6XSSOHPHQWDU\)LJXUH CCLE B FLI1 ChIP-Seq A 3 FRPPRQ 40 (ZLQJ6SHFLILF 2 A673 6.10&  477 Q   690 30    0 UHSHDWV Q  Q  A 20

í Q 

57226 RI**$ 

ORJ)3.0LQRWKHUFDQFHUV í

í A673 6.10& 06& RYHUODS í 0  23 06& ORJ)3.0LQ(ZLQJ6DUFRPD SHDNV SHDNV SHDNV SHDNV

& ./) 67($3 (:6)/,ERXQG 3'(% ! **$AUHSHDWV 3335$ ABI3 '1$-& '863 13<5 .&1( 5%0 DOO**$AUHSHDWUHJLRQV 1.; NEDURXQGWKH766 9$9 ;* .&1$% 4 .&1$ 51) 0<20 3 355/ 355T4 &'$ $'5% 2 '&'& 711, 8*7$ 3+263+2  )(=)

*1*7 ORJ QXPEHURIUHSHDWV A571 0 13<5 /2;+' /,3, 5$; 0 40 60 20        GLVWDQFHIURP766 NE D TCGA 35$0(

5 code cancer normal tumor LAML Acute Myeloid Leukemia 0 151

ORJ )3.0 ACC AdrenocoƌƟcal carcinoma 0 79 0 BLCA Bladder Urothelial Carcinoma 19 411 LGG Brain Lower Grade Glioma 0 529 COAD Colon adenocarcinoma 41 464 5%0 ESCA Esophageal carcinoma 11 162 GBM Glioblastoma muůƟforme 5 168 5 KICH Kidney Chromophobe 24 65 KIRP Kidney renal papillary cell carcinoma 32 289 LIHC Liver hepatocellular carcinoma 50 374

ORJ )3.0 LUAD Lung adenocarcinoma 59 526 0 LUSC Lung squamous cell carcinoma 49 501 DLBC Lymphoid Neoplasm Diīuse Large B-cell Lym 042 /,3, MESO Mesothelioma 0 1 OV Ovarian serous cystadenocarcinoma 0 379 PAAD PancreĂƟc adenocarcinoma 4 178 5 PCPG Pheochromocytoma and Paraganglioma 3 150 PRAD Prostate adenocarcinoma 52 499 READ Rectum adenocarcinoma 1 6 ORJ )3.0 SARC Sarcoma 0 35 0 SKCM Skin Cutaneous Melanoma 1 471 STAD Stomach adenocarcinoma 32 375 /2;+' WXPRU TGCT dĞƐƟĐƵůĂƌ'ĞƌŵĞůůdƵŵŽƌƐ 0156 THCA Thyroid carcinoma 58 510 QRUPDO UCS Uterine Carcinosarcoma 0 56 5 ORJ )3.0 0 / / 29 $&& 8&6 /** *%0 /,+& .,53 .,5& .,&+ %/&$ /$0 67$' LUAD '/%& /86& 7*&7 7+&$ 3$$' (6&$ &(6& 35$' 6$5& &+2 5($' %5&$ 8&(& 3&3* 6.&0 &2$' 0(62 Supplementary Figure 1 (contd) GTEx KLF15 FPKM F PDE1B E STEAP2 ABI3 1.0 MET500 + EwS DNAJC12 PPP1R1A 60 NPY1R VAV1 TNNI3 DUSP26 0.8 50 KLF15 KCNAB3 PRR5L RBM11 KCNE3 ABI3 MYOM2 40 PRR5L NKX2-2 XG RNF182 30 DUSP26 CD79A KCNA2 PRRT4 NPY5R PDE1B TNNI3 20 STEAP2 ARTN RAX PPP1R1A CD79A PRRT4 VAV1 PHOSPHO1 DNAJC12 ADRB3 10 GNGT2 PHOSPHO1 RBM11 DCDC2 MYOM2 ARTN GNGT2

% non-EWS samples NKX2-2 LIPI RNF182 DCDC2 FEZF1 UGT3A2 0 KCNAB3 ADRB3 LOXHD1 0.0 0.2 0.4 0.6 UGT3A2 LIPI with > EWS median expression NPY5R FEZF1 XG KCNA2 NPY1R LOXHD1 Liver Lung Testis Ovary

í Uterus Vagina Spleen Thyroid Bladder Pituitary Prostate Stomach 0 100 200 Pancreas Whole Blood Nerve - Tibial Nerve - Tibial Artery - Artery - Aorta Artery - Brain - Cortex Adrenal Gland Fallopian Tube Kidney - Cortex Colon - Sigmoid Muscle - Skeletal Artery - Coronary Median expression of Amygdala Brain - Cervix - Ectocervix Brain - Cerebellum Colon - Transverse Colon - Cervix - Endocervix Heart - Left Ventricle Minor Salivary Gland Brain - Hippocampus Esophagus - Mucosa ESS32 genes (FPKM) Brain - Hypothalamus Brain - Substantia nigra Esophagus - Muscularis Adipose - Subcutaneous Heart - Atrial Appendage Breast - Mammary Tissue Breast - Mammary Brain - Frontal Cortex (BA9) Adipose - Visceral (Omentum) Adipose - Visceral Brain - Cerebellar Hemisphere Brain - Caudate (basal ganglia) Small Intestine - Terminal Ileum Terminal Small Intestine - Skin - Sun Exposed (Lower leg) Brain - Putamen (basal ganglia) Brain - Spinal cord (cervical c-1) Skin - Not Sun Exposed (Suprapubic) Brain - Anterior cingulate cortex (BA24) Brain - Esophagus - Gastroesophageal Junction Brain - Nucleus accumbens (basal ganglia) G H p < 0.01 MET500 CCLE + EwS GTEX V6p (n=11401) (n=507) (n=980)

#samples 11 496 9 971 172 103 393 227 117 97 320 104 350 214 114 113 108 323 218 32 194 357 84 6 94 97 125 105 133 88 250 72 193 153 247 149 171 57 430 96 304 196 224 286 332 11 96 83 6 106 145 119 63 71 5 10

LOXHD1 0 10 LIPI 0 10 LOXHD1 (log2 _ z-score) RBM11 0 10 CD99 0 Primary Metastasis ( N = 14) Recurrence 10 ( N = 64) ( N = 10)

7 ( N = 18)

PAX Normal Skeletal Muscle 0 10 I qRT-PCR NKX2-2 0 10 1e4 *

1e3 BCL11B 0 10 1e2 GLG1

0 expression mRNA 10

Lung Liver Testis Ovary Spleen Vagina Uterus Thyroid Bladder Pituitary Prostate Stomach Pancreas

pan cancer pan cancer Whole Blood 1 Brain - Cortex Nerve - Tibial Artery - Aorta Artery - Tibial Fallopian Tube Adrenal Gland Kidney - Cortex Colon - Sigmoid LOXHD1 Muscle - Skeletal Ewings Sarcoma Artery - Coronary Brain - Amygdala Ewings Sarcoma Brain - Cerebellum Colon - Transverse Cervix - Ectocervix Cervix - Endocervix Heart - Left Ventricle Brain - Hippocampus Minor Salivary Gland Esophagus - Mucosa Brain - Hypothalamus Esophagus - Muscularis Adipose - Subcutaneous Brain - Substantia nigra Heart - Atrial Appendage Breast - Mammary Tissue 0.1 Brain - Frontal Cortex (BA9) Adipose - Visceral (Omentum) Brain - Cerebellar Hemisphere non-EwS Brain - Caudate (basal ganglia) Skin - Sun Exposed (Lower leg) EwS HeLA RD-ES Brain - Putamen (basal ganglia) Small Intestine - Terminal Ileum Brain - Spinal cord (cervical c-1) (n=4) Skin - Not Sun Exposed (Suprapubic) (n=12)

