USP1 Expression Driven by EWS::FLI1 Transcription Factor Stabilizes Survivin and Mitigates Replication Stress in Ewing Sarcoma

This article is featured in Selected Articles from This Issue, p. 1121

Oncogene expression and genome instability are hallmarks of cancer. Oncogene activation leads to replication stress, characterized by slowing or stalling of replication fork movement, with uncoupling of DNA replication polymerase and helicase activity of the MCM complex, generating long tracts of single-stranded DNA and stalled replication forks. Persistent stalled forks result in fork collapse and DNA breaks, leading to chromosomal alterations and rearrangements. Oncogenes cause replication stress by different mechanisms including deregulated origin licensing and/or origin firing, nucleotide and reactive oxygen species (ROS) metabolism, increased transcription generating R-loops that hinder replication fork movement, and collision of the transcription-replication machinery (1, 2). In normal cells, in response to replication stress, the DNA damage response (DDR) pathway is activated to halt cell-cycle progression and stabilize and repair stalled replication forks. If the burden of genomic alterations is high, the cell-cycle checkpoint pathway promotes apoptosis or cellular senescence. Thus, the DDR pathway functions as a barrier to tumorigenesis, leading to cell-cycle arrest, DNA repair, cell death, or senescence; a dysfunctional DDR pathway promotes genome instability leading to cancer development (1, 3).

USP1 is a well-characterized deubiquitylase (also called ubiquitin protease or deubiquitinase) for PCNA, FANCD2, and FANCI, key players of the DDR pathways, mainly the replication stress response, the Fanconi anemia pathway (FA), and lesion bypass mediated by translesion DNA synthesis (TLS; refs. 4, 5). Moreover, USP1 expression is significantly altered in several tumor types, indicating USP1 regulates additional cellular processes (6). Using different cancer models, studies have now identified additional substrates of USP1. Among these, USP1 stabilizes ID (ID1/ID2/ID3 and ID4) proteins, TBK1, KPNA2, KDM4A, TBLR1, ERα, and TAZ and regulates ubiquitylation mediated signaling of FANCI/D2, PCNA, ULK1, and AKT (7–14).

Ewing sarcoma (EWS) is the second most common malignant pediatric bone cancer, after osteosarcoma. Previous studies have shown that the cytogenetic characteristic of EWS is a reciprocal chromosomal translocation and in-frame gene fusion of EWSR1, at 22q12, with one of five members from the ETS (E26 transformation-specific) gene family of transcription factors, namely FLI1, ERG, ETV1, ETV5, and FEV1 (15–17). EWSR1 encodes for EWSR1 protein, a member of the TET (TLS/FUS, EWSR1, and TAF15), also known as FET, family of proteins (18). The ETS proteins are winged helix-loop-helix transcription factors with a highly conserved 85 amino acid ETS domain that mediates site-specific DNA binding and facilitates protein–protein interactions. In normal cells, members of both the TET and ETS families of transcription factors function in cellular processes required for cellular growth, proliferation, and differentiation (16, 18). In most EWS, the chromosomal translocation and in-frame gene fusion occurs between EWSR1 and FLI1 t(11;22)(q24;q12). The resulting EWS::FLI1 fusion gene encodes the EWS::FLI1 transcription factor that harbors the N-terminal trans-activation domain of EWSR1 and the C-terminal ETS DNA binding domain of FLI1. The two most common types, fusion of EWSR1 exon 7 to FLI1 exon 6 (type 1) and fusion of EWSR1 exon 7 to FLI1 exon 5 (type 2), account for about 60% and 25%, respectively, of EWS::FLI1 fusions. Type 3 EWS::FLI1 arises from fusion of exon 10 of EWSR1 with exon 6 of FLI1. EWS::FLI1 types 1 and 2 underlie more severe disease with metastasis (16, 19, 20). Studies from several laboratories collectively show that EWS::FLI1 oncogenic transcription factor mediates epigenetic alterations regulating gene expression and alternative splicing (20–22). As a transcriptional regulator, EWS::FLI1 deregulates transcription factors, signaling molecules, cell-cycle regulators, and proteins regulating angiogenesis, metabolism, and the DDR pathways (16, 20, 23). Dysregulation of these genes by EWS::FLI1 and their biological functions significantly contributes to drive tumor initiation, disease progression, and metastasis in EWS (20, 23).

Previous studies have shown that EWS cells have high levels of EWS::FLI1 oncogene-induced R-loops and replication stress (24, 25). The precise mechanism by which EWS cells survive and proliferate in elevated levels of replication stress is not known. In this study, we have identified and demonstrated the molecular function of USP1 deubiquitylase in EWS cell survival in response to endogenous replication stress. We show that USP1 is overexpressed in EWS and USP1 is a transcriptional target of EWS::FLI1 transcription factor. USP1 knockdown or inhibition results in delayed S-phase cell-cycle progression and reduced cell proliferation and clonogenic cell survival in the absence of external stress stimuli. Notably, USP1 deubiquitylates Survivin [also known as Baculoviral IAP repeat containing 5 (BIRC5) or apoptosis inhibitor 4 (API4)], regulating Survivin protein stability and thereby the activation of caspase-3/7 and cell death by apoptosis in response to endogenous replication stress. Further, USP1 inhibition sensitizes EWS cells to doxorubicin and etoposide treatment, standard of care chemotherapeutic agents for treating EWS.

A673 (ATCC, RRID:CVCL_0080), U2OS (ATCC, RRID:CVCL_0042), and HEK293T (ATCC, RRID:CVCL_0063) cells were cultured in DMEM supplemented with 10% FBS and 1% penicillin/streptomycin. TC71 (DSMZ, RRID:CVCL_2213), MHH-ES1 (DSMZ, RRID:CVCL_1411), and A4573 (DSMZ, RRID:CVCL_6245) cell lines (kind gift from Dr. Jason Yustein) were cultured in RPMI supplemented with 10% FBS and 1% penicillin/streptomycin. HCT116 (ATCC, RRID:CVCL_0291) cells were cultured in McCoy's 5A supplemented with 10% FBS and 1% penicillin/streptomycin. Human mesenchymal stem cells (hMSC; Lonza, RRID:CVCL_Y510) were cultured in MSCBM (Lonza, PT-3238) supplemented with mesenchymal cell growth supplement (MSGs), l-Glutamine, and GA-1000 SingleQuots (Lonza, PT-4105). ASc52telo (ATCC, RRID:CVCL_U602) cells were cultured in MSCBM (ATCC, PSC-500–030) supplemented with Mesenchymal Stem Cell Growth Kit (ATCC, PSC-500–400). All cell lines were authenticated using short tandem repeat profiling analysis. Cells were maintained in a 37°C incubator containing 5% CO₂. Biological sex of cell line used can be found in Supplementary File S3. Mycoplasma screening of cells is performed at intervals of 30 to 45 days using the SIGMA Mycoplasma PCR detection kit (Catalog no. MP0035). Experiments were performed with cell lines passages 5–15 within 1 week of thawing. For replicate experiments, cells were maintained in culture for 4 weeks at most.

