Identification of a Novel SYK/c-MYC/MALAT1 Signaling Pathway and Its Potential Therapeutic Value in Ewing Sarcoma

Ewing sarcoma is an aggressive soft tissue malignancy of children and adolescents, which is characterized by the chromosomal translocation leading EWS to fuse to FLI1 (1–4). Although about 70% of children with Ewing sarcoma can be cured by surgery and chemotherapy either with or without radiotherapy, only 30% of those with metastasis can be cured (5). Thus, new effective therapies are needed. Many pediatric solid tumors have activation of tyrosine kinases (TK), which play important roles in Ewing sarcoma biology (6, 7). For example, EWS–FLI1 fusion protein promotes the activities of TKs, including FAK, PDGFR, and IGF1R (8–11). Notably, targeting IGF1R by either small-molecule inhibitors or antibodies has enhanced patients' survival in several clinical trials (12).

Spleen tyrosine kinase (SYK) is a nonreceptor TK that is highly expressed in hematopoietic cells and regulates cellular adaptive immune responses (13). SYK also promotes cancer cell survival in leukemia and pediatric retinoblastoma (14). Small-molecule inhibitors of SYK (PRT062607 and GS-9973) have shown antineoplastic properties in these tumor types (15–17).

In this study, two unbiased high-throughput screens, a TK-focused siRNA library (18) and a small-molecule inhibitor library (19), were performed to identify signaling pathways which were valuable for therapeutic interventions. Through a series of functional investigations, we established a novel signaling pathway involving SYK/c-MYC/MALAT1 in the setting of Ewing sarcoma biology, and further showed its potential for therapeutic intervention for this pediatric malignancy.

The following reagents and antibodies were used: GS-9973 (MedKoo Sciences); siRNA pools targeting MALAT1 (Dharmacon); antibodies against SYK (#2712), p-SYK (#2710 for immublotting), p-AKT (#4060), MEK (#4694) and p-MEK (#9122), EZH2 (#5246; Cell Signaling Technology); p-SYK (SAB4503839 for IHC), and β-actin (A5316; Sigma). Other reagents included: Imprint RNA Immunoprecipitation (RIP) Kit (Sigma); Pierce Magnetic ChIP Kit (ThermoScientific); BioT transfection reagent (Bioland Scientific); anti-rabbit IgG and anti-mouse conjugated HRP antibodies (BD Biosciences); Dual-Luciferase Reporter Assay System (Promega). Small-molecule inhibitors were either purchased or generously provided by the sources outlined in Supplementary Table S1.

siRNA screens were performed in SKNMC, A673, TC32, TC71, SKES1, EW8, CADO-ES1, and EWS-502 cell lines. Briefly, 500 cells/well were seeded in 96-well plates and incubated with siRNA transfection mixtures for 96 hours. All siRNAs were from Dharmacon RNAi Technologies and the RaPID Assays (Sigma) and other siRNA experiments were performed as described previously (18). Small-molecule inhibitor screening was performed in SKNMC, A673, TC32, TC71, SKES1, EW8, TTC-446, CADO-ES1 cell lines, and the cell viability was calculated on the basis of an algorithm previously described (19, 20). Briefly, 200 cells/well were seeded in 384-well plates and incubated with indicated small-molecule inhibitors by seven serial of concentrations (three times of dilutions between dosages, ranging from 10 nmol/L to 10 μmol/L) for 72 hours. At the end point, MTS assay was used to measure the cell viability. IC₅₀ values were calculated from a third-order polynomial curve fitting to the data points by an in-house analytic framework as we have described recently (19).

Ewing sarcoma cell lines (SKNMC, A673, TC32, TC71, SKES1, EW8, TTC-446, EWS502, and CADO-ES1) were kindly provided by Dr. Kimberly Stegmaier (Harvard Medical School, Boston, MA) and Dr. Stephen L. Lessnick (University of Utah, Salt Lake City, UT), and were grown in DMEM (Corning) supplemented with 10% FBS, penicillin, and streptomycin. The identity of all cell lines was recently verified by short tandem repeat analysis (21). All cells have been tested with no mycoplasma contamination. Drug treatment and cell viability assays were performed as described previously (22).