Brain - Anterior cingulate cortex (BA24) Esophagus - Gastroesophageal Junction

Brain - Nucleus accumbens (basal ganglia) Supplementary Figure 1: Integrative analysis of ChIP-seq and CCLE, MET500, TCGA and GTEx transcriptomic datasets. (A) Analysis of CCLE dataset; Expression of the 516 genes (step 1 of Fig. 1a) in EwS vs other cell lines. The 89 EwS specific genes, seen in the lower-right quadrant and encircled, are expressed only in EwS cell lines and show <1 FPKM expression in all others. (B) Analysis of FLI1-ChIP-seq data; (left) Overlap analysis between FLI1 enriched ChIP-Seq peaks in A673 (GSM1517562), SK-NM-C (GSM1517537), and EWSR1-FLI1 overexpressed MSCs (GSM2472088, GSM2472102, GSM2472108) yields 1473 conserved FLI1-bound regions, (right) distribution of the number of GGAA microsatellite repeats contained within A673-specific, SK-N-MC-specific, MSC-specific and the 1473 overlapping regions shows pronounced enrichment of the GGAA microsatellites in the overlapping regions. (C) Colormap shows the locations of GGAA microsatellite repeat regions േͳͲͲ kb around the TSS of the ESS32 genes. Symbol overlays represent regions with >= 5 GGAA repeats (blue circles) and contain a FLI1 ChIP-seq enrichment peak (red diamonds). (D) Analysis of TCGA dataset; Boxplot showing nearly zero expression of LOXHD1 and LIPI in all TCGA cancer-subtypes in tumor and normal samples. The expression of RBM11 located in LIPI locus is shown. . Expression of PRAME is shown as an example of gene that is expressed across all cancer-subtypes. Abbreviation of codes for cancer- subtypes and the corresponding numbers of normal and tumor samples are shown alongside. (E) Analysis of MET500 RNA-seq dataset; Plot showing the median expression of the ESS32 genes (step 3 in Fig. 1a) in EwS against the percentage of non-Ewing samples that show expression higher than the median value in EwS. The dotted line marks the cutoff of 1%. (F) Analysis of GTEx datasets; Heatmap showing the expression of the ESS32 genes showing null expression of LOXHD1, LIPI in all tissues except testis. (G) Box plots comparing the expressions of indicated targets across MET500 + EwS (n=507), CCLE (n= 980), and GTEX (n=11401) samples. (H) Box plots showing LOXHD1 expression in the indicated samples. Publicly available Affymetrix dataset GSE34620 comprising 117 EwS samples was analyzed and Log2 z-score is plotted. p <0.01 One-way Anova between the groups. (I) qRT-PCR analysis of LOXHD1 mRNA expression in an independent cohort of EwS tumor tissues, non-EwS tissues (testis is indicated with orange border), HeLa and RD-ES cells were used as negative and positive control, respectively. * p < 0.001, by two-tailed Student’s t test.

Supplementary Figure 2 A B coiled-coil prediction

1.0

0.8

0.6

0.4 t14 score

0.2

0.0 0 200 400 600 800 1000 1200 1400 1600 1800 2000 Amino acid C LOXHD1- NLS coiled coil domain NLS coiled coil HA

anti-HA DAPI Merge

10 μm Supplementary Figure 2: Domain prediction of LOXHD1 protein, and the effect of LOXHD1 overexpression in Hela cells. (A) Coiled-coil structure prediction by COILS (https://embnet.vital- it.ch/software/COILS_form.html), y-axis shows the probability of a 14 amino acid (aa) coiled-coil structure and the x-axis shows the amino acids position of LOXHD1 protein. The predicted coil- coil structure is seen as the peak between aa 653 to aa 677. (B) Nuclear localization signals (NLS) predicted by cNLS Mapper (http://nls-mapper.iab.keio.ac.jp/cgi-bin/NLS_Mapper_form.cgi), NLS with a score larger than 8 is exclusively localized to the nucleus. (C) Functional validation of NLS. Immunofluorescence staining with HA antibody showing nuclear signal in HEK293T cells transfected with plasmid containing HA-tagged NLS-coiled-coil domain of LOXHD1.

Supplementary Figure 3

A B shNT shFLI1 shERG

1.2 1.2 GSE31215 GSE94277 1 1 8 6 0.8 0.8 6

5 expression 0.6 4 expression 0.6 FLI1

4 2 0.4 ERG 0.4

0 Fold Fold LOXHD1 expression (a.u) 0.2 0.2 1 LOXHD1 expression (FPKM) 1 EV EV FLI1 0 0 EWS-FLI EWS-FLI MSC MSC RD-ES CHLA-10 SK-N-MC CADO-ES

5 kb C hg19, chr21:15,577,038-15,592,092

[0-60] FLI1

MSC [0-25] H3K27ac

[0-60] FLI1

[0-25] H3K27ac MSC + EWS-FLI1

FLI1

H3K27ac SK-N-MC

H3K4me3

ESS32 enhancer layered H3K27ac from encode

--(11 GGAA)-- LIPI RBM11 Supplementary Figure 3: LOXHD1 expression in EWSR1-FLI1 overexpressing MSCs and FLI knockdown EwS cell lines. (A) (left) Comparison of LOXHD1 expression profiled using Affymetrix Human Genome U133 Plus 2.0 Array (GSE31215) in MSCs expressing EWSR1-FLI1 (n=4), (right) Comparison of LOXHD1 FPKM expression in MSCs expressing EWSR1-FLI1 (n=4) and wild type FLI1 (n=4), profiled using RNA-seq (GSE94277). (B) Bar graph showing qRT- PCR results of FLI1 (left) and ERG (right) expression with shRNA knockdown of EWSR1-FLI1 and EWSR1-ERG respectively in the indicated EwS cells. (C) Genome browser view showing the de novo enhancer assembly in the vicinity of LIPI and RBM11 locus in EWSR1-FLI1 overexpressing MSCs and SK-N-MC cells.