RNA was isolated from cell lines using Trizol (Life Technologies) following the manufacturer's instructions. Complementary DNA (cDNA) was synthesized using First Strand cDNA Synthesis kit (ThermoFisher) as per reagent protocol. For qPCR analysis, reactions (20 μL) were performed in triplicate using 3 μL of sample DNA, 20 pmol of each primer, and 10 μL of Power SYBR Green PCR Master Mix (ThermoFisher). Forty reaction cycles of 15 seconds at 95°C and 60 seconds at 60°C were carried out on a QuantStudio3 thermocycler (Applied Biosystems). Relative expression was calculated using the comparative C_T method (∆∆C_T) and data was normalized to GAPDH and RNAPII expression. Primer sequences used for the DUB array and reverse transcription-quantitative polymerase chain reaction (RT-qPCR) in this study are provided in Supplementary File S3.

EWS patient tumor samples were procured and processed as described previously (bioRxiv 2023.01.15.524130). Briefly, paraffin-embedded samples from ESFT patients treated at Instituto Ortopedici Rizzoli (IOR), Bologna, Italy, and at the Department of Pathology of the University of Valencia Estudi General (UVEG), Spain from the period comprised between April 1971 and May 2007 were analyzed within the context of two European Translational Research projects [PROTHETS (http://www.prothets.org) and EuroBo-Net (http://www.eurobonet.eu)]. All cases were genetically confirmed as belonging to the ESFT by molecular biology and/or fluorescent in situ hybridization (FISH). Approval for data acquisition and analysis was obtained from the Ethics Committee of the institutions involved in the study. The clinical data were reviewed and stored within a specific database. Characteristics of the tissue microarray (TMA) cohort and relevant clinical information have been previously reported (26). A total of 24 TMAs containing two representative cores for each case (1 mm in diameter) were constructed for IHC analysis. Of 266 samples, 192 samples could be analyzed for USP1 expression. Immunoreactivity was defined as follows: negative, fewer than 5% of tumor cells stained; poorly positive (score 1), between 5% and 10% of tumor cells stained; moderately positive (score 2), between 10% and 50% of tumor cells stained, and strongly positive (score 3), with more than 50% of the tumor cells were stained.

EWS TMA was deparaffinized in xylene, rehydrated in descending alcohols, and treated in a digital pressure cooker containing citrate buffer pH 6.0. Endogenous peroxidase activity was blocked by incubation in 3% hydrogen peroxide for 30 minutes. Sections were rinsed in PBS and blocked for 1 hour in 10% FBS in PBS. Sections were stained with primary monoclonal antibody against USP1 1:100 dilution overnight at 4°C. After rinsing with PBS, sections were incubated for 1 hour with anti-rabbit antibody conjugated horseradish peroxidase secondary antibody. DAB solution was added for 5 minutes. Sections were counter-stained with Mayer's hematoxylin and mounted under cover glasses using a xylene-based mounting medium and images were scanned at the UNMC Tissue Science facility.

For the analysis of public datasets GSE106925 and GSE160898, we first obtained the bigWig files for IGV (version 2.14.1) visualization (27–29). If bigWig files were provided by the original paper, we directly downloaded them for IGV visualization; if bigWig files were not available, we downloaded the BAM files and converted them to bigWig files. Specifically, we first used Homer (30) to convert BAM files into bedGraph files, and then we used the program bedGraphToBigWig (31) to convert bedGraph files to bigWig files. For motif analysis on chromatin immunoprecipitation sequencing (ChIP-seq) data, we first used the findPeaks command in the Homer package to call peaks by comparing the EWS::FLI1 and control groups. Then, we did de novo motif analysis based on the ChIP-seq peaks by using the command findMotifsGenome command in Homer within a 200 basepair window with default options. The found motifs were ranked by P value calculated on the basis of a binomial test against GC%-matched background and only the top motifs were presented. Normalized ChIP-seq signals were obtained by annotating the peaks using the annotatePeaks command in the Homer.

Cells were lysed with NETN buffer at 4°C for 30 minutes. Cell lysates were collected by centrifugation, boiled in 4X Laemmli buffer, and separated by SDS-PAGE. Proteins were transferred to PVDF membrane via semi-dry transfer. Membranes were blocked in 5% milk in 1X TBS/Tween buffer and probed with indicated primary and secondary antibodies.

EWS cells (500 cells/well) were seeded in 96-well plates and treated with DMSO, SJB3–019A, ML-323, doxorubicin or etoposide. Cell proliferation was assessed every 24 hours until 96 hours, and cytotoxicity was assessed following incubation time using the CCK-8 reagent (Dojindo Laboratories) assay as per the manufacturer's protocol.

800 cells were seeded onto 60-mm dishes in triplicate and incubated for 7 to 10 days. Colonies formed were fixed and stained with Coomassie Blue. The numbers of colonies were counted, and the percentage of cell survival was calculated.

EWS cells were seeded at 3,000 cells/well (TC71, 30 cells/μL) or 1,500 cells/well (MHH-ES1, 15 cells/μL) in ultralow attachment 96-well plates (Corning). The serum-free sphere culture media consisted of DMEM/F12 (1:1) (Fisher) supplemented with 1X B27 (ThermoFisher), 4 U/L insulin (Sigma), 4 μg/mL heparin (Stem Cell Tech), 20 ng/mL EGF (Peprotech), 20 ng/mL bFGF (Peprotech), and 1X penicillin/streptomycin. In inhibitor-treated sphere assays, USP1 inhibitor was added to the media at time of plating. Spheres were imaged using an EVOS FL Auto microscope (Life Technologies) 3 days (TC71) or 7 days (MHH-ES1) after plating and spheres were counted using ImageJ.

Cells treated with USP1 inhibitors were pelleted by centrifugation at 1,000 rpm for 5 minutes at 4°C and washed twice with PBS. 150,000 cells were resuspended in 100 μL 1X binding buffer (10X concentrate composed of 0.1 mol/L HEPES pH 7.4, 1.4 mol/L NaCl, and 25 mmol/L CaCl₂ diluted to 1X in sterile water). 5 μL FITC Annexin-V and 10 μL Propidium Iodide was added and incubated for 15 minutes at room temperature protected from light. 400 μL 1X binding buffer was added before FACS analysis. 50,000 events were counted for each sample.

5,000 cells were seeded in 96-well white plate, clear bottom tissue culture plates and treated with indicated USP1 inhibitors. After treatment, caspase activity was measuring using Caspase-Glo activity assay kits for caspase-3/7, caspase-8, or caspase-9 from Promega according to the manufacturer's protocol. For assays using Ac-DEVD-CHO, cells were treated for 1 hour with 750 nmol/L Ac-DEVD-CHO prior to treatment with SJB3–019A or ML-323.

TC71 cells were seeded at 40,000 cells/well in black clear bottom 96-well plates. Cells were treated with ML-323 for 0, 1, or 3 hours and ROS levels were measured using Cellular ROS Detection Assay Kit (Abcam) according to the manufacturer's protocol.

Cells were collected and washed twice in ice-cold PBS and resuspended in homogenization buffer (20 mmol/L HEPES-KOH, pH 7.5, 10 mmol/L KCl, 1.5 mmol/L MgCl₂, 1 mmol/L sodium EDTA, 1 mmol/L sodium EGTA, 1 mmol/L dithiothreitol, 250 mmol/L sucrose, and protease inhibitors). After 30 minutes incubation on ice, cells were homogenized by 40 passes through a 25G needle. Homogenates were centrifuged at 500 × g for 5 minutes at 4°C and the supernatant was centrifuged at 10,000 × g for 20 minutes at 4°C to obtain mitochondria. Supernatant obtained was centrifuged at 100,000 × g for 1 hour at 4°C to obtain the cytoplasmic fraction. The mitochondrial pellet was washed three times in homogenization buffer by centrifuging at 7,000 × g for 5 minutes at 4°C. The mitochondria were solubilized in TNC buffer (10 mmol/L Tris-acetate, pH 8.0, 0.5% Nonidet P-40, 5 mmol/L CaCl₂, and protease inhibitors).

