BRCA1-associated RING domain-1 (BARD1) loss and GBP1 expression enhance sensitivity to DNA damage in Ewing sarcoma

Abstract

Ewing sarcoma is a fusion oncoprotein-driven primary bone tumor. A subset of patients (~10%) with Ewing sarcoma are known to harbor germline variants in a growing number of genes involved in DNA damage repair. We recently reported our discovery of a germline mutation in the DNA damage repair protein BARD1 (BRCA1-associated RING domain-1) in a patient with Ewing sarcoma. BARD1 is recruited to the site of DNA double stranded breaks via the poly(ADP-ribose) polymerase (PARP) protein and plays a critical role in DNA damage response pathways including homologous recombination. We thus questioned the impact of BARD1 loss on Ewing cell sensitivity to DNA damage and the Ewing sarcoma transcriptome. We demonstrate that PSaRC318 cells, a novel patient-derived cell line harboring a pathogenic BARD1 variant, are sensitive to PARP inhibition and by testing the effect of BARD1 depletion in additional Ewing sarcoma cell lines, we confirm that BARD1 loss enhances cell sensitivity to PARP inhibition plus radiation. Additionally, RNA-seq analysis revealed that loss of BARD1 results in the upregulation of GBP1 (guanylate-binding protein 1), a protein whose expression is associated with variable response to therapy depending on the adult carcinoma subtype examined. Here, we demonstrate that GBP1 contributes to the enhanced sensitivity of BARD1 deficient Ewing cells to DNA damage. Together, our findings demonstrate the impact of loss-of function mutations in DNA damage repair genes, such as BARD1, on Ewing sarcoma treatment response.

Keywords: BARD1; DNA damage; Ewing sarcoma; GBP1; PARPi.

FIGURE 1

GSEA analysis and validation of a primary cell line from a Ewing tumor with a BARD1 pathogenic variant. A, Schematic overview of tumor samples associated with analyses in B–F. B, Gene-set enrichment analysis (GSEA) of RNA-seq data comparing the lung relapse of the Ewing tumor with a germline BARD1 pathogenic variant to the original primary/pretreatment biopsy. Genesets significantly impacted (P < 0.05) are included. C, Phase contrast image (400×) of the PSaRC318 Ewing tumor cell line. D, Flow cytometry showing presence of surface CD99 expression in the PSaRC318 cell line. E, Schematic detailing the difference between Type 1 and Type 3 EWS-FLI fusions (top) and Western blot analysis with anti-FLI1 antibody of Ewing sarcoma cell lines with type 3 (PSaRC318) versus type 1 (A673, CHLA9, CHLA10, and TC71) EWS-FL1 fusions. F, Western blot demonstrating BARD1 protein expression in the same Ewing sarcoma cell lines as in E. PSaRC318 cells demonstrate significantly (P < 0.05) less BARD1 expression as compared with other Ewing cell lines. Densitometry values below the blot indicate relative expression values. Experiments in D–F were completed minimally in biological triplicate.

FIGURE 2

Loss of BARD1 enhances Ewing sarcoma cell sensitivity to PARP inhibition. A, Western blot analysis for PARP1 expression in Ewing sarcoma cells. B, p-γH2AX and DAPI immunofluorescence staining of PSaRC318, A673, and CHLA10 cells treated with DMSO or 100 nmol/L talazoparib (Tal). Cells imaged at 630× and p-γH2AX foci were quantified (bottom graphs). C, IncuCyte assay comparing the confluence of PSaRC318 cells treated with DMSO versus 100 nmol/L talazoparib over 1 week. D, qRT-PCR showing BARD1 mRNA expression in A673 or CHLA10 cells treated with control (ctsi) or BARD1 (BARD1si) siRNA. E, Western blot analysis for BARD1 expression in untreated (NT) A673 or CHLA10 cells or cells treated with Ctsi or BARD1si. F, IncuCyte monitoring of cell confluence at increasing concentrations of talazoparib versus DMSO controls in A673 and CHLA10 cells treated with Ctsi versus BARD1si. A673 and CHLA10 cell data is graphed at the 60-hour time point. Normalized expression values from densitometry analyses are included under Western blots. Experiments were completed minimally in biological triplicate. NS, not significant; *, P < 0.01. Error bars, SD.