Supplementary Figure4 $

GGAA deletion 172 bp

Control: 539 bp Deletion: 367 bp

% Neg Ctrl KO clones Neg Ctrl KO clones RDES SKNMC

WT clone5clone6 clone4 clone8 WT clone8clone18 clone12clone16

500bp 500bp

& RDES

clone4 WT gRNA_1 clone4 WT gRNA_2

clone4 WT

SKNMC

clone16 WT gRNA_1

clone16 WT gRNA_2 clone16 WT Supplementary Figure 4: Verifications of the LOXHD1 enhancer knockout single clones. (A) PCR and sequencing strategy for GGAA microsatellite deletion. Primers designed outside of the two gRNAs flanking the GGAA microsatellite; wild-type allele predicted to have 539bp amplicon while the deleted allele is predicted to have 367bp amplicon. (B) DNA gel image of PCR products with the genomic DNA of empty cas9 control cells and the enhancer targeting sgRNA CRISPR- cas9 virus infected cells. SK-N-MC and RD-ES cells were transduced with empty gRNA cas9 virus or cas9 with two gRNAs flanking the GGAA microsatellite, three days post viral transduction cells were subjected for puromycin selection for three days and harvested for genomic DNA for PCR. Arrow indicates the allele with 172bp deletion. (C) Sanger sequencing result of the enhancer knockout RD-ES and SK-N-MC single cell derived clones. Red region shows the deletion between two guide RNAs.

A Supplemetnary Figure5 all geneswithp<0.05 eKD2 RD-ES 7 1034 273 C RD-ES SK-N-MC E

eKD-2 eKD-1 Ctrl 96 KNM RD-ES SK-N-MC Ctrl eKD2 SK-N-MC NME1-NME2 eKO CDC42EP1 SLC25A36 KLHDC7B RPS6KA2 ALDH1L2 NCAPD2 LOXHD1 AMOTL2 CCDC86 CYP2R1 OLFML3 SIPA1L1 LRRC49 KCTD15 MALAT1 ACAD11 CXCL10 ZBTB7B SLC2A1 EEF1A2 ZNF334 EWSR1 PROM1 SRRM2 NCAM1 HOXC6 PNMA2 DOCK6 MAP1B CCNB1 NTNG1 SCN4A CSRP1 LMNB1 ESRP1 ANXA6 NOP56 NUP62 NPTX2 BEST1 SLFN5 ZNF10 DZIP3 TRIB3 HDGF OSTN ASNS MYLK SARS AKNA CNN2 SRRT DHX9 BTG1 PAG1 ASS1 PTX3 H1F0 ATF5 SCD FYN HK2

Relative 590 nm aborbance RD-ES _Ctrl_Rep.1 0.2 0.4 0.6 0.8 1.0 1.2 0 í RD-ES _Ctrl_Rep.2 Ctrl KNM RD-ES SK-N-MC eKD-1 RD-ES _eKD2_Rep.2 Invasion Index í * 0.2 0.4 0.6 0.8 1.0 1.2 1.4 RD-ES _eKD2_Rep.2 0 eKD-2 zscore cas9 ctrl * 0.0 SK-N-MC _Ctrl_Rep.1 SK-N-MC Ctrl eKO eKD-1 0.6 * SK-N-MC _Ctrl_Rep.2 **

eKD-2 * cas9 ctrl 1.2 SK-N-MC _eKD2_Rep.2

SK-N-MC _eKD2_Rep.2 RD-ES

eKO ** D

Time (hr) Wound Healing Rate (%) 100 20 40 60 80 0 B 02448 024 Scratch space DE-tlRD-ES-eKD1/2 RD-ES-Ctrl

Enrichment Score (ES)

Enrichment Score (ES) í í 100 eKD2 í í í eKD2 0.0 0.0 20 40 60 80 0 Healed space ESS32 inSK-NM-C ESS32 inRD-ES *** 48 NES=-2.06 FDR q<0.01 NES=-1.87 FDR q~0 SKNMC RDES Ctrl Ctrl Supplementary Figure 5: LOXHD1 loss impairs EwS signature and tumor phenotype in vitro. (A) Venn diagram and heatmap showing differentially expressed genes in LOXHD1 enhancer knockdown (eKD2) cells. (B) RNA-seq followed by GSEA showing negative enrichment of the ESS32 signature in LOXHD1 eKD2 RD-ES and SK-N-MC cells. LOXHD1 knockdown impairs colony formation ability in EwS cells. (C) left Representative images of a colony formation assay performed by seeding 5000 control and eKD cells in each well of 6-well plates, each group is done with triplicates. right, bar graph showing quantifications of the crystal violet staining of the colonies. LOXHD1 knockdown slows down cell migration in EwS cells. (D) Bar graph of a wound healing assay presented by quantifications of the scratch widths in RD-ES control and eKD cells. LOXHD1 knockdown reduces cell invasion in EwS cells. (E) left, representative images of a Boyden chamber invasion assay with SK-N-MC and RD-ES control and eKO single cell clones. right, bar graph showing quantifications of the invaded cells.

Supplementary Figure 6

A SK-N-MC RD-ES B Hallmark Hypoxia signature 1.2 **** **** 0.75 1.0 0.8 0.50 0.6 0.25 Hallmark Hypoxia NES=2.97 0.4 FDR q~0 Invasion Index 0.2 Enrichment Score (ES) 0.00 0 CTRL eKD-1eKD-2CTRL eKD-1eKD-2 0 5000 10000 15000 Hypoxia Hypoxia Normaxia

C Gene Ontology for 204 hypoxia upregulated genes

GO:BP

response to oxygen levels GO:0070482 response to hypoxia GO:0001666 KEGG response to decreased oxygen levels GO:0036293 HIF-1 signaling KEGG:04066 pyruvate metabolic process GO:0006090 pathway 01234567 monosaccharide metabolic process GO:0005996 -log10p

cellular response to hypoxia GO:0071456

monocarboxylic acid metabolic process WP GO:0032787 Photodynamic therapy-induced WP:WP3614 cellular response to decreased oxygen levels GO:0036294 HIF-1 survival signaling Hereditary leiomyomatosis and WP:WP4206 carboxylic acid metabolic process GO:0019752 renal cell carcinoma pathway Glycolysis and Gluconeogenesis WP:WP534 transport GO:0006810

establishment of localization GO:0051234 Cori Cycle WP:WP1946

oxoacid metabolic process GO:0043436 TP53 Network WP:WP1742

response to abiotic stimulus GO:0009628 01234567

012345 -log10p -log10p

D SK-N-MC E 80 SK-N-MC 60 4h 8h 16h

FPKM DFO 10 ȝM CTRL eKD1 eKD2 CTRL eKD1 eKD2 CTRL eKD1 eKD2 40

HIF1A HIF1Į 20 ȕ-actin 0

Ctrl eKD1 eKD2 Hypoxia Supplementary Figure 6: LOXHD1 knockdown diminishes hypoxic responses in EwS cells. Hypoxia enhances the effects of LOXHD1 knockdown on EwS cell invasion. (A) Bar graph showing quantifications of invasion index of the hypoxia samples in the experiment of Fig. 6a. LOXHD1 proficient SK-N-MC cells present strong hypoxia response. (B) RNA-seq followed by GSEA showing a positive enrichment of Hallmark Hypoxia signature in the hypoxic treated control SK-N-MC cells compared to the normoxia sample. (C) Gene Ontology for the 204 genes induces by hypoxia showing hypoxic responses and HIF-1signaling among the top of the lists. LOXHD1 knockdown in EwS cells does not affect HIF1A transcription. (D) Bar graph of the FPKM of HIF1A in control and two eKD SK-N-MC cells under hypoxia. (E) Immunoblot of HIF1D with DFO treatment at 10 μM for 4, 8 and 12h in control and two eKD SK-N-MC cells. E actin was used as loading control.