Cells were collected and washed in PBS. Cell pellet was resuspended in 3 volumes lysis buffer (10 mmol/L HEPES, pH 7.4, 10 mmol/L KCl, 0.05% Nonidet P-40, and protease inhibitors) and incubated for 20 minutes on an end-to-end rocker at 4°C, followed by centrifugation at 13,300 rpm for 10 minutes at 4°C to obtain the cytoplasmic fraction. The nuclear pellet was washed in lysis buffer by centrifuging at 13,300 rpm for 10 minutes at 4°C. Nuclear pellet was resuspended in 3 volumes low salt buffer (10 mmol/L Tris-HCl, pH 7.4, 0.2 mmol/L MgCl₂, 1% Triton-X 100, and protease inhibitors) and incubated for 15 minutes on an end-to-end rocker at 4°C and then centrifuged at 13,300 rpm for 10 minutes at 4°C to obtain the nuclear fraction.

TC71 cells were transfected with constructs encoding SFB-tagged USP1 FL or C90S and HA-ubiquitin where indicated. Cells were lysed 24 hours later with NETN buffer. The lysates were cleared by centrifugation and then incubated with 20 μL of protein A-beads coated with Survivin antibody on an end-to-end rocker overnight at 4°C. After three washes with NETN buffer, the proteins bound to the beads were eluted by boiling with SDS sample buffer, resolved by SDS-PAGE, and analyzed by Western blotting.

All experiments were independently replicated (biological replicates) at least three times. Technical replicates are listed in the Figure legends. Western blots were quantified using ImageJ software and normalized to the loading control. Results were graphed and analyzed using GraphPad Prism 8 software and data are presented as mean ± SD. Two-tailed paired t test was performed for qPCR, ChIP, colony formation assays, sphere assays, ROS assays, and proliferation assays. Two-way ANOVA and Dunnett multiple comparison test were performed for FACS experiments. Statistical significance was reported as P ≤ 0.05 (*), P ≤ 0.01 (**), P ≤ 0.001 (***), P ≤ 0.0001 (****), or ns (not significant).

Stable short hairpin RNA (shRNA) knockdown cell lines were generated by retroviral induction of shRNA targeted to EWS::FLI1, FLI1, or non-silencing control. Virus was produced in Plat-GP cells using the shRNA plasmid and PIK-VSVG retroviral envelope plasmid. 48 hours post-transfection, virus was collected and used to transduce TC71 and MHH-ES1 cells. GFP or mCherry positive cells were sorted by flow cytometry into 96-well plates and stable clones were selected with puromycin-containing media. Stable shRNA knockdown cell lines were generated by lentiviral induction of shRNA targeted to USP1 or non-targeting control. Virus was produced in HEK293T cells using the shRNA plasmid, psPAX2 (Addgene) and PMD2.G (Addgene) lentiviral packaging plasmids. 48 hours post-transfection, virus was collected and used to transduce TC71 and MHH-ES1 cells. Stable clones were selected for by growing transduced cells in puromycin-containing media. Puromycin concentrations for various cell lines were as follows: TC71, 0.4 μg/mL; MHH-ES1, 0.2 μg/mL. USP1 knockout in TC71 cells was generated using CRISPR-Cas9 technology. TC71 cells seeded in 96-well plate were transfected with TrueGuide sgRNA and Cas9 protein using lipofectamine CRISPRMAX transfection reagent. 48-hour post-transfection limiting dilution was performed and cells were incubated for 96 hours. Pooled cells were screened for USP1 levels by SDS-PAGE and Western blotting and verified by DNA sequencing. shRNA and guide RNA (gRNA) sequences are provided in Supplementary File S3.

Expression data of USP1 and USP43 (Expression Public 22Q4) in EWS cell lines was downloaded from the CCLE included in the Cancer Dependency Map Portal (https://depmap.org/portal; ref. 32). Expression data of USP1 and USP43 in 263 sarcoma patient samples and 2 normal tissue samples was downloaded from Firebrowse (http://firebrowse.org). USP1 expression in EWS patient samples was examined by analyzing publicly available microarray expression data GSE17679 for EWS tumors (n = 64) and compared with expression data GSE9103 for skeletal muscle (n = 18; ref. 33). Log₂FC and P value were calculated using GEO2R and LIMMA packages (34). EWS patient survival data (GSE17679) was downloaded from R2: Genomics Analysis and Visualization Platform (http://r2.amc.nl). All data are freely available online and access is unrestricted and does not require patients’ consent or other permissions. Data were graphed and analyzed using GraphPad Prism 8 software and data are presented as mean ± SD.

ChIP was performed after double cross-linking of cells with disuccinimidyl glutarate and formaldehyde as described. Cross-linked chromatin was immunoprecipitated using 5 μg of a control IgG (Cell Signaling Technology) antibody or α-FLI1 (Abcam) antibody. After immunoprecipitation, 25 μL protein A-beads was added to each sample and incubated for 2 hours on a rocker at 4°C. Beads were then sequentially washed with low salt immune complex wash buffer (0.1% SDS, 1% Triton X-100, 2 mmol/L EDTA, 20 mmol/L Tris-HCl pH 8.1, 150 mmol/L NaCl), high salt immune complex buffer (0.1% SDS, 1% Triton X-100, 2 mmol/L EDTA, 20 mmol/L Tris-HCl pH 8.1, 500 mmol/L NaCl), LiCl buffer (0.25 mol/L LiCl, 1% IGEPAL, 1% Na-Deoxycholate, 1 mmol/L EDTA, 10 mmol/L Tris-HCl pH 8.1), and TE buffer (10 mmol/L Tris-HCL pH 8.1, 1 mmol/L EDTA). After the final wash, DNA was eluted in 250 μL ChIP elution buffer (1% SDS, 0.1 mol/L NaCHO₃) for 20 minutes at 27°C. After elution, beads were pelleted at 10,000 rpm for 5 minutes at room temperature and supernatant was collected into a clean tube. Elution process was repeated once. 20 μL 5 mol/L NaCl was added to 500 μL collected supernatant or 500 μL input control (50 μL sonicated lysate + 450 μL ChIP elution buffer) and incubated overnight at 65°C for reverse cross-linking. The following day, 20 μL 1 mol/L Tris-HCl pH 6.5 + 10 μL 500 mmol/L EDTA + 2.3 μL 20 mg/mL Proteinase K was added to each sample and incubated for 1 hour at 45°C. Samples were cooled to room temperature before adding an equal volume of phenol:chloroform:isopropanol (25:24:1) and mixed vigorously. Samples were centrifuged at 13,000 rpm for 25 minutes at room temperature. The aqueous phase was collected and to it added an equal volume of chloroform:isopropanol (24:1) and mixed vigorously. Samples were centrifuged at 13,000 rpm for 25 minutes at room temperature. The aqueous phase was collected, and DNA was precipitated using two volumes of 100% ethanol, 0.1 volume 3 mol/L NaOAC pH 5.2, and 1.5 μL glycogen. Samples were incubated overnight at −80°C. The following day, samples were washed with 70% ethanol and allowed to dry before resuspending the immunoprecipitated DNA in 20 μL nuclease free water. For ChIP-qPCR analysis, immunoprecipitated purified DNA was then subjected to SYBR green (ThermoFisher) based quantitative PCR. Reactions (20 μL) were performed in triplicate using 2 μL sample DNA, 25 pmol of each primer, and 10 μL Power SYBR Green PRC Master Mix (ThermoFisher). Forty reaction cycles of 15 seconds at 95°C and 60 seconds at 60°C were carried out on a QuantStudio3 thermocycler (Applied Biosystems). Relative enrichment with respect to IgG control based on the 2^−ΔCT method. Primers used for ChIP RT-qPCR are shown in Supplementary File S3.

