KH, Roberts CWM. Targeting EZH2 in cancer. Nature Medicine 2016; 22: 128-134. 122 19 Kogo R, Shimamura T, Mimori K, Kawahara K, Imoto S, Sudo T et al. Long noncoding RNA HOTAIR regulates polycomb-dependent chromatin modification and is associated with poor prognosis in colorectal cancers. Cancer Research 2011; 71: 6320-6326. 20 Liu XH, Liu ZL, Sun M, Liu J, Wang ZX, De W. The long non-coding RNA HOTAIR indicates a poor prognosis and promotes metastasis in non-small cell lung cancer. BMC Cancer 2013; 13: 464. 21 Xue X, Yang YA, Zhang A, Fong KW, Kim J, Song B et al. LncRNA HOTAIR enhances ER signaling and confers tamoxifen resistance in breast cancer. Oncogene 2016; 35: 2746-2755. 22 Padua Alves C, Fonseca AS, Muys BR, de Barros ELBR, Burger MC, de Souza JE et al. Brief report: The lincRNA HOTAIR is required for epithelial-tomesenchymal transition and stemness maintenance of cancer cell lines. Stem Cells (Dayton, Ohio) 2013; 31: 2827-2832. 23 Chen FJ, Sun M, Li SQ, Wu QQ, Ji L, Liu ZL et al. Upregulation of the long noncoding RNA HOTAIR promotes esophageal squamous cell carcinoma metastasis and poor prognosis. Molecular Carcinogenesis 2013; 52: 908-915. 24 Ishibashi M, Kogo R, Shibata K, Sawada G, Takahashi Y, Kurashige J et al. Clinical significance of the expression of long non-coding RNA HOTAIR in primary hepatocellular carcinoma. Oncology Reports 2013; 29: 946-950. 25 Sorensen KP, Thomassen M, Tan Q, Bak M, Cold S, Burton M et al. Long noncoding RNA HOTAIR is an independent prognostic marker of metastasis in estrogen receptor-positive primary breast cancer. Breast Cancer Research and Treatment 2013; 142: 529-536. 26 Esiashvili N, Goodman M, Marcus RB, Jr. Changes in incidence and survival of Ewing sarcoma patients over the past 3 decades: Surveillance epidemiology and end results data. Journal of Pediatric Hematology/Oncology 2008; 30: 425-430. 27 Absalan F, Ronaghi M. Molecular inversion probe assay. Methods in Molecular Biology 2007; 396: 315-330. 28 Tian Y, Morris TJ, Webster AP, Yang Z, Beck S, Feber A et al. ChAMP: Updated methylation analysis pipeline for Illumina BeadChips. Bioinformatics 2017; 33: 3982-3984. 29 Gardiner JD, Abegglen LM, Huang X, Carter BE, Schackmann EA, Stucki M et al. C/EBPbeta-1 promotes transformation and chemoresistance in Ewing sarcoma cells. Oncotarget 2017. 123 30 Lessnick SL, Dacwag CS, Golub TR. The Ewing's sarcoma oncoprotein EWS/FLI induces a p53-dependent growth arrest in primary human fibroblasts. Cancer Cell 2002; 1: 393-401. 31 Barretina J, Caponigro G, Stransky N, Venkatesan K, Margolin AA, Kim S et al. The Cancer Cell Line Encyclopedia enables predictive modelling of anticancer drug sensitivity. Nature 2012; 483: 603-607. 32 Hajjari M, Salavaty A. HOTAIR: An oncogenic long non-coding RNA in different cancers. Cancer Biology & Medicine 2015; 12: 1-9. 33 Delattre O, Zucman J, Plougastel B, Desmaze C, Melot T, Peter M et al. Gene fusion with an ETS DNA-binding domain caused by chromosome translocation in human tumours. Nature 1992; 359: 162-165. 34 May WA, Gishizky ML, Lessnick SL, Lunsford LB, Lewis BC, Delattre O et al. Ewing sarcoma 11;22 translocation produces a chimeric transcription factor that requires the DNA-binding domain encoded by FLI1 for transformation. Proceedings of the National Academy of Sciences of the United States of America 1993; 90: 5752-5756. 35 Hajjari M, Rahnama S. HOTAIR Long Non-coding RNA: Characterizing the locus features by the in silico approaches. Genomics & Informatics 2017; 15: 170177. 124 Figure A.1 HOTAIR expression in Ewing sarcoma. A- HOTAIR is expressed at higher levels in Ewing sarcoma cell lines compared to any other cancer cell line type in the Cancer Cell Line Encyclopedia. B- Genomically complex Ewing sarcoma tumors have higher HOTAIR expression compared to genomically simple tumors. C- The HOTAIR promoter region is unmethylated in complex Ewing sarcoma compared to the simple subtype. 125 Figure A.2 CRISPR knockout of HOTAIR in Ewing sarcoma cells. A- Schematic showing the HOTAIR gene and its exons (numbered 1-7) and the sites targeted by CRISPR. B- Levels of HOTAIR variants 1 and 3 expression by qPCR in single cell clones of A673 and ES8 cells. Each cell type includes wild-type (+/+; no CRISPR treatment) HOTAIR clones, heterozygous (+/-) HOTAIR clones, and knockout (-/-) clones. 126 Figure A.3 HOTAIR deficiency affects Ewing sarcoma cell growth. A- Viability of A673 and ES8 HOTAIR knockout (-/-), heterozygous (+/-), and wild-type (+/+) clones after 96h of growth in normal conditions. B- Number of colonies developed by HOTAIR -/-, +/-, and +/+ A673 and ES8 clones after 14 days of growth in soft agar. C- Proliferation rates of HOTAIR -/-, +/-, and +/+ A673 and ES8 clones. D- Rate of cell migration as measured by the percent of the area of scratch that is filled in over time. 127 Figure A.4 Chemosensitivity of HOTAIR-deficient Ewing sarcoma cells. Viability of HOTAIR -/-, +/-, and +/+ clones after 72 or 96 hours of A- doxorubicin, B- etoposide, and C- HCI-2509 treatment.