B Supplementary Figure7 A

Necrosis (%) Control 25 50 75 0

Control *

eKD2

eKD2

Mean number of mitosis per HPF 10 20 0

Control

eKD2 *** Supplementary Figure 7: LOXHD1 knockdown affects EwS growth in vivo. (A) Bar graph showing the percentage of necrotic regions in the control and LOXHD1 enhancer knockdown SK- N-MC xenograft. (B) Representative images showing mitotic nuclei in the control and LOHXD1 enhancer knockdown xenograft. The bar graph show average mitotic nuclei of all the tumors analyzed.

Materials and method: Cell culture Human EwS cell lines RD-ES (obtained from CLS), SK-N-MC, CHLA-10, CADO-ES, prostate cancer cell lines LNCaP, 22RV1, and osteosarcoma U2OS cell lines (obtained from ATCC) were maintained in RPMI-1640 media (Gibco, 11875093) supplemented with 10% FBS (HYC, SH30910.03) and 1% penicillin-streptomycin (Invitrogen, 15140122). All cells were grown at

37°C in a 5% CO2 incubator. For the hypoxia experiment, the cells were incubated in hypoxia chamber of 1% of O2, 5% CO2 at 37°C for the indicated time. All cell lines were tested negative for mycoplasma contamination and authenticated by STR profiling.

Plasmids To generate pcDNA3-LOXHD1-Myc˖Full Loxhd1 CDS was cloned from P7 mouse organ of Corti using the following primers: NG-138 and NG-64, then re-amplified using primers containing XhoI and EcorI sites respectively, and cloned into pcDNA3. Myc tag was added by using NEBuilder Hifi 2X mixture with primers: PW-83 and PW-84. (Table S2): To generate NLS-coiled- coil-HA constructs, we amplified the sequence of NLS-Coiled-coil from adult mouse testis cDNA, then ligated to pcDNA3.1(+) vector. NheI and NotI sites were used for the construction. HA tag, NheI and NotI sites were added by PCR primers: NG-191-NheI-CC-S and NG-192-NotI-CC-AS (Table S2). pCDNA-HA-HIF1D was a gift from Dr. Frank S. Lee at UPENN, PA.

CRISPR-cas9 mediated LOXHD1 enhancer knockout and enhancer silencing sgRNAs targeting adjacent to the LOXHD1 GGAA microsatellites were designed with the following website: https://portals.broadinstitute.org/gpp/public/analysis-tools/sgrna-design. For the enhancer region knockout, we inserted two pairs of sgRNAs into lentiCRISPR v2 (Plasmid #52961) backbone. For LOXHD1 repression, two independent sgRNAs were cloned into a lentiviral backbone expressing sgRNA and dCas9-KRAB (Plasmid #71236). Lentivirus was packaged using 2nd generation lentiviral packaging systems using the following protocol. 1x106 HEK-293HT cells were seeded in 10cm plates. Next day lentiCRISPR or lenti-dCas9-KRAB plasmids (4 μg) were co-transfected with pVSVg (1 μg) and pSPAX2 (2 μg) using 7 μl Lipofectamine 2000. Media was collected after 48h and 72h of transfection, centrifuged and passed through 0.45 μm filters to clear of any live cells or debris. Virus were then concentrated by 10X with LentiX concentrator (Takara, 631232) aliquoted and stored at -80°C. RD-ES and SK-N-MC cells were infected at a MOI of 10. We selected for positive cells with puromycin at 1μg/ml for three days to obtain stably knock-out/down cells. Cells within 5 passages after puromycin selection were used for all the experiments. shRNA knockdown shRNAs against FLI1, and ERG (Penn Core Facilities) were packaged using second-generation lentiviral packaging systems as described above. CHLA-10, RD-ES and SK-N-MC were transduced with shFLI1 lentivirus and CADO-ES cells were transduced with shERG lentivirus. Four days post infection, cells were harvested for qRT-PCR analysis.

RNA extraction and quantitative RT-PCR Total RNA was isolated from cells using the miRNeasy kit (QIAGEN, 217004) and cDNA was synthesized from 1,000 ng total RNA using SuperScript IV First-Strand Synthesis SuperMix (Life Technologies, 18091200). qPCR was performed using Fast SYBRgreen (Life Technologies, 4385612) on a StepOnePlus Real-Time PCR system (Applied Biosystems). Relative expression was calculated using ΔΔCT values normalized against GAPDH expression. All primers were designed using primer 3 (http://frodo.wi.mit.edu/primer3/) and synthesized by Integrated DNA Technologies.

Chromatin immunoprecipitation (ChIP) qPCR and ChIP sequencing ChIP was performed using iDeal ChIP-seq Kit for Transcription Factors (Diagenode, C01010170) according to manufacturer's protocol. In brief, SK-N-MC and RD-ES cells of the control and knockdown groups were trypsinized, washed, and crosslinked with 1% formaldehyde in culture medium for 10 min at room temperature. Cross-linking was terminated by the addition of 1/10 volume 1.25 M glycine for 5 min at room temperature followed by cell lysis and sonication (Bioruptor, Diagenode), resulting in an average chromatin fragment size of 200bp. Chromatin equivalent to 5×106 cells was isolated and incubated with 5 μg antibody overnight at 4 °C (H3- acetyl K27, H3K4me3, H3K9me3, MED1, and IgG (Diagenode). ChIP DNA was isolated by washing and reversal of cross-linking. The eluted DNA was used for SYBRgreen qPCR. The primer sequences used for ChIP-qPCR are provided in Table S2.

10ng of the ChIP DNA was used for ChIP sequencing library preparation following the protocol of TruSeq ChIP library preparation kit (Illumina, IP-202-1012). Briefly, A single “A” nucleotide was added to the 3’ ends of the blunt-ended ChIP DNA fragments and then ligated to a unique adapter. The ligation products were purified and selected at the size of 250-300bp by 4% NuSieve agarose gel (Lonza) electrophoresis. Size selected DNA was purified and PCR-amplified and quantitated with the Bioanalyzer 2100 (Agilent). Libraries were pooled and ran on Nextseq500 platform (Illumina Inc.) with high throughput single-end reads of 75bases (Cat.TG-160-2005).

ChIP-seq analysis Sequencing reads for both in-house generated and publicly available datasets were uniformly processed using an in-house ChIP-seq analysis pipeline. Reads were quality checked with FASTQC (www.bioinformatics.babraham.ac.uk/projects/fastqc) and aligned to the GRCh37 (release 27) genome using the STAR v2.5.1 aligner (1) with default settings. PCR duplicate reads in the aligned bam files were removed using samtools (2), bam files were converted to CPM normalized bigwig tracks using deeptools (3), and viewed using the Integrated Genomics Viewer(4) or the UCSC genome browser (5). Single end reads were extended up to the fragment length (200bp) along the read direction.