TC71 cell lines expressing a non-targeting shRNA or an shRNA against USP1 were used to analyze proteome quantification. Samples were prepared and tandem mass tag (TMT)-labeled per the manufacturer's protocol (ThermoFisher TMT10plex Mass Tag Labeling Kit). After TMT labeling, acetonitrile was removed by Speedvac, and samples were resuspended in 0.1% trifluoroacetic acid. Sample cleanup with C18 tips was performed per the manufacturer's protocol (Pierce, ThermoFisher). Sample concentrations were requantified (Pierce BCA Protein Assay kit) and combined in equal concentrations. Following combination, samples were dried by Speedvac and fractionated using the ThermoFisher high-pH reverse phase fractionation kit according to the manufacturer's protocol for TMT. Resulting fractions were dried by Speedvac and resuspended in 0.1% formic acid for mass spectrometry (MS) analysis. TMT-MS and data analysis was performed in the Proteomics Core facility at UNMC. Samples were loaded onto trap column Acclaim PepMap 100 75 μm × 2 cm C18 LC Columns (Thermo ScientificTM) at flow rate of 5 μL/min then separated with a Thermo RSLC Ultimate 3000 (Thermo ScientificTM) from 5% to 20% solvent B (0.1% FA in 80% ACN) from 10 to 98 minutes at 300 nL/min and 50ºC with a 120-minute total run time for fractions one and two. For fractions three to six, solvent B was used at 5% to 45% for the same duration. Eluted peptides were analyzed by a Thermo Orbitrap Fusion Lumos Tribrid (Thermo Scientific TM) mass spectrometer in a data dependent acquisition mode using synchronous precursor selection method. A survey full scan MS (from m/z 375- 1500) was acquired in the Orbitrap with a resolution of 120,000. The AGC target for MS2 in iontrap was set as 1 × 104 and ion filling time set as 150 ms and fragmented using CID fragmentation with 35% normalized collision energy. The AGC target for MS3 in orbitrap was set as 1 × 105 and ion filling time set as 200 ms with a scan range of 100 to 500 and fragmented using HCD with 65% normalized collision energy. Protein identification was performed with Proteome Discoverer software version 2.1 (Thermo Fisher Scientific) by searching MS/MS data with Sequest against the UniProt human protein database (74600 sequences; downloaded 04/2020). The search was set up for full tryptic peptides with a maximum of 2 missed cleavage sites. Oxidation, TMT6plex of lysine, and phosphorylation of serine/threonine/tyrosine were included as variable modifications and carbamidomethylation and TMT6plex of the amino terminus were set as fixed modifications. The precursor mass tolerance threshold was set at 10 ppm and fragment mass error of 0.6 Da with a minimum peptide length of 6 and a maximum peptide length of 144. The significance threshold of the ion score was calculated on the basis of a false discovery rate calculated using the percolator node. Quantification was performed on unique peptides considering protein groups for peptide uniqueness. Protein abundances were normalized by total peptide amount and scaled on channels average. Gene names were obtained from the description tab in the Proteome Discoverer output file. Gene ontology pathway analysis was performed using DAVID Bioinformatics Database 6.8 using the functional annotation tool. TMT data is deposited to the ProteomeXchange Consortium and processed data is provided in Supplementary File S2.

6 × 10⁵ cells were plated in 100-mm dishes. The following day, thymidine (Sigma) was added to a final concentration of 2 mmol/L and cells were incubated at 37°C for 16 hours. At the end of the first thymidine block, the cells were washed twice with PBS and allowed to grow in fresh medium for 9 hours. At the end of the first release, thymidine was added to a final concentration of 2 mmol/L and cells were incubated at 37°C for 16 hours. At the end of the second thymidine block, cells were washed twice with PBS and allowed to grow in fresh medium until harvested at specific timepoints post release from double thymidine block. 1.1 × 10⁶ cells were fixed in ice-cold 70% ethanol. Samples were washed in 1 mL PBS and resuspended in 1 mL Telford Reagent (16.81 mg EDTA, 500 μL Triton X-100, 500 mL PBS) supplemented with 27 μg/mL RNase A (Roche Diagnostics) and 40 μg/mL propidium iodide (Sigma), incubated at 37°C for 30 minutes protected from light and analyzed by FACS.

1 × 10⁶ cells were seeded in 100-mm plates and treated with 100 μg/mL cycloheximide (Sigma) for the indicated time points. Cells were lysed in NETN buffer supplemented with 1 μmol/L sodium orthovanadate (Sigma) on an end-to-end rocker at 4°C for 30 minutes and lysates were collected by centrifugation at 13,000 rpm for 15 minutes.

1 × 10⁶ cells were seeded in 100-mm plates. The following day, cells were treated with 10 μmol/L MG132 (Sigma) for 4 hours. Cells were lysed in NETN buffer on an end-to-end rocker at 4°C for 30 minutes and lysates were collected by centrifugation at 13,000 rpm for 15 minutes.

Antibodies, sources of reagents, RRIDs and primer sequences are provided in Supplementary File S3.

Plasmid DNA constructs and cell lines generated in this study will be made available via material transfer agreement. The MS proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE (35) partner repository with the data set identifier PXD041033. The data analyzed in this study were obtained from Gene Expression Omnibus at GSE106925, GSE17679, GSE160898, and GSE9103.

The human genome encodes for 100 DUBs, of which 78 function as active ubiquitin proteases. The role of DUBs in EWS pathogenesis is not well studied. To gain insight into the expression profile of DUBs in EWS, we performed a comparative real-time quantitative PCR (RT-qPCR) array of all DUB encoding genes in a panel of EWS::FLI1 fusion harboring EWS cell lines TC71 (type 1), A673 (type 1), MHH-ES1 (type 2), and A4573 (type 3) and bone marrow–derived hMSCs, the EWS cell of origin (Fig. 1A, left; Supplementary Fig. S1A, and File S1) and observed at least 10-fold overexpression of several DUBs in EWS cell lines compared with hMSC. However, among all upregulated DUBs, USP1 and USP43 were upregulated over 38-fold in all four EWS cell lines compared with hMSC (Fig. 1A, right). RNA sequencing (RNA-seq) data of sarcoma tumor samples from The Cancer Genome Atlas (TCGA) dataset revealed USP1, but not USP43, is overexpressed in sarcoma tumors compared with normal tissue (Fig. 1B). Comparing the mRNA expression of USP1 and USP43 in EWS cell lines from DepMap portal (32), USP1 is overexpressed in EWS cell lines compared with USP43 (Fig. 1C). Focusing on USP1, we confirmed the DUB array data by performing RT-qPCR analysis using three individual primer sets targeting the USP1 transcript (Supplementary Fig. S1B). RT-qPCR confirmed increased levels of USP1 transcripts in EWS cell lines (A673, TC71, MHH-ES1, and A4573) compared with bone marrow–derived hMSC (Fig. 1D) and ASc52telo cells (adipose-derived mesenchymal stem cells; Supplementary Fig. S1C). The increase in USP1 mRNA expression (Fig. 1D; Supplementary Fig. S1C) correlated with an increase in USP1 protein levels in all four EWS cells compared with hMSC and ASc52telo cells (Fig. 1E). Notably, analysis of microarray data of EWS patient samples (GSE17679) showed high USP1 expression in EWS patient samples compared with skeletal muscle control (Fig. 1F; ref. 33). Examining patient survival data available at R2 EWS database (GSE17679), we found that high USP1 expression correlated with poor event-free (Fig. 1G, left and bottom) and overall patient survival (Fig. 1G, right and bottom; ref. 33). IHC analysis for USP1 expression on an EWS TMA (26) revealed nuclear USP1 staining (Fig. 1H, top) with high USP1 expression in 155 samples (∼81%) with a staining intensity score of 2 (n = 76) or 3 (n = 79) in a cohort of 192 total samples (Fig. 1H, bottom). Together, these data demonstrate high USP1 expression in EWS. The function of USP1 overexpression in EWS disease initiation, progression, metastasis, and/or drug resistance is not known. Therefore, we focused our investigation on delineating the role of USP1 in EWS pathogenesis.

