Genetic and epigenetic characterization of sarcoma stem cells across subtypes identifies EZH2 as a therapeutic target - PubMed (original) (raw)

Genetic and epigenetic characterization of sarcoma stem cells across subtypes identifies EZH2 as a therapeutic target

Edmond O'Donnell 3rd et al. NPJ Precis Oncol. 2025.

Abstract

High-grade soft tissue sarcomas (STS) are a heterogeneous and aggressive set of cancers. Failure to respond anthracycline chemotherapy, standard first-line treatment, is associated with poor outcomes. We investigated the contribution of STS cancer stem cells (STS-CSCs) to doxorubicin resistance. We identified a positive correlation between CSC abundance and doxorubicin IC50. Utilizing patient-derived samples from five sarcoma subtypes we investigated if a common genetic signature across STS-CSCs could be targeted. We identified Enhancer of Zeste homolog 2 (EZH2), a member of the polycomb repressive complex 2 (PRC2) responsible for H3K27 methylation as being enriched in CSCs. EZH2 activity and a shared epigenetic profile was observed across subtypes and targeting of EZH2 ablated the STS-CSC population. Treatment of doxorubicin-resistant cell lines with tazemetostat resulted in a decrease in the STS-CSC population. These data confirm the presence of shared genetic programs across distinct subtypes of CSC-STS that can be therapeutically targeted.

© 2025. The Author(s).

PubMed Disclaimer

Conflict of interest statement

Competing interests: The authors declare no competing interests.

Figures

Fig. 1

Fig. 1. Acquired doxorubicin resistance correlates with increased soft tissue sarcoma cancer stem cell abundance and phenotype.

A A doxorubicin-resistant (doxoR) derivative line of WT GCT (undifferentiated pleomorphic sarcoma) cells was generated through serial passage in increasing concentrations of doxorubicin. H&E stains of representative subcutaneous tumors generated from WT and DoxoR SW872 cells. B H&E stains of representative subcutaneous tumors generated from WT and DoxoR cells SW872 cells (dedifferentiated liposarcoma). C Cell-titer glo viability assays for doxorubicin dose-response in WT and doxoR derivative GCT cells. Data points are the mean ± SD of triplicate wells and are representative of three different experiments. D A doxorubicin-resistant (doxoR) derivative line of WT SW872 cells was generated through serial passage in doxorubicin and exhibited an increased IC50. Data points are the mean ± SD of triplicate wells and are representative of three different experiments. E Representative photographs of SW872 WT and DoxoR cells after 48 hours of treatment with either DMSO (vehicle) or 300 nM doxorubicin. bars = 250 nm. F WT and DoxoR GCT cells were subjected to soft-colony formation assays and the abundance and size of colonies was counted with imageJ software. Each column is a biological replicate. Colony size is shown in units as pixels (px2). G Aldefluor assay showing increased bright population in doxoR (right) cells vs WT GCT cells (left). The DEAB controls are shown in red (n = 3). H WT and DoxoR SW872 cells were subjected to soft-colony formation assays and the abundance and size of colonies was counted with imageJ software. Colony size is shown in arbitrary units as pixels2. I Aldefluor assay showing increased bright population in doxoR (right) cells vs WT SW872 cells (left) (n = 3). *P < 0.05, **P < 0.05.

Fig. 2

Fig. 2. Isolation and RNA-sequencing-based analysis of soft-tissue sarcoma cancer stem cells with Aldefluor sorting reveals a shared gene signature.

A Sorting scheme of Aldefluor bright and dim sorted cells among five soft tissue sarcoma cell lines. A representative soft-agar colony forming assay from sorted GCT cells shows the increased propensity for anchorage-independent growth in Aldefluor bright cells. B Table showing the percent Aldefluor positive population among the five different cell lines. LMS=leiomyosarcoma (SKLMS1), RH=rhabdomyosarcoma, MFS=myxofibrosarcoma, UPS=undifferentiated pleomorphic sarcoma (GCT), LPS=dedifferentiated liposarcoma (SW872). B Percent Aldefluor positive (bright) cells for each of the different lines tested. C Venn diagram showing the overlap of differentially expressed genes meeting a significance of p < 0.05. 36 genes were upregulated and 1 gene (BRCA1) was downregulated. D Table of the commonly upregulated genes shown in panel C subdivided according to fold change in expression. E Gene ontology analysis of the 37 shared, differentially regulated genes across the five sarcoma types.

Fig. 3

Fig. 3. Identification of a shared EZH2 signature among soft tissue sarcoma cancer stem cells.

