DNA demethylating agents suppress preclinical models of synovial sarcoma
Nobuhiko Hasegawa, Nezha S. Benabdallah, Kyllie Smith-Fry, Li Li, Sarah McCollum, Jinxiu Li, Caelen A. Jones, Lena Wagner, Vineet Dalal, Viola Golde, Anastasija Pejkovska, Lara Carroll, Malay Haldar, Seth M. Pollack, Scott W. Lowe, Torsten O. Nielsen, Ana Banito, Kevin B. Jones
Nobuhiko Hasegawa, Nezha S. Benabdallah, Kyllie Smith-Fry, Li Li, Sarah McCollum, Jinxiu Li, Caelen A. Jones, Lena Wagner, Vineet Dalal, Viola Golde, Anastasija Pejkovska, Lara Carroll, Malay Haldar, Seth M. Pollack, Scott W. Lowe, Torsten O. Nielsen, Ana Banito, Kevin B. Jones
View: Text | PDF
Research Article Genetics Oncology

DNA demethylating agents suppress preclinical models of synovial sarcoma

Abstract

Synovial sarcoma is an aggressive soft-tissue cancer driven by the chimeric SS18::SSX fusion oncoprotein, which disrupts chromatin remodeling by combining two antagonistic transcriptional regulators. SS18 participates in BAF complexes that open chromatin, while the SSX genes are cancer-testis antigens that interface with chromatin decorated with monoubiquitinated histone H2A placed by polycomb repressive complex activity. Because KDM2B brings polycomb repressive complex to unmethylated CpG islands, it is plausible that methylation directly determines the distribution of SS18::SSX to target loci. Given that synovial sarcoma is also characterized by a peculiarly low DNA hypomethylation profile, we hypothesized that further disturbance of DNA methylation would have a negative impact on synovial sarcoma growth. DNMT1 disruption by CRISPR/Cas9 targeting or pharmacological inhibition with cytidine analogs 5-aza-2′-deoxycytidine (decitabine) and 5-azacytidine led to decreased genome-wide methylation, redistribution of SS18::SSX, and altered gene expression profiles, most prominently including upregulation of tumor suppressor genes, immune-related genes, and mesenchymal differentiation-related genes. These drugs suppressed growth of synovial sarcoma cell lines and drove cytoreduction in mouse genetic models. DNMT1 inhibitors, already approved for treating myelodysplastic syndromes, warrant further clinical investigation for synovial sarcoma as repurposed, targeted treatments exploiting a vulnerability in the intrinsic biology of this cancer.

Authors

Nobuhiko Hasegawa, Nezha S. Benabdallah, Kyllie Smith-Fry, Li Li, Sarah McCollum, Jinxiu Li, Caelen A. Jones, Lena Wagner, Vineet Dalal, Viola Golde, Anastasija Pejkovska, Lara Carroll, Malay Haldar, Seth M. Pollack, Scott W. Lowe, Torsten O. Nielsen, Ana Banito, Kevin B. Jones

×

Figure 7

KDM2B redistributes upon DNMT inhibition in synovial sarcoma.

Options: View larger image (or click on image) Download as PowerPoint
KDM2B redistributes upon DNMT inhibition in synovial sarcoma. (A) Schema...
(A) Schematic diagram illustrating the possibility of SS18::SSX redistribution with increased regions of hypomethylation after DNMT inhibitor treatment. (B) Left, heatmaps of the log2-transformed fold change KDM2B or SS18-SSX (endogenously HA tagged) calibrated ChIP signals in HS-SY-II cells treated with DMSO or 500 nM decitabine over the total SS18-SSX peaks (DMSO and decitabine, n = 24,444). Rows correspond to ±10 kb regions across the midpoint of each enriched region, ranked by increasing signal. Right, heatmaps for SS18::SSX1 calibrated ChIP-Seq over decreased SS18-SSX peaks (n = 3,699) or gained peaks (n = 4,689). Rows correspond to ±10 kb regions across the midpoint of each enriched region, ranked by increasing signal. (C) Enrichment plots for mouse tumor ChIP-Seq for KDM2B and SS18::SSX across the transcription start sites of previously confirmed target genes of the fusion after vehicle, decitabine, or 5-AZA treatment. (D) ChIP-Seq track at 2 (rare) example target genes that showed the pattern of diminished KDM2B and fusion ChIP-Seq, as well as RNA-Seq.