FIGURE 3

BARD1 loss enhances Ewing sarcoma cell apoptosis in response to niraparib plus radiation. A, Relative apoptosis (caspase 3/7 activity) data from IncuCyte assays showing the effect of 0.5 μmol/L niraparib (Nir) versus DMSO control plus either 0 or 2 Gy radiation on PSaRC318 cells. B, Confluence data from IncuCyte assays showing the effect of 0.5 μmol/L niraparib versus DMSO control plus either 0 or 2 Gy radiation on PSaRC318 cells. C and D, A673 cells were treated with control (Ctsi) or BARD1 (BARD1si) siRNA, niraparib (at doses indicated) versus DMSO control, and either 0 or 2 Gy radiation and monitored via IncuCyte apoptosis assay (C) or confluence assay (D). For these experiments, cells were seeded in the presence of niraparib and radiation was performed at 12–15 hours. Relative apoptosis (caspase 3/7 dye activity) is calculated as green fluorescence in μm² divided by confluence. Experiments were completed minimally in technical and biological triplicates. *, P < 0.01 as determined by ANOVA analysis with Tukey multiple comparisons test. Error bars, SD.

FIGURE 4

Impact of BARD1 loss on the Ewing sarcoma cell transcriptome. A, Volcano plot of genes up-/downregulated upon loss of BARD1 as compared with A673 Ewing sarcoma cells transfected with Ctsi in three biological replicates. Horizontal dashed lines denote P = 0.05 (5 on −log₁₀ scale). Vertical dashed lines denote a value of 1 on the log₂ scale. BARD1 is circled in blue and highlighted as to verify that it is significantly downregulated upon RNA-seq analysis. In addition, the inset image demonstrates RT-PCR analysis of BARD1 expression as a second means by which to validate reduction of BARD1 expression in these RNA samples (×3 biological replicates) as compared with Ctsi-treated cells, *, P < 0.05; error bars, SD. B, List of most significantly (P_adj) upregulated and downregulated genes with a log₂ fold change of 2 or greater when comparing cells treated with BARD1 siRNA versus Ctsi. C, Pathway analysis (C2 and Hallmark genesets) of BARD1 siRNA-treated cells as compared with Ctsi-treated cells. NES, normalized enrichment score; P_adj, adjusted P value.

FIGURE 5

Guanylate-binding protein 1 (GBP1) contributes to Ewing cell sensitivity to DNA damage noted upon loss of BARD1. A, IHC analysis of GPB1 expression in the PSaRC318 patient tumor and in two additional independent Ewing tumors. Images provided at both low (200×) and high (1,000×) power. B, Western blot for GBP1 in Ewing sarcoma cell lines. Numbers under the blot indicate normalized expression as determined by densitometry analysis. C, Normalized GPB1 mRNA expression of PSaRC318 cells treated with control (ct) or GBP1 siRNA (si). *, P < 0.0001. D, PSaRC318 cells were transfected with control or GBP1 siRNA for 72 hours. Cells were then seeded into 96-well plates in quadruplicate, allowed to adhere and then radiated (dose = 1 Gy). Live-cell IncuCyte monitoring of cell confluence and apoptosis (caspase 3/7 activity) was monitored. The graph displays the relative apoptosis (apoptosis/confluence) in control versus GBP1 siRNA-treated cells over time. *, P < 0.05. E, PSaRC318 cells treated with siRNA and seeded as in D and then treated with DMSO, 0.5 μmol/L niraparib (Nir), or 0.75 μmol/L niraparib. The graph displays the relative apoptosis (apoptosis/confluence) over time. Representative IncuCyte images at 48 hours are included (right). NS, not significant; *, P < 0.05. Experiments completed minimally in biological triplicate. Error bars, SD.