A Expression data from RNA-Seq was analyzed by gene set enrichment analysis, and enrichment plot of one of the top-scoring hits is shown along with generative heatmap. B Survival analysis (oncolnc.org) showing overall survival among available sarcoma cases within the TCGA database, low expression represents the lowest quartile, while highest expression represents a composite of the three higher quartiles, Log-rank P = 0.05. C Circos plot of genome wide ATAC-seq data generated from sorted soft-tissue sarcoma bright and dim cells. More ATAC-seq signal (accessible chromatin) in bright vs dim cells directs outward, while decreased chromatin access in bright vs dim cells directs inward. The center plot shows an enlargement of the track data from chromosome 1. D ATAC-seq data from the promoter region of CDK1, one of the differentially identified genes by RNA-seq. The window region is 50 kb. E Increased chromatin accessibility at transcription start sites (TSS) correlating with the molecular target of EZH2. The plot on the right shows ATAC-seq signal as a ratio of bright to dim in WT GCT (UPS) cells was plotted for all canonical TSS in a 4 kb window, and ordered from low (top) to high (bottom). The plot on the left shows H3K27-me3 Chip-Seq data from a publicly available HepG2 dataset plotted according to the same rank-order as the right plot. Increasing black represents more Chip-Seq signal.

Fig. 4

Fig. 4. Inhibition of EZH2 with the small molecular inhibitor Tazemetostat abrogates the cancer stem cell phenotype.

A Tazemetostat does not significantly inhibit cell viability during short term treatment. Data are the mean ± SD of three biological replicates, and are representative of three independent experiments. There is no statistically significant difference between the groups. B Representative soft-agar assay of WT GCT cells treated with vehicle or Tazemetostat (25 uM) (right) and quantification with imageJ (left). C Treatment schema of WT GCT cells subjected to long-term culture with vehicle (A, B) or Tazemetostat for continuous exposure (C, E) or on-off drug exposure (D, F). D Aldefluor assay of WT GCT cells treated with vehicle or Tazemetostat over a 3 week period. DEAB controls for each experimental arm are shown in the top row, for which the corresponding gates are directly maintained in the bottom row. The percentage of Aldefluor bright cells relative to DEAB controls is shown in brackets. E Continuous treatment of GCT cells with either Tazemetostat or two additional EZH2 inhibitors EI1 and PF-06726304 decreases the abundance of bright cells in the Aldefluor assay. *p < 0.001 for all three treatment groups compared to control. F Western blot of H3K27me3 expression in GCT cells following treatment with vehicle or the EZH2 inhibitors Tazemetostat, EI1 and PF-06726304. Beta actin is shown as a loading control.

Fig. 5

Fig. 5. Co-treatment of soft tissue sarcoma cell lines with Tazemetostat and doxorubicin rescues treatment-acquired chemotherapeutic resistance.

A SW872 (LPS) and derivative doxoR cells treated for 96 hours with doxorubicin and Tazemetostat alone or in combination. B GCT (UPS) and derivative doxoR cells treated for 96 hours with doxorubicin and Tazemetostat alone or in combination. Data for A, B are the mean ± SD of three biological replicates analyzed by ANOVA, * p < 0.05, *** p < 0.001 **** p < 0.0001. C Viability assays of unique, independently generated soft tissue sarcoma cell lines from oncogene-mediated forward transformation of mesenchymal stem cells in vivo treated with doxorubin at the indicated concentrations alone or in combination with 25 uM Tazemetostat. Data for A-C are the mean ± SD of three biological replicates analyzed by ANOVA, * p < 0.05, *** p < 0.001 **** p < 0.0001.

Update of

References

    1. WHO Classification of Tumours Editorial Board (2020). Soft Tissue and Bone Tumours WHO Classification of Tumours 5th Edition, Volume 3.
    1. Bovée, J. V. M. G. & Hogendoorn, P. C. W. Molecular pathology of sarcomas: concepts and clinical implications. Virchows Arch.456, 193–199 (2010). -DOI -PMC -PubMed
    1. Nacev, B. A. et al. The epigenomics of sarcoma. Nat. Rev. Cancer20, 608–623 (2020). -DOI -PMC -PubMed
    1. Helman, L. J. & Meltzer, P. Mechanisms of sarcoma development. Nat. Rev. Cancer3, 685–694 (2003). -DOI -PubMed
    1. Valent, P. et al. Cancer stem cell definitions and terminology: the devil is in the details. Nat. Rev. Cancer12, 767–775 (2012). -DOI -PubMed

Grants and funding

LinkOut - more resources