Enrichment analysis of ChIP-seq data The enrichment peaks for the various transcription factors and histone marks were computed using MACS2(6), with default settings. Regions found to ubiquitously enriched across a number of next- generation sequencing experiments (7), also known as the blacklisted peaks (https://sites.google.com/site/anshulkundaje/projects/blacklists), were excluded in all subsequent analysis.

Overlap of enrichment peaks Overlap analysis of enrichment peaks in different samples was performed using an in-house python script. In our analysis, two peaks were taken to be overlapping even if they were a single base.

RNA-seq library preparation and sequencing After the treatments, total RNA was isolated using miRNeasy kit (QIAGEN) and the quality of the RNA was analyzed by Bio-analyzer (Agilent) using RNA nano chip. After confirming that all of the RNA samples have RNA integrity number (RIN) more than 8, RNA-seq libraries were constructed using the TruSeq sample Prep Kit V2 (Illumina) according to the manufacturer’s instructions. Briefly, 1 μg of purified RNA was poly-A selected and fragmented with fragmentation enzyme. After first and second strand synthesis from a template of poly-A selected/fragmented RNA, other procedures from end-repair to PCR amplification were performed according to the instructions given in the protocol. Libraries were purified and validated for appropriate size on a 2100 Bioanalyzer DNA 1000 chip (Agilent Technologies, Inc.). The DNA library was quantitated using Qubit and normalized to 4 nM prior to pooling. Libraries were pooled in an equimolar fashion and diluted to a final concentration of 1.8 pM. Library pools were clustered and run on Nextseq500 platform (Illumina Inc.) with single-end reads of 75 bases (Cat.TG-160- 2005).

RNA-seq analysis Single-end sequencing reads were demultiplexed using Illumina bcl2fastq, quality checked using FASTQC (www.bioinformatics.babraham.ac.uk/projects/fastqc), and aligned to the GRCh37 (release 27) genome using the STAR v2.5.1 aligner,(1) with default settings. The read statistics generated by STAR v2.5.1 was used to ensure that the aligned reads in all samples were over 90% of the total reads. The transcripts were assembled using cufflinks and the count and FPKM tables were computed using cuffnorm (8). Principal component analysis on the FPKM values was used to cluster the samples for further quality check.

Differential Gene expression analysis Genes differentially expressed between any two sets of treatment conditions were determined using DESeq2 (9) (https://bioconductor.org/packages/release/bioc/html/DESeq.html), a statistical tool that employs shrinkage estimates to compute fold changes. In all our calculations, the raw RNA-seq read counts from biological duplicates, for each treatment condition, was used as the input for DESeq2. Heatmaps for differentially expressed genes were generated using in-house python scripts.

Gene Ontology analysis We determined the Gene Ontology (GO) associated with a given set of genes using the python api for g:Profiler (https://biit.cs.ut.ee/gprofiler). The results were plotted using in-house python scripts.

Gene set enrichment analysis (GSEA) Functional class scoring(10) of the all differentially expressed genes, against a given gene signature, was performed using the GSEA tool (11), developed by the Broad Institute (https://software.broadinstitute.org/cancer/software/gsea/wiki/index.php/Main_Page).

Immunoblot Analysis For immunoblot analyses, cells were lysed in RIPA buffer (Boston Bioproducts, BP-115DG) supplemented with protease inhibitor (Pierce, A32965). Samples for HIF1D immunoblot were harvested in 1x Laemmli sample buffer (Bio-Rad, 1610737). Lysates were boiled in SDS sample buffer (Invitrogen) and 30-50 μg of protein was separated by SDS-PAGE and loaded onto a PVDF membrane. Membranes were blocked for one hour in blocking buffer (Tris-buffered saline, 0.1% Tween (TBS-T), 5% non-fat dry milk) and incubated overnight at 4°C in primary antibody. Blots were washed with TBS-T and incubated with HRP-conjugated secondary antibody for one hour at room temperature. Blots were washed again with TBS-T and visualized after incubation with chemiluminescent substrate (GE Healthcwere). The antibodies used in the study are provided in Table S3.

Immunoprecipitation For immunoprecipitation assay, HEK293T cells were transfected with pCDNA-HA-HIF1D (gift from Dr. Frank S. Lee, UPENN) and pCDNA-Myc-LOXHD1 at a 1 to 2 ratio. 48h post- transfection, cells were incubated in 1% of O2 hypoxia chamber for 6h followed by lysis in IP buffer (20 mM Tris pH7.5, 150 mM NaCl, 1% Triton-X 100, Protease Inhibitor) and sonication. Whole cell lysates (500 Pg) were pre-cleaned by incubation with protein G Dynabeads (Life Technologies) for 2h on a rotator at 4°C. 5 μg antibody was added to the pre-cleared lysates and incubated on a rotator at 4°C overnight. Protein G Dynabeads were then added for 2h. Beads were washed four times in IP buffer, containing 300 mM NaCl, and resuspended in 40 μL of 1x Laemmli sample buffer and boiled at 95°C for 5 min for separation of the protein and beads. Samples were then analyzed by SDS-PAGE and western blotting as described above.

Immunofluorescent staining For general staining 100,000 cells were seeded on a coverslip which fits in wells of 24-well plate, the next day, cells were fixed with 4% of paraformaldehyde for 10min at RT. Cells were then washed with PBS and blocked with 10% goat serum and 0.5% Triton 100 in PBS at RT for 60min. Primary antibody of LOXHD1 and/or HA was diluted at 1:500 in 10% of goat serum in PBS and incubated with the cells at 4°C overnight. On the next day, secondary antibody goat anti-Rabbit Alexa Fluor 568 (Thermo Fisher, A-11011) and/or anti-Mouse Alexa Fluor 488 (Thermo Fisher, A28175) was diluted at 1:1000 and incubated at RT for one hour. Coverslips were then mounted and stained for DAPI with anti-fade mounting medium (Vector, H-1200). Images were acquired and processed on a Zeiss confocal microscope (LSM 880).

For F-actin staining, cells were plated in the same way, stained with Phalloidin-iFluor 488 at 1:1000 at room temperature for 1h. Images were acquired and processed on a Zeiss confocal microscope (LSM 880). Cell surface area were measured with ImageJ.

For HA-NLS-coiled coil staining in 293T cells: 293T cells (ATCC, CRL-3216) were plated at a density of 8X104 in per well of a 24-well plate with Ploy-L-Lysine (Sigma-Aldrich, P8920-100ML) coated cover slips the day before transfection. 500ng plasmids were transfected by using FuGENE HD regent. 24h and 48h after transfection, the cells were fixed with 4% PFA for 10 min, RT. The cells were permeabilized and blocked by incubation in blocking buffer (4% BSA, 0.5% saponin in PBS) for 1h at room temperature and then incubated with 1st antibody (rat anti-HA, 3F10, Roche, 1:200 in blocking buffer) overnight at 4°C. The next day, the cover slips were washed with PBS for three times and then incubated in 2nd antibody (Alexa Fluor 488, Goat anti-rat IgG, Thermo Fisher, A11006, 1:300) and DAPI (Sigma-Aldrich, D8417, 30ng/ml) diluted with blocking buffer for one hour at room temperature, followed by 3 times of washing with PBS. Lastly, coverslips were mounted with Prolong Gold anti-fade mounting medium (Invitrogen, P36934). The images were acquired using Zeiss LSM 880 confocal microscope.