The EWS::FLI1 fusion gene encodes for an oncogenic transcription factor that drives the gene signature underlying EWS. To examine if USP1 expression is upregulated by EWS::FLI1, we generated EWS::FLI1 knockdown cells using shRNA targeting the type specific EWS::FLI1 fusion (shEWS::FLI1–1) and FLI1 3' UTR (shEWS::FLI1–2) sequences in TC71 (type 1) and MHH-ES1 (type 2) cell lines. RT-qPCR analysis showed USP1 expression was decreased in TC71 (Fig. 2A, left) and MHH-ES1 (Fig. 2A, right) upon EWS::FLI1 knockdown compared with control. CDKN1A (p21) was upregulated and NR0B1 (DAX1) was downregulated upon EWS::FLI1 knockdown as previously reported (Fig. 2A; refs. 36, 37). The decrease in USP1 mRNA was translated at the protein level in TC71 (type 1; Fig. 2B, left) and MHH-ES1 (type 2; Fig. 2B, right) EWS::FLI1 knockdown cells compared with control, suggesting that USP1 expression is regulated by the EWS::FLI1 transcription factor, independent of the specific EWS::FLI1 fusion type in EWS. To determine if USP1 is a transcriptional target of EWS::FLI1, we performed ChIP of EWS::FLI1 in control and EWS::FLI1 knockdown TC71 (Fig. 2C, top) and MHH-ES1 (Fig. 2C, bottom) cells. ChIP RT-qPCR analysis showed that EWS::FLI1 bound the USP1 promoter in control cells, and binding was lost upon EWS::FLI1 knockdown. NR0B1, a known transcriptional target of EWS::FLI1 was used as a control (Fig. 2C). We examined EWS::FLI1 association on the USP1 promoter by analyzing published ChIP-seq data (GSE106925) for FLI1 (EWS::FLI1), H3K27Ac, H3K24me3, and H3K4me1 in MSCs expressing vector control or EWS::FLI1 expression construct (28). We observed increased association of EWS::FLI1 on the USP1 promoter in MSCs expressing EWS::FLI1 compared with the control (Fig. 2D). EWS::FLI1 binds GGAA microsatellite repeats to regulate the expression of its target genes (28, 38). Motif analysis of ChIP-seq peaks revealed an enrichment of GGAA/TTCC microsatellite repeats (Fig. 2E), indicating that high USP1 in EWS is driven by EWS::FLI1. In addition to EWS::FLI1, binding motifs for Nrf2, ERG(ETS), and Sox18 were identified, suggesting that other transcription factors expressed at normal levels may regulate USP1 at the transcript level in non-EWS cells. Further, no significant alterations in histone marks were observed on the USP1 promoter in MSCs expressing EWS::FLI1 compared with MSC control, suggesting that overexpression of EWS::FLI1 increased its binding to USP1 promoter but did not alter overall binding activity of other transcription factors on USP1 promoter region and does not directly alter histone marks on USP1 promoter. To address if the ETS DNA binding domain of FLI1 in EWS::FLI1 is required for EWS::FLI1 association on the USP1 promoter, we analyzed published CUT&RUN and ATAC sequencing (ATAC-seq) datasets (GSE160898) of EWS::FLI1 from A673 EWS::FLI1 knockdown (EF KD) and knockdown cells reconstituted with EWS::FLI1 (EF) or FLI1 ETS DNA binding domain (EF DBD) (27). CUT&RUN signals revealed that loss of EWS::FLI1 occupancy on the USP1 promoter in EWS::FLI1 knockdown cells is rescued by EF and EF DBD in A673 rescue cells (Fig. 2F), indicating that ETS DNA binding domain mediates EWS::FLI1 binding to the USP1 promoter. Further, ATAC-seq signal showed that chromatin accessibility on the USP1 promoter was unaltered in A673 EF knockdown and rescue cells. This agrees with the previous study showing the loss of EWS::FLI1 is not always associated with changes in the open chromatin state of EWS::FLI1 targets. Alternatively, partial EWS::FLI1 knockdown was insufficient to reflect chromatin changes in this system (27). Collectively, our data shows USP1 is upregulated in EWS cell lines and USP1 is a transcriptional target of EWS::FLI1.

To determine the functional relevance of high USP1 expression in EWS, we examined the effect of USP1 knockdown on cell proliferation and the ability of the cells to undergo unlimited cell divisions by clonogenic cell survival assay. USP1 knockdown in TC71 (Fig. 3A, left) and MHH-ES1 (Fig. 3A, right) led to a dramatic reduction in the number of colonies formed in the absence of external stress stimuli, supporting a requisite role for USP1 in EWS cell growth and proliferation. Notably, we also observed a dramatic reduction in the number of spheroids formed in USP1 knockdown TC71 (Fig. 3B; Supplementary Fig. S2A, left) and MHH-ES1 (Fig. 3B; Supplementary Fig. S2A, right) cells compared with control, suggesting that USP1 promotes cancer stem/progenitor cell proliferation and self-renewal. To examine if the deubiquitylase activity of USP1 is required to regulate cell proliferation, we examined cell proliferation of TC71 and MHH-ES1 cells upon treatment with ML-323 or SJB3–019A, well characterized reversible and irreversible USP1 inhibitors, respectively (39, 40). TC71 (Supplementary Fig. S2B) and MHH-ES1 (Supplementary Fig. S2C) cells were sensitive to treatment with increasing concentration of SJB3–019A (Supplementary Fig. S2B and S2C, left) with 100-fold lower IC₅₀ concentration compared with ML-323 (Supplementary Fig. S2B and S2C, right). IC₂₅ and IC₇₅ concentrations were calculated using the experimentally determined IC₅₀ and Hill slope (Supplementary Fig. S2D). TC71 cells were hypersensitive to ML-323 compared with U2OS and HCT116 cells, non-EWS cell lines (Supplementary Fig. S2E). Importantly, we observed a dramatic decrease in TC71 (Fig. 3C and D, left) and MHH-ES1 (Fig. 3C and D, right) cell proliferation over a period of 96 hours upon treatment with either USP1 inhibitor SJB3–019A (Fig. 3C) or ML-323 (Fig. 3D). In addition, similar to USP1 knockdown, inhibition of USP1 by SJB3–019A led to a significant reduction in the number of colonies formed in TC71 (Supplementary Fig. S2F, left) and MHH-ES1 (Supplementary Fig. S2F, right). Further, we observed a dramatic reduction in the number of spheroids formed in both TC71 (Fig. 3E and F; Supplementary Fig. S2G, and S2H, left) and MHH-ES1 (Fig. 3E and F; Supplementary Fig. S2G, and S2H, right) upon USP1 inhibition by either SJB3–019A (Fig. 3E; Supplementary Fig. S2G) or ML-323 (Fig. 3F; Supplementary Fig. S2H).