Cell apoptosis was measured using propidium iodide (PI) and Annexin V (BD Biosciences) double staining and was assessed on a LSRII flow cytometer (BD Biosciences). Cell-cycle analysis was performed by PI staining (Sigma-Aldrich) for DNA content followed by flow cytometric analysis. All flow cytometry data were analyzed using FlowJo software (Tree Star).

Cells were seeded in 12-well plates and treated with the indicated doses of drugs. Two weeks after the treatment, colonies were fixed with 70% ethanol, stained with 0.1% methylene blue (Sigma), and positive colonies were counted using open CFU software (http://opencfu.sourceforge.net/).

Protein lysates were resolved by SDS-PAGE, transferred to polyvinylidene difluoride membrane (Merck Millipore), and followed by immunoblotting procedures as described previously (22). IHC was performed as described previously (23). Briefly, snap-frozen sections were fixed in 100% acetone at 4°C, and p-SYK antibody was applied using standard immunoperoxidase techniques in a Sequenza semiautomatic stainer (Thermo Scientific). p-SYK expression was scored using H-score method as described previously (21).

A total of 2 × 10⁶ cells for each chromatin immunoprecipitation (ChIP) reaction were fixed with formaldehyde and ChIP experiments were performed using MAGnify Chromatin Immunoprecipitation System (Invitrogen) according to the manufacturer's instructions. The precipitated DNA samples were quantified by qRT-PCR, and the data were expressed as the percentage of input DNA. Primers used for the ChIP-q-PCR are listed in Supplementary Table S2. For RNA-IP, lysate prepared from 2 × 10⁶ cells were immunoprecipitated using 5 μg of either normal rabbit IgG, or anti-EZH2 antibody according to the manufacturer's instructions. Immunoprecipitation of EZH2-recruited RNA was measured by qRT-PCR using GAPDH as a control gene.

The LentiCRISPRv2 virus vector was obtained from Addgene, and the guide RNAs were designed on the basis of the Optimized CRISPR Design service engine (http://crispr.mit.edu/). Vector subcloning was described previously (24). Guide RNA sequences and primers for vector subcloning are listed in Supplementary Table S2.

Lentiviral shSYKs and control vectors were generous gifts from Dr. Kimberly Stegmaier (25). Lentiviral shc-MYC and control vectors were obtained from Addgene (26). Cells were transduced with viral particles in the presence of 8 μg/mL polybrene for 16 hours followed by replacement of the lentivirus-containing media with fresh media. Two days after infection, puromycin (2 μg/mL) was added for 3 days to eliminate uninfected cells.

MALAT1 promoter regions (1010 bp, −700 to +310 bp) were amplified using DNA from TC71 cells as templates, and the PCR products were cloned into pGL3B luciferase vector (Promega). This vector was transfected into cells, luciferase activity was assessed 48 hours after transfection using the Dual-Luciferase Reporter Assay System according to the manufacturer's instructions (Promega), and the ratio of Firefly/Renilla luciferase activities (RLU) was calculated. Primers for the construction of the luciferase expressing vector (pGL3B-MALAT1) are listed in Supplementary Table S2.

Total RNA was extracted using RNeasy Isolation Kit (Qiagen). qRT-PCR was performed with standard procedures as described previously (21). Briefly, cDNA was generated using qScript cDNA Synthesis Kit (Quanta Biosciences) and qRT-PCR was performed on CFX96 qPCR System (Bio-Rad). Expression of each gene was normalized to GAPDH and quantified using 2^−Δ(Ct) method. Primers for qRT-PCR are listed in Supplementary Table S2.

cDNA microarray and data analysis were performed as described previously (21). Briefly, RNA was reverse transcribed and hybridized on Affymetrix GeneChip HGU133 plus 2. The whole transcriptome expression data were obtained using the robust multichip average method (https://www.bioconductor.org). Gene-set enrichment analysis (GSEA) was performed using GSEA v2.07 tool (http://www.broad.mit.edu/gsea/) with msigdb.v4.0 to identify significantly enriched gene sets. cDNA microarray data are available at the GEO repository website (GSE93677).

