The patient was a 22-year-old Japanese man that presented with a painful tumor measuring 5.3 cm that was located in the left perineum. Magnetic resonance imaging (MRI) and positron emission tomography were performed suspecting a malignant soft tissue tumor. Computed tomography revealed a non-metastatic lesion ( Figures 1A–C ). A needle biopsy was performed and revealed a distal type epithelioid sarcoma. The epithelioid cells with acidic vesicles and eccentric nucleus showed infiltrative growth ( Figure 1D ). Immunoreactivity for cytokeratin (AE1/AE3), vimentin, and CD34 was positive but integrase interactor 1 (INI-1) was negative for the tumor cells ( Figures 1E, F ). Wide resection of the bladder, prostate, penis, and rectum was performed, and creation of a stoma and urinary diversion were performed. After adjuvant radiotherapy with 66Gy/33fr, 3 cycles of adjuvant chemotherapy consisting of doxorubicin (60 mg/m²) and ifosfamide (10 g/m²) were administrated. The patient was in continuous disease-free (CDF) for more than 1 year.

The patient was a 50-year-old Japanese man that presented with a painful tumor measuring 11 cm that was located in the right proximal thigh. Magnetic resonance imaging (MRI) was performed suspecting a malignant soft tissue tumor. Computed tomography revealed a multiple subcutaneous metastatic lesions ( Supplementary Figure 3A ). A open biopsy was performed and revealed a distal type epithelioid sarcoma. The polygonal to spindle shaped cells with giant and polymorphic nucleus showed infiltrative growth with wide lesion of inflammation and necrosis ( Supplementary Figure 4A left). Immunoreactivity for pan-keratin, vimentin was positive and CD34 was faintly stained in tumor cells. But integrase interactor 1 (INI-1) was negative for the tumor cells ( Supplementary Figures 4B, C ). Wide resection of the tumor and proximal femur was performed, after that proximal femoral replacement was performed. After the surgery we planed adjuvant chemotherapy, but the patient did not want chemotherapy and chose the best supportive care. The patient was in dead of disease (DOD) 2 months after surgery.

The basal medium consisted of Advanced DMEM/F12 (Thermo fisher Scientific) supplemented with HEPES (10 mM, Thermo fisher Scientific), GlutaMAX (1x, Thermo fisher Scientific), and Penicillin-Streptomycin-Glutamine (1x, Thermo fisher Scientific).

The organoid culture medium, which was used for the organoid culture of human prostate cancer, consisted of the basal medium supplemented with nicotinamide (10 mM, Sigma), N-acetylcysteine (1 mM, Wako), A-83-01 (500 nM, Wako), B-27 (1x, Gibco), recombinant human EGF (50 ng/mL, Sigma), gastrin I human (10 nM, Peprotech), recombinant human Noggin (100 ng/mL, Peprotech), recombinant human R-spondin 1 protein (500 ng/mL, R&D), SB-202190 (10 μM, Chemscene), and 10% Afamin/Wnt3a CM (MBL) ( Table S1 ).

We prepared the collagen gel matrix (Cellmatrix type I-A, Nitta Gelatin Inc.) according to the protocol of organoid culture procedure as previously described (26, 27). In detail, on ice, we mixed Cellmatrix I-A with 10 X concentrated sterile culture medium (Ham’s F-12) and sterile reconstitution buffer (2.2g NaHCO3 in 100ml of 0.05N NaOH and 200mM HEPES) at a ration of 8:1:1. During the mixing, the creation of bubbles was avoided. The mixed solution was kept on ice (at 4°C) to prevent gel formation.

For the preparation of the culture dish, Millicell culture plate inserts (PICM03050, Millicell-CM, Millipore), with permeable and membranous bottoms, were placed in a 60 mm tissue culture dish. To creat the bottom layer, 1 ml of the prepared reconstituted collagen solution was added to the inserts under sterile conditions. The culture dish with the inserts was incubated until solidification in an incubator at 37°C for 30 min ( Supplementary Figures 1A, B ). During the incubation, tissue preparation was performed as described below.