To gain further insight, we synchronized control and USP1 knockdown TC71 cells at the G1-S boundary using double thymidine block and followed cell-cycle progression over a period of 24 hours. Intriguingly, USP1 knockdown cells showed a dramatic delay in both S-phase cell-cycle progression and entry into S-phase as the cells progressed through another round of the cell cycle compared with control (Fig. 4A). Notably, delay in progression through S-phase did not trigger G₂–M or G₁ arrest in USP1 knockdown cells (Fig. 4A). Previous studies have shown that USP1 deubiquitylates mono-ubiquitinated PCNA (ub-PCNA) and mono-ubiquitinated FANCD2 (ub-FANCD2) in response to replication stress due to DNA damage during DNA replication (4, 5, 41). Thus, it is possible that loss of USP1 function leads to the retention of ub-PCNA and ub-FANCD2 on chromatin following endogenous DNA damage, thereby hindering DNA replication fork progression resulting in delayed progression through S-phase. This may explain the dramatic decrease in cell proliferation but does not explain the reduced ability of cells to grow and reproduce into a clonal population upon USP1 knockdown or inhibition (Fig. 3; Supplementary Fig. S2F), leading us to ask if USP1 inhibition resulted in cell death by apoptosis. Apoptosis in TC71 (Fig. 4B and C) and MHH-ES1 (Supplementary Fig. S3A and S3B) cells treated with DMSO, SJB3–019A (Fig. 4B; Supplementary Fig. S3A) or ML-323 (Fig. 4C; Supplementary Fig. S3B) USP1 inhibitors at IC₅₀ concentrations was examined by Annexin V staining. We observed an increase in early and late apoptosis upon USP1 inhibition compared with control (Fig. 4B and C; Supplementary Fig. S3A and S3B). Similarly, USP1 knockdown (Supplementary Fig. S3C) and knockout cells (Supplementary Fig. S3D) showed an increase in apoptotic fraction compared with the control. Taken together, our data shows that downregulation or inhibition of USP1 delays cell-cycle progression and induces cell death by apoptosis leading to decreased EWS cell proliferation and clonogenic cell survival in the absence of any exogenous DNA damage.

A previous study showed that USP1 is overexpressed in osteosarcoma cells and promotes cell proliferation and stem cell maintenance by regulating the stability of ID1, ID2, ID3, and ID4 proteins (ID proteins; ref. 7). Studies have also shown that ID2 is a transcriptional target of EWS::FLI1 and EWS::FLI1 regulates the expression of other ID genes via miRNA (42). Therefore, we hypothesized that, in EWS, EWS::FLI1 regulates ID protein mRNA expression, while USP1 controls ID protein turnover. However, USP1 depletion did not result in a decrease in ID protein levels in EWS cells (Supplementary Fig. S4A). This can be attributed to the fact that the genetic, and thereby the protein, landscape of EWS is distinct from that of osteosarcoma, as EWS is driven by the oncogenic transcription factor EWS::FLI1 which is not expressed in osteosarcoma. Therefore, to identify substrate(s) of USP1 that regulate EWS cell proliferation (Fig. 3), we performed TMT-MS analysis of TC71 control and USP1 knockdown cells. TMT-MS analysis revealed 3,154 proteins with ≥ 2 peptides, of which 70 proteins were differentially regulated in shUSP1/Control ≤ 1.5-fold (Fig. 5A; Supplementary File S2). Pathway analysis of the differentially regulated proteins revealed that USP1 knockdown alters proteins that participate in cell division, mitosis, cell cycle, host–virus interaction, and DNA repair processes (Fig. 5B). Our data showed that USP1 depletion resulted in decreased cell proliferation and increased cell death by apoptosis (Figs. 3 and 4). Therefore, we examined differentially regulated proteins known to function in cell division, mitosis, and cell-cycle pathways from the TMT analysis (Fig. 5B) and found that among all the targets examined (Fig. 5C), Survivin/BIRC5 protein levels decreased in USP1 knockdown (Fig. 5D), USP1 knockout (Supplementary Fig. S4B), and USP1 inhibitor treated cells (Supplementary Fig. S4C) compared with the control. EWS::FLI1 and PRC1, a candidate target identified in TMT-MS as differentially regulated, were not affected by USP1 knockdown (Fig. 5D; Supplementary Fig. S4B). Further, Survivin transcript levels remained unchanged upon USP1 or EWS::FLI1 depletion (Supplementary Fig. S4D), indicating that USP1 regulates Survivin protein levels. Notably, the decrease in Survivin protein levels in USP1 inhibitor treated (Fig. 5E) or knockout cells (Supplementary Fig. S4E) correlated with an increase in the levels of cleaved PARP1 (c-PARP), indicating cell death by apoptosis. Similar to previous studies, we found that USP1 inhibition decreased USP1 protein levels (43). Cycloheximide (CHX) chase analysis in control and USP1 knockdown (Fig. 5F) or knockout (Supplementary Fig. S4F) TC71 cells showed a decrease in steady state USP1 protein levels and confirmed faster Survivin protein turnover upon USP1 depletion. PRC1 stability, which is not regulated by USP1, did not change upon USP1 depletion (Fig. 5F; Supplementary Fig. S4F). Further, we observed that Survivin protein stability is regulated by the UPS, as revealed by the accumulation of Survivin protein in control or WT cells treated with the proteasome inhibitor MG132 (Fig. 5G; Supplementary Fig. S4G). Survivin protein was destabilized in USP1 knockdown (Fig. 5G), and knockout (Supplementary Fig. S4G) cells compared with control. Residual Survivin protein was stabilized in MG132 treated USP1 knockdown (Fig. 5G) and knockout (Supplementary Fig. S4G) cells compared with DMSO-treated cells. Because USP1 depletion reduced Survivin protein levels, we asked whether Survivin is a substrate of USP1 deubiquitylase. We examined Survivin ubiquitylation in cells expressing USP1 and USP1 C90S catalytic dead mutant and found that USP1, but not USP1 C90S, was able to decrease Survivin ubiquitylation. Notably, overexpression of USP1 C90S catalytic inactive mutant decreased endogenous Survivin protein levels (Fig. 5H). We observed a marginal reduction in Survivin with USP1 C90S mutant which could be due to only partial loss of USP1 activity, as the catalytic core of all UBP family of DUBs is comprised of both Cys (nucleophile) and His (proton acceptor) residues (44). Collectively, our data shows that USP1 deubiquitylates Survivin and USP1 depletion or inhibition destabilizes Survivin protein in cells. As this manuscript was under preparation, a recent publication showed that Survivin interacts with USP1 and USP1 deubiquitylates Survivin protein (45).