Animal experiments were approved by the Cedars-Sinai Institutional Animal Care and Use Committee (IACUC). Seven-week-old female athymic nude mice (Crl:NU(NCr)-Foxn1^nu) were obtained from Charles River Laboratories (San Diego, CA) and inoculated subcutaneously in both flanks with a suspension of TC71 cells (2.0 × 10⁶) in Matrigel. Five days after injection, when the tumor xenografts were noted to be growing, mice were randomly divided into two groups and orally treated with either vehicle [0.6% (w/v) aqueous Pluronic F-68, n = 8] or GS-9973 (20 mg/kg, n = 8), thrice weekly × 4 weeks. Tumor volumes were measured with calipers, and were calculated using the following formula: volume (mm³) = [width (mm)]² × length (mm)/2. Mice were sacrificed at the end of the fifth week from the date of cell injection, and the tumors were dissected and weighted.

c-MYC ChIP-seq uniform peaks on MALAT1 genomic region was analyzed using CistromeFinder system (http://cistrome.org/finder) and ENCODE project (27) on 10 human cells. H3K27ac histone mark was analyzed using ENCODE project. The results were visualized further by UCSC Genome Browser (https://genome.ucsc.edu/cgi-bin/hgGateway).

The mRNA expression level of MALAT1 from various types of primary cancer tissues and cell lines were examined by analyzing Expression Project for Oncology ExpO dataset (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE2109) and Cancer Cell Line Encyclopedia (CCLE, http://www.broadinstitute.org/ccle; ref. 28), respectively. Differences between two groups were analyzed using either paired or unpaired two-tailed Student t test. One-way ANOVA was used for comparisons among multiple groups (*, P < 0.05; **, P < 0.01; ***, P < 0.001). Overlaps of gene lists were identified using program VENN (http://bioinformatics.psb.ugent.be/webtools/Venn/). Statistical significance of overlapping was determined using χ² tests with one degree of freedom and Yates correction as previously described (29).

To screen unbiasedly for potential novel oncogenic kinases in Ewing sarcoma, we tested eight Ewing sarcoma cell lines by transfecting a kinase-specific siRNA library (set of four siRNAs targeting each gene) against 94 known kinases on a gene-per-gene basis followed by the measurement of cell viability. As shown in Fig. 1A, the top 20 genes ranked by median inhibition rate values (MV) included a number of known oncogenic kinases in Ewing sarcoma, such as ROR1, KIT, FGFR, PTK families (PTK-2 and 7), Src families (FGR and SRMS), ERBB4, and EGFR (9, 30–37), which strongly supported the methodology and effectiveness of our high-throughput screen. Importantly, we identified novel targets including SYK and related proteins (SYK and ZAP70), BTK, ABL1, FLT4, and MATK. Of interest, MV of siSYK ranked the second and significantly suppressed the growth of three of the Ewing sarcoma cell lines examined (Fig. 1A and B).