The remaining surgically resected tissue was used for the experiments following the guidelines of the Institutional Review Board for Clinical Research after obtaining the patient informed consent. After measuring the weight of the tumor tissues with scales, the tumor tissues were washed four times with iced phosphate-buffered saline (PBS) in 50 ml conical tubes by tipping over. After that, they were minced with sterilized scissors and scalpel on the 60 mm dish for 15 minutes at room temperature. The minced tumor tissues were collected with 3 ml of Dulbecco’s Modified Eagle’s Medium (DMEM; GIBCO, Grand Island, NY, USA) using a 5-ml pipette into a 10 ml conical tube and centrifuged at 600g for 5 min. The cell suspensions were incubated with 50 μg/ml of Liberase TH (Sigma-Aldrich, St. Louis, MO, USA) for 15 min at 37°C. The supernatant was discarded, and the cells were incubated in fetal bovine serum (FBS; Sigma) at 37°C for 15min to inactivate Liberase TH. The cells were suspended into 30 ml DMEM and passed through a 1 mm nylon mesh (Becton Dickinson Falcon, Franklin Lakes, NJ, USA) to remove the debris when needed. After four times washing with basal medium, the prepared cells were suspended into the upper layer of the prepared collagen and 1.5 ml of organoid culture medium were added into the culture dish on the outside insert ( Supplementary Figures 1A, B ). The organoids were cultured at 37°C with 95% air, 5% CO2, and 100% humidity. The organoid culture medium was changed twice a week.

The organoids were passaged after two to four weeks after the organoids culture. The upper and bottom layers were transferred into another 60 mm culture dish with tweezers and the organoids with the layers of collagen were minced with sterilized scissors and scalpel on the dish for 15 minutes at room temperature. After incubation with 50 μg/ml of Liberase TH for 15 min at 37°C, the cells were centrifuged at 600g for 5 min and supernatant was discarded, and they were incubated in FBS at 37°C for 15min to inactivate Liberase TH. After two times washing with basal medium, the prepared cells were cultured in the upper layer of the prepared collagen at 37°C with 95% air, 5% CO2, and 100% humidity.

The minced organoids obtained during the passage were frozen in Stem Cell Banker (Amsbio ZENOAQ, 11890) and stored in -80°C freezer. During reculture of the preserved organoids, the cryotube was warmed in water at 37°C for 5 min. After washing twice with the basal medium, the prepared cells were cultured in the upper layer of the prepared collagen at 37°C with 95% air, 5% CO2, and 100% humidity.

EPS tumor tissues were prepared using the above method, and the tumor cells were cultured with ALI organoid culture. After several serial passages, the cultured cells were xenografted in a NOD-scid IL2Rgnull (NSG) mouse. The developed tumor was resected from the mouse, and the same steps of tissue preparation were performed as described above. The ALI organoid culture was repeated for the serial passages and xenograft in NSG mice. When the tumor developed, the PDO was established from EPS ( Figure 2A ).

The total RNA was extracted using RNeasy Plus Universal Mini Kit (Qiagen) according to the manufacturer’s protocol, and the determined purity was A260 to A280. For cDNA synthesis, the Superscript IV VILO (Invitrogen) was used according to the manufacturer’s instructions. RT-PCR was performed with PrimeSTAR MAX DNA polymerase (Takara). The primers, designed according to Ref. 32 for INI1, are listed in the supplementary material ( Table S2 ).

The total DNA was extracted using DNeasy Blood & Tissue Kit (Qiagen) according to the manufacturer’s protocol.

EWS-FLI1 cDNA or DNA was identified by polymerase chain reaction (PCR) using a primer set of EWS-FLI1 or genomicEWS-FLI1 ( Table S2 ). For sequence analysis, the reverse-transcriptase (RT) PCR-amplified EWS-FLI1 cDNA or genomic DNA fragments were analyzed on 1.5% agarose gels, purified using a Qiagen gel extraction kit (Qiagen, Hilden, Germany), and directly sequenced in Eurofins Genomics with forward or reverse primers mentioned above. BLAST software (http://blast.ncbi.nlm.nih.gov/Blast.cgi) was used for computer analysis of sequence data.

