Patient cohorts

This study included 28 patients who had a diagnosis of ATRT from 2007 to 2022 at Taipei Medical University Hospital (TMUH) and Taipei Veterans General Hospital (VGH) and were aged < 18 years at diagnosis. This study was conducted in accordance with the Declaration of Helsinki, and all patients provided written informed consent. Patient information was completely anonymized. The study protocol was approved by the Institutional Review Board of Human Subjects Research Ethics Committee of the Taipei Medical University and Taipei VGH, Taiwan (IRB approval numbers: N201901033 and 2019-02-010C, respectively).

Cell culture

Human ATRT cell lines (BT12, CHLA266, CHLA02, CHLA04, and Re1P6) were cultured as previously described [30]. HEK293T cell line was cultured in DMEM medium (Thermo Fisher Scientific, MA, USA cat# 12439054) with 10% FBS. All the mediums were added 1% antibiotic-antimycotic (Santa Cruz Biotechnology, CA, USA, cat# sc-3690), and all cell lines were incubated at 37 °C in humidified incubator containing 5% CO2. Cells were confirmed to be Mycoplasma-free using the Mycoplasma PCR Detection Kit (abm, Canada).

Lentiviral transduction and cell transfection

To generate stable gene knockdown cell lines, short hairpin (sh)RNA lentiviral expression system was used. The lentiviral vector pLKO.1-puro carrying shRNA sequences shRRM2#1 (TRCN0000038962 target sequences GCTCAAGAAACGAGGACTGAT), shRRM2#2 (TRCN0000286353 target sequences GCAGACAGACTTATGCTGGAA), and shLuc (target sequence CTTCGAAATGTCCGTTCGGTT). The pLKO.1-shRNA and packaging plasmids pCMV-ΔR8.91, pMD.G were purchased from the National RNAi Core Facility (Academia Sinica, Taiwan). Construct contain short hairpin RNA shRRM2#1, shRRM2#2 or shLuc were produced in HEK293T cells with packaging plasmids pCMV-ΔR8.91 and pMD.G using transfection reagent Polyjet (SignaGen Laboratories, MD, USA, cat# SL100688). The medium supernatant containing the lentivirus was collected, filtered through a 0.45 µm filter, and supplemented with 8 µg/mL polybrene to infected in ATRT cells. Cells were incubated for 24 h at 37 °C, then replaced by complete medium to allowed cells recover for 48 h. Stable knock-down cell lines were selected by puromycin (Gibco, USA, cat# A1113803) for at least 72 h. Knockdown efficiency was assessed by immunoblotting and cells transfected with shLuc were used as control group.

Cell viability (IC50) assay

COH29 (MedChemExpress, NJ, USA, HY-19931) was dissolved in dimethyl sulfoxide (DMSO) (Sigma-Aldrich, Germany, D2650) then diluted to serial concentration in medium. BT12, CHLA266, and Re1P6 cells were seeded into 96-well plate with 5000 cells per well, 6 replicate wells per plate. Cell growth inhibition assay was assessed with CCK8 kit (Abcam, UK, ab228554) by a microplate reader after 48 h of exposure to a serial dilution of COH29 (0.5–100 µM). The half-maximal inhibitory concentration (IC50) of COH29 in BT12, Re1P6, and CHLA266 cells was analyzed using AT Bioquest tool (https://www.aatbio.com).

Cell proliferation assay

Cells were seeded into 96-well plates with 1000, 2500 and 2500 cells per well for BT12 - CHLA266 - Re1P6 respectively, 6 replicate wells per plate. Cells were treated with different dilutions of COH29: 4-8-16-20 µM (for BT12) or 3.5-7-14 µM (for Re1P6 and CHLA266), and 0.1% DMSO as control. Cell growth was assessed every 24 h, then 20µL CCK8 reagent was added to each well before measuring the optical density at 460 nm by microplate reader.

Colony formation assay

Stable knockdown cells or parental cells were seeded into 6-well plates with 2000, 4000 and 8000 cells per well for BT12, CHLA266 and Re1P6, respectively. For drugs treatment, medium with COH29: 2-4-6 µM (for BT12), 3-5-7 µM (for Re1P6), 1-3-5 µM (for CHLA266), or 0.1% DMSO were changed one day after seeding parental cells. Colonies formed after 10 days (for BT12) and 20 days (for CHLA266 and Re1P6) were washed with PBS 1X, fixed with 4% formaldehyde (Macron, cat# H121-08), and stained with 0.5% crystal violet.