To complement the siRNA library screen, we performed an independent high-throughput approach testing a panel of 116 small-molecule inhibitors that are either FDA approved or in clinical trials. This compound library included inhibitors targeting well-established kinase pathways (e.g., RTK-MAPK, JAK, PI3K/AKT, PKC, IκK, AURK, and CDKs), as well as a variety of non-kinase prosurvival factors including HSP70/90 proteins, BCL2 and BET families, WNT/β-catenin, and sonic hedgehog pathways, and so on (Supplementary Table S1). The IC₅₀ was determined for each compound across eight Ewing sarcoma cell lines. The specific anti–Ewing sarcoma effects of the compounds were further determined by the comparison of the IC₅₀ values in Ewing sarcoma cell lines to those in 151 bone marrow aspirate samples tested using the same platform that we previously published (Fig. 1C; Supplementary Table S3; ref. 19). For those compounds newly included in this study, comparably low IC₅₀s, based on their previous reports in other cancers, were also listed (Supplementary Table S4; refs. 16, 38–41). As shown in Fig. 1C, the targets of top ranked chemicals included protein kinase C (PKC), AURK, PI3K, B-RAF/VEGFR, EGFR/JAK2, and SYK. As SYK stood out as a novel candidate in both screens, our further studies focused on this kinase in Ewing sarcoma. Immunoblotting assays showed that although total SYK protein was expressed relatively weakly in Ewing sarcoma, the kinase was prominently phosphorylated in Ewing sarcoma cells compared with leukemic cells which have been shown to have hyperactivation of SYK (ref. 42; Fig. 1E). In addition, IHC staining of p-SYK on 35 primary Ewing sarcoma samples revealed that 40% of samples showed strong phosphorylation of this kinase (Fig. 1D; Supplementary Fig. S2; and Supplementary Table S5).

To determine the biological significance of SYK in Ewing sarcoma, it was inhibited through a variety of genetic and chemical approaches. Consistent with the earlier screening results, knockdown of SYK by two independent shRNAs significantly decreased the proliferation of Ewing sarcoma cell lines (Fig. 2A). Either of the two selective SYK inhibitors (PRT062607 and GS-9973), dose-dependently inhibited the viability of Ewing sarcoma cells (mean IC₅₀s approximately 3.8 and 4.6 μmol/L, respectively; Fig. 2B). In addition, through blocking the phosphorylation of SYK (Fig. 2D), these inhibitors markedly suppressed AKT phosphorylation (Fig. 2F), clonogenic growth (Fig. 2C), and induced massive cell death (Fig. 2E) as well as cell-cycle arrest (Supplementary Fig. S3). Importantly, ectopic expression of the myristoylated (constitutively active) form of SYK promoted cell proliferation (Fig. 2G and H), enhanced phosphorylation of both AKT and MEK (Fig. 2I, left), and completely or partly rescued the cytotoxicity triggered by either shRNA knockdown or chemical inhibition of SYK (Fig. 2I–K). To confirm further the ontarget effect of these loss-of-function experiments, CRISPR/Cas9–mediated gene editing was used to disrupt endogenous SYK expression, which again significantly decreased Ewing sarcoma cell proliferation (Fig. 2L and M), and was rescued by the restoration of active SYK (Fig. 2M).

We next examined the anti–Ewing sarcoma property of the SYK inhibitor (GS-9973) in a murine xenograft model. As shown in Fig. 3A and B, GS-9973 treatment significantly suppressed tumor growth. GS-9973 also potently decreased the levels of p-SYK, p-AKT, and p-MEK in the tumor cells (Fig. 3C), in accordance with our observations obtained in vitro.

To understand better the role of SYK in Ewing sarcoma, we performed whole transcriptome profiling of TC71 cells after SYK inhibition through either shRNA knockdown or exposure to GS-9973 (2 μmol/L, 24 hours), and compared the results to the control cells. SYK inhibition globally altered gene transcription; and GSEA analysis revealed that genes associated with genomic-unstable Ewing sarcoma phenotype (43), cell cycle (44, 45), and NFκB pathway (46) were significantly enriched in SYK knockdown cells (Supplementary Fig. S1A–S1C), and genes associated with cell metastasis (47) were significantly enriched in cells treated with the SYK inhibitor (Supplementary Fig. S1D). Venn diagram analysis showed that 426 and 1,126 genes were codownregulated and coupregulated, respectively, by both shRNA- and GS-9973–mediated approaches (P < 0.00001, χ² tests with one degree of freedom and Yates correction; Fig. 4A; Supplementary Table S6, random qRT-PCR validation results were shown in Fig. 4C and D). Interestingly, we noticed a long noncoding RNA, known as Metastasis Associated Lung Adenocarcinoma Transcript 1 (MALAT1), among the top-ranked downregulated genes (Fig. 4B). MALAT1 has been implicated in multiple physiologic processes, and it is highly expressed in many types of cancers (48). Analysis of a Pan-cancer expression microarray dataset (49) showed that MALAT1 expression was ranked third highest in Ewing sarcoma among all 18 different types of sarcomas (Fig. 4E; Supplementary Table S7; ref. 49). In addition, CCLE dataset showed that MALAT1 expression was the eighth highest in Ewing sarcoma among 37 different types of cancer cell lines (Fig. 4F; Supplementary Table S8). These results together suggest that MALAT1 is highly expressed both in Ewing sarcoma tumor tissues and cell lines. To validate that MALAT1 expression is dependent on SYK-activated signaling, we inhibited SYK either by shRNA knockdown (Fig. 4G), CRISPR/Cas9 knockout (Fig. 4H), or small-molecule inhibitors (2 μmol/L, 24 hours; Fig. 4I), and confirmed the consistent and marked decrease of MALAT1 levels. In contrast, ectopic expression of active SYK enhanced the expression levels of MALAT1 (Fig. 4J).