Colony and sphere formation assay For colony formation assay, 5,000 cells were plated in one well of the 6-well plates, in two weeks cells were fixed and stained with 0.5% of crystal violet, quantification was done by de-staining of crystal violet and measure absorbance at 560nm. For sphere formation assay, 500 cells were suspended in 100μL 50% of Matrigel of full RPMI culture media and spread on the edge of wells in a 24-well plate, we then fill the wells with 1 ml of full RPMI culture media. In three weeks, spheres were counted, and the sizes of the spheres were measured with ImageJ. Triplicates were carried in all the above described experiments; student t-test was used for statistical analysis. Cell aggregation assay 24-well plates were coated with 100 μL of 3% of poly-HEMA and dried in cell culturing hood overnight. 10,000 cells in single cell suspension were seeded in the wells, pictures were taken at time 0 and 16hr after seeding. Aggregation sizes were measured with ImageJ, student t-test was used for statistical analysis.

Wound healing assay Cells were grown as monolayer in 6-well plate, scratches were made with 1000 μL tips. Pictures of the same area were taken at 24 and 48h. Gap between scratches were measured by ImageJ.

Matrigel invasion assay Stably knockout/down cells were trypsinized and 100,000 cells were suspended in 500μL serum- free RPMI medium and added into Matrigel coated invasion chambers (Corning, 354480). Groups of hypoxic and normoxic samples were performed side by side. The bottom of the chamber was filled with RPMI containing 20% serum as chemo attractant. Cells that had degraded the matrix and migrated through the porous membrane (8 μm pore size) to the other end after a period of 24h or 48h were fixed and stained with crystal violet (0.5%), and images were captured using phase contrast microscopy. The same cells were seeded into chambers with no Matrigel coating as migration control. % invasion was determined by numbers of cells invaded divided by numbers of cells migrated. Invasion index was calculated by % invasion test cells / % invasion control cells. Triplicates were carried in all the above described experiments; we take six images for each well and count the cell numbers. Student t-test was used for statistical analysis.

Chicken chorioallantoic membrane assay for tumor cell intravasation The CAM assay for tumor cell intravasation was performed as previously described (12). Briefly, fertilized chicken eggs were incubated in a rotary humidified incubator at 38°C for 10 days. A small hole was drilled through the eggshell into the air sac and another hole was drilled near the allantoic vein that penetrates the eggshell, keeping the chick chorioallantoic membrane (CAM) intact. The CAM was dropped by applying mild vacuum to the hole over the air sac. Subsequently, a cutoff wheel (Dremel) was used to cut a 1 cm2 window encompassing the second hole near the allantoic vein to expose the underlying CAM. Cells were prepared for implantation by trypsinizing and resuspending in media (without FBS) at the density of 2 × 106 cells/50 μL. The CAM was gently abraded with a sterile cotton swab to provide access to the mesenchyme and 50 μL cell suspension was implanted on top of it. The windows were sealed, and the eggs returned to a stationary incubator. The eggs remained in the incubator for 72h, after which the egg was cut along the long circumference, and the upper half (with the inoculum) and the content of the egg was discarded; the CAM that lines the cavity of the eggshell was lifted and snap-frozen. Genomic DNA from the lower CAM was extracted using PureGene Genomic DNA isolation kit (Gentra-Qiagen) following the manufacturer’s protocol and used as a template for human Alu sequence amplification by q-PCR to quantitate the difference in the number of cells between control group and LOXHD1 KD group.

Zebrafish migration assay All procedures on zebrafish (Danio rerio) were approved by the Institutional Animal Care and Use Committee of the University of Pennsylvania. Fertilized zebrafish eggs were incubated at 28 °C in E3 solution and raised using standard methods. Embryos were transferred to E3 solution containing 5 μg/ml protease and 0.2 mM1-phenyl-2-thio-urea (PTU, Sigma) at 24 h post- fertilization to dechorionate the fish embryos and prevent pigmentation, respectively. At 48 h post- fertilization, zebrafish embryos were anesthetized with 0.03% tricaine (Sigma) and transferred to an injection plate made with 1.5% agarose gel for microinjection. Approximately 200–400 mCherry tagged EWS cells suspended in conditioned media supplemented with 0.5 mM EDTA were injected into the perivitelline space of each embryo using a XenoWorks Digital Microinjector (Sutter Instrument). Pre-pulled micropipettes were used for the microinjection (Tip ID 50 μm, base OD 1 mm, Fivephoton Biochemicals). After injection, the fish embryos were immediately transferred to PTU-E3 solution. Injected embryos were kept at 33 °C and examined every day to monitor tumor cell migration using an Olympus Ix81 widefield microscope.

EwS mouse xenograft 3.5×106 SK-N-MC EwS cells were injected in a 1:1 mix of cells suspended in PBS with Geltrex Basement Membrane Mix (Thermo Fisher) in the right flank of 10–12 weeks old NOD/Scid/gamma (NSG) mice. Tumor diameters were measured every second day with a caliper and tumor volume was calculated by the formula L×l2/2. Once the first tumor of the control group reached an average volume of 1,500 mm3, all animals of the experiment were sacrificed by cervical dislocation. Other humane endpoints were determined as follows: Ulcerated tumors, loss of 20% body weight, constant curved or crouched body posture, bloody diarrhea or rectal prolapse, abnormal breathing, severe dehydration, visible abdominal distention, obese Body Condition Scores (BCS), apathy, and self-isolation. Animal experiments were approved by local authorities and conducted in accordance with ARRIVE guidelines, recommendations of the European Community (86/609/EEC), and UKCCCR (guidelines for the welfare and use of animals in cancer research).

Statistical analysis Statistical analysis was performed using GraphPad Prism 6 software. For individual comparisons, unpaired Student t test was used and P < 0.05 were considered significant. Statistical significance for Kaplan–Meier analysis was determined by the log-rank (Mantel–Cox) test.