We observed that USP1 depletion or inhibition decreases cell proliferation (Fig. 3) with an increase in cell death by apoptosis (Fig. 4; Supplementary Fig. S3). Additionally, we found a decrease in Survivin protein and increase in cleaved-PARP1 levels upon USP1 inhibition or depletion (Fig. 5E; Supplementary Fig. S4E). Therefore, we examined whether Survivin destabilization correlated with apoptosis induction upon USP1 inhibition. We performed a time course experiment wherein TC71 (Fig. 6A and B, top) and MHH-ES1 (Fig. 6A and B, bottom) cells were treated with DMSO or USP1 inhibitors SJB3–019A (Fig. 6A) or ML-323 (Fig. 6B) at IC₅₀ concentrations for 0, 6, 24, and 48 hours. Apoptosis was examined by Annexin V staining and Survivin protein levels were monitored by Western blotting. We observed a step wise increase in the percentage of cells in early and late apoptosis at 24 hours and 48 hours posttreatment with SJB3–019A (Fig. 6A) or ML-323 (Fig. 6B) in both TC71 and MHH-ES1 cells. Actin has been shown to be a substrate for cleavage by caspases in mammalian cells during apoptosis (46). We found that the dramatic induction of apoptosis also led to a decrease or loss of β-actin levels 48 hours post USP1 inhibition by SJB3–019A and ML-323, respectively. Notably, Survivin levels decreased and cleaved-PARP1 protein levels increased along with the percentage of cells in early and late apoptosis in SJB3–019A or ML-323 treated TC71 and MHH-ES1 cells (Fig. 6A and B). These findings suggest that destabilization of Survivin protein may mediate the induction of apoptosis observed upon USP1 inhibition.

Previous studies have shown that Survivin inhibits caspase-3/7 activation (47). Therefore, we predict USP1 knockdown, or inhibition will result in destabilization of Survivin, leading to the activation of caspase-3/7 in the absence of external stress stimuli. Indeed, we observed that USP1 knockdown or inhibition by SJB3–019A or ML-323 resulted in the activation of caspase-3/7 in TC71 (Fig. 6C, left) and MHH-ES1 (Fig. 6C, right) cells. Notably, induction of caspase-3/7 activity observed upon USP1 inhibition was prevented by co-treatment with Ac-DEVD-CHO, a caspase-3/7 inhibitor, in TC71 (Fig. 6D, left) and MHH-ES1 (Fig. 6D, right). The intrinsic and extrinsic apoptosis pathways signal the activation of caspase-3 by the activation of caspase-9 and caspase-8, respectively (48). Woo et al. recently showed that USP1 regulates Survivin and DR5 protein stability and promotes TRAIL-induced apoptosis (45). TRAIL-induced apoptosis is mediated via recruitment and activation of caspase-8 (49). However, we did not observe caspase-8 activation upon USP1 knockdown or inhibition in TC71 (Fig. 6E, left) or MHH-ES1 (Fig. 6E, right) cells. We next examined whether caspase-3/7 activation upon USP1 depletion or inhibition is signaled via the intrinsic apoptosis pathway mediated by caspase-9. Indeed, we found that caspase-9 was activated in USP1 knockdown and SJB3–019A or ML-323 treated TC71 (Fig. 6F, left) and MHH-ES1 (Fig. 6F, right) cells compared with control in the absence of external stress stimuli.

The intrinsic apoptosis pathway is activated by intracellular stress signals such as oxidative stress, ischemia, and endogenous DNA damage that signal the release of cytochrome c from the mitochondria leading to subsequent activation of caspase-9 and caspase-3/7 (48). USP1 is a well characterized regulator of replication stress and DNA repair pathways (6). We observed a delay in S-phase cell-cycle progression in USP1 knockdown cells. Therefore, we anticipate replication stress induced upon USP1 depletion will signal the intrinsic apoptosis pathway with the release of cytochrome c into the cytoplasm. In response to replication stress and DNA damage, ATR and ATM kinases are phosphorylated, which in turn phosphorylate CHK1 and CHK2 effectors to recruit downstream DNA repair proteins (3). We observed an increase in p-CHK1 and p-CHK2 levels in ML-323 and SJB3–019A treated cells compared with control (Fig. 7A). H2AX phosphorylation, a marker for DNA breaks, was increased in cells treated with USP1 inhibitors ML-323 (Fig. 7B, top) and SJB3–019A (Fig. 7B, bottom). We also observed a dramatic increase in the levels of endogenous ROS upon USP1 inhibition compared with control (Fig. 7C). Notably, we observed an increase in cytochrome c release from mitochondria into the cytoplasm upon USP1 inhibitor treatment (Fig. 7D) and USP1 knockout (Fig. 7E) compared with control in the absence of exogenous DNA damage.

The standard-of-care chemotherapy regimen for EWS includes vincristine, ifosfamide, doxorubicin, and etoposide (VP16). Doxorubicin and etoposide generate DNA-protein cross-links (DPC) and ifosfamide induces interstrand or intrastrand cross-links (ICL) in the genome. DPCs and ICLs are toxic lesions that hinder the progression of the transcription and replication fork and upon fork collision cause replication stress and DNA breaks in the cells. USP1 participates DNA damage repair, including ICL repair, during DNA replication (6). We also observed that USP1 inhibition exacerbates endogenous replication stress and DNA damage (Fig. 7A and B). Thus, we hypothesize that USP1 inhibition can sensitize cells to replication stress inducing agents. We determined the IC₅₀ concentration for doxorubicin (Supplementary Fig. S5A) and etoposide (Supplementary Fig. S5B) treatment for the TC71 cell line. Cytotoxicity assay in TC71 cells treated with DMSO, IC₅₀ doxorubicin, or IC₅₀ USP1 inhibitor alone and combined treatment of doxorubicin and USP1 inhibitor for 24 hours revealed that cells were sensitive to both doxorubicin and USP1 inhibition alone compared with DMSO-treated cells (Fig. 7F). Notably, TC71 cells were hypersensitive to co-treatment with doxorubicin and USP1 inhibitor SJB3–019A or ML-323 compared with single agent alone (Fig. 7F). Similarly, TC71 cells were hypersensitive to co-treatment with etoposide and USP1 inhibitor SJB3–019A or ML-323 compared with single agent alone (Fig. 7G). Collectively, these data show that in the absence of USP1 activity, failure to repair endogenous DNA damage and prolonged replication stress due to aberrant cell proliferation leads to the activation of the DNA damage signaling cascade, an increase in endogenous ROS levels, and the subsequent release of cytochrome c from the mitochondria. This signals the activation of caspase-9 and caspase-3/7 resulting in cell death by apoptosis, decreased cell proliferation, and decreased clonogenic cell survival (Fig. 7H). Moreover, USP1 inhibition sensitizes EWS cells to replication stress inducing agents. Thus, our data shows that high USP1 expression and replication stress are vulnerabilities that can be exploited for EWS therapy.

Replication stress or replication-associated DNA damage leads to spontaneous DNA breaks and genome instability. In response to genome instability, cells activate the DDR pathways to repair damaged DNA. If DNA damage is unrepaired or mis-repaired, the increased burden of genome instability is sensed by the apoptosis and/or cellular senescence pathways which prevent cells from becoming cancerous (3). EWS tumors have significantly elevated levels of replication stress, indicating that genome maintenance pathways are deregulated in EWS (24, 25). The mechanism by which genome maintenance pathways are deregulated in different cancer types is largely determined by the genetic landscape of the cancer. Thus, understanding how the genome maintenance pathways, namely apoptosis and the DDR pathways, are deregulated in EWS will help understand disease initiation and progression and identify targets and strategies to improve therapy outcomes.