We next examined the role of MALAT1 in Ewing sarcoma through siRNA-mediated knockdown, and noticed that it potently inhibited cell proliferation (Fig. 5A and B). Furthermore, silencing MALAT1 robustly induced cell apoptosis (Fig. 5C) and G₁ cell-cycle arrest (Fig. 5D) in Ewing sarcoma cells, with concomitant downregulation of cyclin D1 level and upregulation of p27^kip1 and p21^cip1 levels (Fig. 5E).

We next sought to determine the mechanisms underlying the upregulation of MALAT1 by signaling mediated by SYK. MALAT1 was recently reported to be transcriptionally activated by c-FOS in renal cell carcinoma (50). In addition to c-FOS, bioinformatics analysis predicted that 20 additional transcription factors, including c-MYC, had the potential to bind to the promoter region of MALAT1 (50). Interestingly, our microarray data showed inhibition of SYK significantly downregulated the expression of c-MYC (Supplementary Table S6). GSEA analysis further identified that c-MYC downstream targets (51) were significantly enriched in those genes altered by SYK inhibition (Supplementary Fig. S1E). Given that SYK activation was observed to induce expression of c-MYC in hematopoietic cell lines (52), we hypothesized that a novel signaling axis involving SYK/c-MYC/MALAT1 might be operative in Ewing sarcoma cells. To address this hypothesis, we analyzed public c-MYC ChIP-seq data using both CistromeFinder and ENCODE projects. Of note, two major c-MYC–binding peaks spanning MALAT1 promoter were identified across different cell types (Fig. 6A; Supplementary Table S9), supporting the positive regulation of c-MYC on MALAT1 transcription. In addition, strong H3K27ac signals (predictive of active enhancer and promoter) were detected to be surrounding MALAT1 promoter. Both c-MYC–binding peaks almost perfectly fitted into H3K27ac “dip” (Fig. 6A) strongly suggesting that c-MYC occupies the MALAT1 promoter and recruits cofactors and histone modifiers to activate MALAT1 transcription. To validate this observation experimentally, ChIP-qPCR and dual luciferase reporter assays were performed. A set of primer pairs was designed to tile across the promoter regions (−700 to +310 bp) of MALAT1 (Fig. 6B). ChIP-qPCR results revealed that c-MYC bound more avidly to the promoter of MALAT1 in Ewing sarcoma cells compared with either IgG or FLI1 antibody (recognizing EWS-FLI1) negative controls (Fig. 6C). In addition, the transcriptional activity of MALAT1 promoter was confirmed by the luciferase reporter assay (Fig. 6D and E). Importantly, ectopic expression of c-MYC significantly induced the promoter activity, whereas silencing of either c-MYC or SYK produced an opposite effect (Fig. 6D and E; Supplementary Fig. S4). In addition, the suppressive effect of SYK inhibition was rescued by ectopic expression of c-MYC (Fig. 6E). Consistent with this model, silencing of SYK downregulated the expression of both c-MYC and MALAT1 (Fig. 6F), and c-MYC knockdown significantly inhibited MALAT expression (Fig. 6G and H). These results indicate that SYK-mediated signaling enhanced the expression of c-MYC, which further transcriptionally activated MALAT1 through direct binding to its promoter.