Immunohistochemical studies were performed to clarify whether the phenotype of the cultured organoids and developed tumors matched with that of the original tumors. The specimens of the tumors and organoids were fixed in 10% neutral, buffered with formalin, embedded in paraffin, sectioned at 4 μm thickness, and stained with hematoxylin and eosin. The paraffin embedded sections were deparaffinized and dehydrated. The antigens were retrieved at 95°C for 15 min in PH6.0 citric acid buffer. After blocking the endogenous peroxidase activity for 15 min with VECTASTAIN Elite ABC Universal PLUS Kit Peroxidase (VECTOR PK-8200), the sections were reacted for 1 h with 2.5% Normal Horse Serum at room temperature to prevent nonspecific binding. They were then incubated overnight with primary antibody in a blocking solution at 4°C. After six subsequent PBS washings, the appropriate prediluted biotinylated secondary antibodies (1:500, BD Biosciences) were added for 30 min. The sections were incubated for 30 minutes with VECTASTAIN Elite ABC Reagent, following by DAB staining (ImmPACT DAB EqV solutions) until the appropriate stain intensity develops (1-2 min). The slides were then washed and counterstained with Gill No. 3 hematoxylin (Sigma) for 45 s. As a negative control, staining was carried out in the absence of primary antibodies. The histological analyses were checked by a pathologist in our institute. The primary antibodies against Pan-Keratin (C11) (1:500, Cell Signaling Technology, USA), Vimentin (D21H3) Xp (1:200, Cell Signaling Technology, USA), CD34 (B-6) sc-74499 (1:100, Santa Cruz), INI-1 (D8M1X) (1:1000, Cell Signaling Technology, USA), and PCNA (D3H8P) XP (1:5000, Cell Signaling Technology, USA) were used.

Cultured EPS organoids or the developed tumor in the mouse were used for the proliferation assay. Cell suspension was prepared according to the organoid passage or tissue preparation procedures. The collagen gel matrix was prepared according to the protocol of the organoid culture procedure. We scaled down the experiment because the prepared cell number and repetition of the experiments ( Figures 6B, C ) were limited. For the preparation of the culture dish, Millicell culture plate inserts (PICM 01250, Millicell-CM, Millipore) were placed in a culture plate with 24 wells (Thermo). In order to form the bottom layer, 150 μl of the prepared reconstituted collagen solution was added to the inserts under sterile conditions. The culture plate with the inserts was incubated until solidification in an incubator at 37°C for 30 min. During the incubation, the steps of tissue preparation described above were performed. The prepared cells were counted and cultured in the upper layer of the prepared 150 μl of collagen with 200 μl of organoid culture medium at 37°C with 95% air, 5% CO2, and 100% humidity for 72h.

Cell counting Kit-8 assays were performed to measure the cell viability. We measured the viability of the cells at 0 h and 72 h after the start of organoid culture. The ratio of cell viability was calculated.

3-(4,5-Dimethyl-2-thiazolyl)-2,5-diphenyltetrazolium Bromide (MTT) assays were used to detect live organoids 72 hours after treatments with doxorubicin (Wako) at concentrations of 0.015–15M or paclitaxel (Wako) at concentrations of 0.04-4μM. We detected the live organoids by adding 100 μl of 5 mg/ml MTT solution in PBS and incubated for 4 h. Only live organoids could be detected as blue spots and counted the spot areas with ImageJ according to the manufacturer’s protocol.

The lentiviral packaging mixture (pLP1, pLP2, and pLP/VSVG) were purchased from Invitrogen. The GFP expression vector, pLentiCMV GFP Puro (658–5), and lentivirus packaging mixture were co-transfected into 293FT cells using PEI-MAX instead of Lipofectamine 2000 according to Invitrogen’s instruction. The lentivirus containing supernatants were collected 48h after transfection and used to infect the target cells. The target OICI-EPS-0530 cells were cultured in an adherent culture dish prepared from the organoids. Furthermore, 24h after infection, the culture medium was replaced with fresh organoid culture medium. The cells were then cultured with ALI organoid culture method and applied to each assay.