Migration assay

Wound-healing assay and transwell assay were applied for detecting cell migration ability. For the transwell assay, 105 cells in 500µL serum-free medium were seeded into transwell upper chamber (Corning, MA, USA, cat# 353097) and placed in 24-well plates with 500µL medium containing 10% FBS. For the drug treatment experiment, 105 cells in 250µL serum-free medium were seeded in the upper chamber for 4 h before adding 250µL serum-free medium containing 2X concentration of COH29 or 0.2% DMSO. After 16-20-24 h incubation for BT12, CHLA266 and Re1P6 respectively, non-migrated cells in the upper chamber were carefully removed using cotton swab. The migrated cells were fixed with 4% formaldehyde and stained with 0.5% crystal violet. Finally, the samples were observed for recording images of cell migration. Ten random fields of each insert were photographed at 20X magnification and quantified using ImageJ.

Wound-healing assay was proceeded using commercial chambers (ibidi, Germany, cat# 80209). Firstly, chambers were fixed in 6 cm dishes, then 100µL of 2 × 104 cells were seeded in to each well of chamber. After 24 h for cell attachment, the chambers were gently taken out and washed with PBS. For the knockdown experiment, 4 mL of culture medium was added to further cultivating the cell culture. For the drugs treatment experiment, cells were treated with medium containing COH29: 8-16-20 µM (for BT12) or 3.5-7-14 µM (for Re1P6 and CHLA266), or 0.1% DMSO. Images were acquired at 0-24-48-72 h. Finally, the average wound gap between wound edges were calculated for cell movement. Images were obtained from 6 different fields using microscope, wound closure and migration efficiency were analyzed using ImageJ.

Western blot

Cell lysates were incubated in RIPA buffer plus EDTA-free Protease Inhibitor Cocktail (Roche Applied Science, Germany, cat# 4693132001) for 30 min, then centrifuged at 13000xg for 20 min at 4 °C. The supernatant fluid was collected to measuring protein concentration using BCA Protein Assay Kit (Thermo Fisher Scientific, USA, cat# 23225). Human normal brain tissue lysate was procured from GeneTex, Inc (cat# GTX28771). Equal amounts of samples were loaded for western blot analysis in 4–15% SDS-PAGE gel. Proteins were transferred on a nitrocellulose membrane (0.45 µm NC), blocked with 5% non-fat mill in PBS 1X with 0.1% Tween 20 (PBST) and probed with primary antibodies overnight at 4 °C. The primary antibodies used were: Beta actin (Epitomics, cat# S0861), RRM2 (GeneTex, cat# GTX103193), TOP2A (GeneTex, cat# GTX35137), TTK (Proteintech, cat# 10381-1-AP), PBK (GeneTex, cat# GTX60560), KIF20A (Santa Cruz Biotechnology, cat# SC-374508), TGFB3 (Abclonal, cat# A8460), MUC1 (Proteintech, cat# 19976-1-AP), Caspase 9 (Abclonal, cat# A13682), Cleaved Caspase 9 (Cell Signaling Technology, cat# 95016), Caspase 7 (Cell Signaling Technology, cat# 9492), Cleaved Caspase 7 (Cell Signaling Technology, cat# 9491), Caspase 3 (Cell Signaling Technology, cat# 9662), Cleaved Caspase 3 (Cell Signaling Technology, cat# 9661), Survivin (Abclonal, cat# A1551), MCL1 (Abclonal, cat# A0250), PARP (Cell Signaling Technology, cat# 9532), BAD (Cell Signaling Technology, cat# 9292), BAX (GeneTex, cat# GTX61026), Cleaved PARP (Cell Signaling Technology, cat# 9541), H2AX (GeneTex, cat# GTX108272), H2AX-γ (GeneTex, cat# GTX127340), BRCA1 (Abclonal, cat# A11549), RAD51 (Abclonal, cat# A6268). After washing with PBST, the membranes were incubated with corresponding HRP conjugated Goat anti-Mouse or HRP conjugated Goat anti-Rabbit (Jackson ImmunoResearch Inc). The membranes were washed with PBST (three times for 10 min) before visualized by enhanced chemiluminescence ECL (PerkinElmer, MA, USA).

Apoptosis assay

Apoptosis assay was performed using an eBioscience Annexin V Apoptosis Detection Kit FITC (Invitrogen, MA, USA, cat# 88800574) based on the manufacturer’s manual. Briefly, cells were collected and washed, followed by centrifugation then re-suspended in 1X binding buffer. Then, 5µL fluorochrome-conjugated Annexin V was added in 100µL of the cell suspension and incubated for 15 min at room temperature. Finally, the cells were washed, centrifuged and re-suspended in 200µL 1X binding buffer before adding 5µL propidium iodide staining solution (PI). The stained cells were analyzed using FACS Canto II cytometer (BD Biosciences, NJ, USA).