Ubiquitylation is a key post-translational mechanism that regulates the DDR pathways and apoptosis (50). Deregulation of DUBs and their functional significance in EWS is not well studied. In this study, we examined the altered expression of DUBs in EWS and found that USP1 is one among the most highly expressed DUBs in EWS cells. USP1 is a key regulator of replication stress and DNA damage repair pathways through the deubiquitylation of PCNA and FANCD2, respectively. Upon DNA damage, USP1 self-inactivates through autocleavage, which enables its own degradation and, in turn, upregulates ub-PCNA and ub-FANCD2 on chromatin, facilitating DNA lesion bypass by TLS and DNA repair by the FA pathway, respectively (4, 6). Our study shows that EWS::FLI1 drives USP1 expression in EWS. USP1 deubiquitylase stabilizes Survivin and promotes cell survival in response to endogenous replication stress and DNA damage. USP1 knockdown or inhibition promotes EWS cell death by apoptosis in the absence of external DNA damage. Notably, USP1 inhibition sensitizes EWS cells to doxorubicin and etoposide, replication stress inducing agents used to treat patients with EWS. The dependency of EWS cells on high USP1 expression for cell survival has revealed USP1 as a potential therapeutic target for EWS.

EWS cells have high levels of endogenous DNA damage and replication stress and are sensitive to ATR and CHK1 inhibitors. Elevated replication stress is attributed to an increased number of R-loops promoted by EWS::FLI1 mediated CDK9 activation of RNAPII (24, 25). How do EWS cells tolerate elevated levels of replication stress? On the basis of our study and studies in the field, we propose a mechanism by which EWS cells evade apoptosis in response to endogenous replication stress and DNA damage (Fig. 7H). The EWS::FLI1 oncogene is the key driver of EWS (17). We show that EWS::FLI1 transcription factor associates with the USP1 promoter, resulting in high USP1 mRNA expression in EWS cells (Fig. 2). High mRNA expression correlates with high USP1 protein levels in EWS (Fig. 1). USP1 is a Survivin deubiquitylase and stabilizes Survivin protein levels (Fig. 5). Survivin, a member of the inhibitor of apoptosis family of proteins, is known to prevent the activation of caspase-3/7. USP1 downregulation or inhibition destabilizes Survivin and induces caspase-3/7 activity in the absence of external DNA damage stimuli (Fig. 6). Taken together, high USP1 levels, driven by EWS::FLI1, stabilize Survivin protein and promotes EWS cell survival by inhibiting the activation of death effector caspases-3/7 in the presence of endogenous replication stress and DNA damage (Fig. 7H, left). Due to elevated replication stress in EWS, we hypothesized that Ewing cells would be highly sensitive to USP1 inhibition (Fig. 7H, right). Indeed, we observed that USP1 inhibition led to a significant reduction in cell proliferation (Fig. 3) and induction of cell death by apoptosis, in the absence of exogenous DNA damage (Fig. 6). EWS cells treated with USP1 inhibitor, SJB3–019A or ML-323, showed an increase in replication stress and endogenous DNA damage markers (Fig. 7A and B), ROS levels (Fig. 7C), and increased release of cytochrome c from the mitochondria into the cytoplasm (Fig. 7D and E), signaling the activation of caspase-9 and subsequent activation of caspase-3/7 (Fig. 6). USP1 inhibition also destabilizes Survivin, thereby attenuating the inhibitory effect of Survivin on caspase-3/7 activation (Fig. 7H, right). Thus, USP1 knockdown or inhibition exacerbates endogenous replication stress and DNA damage which is sensed by the intracellular apoptosis pathway leading to the activation of caspase-3/7 and cell death by apoptosis (Fig. 7H, right).

Although USP1 is well-studied among the DUB family of proteins, transcriptional regulation of USP1 is not known. We show that USP1 expression in EWS is driven by EWS::FLI1 oncogenic transcription factor. ChIP, CUT&RUN, and ATAC analysis revealed the ETS domain in EWS::FLI1 increases its occupancy on the USP1 promoter (Fig. 2C–F). Motif analysis revealed that in addition to the EWS::FLI1 binding motif, the USP1 promoter harbors Nrf2 and ERG-ETS binding sites (Fig. 2E). Thus, in normal cells and non-EWS::FLI1 expressing cancer cell lines, USP1 expression may be regulated by transcription factors such as Nrf2, ERG, and others which remains to be investigated. We also observed a delay in S-phase cell-cycle progression upon USP1 depletion in EWS cells (Fig. 4). Loss or inhibition of USP1 led to the accumulation of ub-PCNA and ub-FANCD2 on chromatin leading to stalling or slowing of the replication fork (4, 5). Recently, it was shown that USP1 inhibition leads to USP1-trapping on chromatin and this DNA-protein complex led to increased replication fork stalling and fork termination events (51). In addition to these effects, other unidentified USP1 substrates that regulate replication fork progression cannot be ruled out and will be investigated in follow up studies. Because EWS cells have high USP1 levels, USP1 inhibition will result in a high burden of replication stress leading to cell death by apoptosis even in the absence of external DNA damage. Notably, we observed that EWS cells displayed hypersensitivity to combinatorial treatment of doxorubicin or etoposide and USP1 inhibitor compared with single agents alone. We anticipate a similar response upon USP1 inhibition in other cancer types with high levels of endogenous replication stress. Thus, USP1 and the replication stress response pathway are potential targets for cancer therapy, which will be the focus of future studies.

N.T. Woods reports grants from NIH during the conduct of the study. J.A. Lopez-Guerrero reports grants from Generalitat Valenciana, H2020-FETOPEN; personal fees from Diaceutics, GSK; and personal fees from AstraZeneca outside the submitted work. H. Band reports grants from Department of Defense Breast Cancer Research Program; and other support from Nimbus Therapeutics outside the submitted work. G. Ghosal reports grants from NCI, National Institute of General Medical Sciences (NIGMS); and grants from UNMC Pediatric Cancer Pilot Award during the conduct of the study. No disclosures were reported by the other authors.

H.J. Mallard: Conceptualization, resources, formal analysis, methodology, writing–original draft, writing–review and editing. S. Wan: Formal analysis, investigation. P. Nidhi: Resources, investigation. Y.D. Hanscom-Trofy: Resources, investigation. B. Mohapatra: Investigation. N.T. Woods: Formal analysis, investigation. J. Lopez-Guerrero: Resources. A. Llombart-Bosch: Resources. I. Machado: Resources. K. Scotlandi: Investigation. N.F. Kreiling: Resources, investigation. M.C. Perry: Resources, investigation. S. Mirza: Resources. D.W. Coulter: Resources. V. Band: Resources. H. Band: Resources. G. Ghosal: Conceptualization, resources, supervision, funding acquisition, methodology, writing–original draft, writing–review and editing.

This work was supported by NCI R01 (CA263504–01A1), NIGMS R01 (GM141232–01A1), NCI K22 Career Development Award (CA188181), and UNMC Pediatric pilot grant awarded to G. Ghosal.

We thank all members of the Ghosal lab for advice and technical assistance. We thank all the members in UNMC Core facilities- Proteomics and Systems Biology Core for assistance with TMT analysis, Genomics Core for DNA sequencing, Tissue Sciences facility for slide scanning, and Flow Cytometry Research facility for FACS analysis.

The publication costs of this article were defrayed in part by the payment of publication fees. Therefore, and solely to indicate this fact, this article is hereby marked “advertisement” in accordance with 18 USC section 1734.