Table S2. List of Oligonucleotide Primers used in this study

CRISPR gRNAs Sequences 5’ – 3’ GGAA Enhancer KO CACCGAGAGAAATTAAAAACAAACA gRNA_1 FWD GGAA Enhancer KO AAACTGTTTGTTTTTAATTTCTCTC gRNA_1 REV GGAA Enhancer KO CACCGGAAAAACATCTGCAAGCATC gRNA_2 FWD GGAA Enhancer KO AAACGATGCTTGCAGATGTTTTTCC gRNA_2 REV GGAA dcas9 gRNA_1 CACCGAGAGAAATTAAAAACAAACA FWD GGAA dcas9 gRNA_1 AAACTGTTTGTTTTTAATTTCTCTC REV GGAA dcas9 gRNA_2 CACCGGTAGAGATGACAGGAGTAAA GGAA dcas9 gRNA_2 AAACTTTACTCCTGTCATCTCTACC Cloning NG-138 ATGATGGCCCAGAAGAAGAAGCGGAG NG-64 ACACCCTGCAGCAAGTCCCAACC PW-83 GAACAAAAACTCATCTCAGAAGAGGATCTTGAGAATT CCACCACACTG PW-84 TCTGAGATGAGTTTTTGTTCAACGGCCGCGACAGACG GGAAGAGCTC NG-191-NheI-CC-S CAGCTGGCTAGCACCATGGTGTGGCTGC GGCACCTGGTG NG-192-NotI-CC-AS TTGCGGCCGCTCAAGCGTAATCTGGAACATCGTA TGGGTAAACGGCCGCAATCACCTCCTGCATCCCT GGCC Genomic PCR validation GGAA enhancer FWD AAGTGGAACTCAGTGTGGAACA GGAA enhancer REV GCAGGGCACAGAACAGGTACCT SYBR Green qPCR LOXHD1 FWD TAGTGACCAGGCTGGGACTTG LOXHD1 REV GCTTCTCCACTTCTATCCCCT FLI-1 FWD TTAAGGAGGCTCTGTCGGTG FLI-1 REV GAGGGGGTTGATCTTGTGGG CCK FWD AGGGTATCGCAGAGAACGGA CCK REV GGGCCTGCTGGATGTATCTT GAPDH FWD TGCACCACCAACTGCTTAGC GAPDH REV GGCATGGACTGTGGTCATGAG CHIP qPCR GGAA microsatellite FWD AAACAAATAGCCTGCCCATCAG GGAA microsatellite REV CCCTCCTTCCTTCCGTGTTT LOXHD1 TSS FWD CTCAGGTTCCCGCAGGTGT LOXHD1 TSS REV GGGGCATCATTCTGTCGGC Non-Specific control FWD ATCCCCCACAACTCCACCTA Non-Specific control REV ACAGGTAGCAACGAACTGGG Human Alu Taqman qPCR Human Alu FWD GTCAGGAGATCGAGACCATCCT Human Alu REV AGTGGCGCAATCTCGGC Human Alu Taqman probe 5′-6-FAM-AGCTACTCGGGAGGCTGAGGCAGGA- TAMRA-3′

Table S3. List of Antibodies used in this study

Antibody Use Supplier Catalog Number H3K27ac ChIP Active motif 39133 H3K4me3 ChIP Millipore 07-473 LOXHD1 IB/IF Grillet N, et al. (2009) FLI-1 antibody IB Abcam ab15289 GAPDH-HRP IB Cell Signaling 3683 HIF1D IB Cayman Chemical 10006421 Company HA IB/IF Cell Signaling Technology 3724S Myc-tag IB/IP Cell Signaling Technology 2276S Phalloidin-iFluor 488 IF Abcam ab176753 Reagent Goat anti mouse IgG 594 IF Fisher Scientific A11032 Goat anti rabbit IgG 488 IF Fisher Scientific A11008 Goat anti mouse IgG 488 IF Fisher Scientific A11029 Goat anti rabbit IgG 568 IF Fisher Scientific A11011 Mouse IgG IP Diagenode K01641008 Rabbit IgG IP Cell Signaling Technology 2729S

Reference

1. A. Dobin et al., STAR: ultrafast universal RNA-seq aligner. Bioinformatics 29, 15-21 (2013). 2. H. Li et al., The Sequence Alignment/Map format and SAMtools. Bioinformatics 25, 2078-2079 (2009). 3. F. Ramírez et al., deepTools2: a next generation web server for deep-sequencing data analysis. Nucleic Acids Res 44, W160-165 (2016). 4. J. T. Robinson et al., Integrative genomics viewer. Nat Biotechnol 29, 24-26 (2011). 5. W. J. Kent et al., The human genome browser at UCSC. Genome Res 12, 996-1006 (2002). 6. Y. Zhang et al., Model-based analysis of ChIP-Seq (MACS). Genome Biol 9, R137 (2008). 7. E. P. Consortium, An integrated encyclopedia of DNA elements in the human genome. Nature 489, 57-74 (2012). 8. C. Trapnell et al., Transcript assembly and quantification by RNA-Seq reveals unannotated transcripts and isoform switching during cell differentiation. Nat Biotechnol 28, 511-515 (2010). 9. M. I. Love, W. Huber, S. Anders, Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol 15, 550 (2014). 10. A. L. Tarca, G. Bhatti, R. Romero, A comparison of gene set analysis methods in terms of sensitivity, prioritization and specificity. PLoS One 8, e79217 (2013). 11. A. Subramanian et al., Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci U S A 102, 15545- 15550 (2005). 12. I. A. Asangani et al., Characterization of the EZH2-MMSET histone methyltransferase regulatory axis in cancer. Mol Cell 49, 80-93 (2013).

Table S1. Details on ESS32 genes

Gene ontology Status in ESS32 # Full name (biological Function Ewing sarcoma Gene list process) literature

Encodes for a transcriptional regulator that Todd M. binds to, among other promoter regions, the positive regulation Stevens, CLCNKA promoter. It is known to inhibit Kruppel Like of transcription by International 1 KLF15 MEF2A and GATA4, thereby playing a role Factor 15 RNA polymerase Journal of in controlling cardiac hypertrophy. It has II [GO:0045944] Surgical also been elucidated as a negative regulator Pathology, 2018 of TP53 acetylation.

Cyclic nucleotide phosphodiesterase with a dual-specificity for the second messengers Phosphodiester apoptotic process 2 PDE1B cAMP and cGMP, which are key regulators None ase 1B [GO:0006915] of many important physiological processes. Has a preference for cGMP as a substrate

Inês M. Gomes, Thomas G. P. STEAP2 Metalloreductase that has the ability to copper ion import Molecular Grunewald, 3 STEAP2 Metalloreducta reduce both Fe(3+) to Fe(2+) and Cu(2+) to [GO:0015677] Cancer Biology of the se Cu(1+). Research, 2012 Cell, 2012

The encoded protein is known to inhibit regulation of cell ABI Family ectopic metastasis of tumor cells and cell 4 ABI3 migration None Member 3 migration via interaction with p21-activated [GO:0030334] kinase.

DnaJ Heat Encodes for a member of a subclass of the Shock Protein HSP40/DnaJ protein family, which are 5 DNAJC12 None None Family (Hsp40) known to associate with complex assembly, Member C12 protein folding, and export.