Comet assay

Re1P6 and CHLA266 cells were seeded in 6 cm dishes and incubated overnight at 37 °C. The next day, cells were treated with COH29 (14-25-50 µM) or 0.1% DMSO for 10 h. Cells were collected and the alkaline comet assay was applied according to the manufacturer’s instructions of Comet Assay kit (R&D System, Bio-Techne, MN, USA, cat# 4250-050-K). The individual cells or comets were viewed and photographed using a fluorescent microscope (Zeiss) equipped with an DAPI filter. To evaluate DNA damage, a total of 100 individual cells per sample were used to calculate the tail DNA percentage. The photographs were analyzed using Comet Score 2.0.0.38 software (TriTek Corp.).

Histological stains and immunohistochemistry

Tissue specimens were embedded in paraffin after fixing in 4% paraformaldehyde. Further, thin sections of tissue embedded in paraffin were cut and stained with hematoxylin and eosin (HE) or used for the immunohistochemistry (IHC) analysis. IHC assay was performed according to the manufacturer’s instructions (Novolink Polymer Detection Systems, Leica, Germany). Briefly, tissue specimens were de-paraffinized in xylene, and rehydrated in serial ethanol solutions. Antigen retrieval was achieved by incubation in antigen retrieval buffer (pH 6.0) in a pressure cooker for 10 min at high pressure. The first antibodies RRM2 (Sigma-Aldrich, cat# HPA056994), Cleaved Caspase 3 (Cell Signaling Technology, cat# 9661) were incubated overnight at 4 °C. Then the sections were incubated in the appropriate secondary antibody and processed with DAB Chromogen before staining with hematoxylin. Normal brain control tissue slides (HuFPT017), and appropriate positive control (colon tissue - HuFPT036) and negative controls (liver tissue - HuFPT074) from US Biomax, Inc (MD, USA) were included in the IHC staining. Images were captured and analyzed using Motic DSAssistance 4 K (Motic).

RNA-sequencing of cell lines treated with COH29

BT12 and Re1P6 cell were treated with COH29 (16 µM and 14 µM respectively), then total RNA of indicated samples were isolated using the Trizol Reagent (Invitrogen, MA, USA) and quantified in a NanoDrop (ThermoFisher Scientific, USA). The samples were sent to the Biotools Microbiome Research Center (Taiwan) for library preparation and sequencing.

Orthotopic ATRT xenograft model

The orthotopic ATRT xenograft model was performed according to previously described procedures [30]. This research has been approved by TMU Ethic Committee for ATRT orthotopic xenograft mice model (LAC-2020-0219 and LAC-2021-0517). Tumor formation confirmed by MRI was performed at day 20 after injection (pre-treatment). From day 21, the mice with tumor formation confirmed by MRI were randomly divided into control group (treated with Kolliphor HS15 in saline; n = 8) and COH29 treatment group (treated with COH29 (400 mg/kg) by oral gavage every day in 3 weeks; n = 8). The post-treatment brain MRI was obtained at day 28. Mice were monitored the irreversible neurological deficits and body weight daily. Tumor volumes were calculated based on post-contrast T1-weighted sequences using Image J. All animal care and experimental studies were performed according to the guidelines and approval of the TMU Animal Center.

Gene expression profiles and clinical data analysis

Data for clinical cohorts published by other studies were analyzed on R2 Genomics Analysis and Visualization Platform (https://hgserver1.amc.nl), and PedcBioPortal for Integrated Childhood Cancer Genomic (https://pedcbioportal.org). Genetic dependencies and gene expression of cell lines were analyzed by Depmap Portal tool (https://depmap.org). ATRT patients data and ATRT-PDX data have been described in our previous study [30]. RNA-seq and clinical data were analyzed in Biotools RNA-seq v1.6.4 and in R environment. Differentially expressed genes threshold (FDR < 0.05) was used as a cutoff for differential expression assessment. The RNA-Seq data of the 28 patients are available in the Gene Expression Omnibus (GSE218948).

Statistical analysis

All experiments were performed in triplicate and analyzed using GraphPad Prism version 8.0.1 software. Data were expressed as mean ± standard deviation (SD) or standard error of the mean (SEM), and a p value ≤ 0.05 was considered to be statistically significant. Statistical parameters were described in the figures and figure legends.