Taken together, our results suggest that a novel oncogenic signaling involving SYK/c-MYC/MALAT1 contribute to the malignancy of Ewing sarcoma, which provide potential therapeutic targets to treat this malignancy (Fig. 7).

Although several potential molecular targets have been identified in Ewing sarcoma, their clinical trials aimed at these targets have yet to show success (53). Through high-throughput screens using both siRNA and small-molecule inhibitor libraries, we confirmed a number of previously reported Ewing sarcoma actionable candidates such as ERBB4, EGFR, ROR1, KIT, and FGFRs (9, 30–37). Most importantly, we identified SYK as an important progrowth kinase in Ewing sarcoma through unbiased integrative approaches.

Chronic lymphocytic leukemia (CLL) is strongly addicted to SYK-mediated prosurvival pathway and is considered as one of the most sensitive malignancies for SYK inhibition. An in vitro study of PRT062607 shows that 64% (27 of 42 samples) of CLL primary cells had IC₅₀s higher than 3 μmol/L (16). Another SYK inhibitor, GS-9973, is currently under assessment in a phase II CLL clinical trial (NCT01799889). In vitro viability study of 14 primary CLL samples showed that GS-9973 had a mean IC₅₀ of 3.7 μmol/L (17, 54). Our results demonstrated a similar sensitivity to these two SYK inhibitors compared with CLL cells. We further noted that forced activation of SYK rescued the anti–Ewing sarcoma effects of the SYK inhibitors. Taken together, our data indicate that targeting SYK by selective small-molecule inhibitors holds the potential for treating Ewing sarcoma.

Our immunoblotting result noted that albeit total SYK protein levels were low in Ewing sarcoma cells, the kinase was prominently phosphorylated in these cells compared with leukemic lines which are known to have hyperactivation of SYK (ref. 42; Fig. 1E). These results were further supported by IHC staining of p-SYK on 35 primary Ewing sarcoma tissues, which revealed that 40% of samples had strong phosphorylation of SYK (Fig. 1D; Supplementary Table S5). How does SYK become hyperactivated in Ewing sarcoma cells? One possible mechanism is through SRC family kinase-dependent activation (13). Attas and colleagues showed that macrophages derived from hck^−/−fgr^−/−lyn^−/− mice, which lose Src-family kinases in myeloid leukocytes, markedly reduced SYK activation (55). Interestingly, recent studies reported that SRC kinases are also highly activated in Ewing sarcoma cells, and inhibition of SRC efficiently decreased Ewing sarcoma cell growth (9, 36), which was confirmed by our high-throughput approaches (Fig. 1A and C).

Activated SYK results in phosphorylation of tyrosine residues of downstream targets including PKC, ERK, AKT, and NF-κB in hematopoietic cells (14). In agreement with these observations, our high-throughput screens found that Ewing sarcoma cells were also dependent on PKC, ERK, and AKT for survival. Moreover, cDNA microarray analysis suggested that NF-κB pathway was suppressed by SYK inhibition in Ewing sarcoma cells. In addition, our immunoblotting assays showed that SYK inhibitors decreased the phosphorylation of both AKT and ERK. Together, these results suggest that SYK activates multiple important prosurvival pathways in Ewing sarcoma, and future studies will determine the biological significance underlying these signals activated by SYK.