Daniel H Protein Wen Luo, Wai, Phosphatase 1 intracellular signal Wen Luo, Nature International 6 PPP1R1A Regulatory transduction Inhibitor of protein-phosphatase 1 Nature Oncotarget, Journal of Inhibitor [GO:0035556] Oncogene, 2018 2020 Oncology, Subunit 1A 2002

adenylate cyclase- inhibiting G- Jason U. Tilan, Neuropeptide protein coupled Receptor for neuropeptide Y and peptide 7 NPY1R Oncotarget, Y Receptor Y1 receptor signaling YY. 2013 pathway [GO:0007193] VAV proteins are guanine nucleotide Rodolphe Vav Guanine exchange factors (GEFs) for Rho family Guinamard, regulation of Nucleotide GTPases, which activate downstream Scandinavian 8 VAV1 GTPase activity Exchange pathways that lead to actin cytoskeletal Journal of [GO:0043087] Factor 1 rearrangements and transcriptional Immunology, changes. 1997

Encodes for a tyrosine phosphatase, and has the ability to dephosphorylate both Dual protein tyrosine and serine/threonine residues. 9 DUSP26 Specificity dephosphorylation Protein product may regulate neuronal None Phosphatase 26 [GO:0006470] proliferation. This gene has been described as both a tumor suppressor and an oncogene.

Potassium Voltage-Gated ion Encodes for protein that forms a Channel transmembrane heterodimer with the potassium voltage- 10 KCNAB3 None Subfamily A transport gated channel, shaker-related subfamily of Regulatory [GO:0034220] proteins. Beta Subunit 3

Tissue-specific splicing factor with Andrew J. RNA Binding cell differentiation potential implication in the regulation of Annalora, 11 RBM11 Motif Protein [GO:0030154] alternative splicing during neuron and germ Oncotarget, 11 cell differentiation. 2018

Potassium negative Voltage-Gated regulation of Regulates neurotransmitter release, heart Channel voltage-gated rate, insulin secretion, neuronal excitability, 12 KCNE3 None Subfamily E potassium channel epithelial electrolyte transport, smooth Regulatory activity muscle contraction, and cell volume. Subunit 3 [GO:1903817]

Binds to myosin, titin, and light muscle contraction meromyosin. Shares genetic homology to 13 MYOM2 Myomesin 2 None [GO:0006936] fibronectin type III and immunoglobulin C2 domains.

negative regulation of Associates with the mTORC2 complex, Proline Rich 5 14 PRR5L protein thereby regulating cellular processes such None Like phosphorylation as survival and cytoskeletal organization. [GO:0001933]

Mitchel J. positive regulation Protein-coding gene that contains a Richard Machiela, Leah A. NK2 of transcription by homeobox domain and has possible role in Smith, 15 NKX2-2 Nature Owen, PLoS Homeobox 2 RNA polymerase the morphogenesis of the central nervous Cancer Cell, Communication One, 2008 II [GO:0045944] system. 2006 s, 2018

homotypic cell- Xg 16 XG cell adhesion Encodes for the XG blood groupOphélie antigen Meynet, Cancer Research, 2010 Glycoprotein [GO:0034109] Encodes for E3 ubiquitin-protein ligase. Mediates the ubiquitination of ATP6V0C protein and marks it for degradation through the Ring Finger 17 RNF182 ubiquitination ubiquitin-proteasome pathway. Inhibits the None Protein 182 [GO:0016567] TLR triggered innate immune response via ubiquitination and subsequent degradation of NF-kappa-B component RELA.

Potassium Voltage-Gated potassium ion A voltage-gated potassium channel found 18 KCNA2 Channel transport primarily in the brain, central nervous None Subfamily A [GO:0006813] system, and the cardiovascular system. Member 2

Proline Rich A protein-coding gene associated with the 19 PRRT4 Transmembran None None disease, Zellweger Syndrome. e Protein 4

Congyi Lu, cardiac left Jason U. Tilan, Jason Tilan, The Journal Neuropeptide ventricle Receptor for neuropeptide Y and peptide 20 NPY5R Oncotarget, Neuropeptide of Biological Y Receptor Y5 morphogenesis YY. 2013 s, 2016 Chemistry, [GO:0003214] 2011

Part of the Troponin I subfamily of genes. It Troponin I3, muscle contraction 21 TNNI3 encodes for the TnI-cardiac protein which None Cardiac Type [GO:0006936] is only expressed in cardiac muscle tissues.

neuroblast Encodes for the ligand that activates the 22 ARTN Artemin proliferation None GFR-alpha-3-RET receptor complex. [GO:0007405]

CD7B-Cell Metin Metin Antigen A B lymphocyte antigen receptor that Ozdemirli, Ozdemirli, Receptor B cell receptor works in conjunction with CD79B to David R. Lucas, The Nature 23 CD79A Complex- signaling pathway initiate the signal transduction cascade Anatomic American Modern Associated [GO:0050853] activated by the binding of an antigen to the Pathology, 2001 Journal of Pathology, Protein Alpha B-cell antigen receptor complex. Surgical 2001 Chain9a Pathology,

bone A phosphatase that has a high specificity Phosphoethano mineralization for phosphoethanolamine (PEA) and PHOSPHO lamine/Phosph 24 involved in bone phosphocholine (PCho). Plays a role in None 1 ocholine maturation generating inorganic phosphate for bone Phosphatase 1 [GO:0035630] mineralization.

adenylate cyclase- modulating G The protein product is part of the beta Andreas Adrenoceptor protein-coupled adrenergic receptor family, and is involved Kirschner, 25 ADRB3 Beta 3 receptor signaling in the regulation of lipolysis and Oncotarget, pathway thermogenesis. 2016 [GO:0007188] Doublecortin cellular defense Encodes for protein that plays a role in the 26 DCDC2 Domain response inhibition of canonical Wnt signaling None Containing 2 [GO:0006968] pathway.

G Protein G protein-coupled Encodes for protein that plays a crucial role Subunit receptor signaling 27 GNGT2 in cone phototransduction, and is None Gamma pathway specifically localized in cones. Transducin 2 [GO:0007186]

Encodes for a transcriptional repressor FEZ Family neuron migration protein, and is involved in the axonal 28 FEZF1 None Zinc Finger 1 [GO:0001764] projection and proper termination of olfactory sensory neurons.

UDP-glucuronosyltransferases catalyze UDP cellular response phase II biotransformation reactions in Glycosyltransfe 29 UGT3A2 to genistein which lipophilic substrates are conjugated None rase Family 3 [GO:0071412] with glucuronic acid to increase water Member A2 solubility and enhance excretion.

Lipoxygenase calcium ion Homology transmembrane Involved in hearing. Required for normal 30 LOXHD1 Domain- None transport function of hair cells in the inner ear. Containing [GO:0070588] Protein 1

a potent bioactive lipid mediator) and fatty acid. Does not hydrolyze other phospholipids, Dorothea E. like Benjamin J. Mahlendorf, phosphatidylserine Hydrolyzes specifically phosphatidic acid Juergen L. Foell, Schmiedel, Cancer 31 LIPI Lipase I (PS), (PA) to produce 2-acyl lysophosphatidic Pediatric Blood Molecular Biology & phosphatidylcholi acid (LPA & Cancer, 2008 Biology Therapy, ne (PC) and Reports, 2011 2013 phosphatidylethan olamine (PE) or triacylglycerol (TG). {ECO:0000269|Pu bMed:12963729}.

Contains a homeobox domain, and encodes Retina And for a transcription factor known to have visual perception 32 RAX Anterior Neural functions in eye development. Required for None [GO:0007601] Fold Homeobo retinal cell fate determination and regulates stem cell proliferation.