All the animal experiments were approved by the institutional committee of animal care of the Osaka International Cancer Institute, and all the animals were euthanized with diethyl ether at the end of the experiments. The 6-week-old NSG mice were purchased from Japan SLC (Shizuoka, Japan). The upper and bottom layers were transferred into another 60 mm culture dish with tweezers and incubated the organoids with 50 μg/ml of Liberase TH for 15 min at 37°C. The cells were centrifuged at 600g for 5 min and the supernatant was discarded. The organoids were suspended into 50 μl PBS. After mixing with 50 μl of growth factor reduced Matrigel (BD Biosciences), the mixed solution was kept on ice (at 4°C) until xenograft. Small incisions on both sides of the back in the NSG mice were made, and the prepared organoid suspension were injected using a syringe without a needle. The incisions were stiched thereafter. The mice were inspected once a week and were sacrificed when the total tumor burden reached 1-1.5 cm³. The tumors were then resected and used for organoid culture, RNA collections, and immunohistochemical studies.

The total RNA of samples were sent to Seibutsu Giken Inc (Kanagawa, Japan) for next generation sequencing (NGS). The concentrations of the RNA samples were measured using the Quantus Fluorometer and Quanti Fluor RNA system (Promega). Quality check of the RNA was performed with 5200 Fragment Analyzer System and Agilent HS RNA Kit (Agilent Technologies). The NGS library was prepared by MGIEasy RNA Directional Library Prep Set (MGI Tech Co Ltd) according to the manufacturer’s protocol. The concentrations of the library were measured with Synergy LX and Quanti Fluor dsDNA System. Quality check of the library was performed with Fragment Analyzer and dsDNA915ReagentKit (Advanced Analytical Technologies). A circular DNA was made from the library by MGIEasy Circularization Kit (MGI Tech Co Ltd). DNA Nanoball (DNB) was made with DNBSEQG 400 RS High throughput Sequencing Kit (MGI Tech Co Ltd) according to the manufacturer’s protocol. Sequencing was performed with DNBSEQ-G400 for at least 2 × 100 bp read. The adapters and primers from the delivered fastq files were removed using cutadapt (ver. 1.9.1). Short read sequences under 20 and low-quality score reads (under 40) were removed with a sickle (ver. 1.33). HISAT2 (ver. 2.2.1) was used to read the alignment and mapping of the GRCh38 reference genomes. After reading and writing the alignment data in the SAM and BAM formats with the SAMtools, the data were sorted and indexed. The read counts per gene were obtained using the featureCounts (ver. 2.0.0). The relative expression level of each gene was normalized by the RPKM (Reads per Kilobase Million) and Transcripts Per Million methods. We analyzed and visualized the genetic characteristics and transcriptome profiling with iDEP and RaNAseq (33, 34). The presence of fusion gene was analyzed by Seibutsu Giken Inc.

Statistical significance was set at P < 0.05. Statistical analyses were performed using the Student’s t test. Ethical approval for this study was obtained from the institutional review board of our institute.

We cultured the EPS cells using the ALI organoid culture method and labeled the organoid lines as OICI-EPS-0530 and OICI-EPS-0486. The growth with time was stable in the serial passages. The organoid morphology was distorted into a spherical shape and small clumps of cells, and disperse cells were observed ( Figure 2B and Supplementary Figure 3B ). The size of the OICI-EPS-0530 and OICI-EPS-0486 organoid increased in a short period and their viability increased by approximately 3.3-fold and 2.3-fold in 72 h, respectively ( Figures 2C, D and Supplementary Figures 3B, C ). In addition, the organoids were easily broken during the pipetting phase of the passage procedures, implying that the cell-cell adhesion in the organoids was not strong.

Thirty-six days after xenograft of the OICI-EPS-0530 organoids (PDO1) into a NSG mouse, the tumor developed (ODX1) ( Figure 2E ). The tumors were then resected, minced according to the tissue preparations, and cultured using the ALI organoid methods (PDO2). These replated organoids showed similar morphological features to the original organoids from the patient tumor tissues. Fifty-three days after xenograft of the re-cultured organoids into 3 NSG mice, the tumors developed in all the xenografted mice (ODX2) ( Figure 2E ). This means that the OICI-EPS-0530 had high tumorigenesis. The cells prepared from the ODX2 were able to grow in the ALI organoid cultures. In the same manner, 27 days after xenograft of the OICI-EPS-0486 organoids (PDO) into a NSG mouse, the tumor developed (ODX) ( Supplementary Figure 3D ).With the RT-PCR, no INI-1 expression was detected in the OICI-EPS-0530 tumors similar to OICI-EPS-0486 ( Supplementary Figure 3A ), in contrast to Hela and an Ewing sarcoma cell line, Kamui-EWS, which was used as a positive control ( Figure 2F and Supplementary Figures 2A–E ).