Our work demonstrated that a LncRNA, MALAT1, is deregulated by SYK-mediated signaling. MALAT1 is a known oncoRNA overexpressed in many different cancer types (48, 56, 57). We observed that MALAT1 knockdown induced G₁ arrest in Ewing sarcoma cells (Fig. 5E). In agreement with our findings, depletion of MALAT1 caused G₁ arrest with the upregulation of p53, p27^kip1, and p21^cip1 in human fibroblasts (58). Most recently, Malakar and colleagues showed that MALAT1 transcriptionally activated cyclin D1 in hepatocellular carcinoma cells (59). These data suggest that MALAT1 controls G₁ cell-cycle transition in multiple cell types by modulating the expression of important cell-cycle regulators.

Although the mechanism underlying cancer-specific overexpression of MALAT1 is poorly understood, previous bioinformatic analysis predicted that a number transcription factors, including c-FOS and c-MYC, could interact with the promoter region of MALAT1 (50). Importantly, our analysis of publicly available c-MYC ChIP-seq data identified strong c-MYC occupancy on MALAT1 promoter region. ChIP-qPCR and luciferase reporter assays confirmed that c-MYC directly interacted with this region, and enhanced the transcriptional activity of this lncRNA. Interestingly, c-MYC is a SYK-regulated gene in hematopoietic cells (52, 60), and our results also showed that silencing of SYK downregulated c-MYC expression in Ewing sarcoma cells. Moreover, luciferase assay demonstrated that forced expression of c-MYC promoted the transcription of MALAT1 regardless of SYK activity. Together, these results support the model that SYK elevates the transcription of MALAT1 through regulating c-MYC.

EZH2 is a histone methyl transferase subunit of the Polycomb repressor complex, which is either recurrently mutated or highly expressed in many cancers (61). Recent studies have shown that EZH2 helps EWS-FLI1 to drive tumor growth and metastasis in Ewing sarcoma cells (62). MALAT1 interacts with EZH2, and the oncogenic activities of MALAT1 were inhibited by EZH2 depletion in renal cancer (50). Therefore, we hypothesized that MALAT1 and EZH2 might also cooperate in Ewing sarcoma cells. Importantly, RNA-IP assay revealed that MALAT1 interacted directly with EZH2 in both TC71 and TC32 cells (Supplementary Fig. S5). Although further investigations are needed to elucidate the biological significance of this interaction, our findings highlight a strong oncogenic role of MALAT1 in Ewing sarcoma biology.

In summary, through integrative molecular and cellular approaches, the current study elucidates a novel oncogenic signaling axis involving SYK, c-MYC, and MALAT1 in the setting of Ewing sarcoma biology (Fig. 7). Importantly, our results strongly suggest that targeting SYK-mediated signaling may serve as a promising therapeutic strategy for the treatment of Ewing sarcoma patients.

No potential conflicts of interest were disclosed.

Conception and design: H. Sun, H.J. Lim, S. Gery, H.P. Koeffler

Development of methodology: H. Sun, J.W. Said

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): H. Sun, D.-C. Lin, Q. Cao, B. Pang, V.K.M. Lee, J.W. Said, M. Chow, J.W. Tyner

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): H. Sun, B. Pang, J.W. Said, A. Mayakonda, J.W. Tyner, H.P. Koeffler

Writing, review, and/or revision of the manuscript: H. Sun, D.-C. Lin, D.D. Gae, D.-C. Lin, C. Forscher, J.W. Tyner

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): D.-C. Lin, S. Gery, M. Chow

Study supervision: D.-C. Lin, H.P. Koeffler

Minor edits in figure clarity: D.D. Gae

We thank Dr. Kimberly Stegmaier and Dr. Stephen L. Lessnick for their generous help with the reagents and cell lines. This research was supported by the Alan B. Slifka Foundation, the National Research Foundation Singapore and the Singapore Ministry of Education under the Research Centres of Excellence initiative, and the Singapore Ministry of Health's National Medical Research Council under its Singapore Translational Research (STaR) Investigator Award.

D.-C. Lin was supported by Donna and Jesse Garber Awards for Cancer Research and the National Center for Advancing Translational Sciences UCLA CTSI Grant UL1TR000124. This research is also partially supported by the RNA Biology Center, Cancer Science Institute of Singapore, NUS (MOE2014-T3-1-006).

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