Growth Inhibition and Induction of Innate Immune Signaling of Chondrosarcomas with Epigenetic Inhibitors

Abstract

Chondrosarcomas are inherently resistant to chemotherapy and radiotherapy, pointing to an unmet need for new treatment options. Immune checkpoint inhibitors, which have shown remarkable promise in multiple solid cancer types, have limited efficacy in chondrosarcomas. Mutations in IDH1/2 genes, which result in progressive increases in DNA and histone methylation, are observed in 50% of conventional chondrosarcomas, suggesting that epigenetic dysregulation represents a potential barrier for tumor progression and target for therapeutic intervention. Here, we demonstrated that combined treatment of FDA-approved inhibitors of DNA methyltransferases (DNMTs) 5-aza-2'-deoxycytidine (5-aza), and histone deacetylases (HDACs) suberanilohydroxamic acid (SAHA) impaired the proliferation of chondrosarcoma cell lines in vitro and in xenograft studies. Transcriptomic analysis reveals that chondrosarcoma cells treated with 5-aza and SAHA markedly elevated the expression of IFN-stimulated genes including PD-L1, indicating that these epigenetic drugs induced a potent innate immune response. We demonstrated that 5-aza and SAHA resulted in both genomic and epigenomic instability, as shown by elevated DNA damage response and derepression of retrotransposons, respectively, which in turn activated pattern recognition receptors (PRRs) and the downstream IFN signaling pathways. Importantly, the cytotoxic effects of 5-aza and SAHA can be rescued by depletion of PRRs such as cGAS and MAVS, and potentiated by depletion of the RNA-editing enzyme ADAR1. Together, our results demonstrate preclinical activity of combined DNMT and HDAC inhibition against chondrosarcomas and suggest that targeted epigenetic therapies could represent a new therapeutic approach in the treatment of chondrosarcomas, and this is being tested in an ongoing clinical trial (NCT04340843).

Conflict of interest statement

Conflict of Interest Disclosure: The authors declare no conflict of interest.

©2021 American Association for Cancer Research.

Figures

Fig. 1:. Combined DNMT and HDAC inhibition impairs CS proliferation and induces cell death in vitro.

(A) Human CS cell lines (CH2879, JJ012, CS1, and SW1353) were treated with DMSO (no drug, ND), 5-aza 250 nM daily, SAHA 500 nM daily or the combination (A+S) for 5 days and cell viability was assessed using Cell Counting Kit-8 (CCK-8) assay. **, P<0.01; ***, P<0.005; ****, P<0.001. (B) CS cell lines were treated with DMSO (no drug, ND), 5-aza 250 nM, SAHA 500 nM, or the combination for 48hr and lysates were collected for immunoblotting with indicated antibodies with GAPDH as loading control. (C) Cell cycle analysis by flow cytometry was performed on CS cell lines treated with DMSO (no drug, ND), 5-aza 250 nM daily, SAHA 500 nM daily or the combination (A+S) for 48hr. The percentage of cells in each cell cycle stage, including subG1 that contains apoptotic cells, is shown as bar graphs.

Fig. 2:. Combined DNMT and HDAC inhibition inhibits CS tumor growth in vivo.

(A) Tumor growth of JJ012 mouse xenografts treated with indicated drugs is shown (IP injection daily, 5 mice per group). *, P<0.05; ***, P<0.005. (B) At day 90, tumors harvested from the xenograft studies were grinded, lysed and subjected to immunoblotting with indicated antibodies with GAPDH as loading control.

Fig. 3:. Upregulated expression of interferon-stimulated genes upon HDAC and DNMT inhibition.

(A) Top enriched MSigDB Hallmark pathways from gene set enrichment analysis of top 500 upregulated and downregulated genes after 5-aza and SAHA treatment. *, P<0.05, **, P<0.01, ***, P<0.001. (B) Left: violin plots showing the standardized expression levels (z-scores) of 38 ISGs in CS1 cells treated with DMSO (no drug, ND) or 5-aza 250 nM and SAHA 500 nM for 48hr, measured by RNA-seq. Right: heatmap showing standardized expression (z-score) of each of the 38 ISGs in CS1 cells treated with ND or 5-aza 250 nM and SAHA 500 nM. (C) CS1 and JJ012 cell lines were treated with DMSO (no drug, ND) or 5-aza 250 nM and SAHA 500 nM for 48 hr and expression of select ISGs was measured with RT-qPCR and normalized to untreated controls. **, P<0.01; ***, P<0.005; ****, P<0.001.

Fig. 4:. Combined DNMT and HDAC inhibition induces genomic and epigenomic instability and activation of interferon signaling pathways.

(A) A schematic illustration of interferon response pathway following DNA damage and retrotransposon activation. (B) Left: violin plots showing the standardized expression levels (z-scores) of retrotransposons in CS1 cells treated with DMSO (no drug, ND) or 5-aza 250 nM and SAHA 500 nM for 48hr, measured by RNA-seq. Right: table showing fold change (Aza+SAHA over ND) of expression of various retrotransposon families. *, P<0.05; ***, P<0.005; ns, not significant. (C) CS cell lines (CH2879, JJ012, CS1, and SW1353) were treated with DMSO (no drug, ND), 5-aza 250 nM, SAHA 500 nM, or the combination for 48hr and lysates were collected for immunoblotting with indicated antibodies with GAPDH as loading control. (D) JJ012 and CS1 cells were treated with DMSO (no drug, ND), 5-aza 250 nM, SAHA 500 nM, or the combination for 48hr and lysates were collected for immunoblotting with indicated antibodies with GAPDH as loading control.

Fig. 5:. Depletion of PRRs rescues the cytotoxicity of HDAC and DNMT inhibition to CS cells.

(A and B) CS1 cells were transfected with 100 nM siRNA against MAVS for 48hr and re-seeded and treated with DMSO (no drug, ND), 5-aza 250 nM, SAHA 500 nM, or the combination (A+S). Cell viability (A) was assessed 6 days later and lysates were collected 48hr later for immunoblotting with indicated antibodies with GAPDH as loading control (B). ***, P<0.005; ****, P<0.001. (C and D) CS1 cells were transfected with 100 nM siRNA against cGAS for 48hr and re-seeded and treated with DMSO (no drug, ND), 5-aza 250 nM, SAHA 500 nM, or the combination (A+S). Cell viability (C) was assessed 6 days later and lysates were collected 48hr later for immunoblotting with indicated antibodies with GAPDH as loading control (D). ***, P<0.005; ****, P<0.001. (E) CS1 cells were transfected with 100 nM siRNA against ADAR for 48hr and re-seeded and treated with DMSO (no drug, ND), 5-aza 250 nM, SAHA 500 nM, or the combination (A+S). Cell viability was assessed 6 days later and shown as bar graphs. Insert shows a western blot demonstrating reduced ADAR expression after siRNA treatment. ***, P<0.005.