Hematoxylin and eosin staining of the OICI-EPS-0530 xenograft tumors (ODX1 and ODX2) demonstrated a heterogeneous histological appearance resembling the original tumor ( Figure 3A ). They showed a multinodular proliferation of eosinophilic epithelioid and spindle-shaped tumor cells. The immunoreactivity for pan-keratin, vimentin, CD34 and PCNA was positive, but INI-1 was negative in the ODX1 and ODX2 tumors similar to the original EPS tumor tissue ( Figures 3B, C ). In addition, hematoxylin and eosin staining of the OICI-EPS-0530 organoids showed distorted spherical morphology and peripheral cells of the organoids were epithelioid and spindle-shaped tumor cells ( Figure 4A ). The immunoreactivity for pan-keratin, vimentin, CD34 and PCNA was positive, but INI-1 was negative in the organoids similar to the original EPS tumor tissue and xenograft tumors ( Figures 4B, C ). Almost similar results were obtained from OICI-EPS-0486 except for faint staining of CD34 ( Supplementary Figure 4 ). These data demonstrated that the morphological characteristics of the OICI-EPS-0530 and OICI-EPS-0486 were preserved during organoid culture and xenograft.

We conducted RNA-seq of paired tumors and normal samples from patients with EPS and developed the tumors in mice (ODX1 and ODX2). The gene expressions of each sample were visualized in heatmap, scatter plot and principal component analysis ( Figures 5A–C ). The gene expression profiles of the normal tissue did not correlate with those of the original tumors, ODX1 and ODX2 ( Supplementary Figures 3A–C ). The gene profile of the original tumor showed moderate correlations with the profiles of ODX1 and ODX2. The gene profiles between ODX1 and ODX2 strongly correlated. These results suggested that the gene expression signatures of the OICI-EPS-0530 tumors were not completely identical to those of the primary EPS tumor. In contrast, the gene expression signatures were preserved in the PDOs and ODXs. No fusion gene was detected in all RNA samples.

The transcriptome analysis showed that 10116 and 1732 genes were upregulated and downregulated in EPS ODX compared to the normal tissue at fold changes > 2, respectively ( Supplementary Figure 4A and Tables S3, and S4). We performed Kmeans Enrichment analyses for gene functional and ontological characterization ( Supplementary Figure 4B and Tables S5, S6). The gene expressions were clustered in four groups and cluster C was enriched in both the original EPS tumor and ODXs. In the cluster, 225 genes and 15 biological pathways were nominated, including the proteoglycans in the cancer (11 genes), cell cycle (8 genes), P53 signaling pathway (6 genes).

It was essential to determine whether the established EPS organoid line could be used with various experiments. Firstly, we induced GFP with lentivirus infection for OICI-EPS-0530 organoid cells and observed the re-formation of the organoids and expression of GFP in the EPS organoids ( Figure 6A ). This means that the established organoid line could be used for RNA interference (RNAi) experiments. We performed drug treatment experiments, using doxorubicin or paclitaxel for the organoid in vitro, and found that 1.5-15μM of doxorubicin and 0.04-4μM of paclitaxel significantly suppressed the organoid growth of OICI-EPS-0530. In addition, 1.5-15μM of doxorubicin showed similar efficacy for the organid growth and 0.04-4μM of paclitaxel moderately suppressed the organoid growth of OICI-EPS-0486. It demonstrated that OICI-EPS-0530 and OICI-EPS-0486 could be used for drug screening experiments. In addition, OICI-EPS-0530 and OICI-EPS-0486 were resistant to doxorubicin, which was the first-line drug for soft tissue sarcoma, and OICI-EPS-0530, but not OICI-EPS-0486, was sensitive to paclitaxel, which was a widely used taxane anticancer drug, according to the database of drug sensitivity and Genomics of Drug Sensitivity in Cancer ( Figures 6B, C , Supplementary Figures 7A, B and Table S7 ) (35).