Targeting Therapeutic Resistance and Multinucleate Giant Cells in CCNE1-Amplified HR-Proficient Ovarian Cancer

Abstract

Approximately 20% of high-grade serous ovarian cancers (HGSOC) have CCNE1 amplification. CCNE1-amplified tumors are homologous recombination (HR) proficient and resistant to standard therapies. Therapy resistance is associated with increased numbers of polyploid giant cancer cells (PGCC). We sought to identify new therapeutic approaches for patients with CCNE1-amplified tumors. Using TCGA data, we find that the mTOR, HR, and DNA checkpoint pathways are enriched in CCNE1-amplified ovarian cancers. Furthermore, Interactome Mapping Analysis linked the mTOR activity with upregulation of HR and DNA checkpoint pathways. Indeed, we find that mTOR inhibitors (mTORi) downregulate HR/checkpoint genes in CCNE1-amplified tumors. As CCNE1-amplified tumors are dependent on the HR pathway for viability, mTORi proved selectively effective in CCNE1-amplified tumors. Similarly, via downregulation of HR genes, mTORi increased CCNE1-amplifed HGSOC response to PARPi. In contrast, overexpression of HR/checkpoint proteins (RAD51 or ATR), induced resistance to mTORi. In vivo, mTORi alone potently reduced CCNE1-amplified tumor growth and the combination of mTORi and PARPi increased response and tumor eradication. Tumors treated with mTORi demonstrated a significant reduction in ALDH+ PGCCs. Finally, as a proof of principle, we identified three patients with CCNE1 amplified tumors who were treated with an mTORi. All three obtained clinical benefits from the therapy. Our studies and clinical experience indicate mTORi are a potential therapeutic approach for patients with CCNE1-amplified tumors.

Figure 1.. CCNE1 Amplification/overexpression is associated with PGCCs and induction of HR and DNA checkpoint pathway gene expression.

A. CCNE1 IHC in an ovarian cancer TMA showing PGCCs are more prevalent in CCNE1 high (CCNE1_OE, >90% nuclei intensely positive) tumors compared to CCNE1 low (<50% nuclei positive). B. Percentage of PGCCs in CCNE1-overexpressing and control ovarian tumors. N=10 samples/group were analyzed. p value was calculated using un-paired t-test. C. Gene pathways significantly co-expressed with CCNE1 mRNA in the TCGA dataset. D. CCNE1 mRNA expression of the indicated genes in three CCNE1-amplified HGSOC lines (NIHOVCAR3, OVCAR4, and COV318, red) compared with CCNE1-WT (HEY1, black) and CCNE1-Hetdel (COV504, green) control cell lines. E-G. mRNA expression (top panels) and correlations (bottom panels) of RAD51, ATR and CHK2, respectively, in the cells as in panel D. At least three independent experiments, with technical duplicates, were performed for qRT-PCR analysis. The expression of each gene was normalized to HPRT. The level of each gene in COV504 was set as 1. CCNE1-amplified (red), WT (black) and Hetdel (green). Pearson correlation coefficient was calculated using GraphPad Prism.

Figure 2.. Induction of HR and DNA checkpoint pathway gene expression in CCNE1-amplified lines is linked with the mTOR pathway.

A. Western blot of HR, NHEJ, and checkpoint pathway protein expression in the indicated control and cisplatin-treated (1μg/ml for 48hrs). inexpression was normalized with expression in COV504 set as 1 and ACTB protein used as a loading control. B. Reactome mapping of proteins and their associated pathways and their correlation with CCNE1 protein expression (https://reactome.org) (29). C. Western blot analysis of HR/checkpoint proteins in control, mTORi (RAD001, 20nM), PARPi (Olaparib, 10μM), or mTORi+PARPi treated cell lines. Experiments were repeated at least 2 times. Di. Immunofluorescence staining of RAD51 and γH2AX in CCNE1-amplified NIHOVCAR3 treated with mTORi inhibitors (RAD001 (10nM) or Torin1 (10nM)) or cisplatin (300ng/ml), and ii. Percentage of RAD51-positive cells in the indicated treatment groups evaluated in 10-high power fields, quantified using Image J. p values were calculated using a student’s t-test. E. (i) Schematic of DR-GFP assay used to measure homologous recombination. I-SceI expression induces a DNA double-strand break resulting in a non-functional copy of GFP (SceGFP). DNA Recombination, using the incomplete GFP (iGFP) fragment as repair template, restores GFP expression by. (ii) Quantification of the percentage of GFP+ cells following I-SceI transfection with or without mTORi (100 nM Torin 1). The experiment was repeated three times and statistical significance was calculated using an unpaired two-tailed t-test with Welch’s correction, error bars indicate standard deviation. F. i. Western blot of RAD51 expression following RAD51 siRNA knockdown. ii. Fold change in cell growth 48hr after scrambled control/Rad51 siRNA knockdown (25nm FlexiTube siRNA (Qiagen)) or mTORi (Torin 10nm) treatment in NIHOVCAR3 and COV504 cells. p value was calculated using paired t-test.

Figure 3.. CCNE1-amplified cells are preferentially responsive to mTOR inhibitors in an HR/checkpoint protein-dependent manner.

A. Realtime cell growth of CCNE1-AMP NOHOVCAR3 and control lines (HEY1 and COV504) with control and mTORi (Torin1, 20nM) treatment. Mean and standard deviation were calculated using four images each of three replicates, the experiments were repeated twice. B. Normalized ATR mRNA expression in control and ATR shRNA-knockdown in NIHOVCAR3, CCNE1-amplified HGSOC line. Individual value represents data from three independent experiment with two replicates each. C. Dose-response of ATR-knockdown NIHOVCAR3 cells to mTORi (Torin1, 0–80nM). D. Representative (i) Western Blotting and (ii-iii) quantification demonstrating ATR and RAD51 protein level expression in NIHOVCAR3 cells transfected with the ATR or RAD51, respectively (3 independent repeats). E. Normalized cell viability of mTORi (Torin1,10 nM) treated CCNE1-amplified cells overexpressing ATR or RAD51. F. Live cell growth of CCNE1-amplified NIHOVCAR3 lines overexpressing ATR or RAD51 compared to control in response to mTORi (Torin 1,10 nM). Mean and standard deviation were calculated using four images each of three replicates. p values were calculated using the t-test (between groups) and ANOVA analysis (among groups).

Figure 4.. Impact of mTOR or PARP inhibition on CCNE1-amplified ovarian tumor growth and DNA damage response.

A. Tumor growth (i) and Percentage of tumors initiated (ii) in CCNE1-amplified NIHOVCAR3 in vivo in xenograft mouse model in response to mTORi (RAD001, 10mg/kg, 2 days/week,), PARPi (Olaparib, 50mg/kg, 5 days/week), or dual treatment. N=10/group. B. Tumor growth (i) and Percentage of tumors initiated (ii) in CCNE1-amplified OVCAR4 in vivo in xenograft mouse model in response to mTORi (RAD001), PARPi (Olaparib), or dual treatment as described in A. N=10/group. Paired t-test was used for p value between two groups. C. Tumor weight (i) and Percentage of tumors initiated (ii) after 5 months off therapy in mTORi and dual mTORi plus PARPi treatment groups as in B. D. Response of PARPi resistant tumor to mTORi alone or in combination with PARPi. N=2/group. p values were calculated using the student t-test (between groups) and ANOVA analysis (among groups). E. Representative IF imaged of p-CHEK2 and γH2AX level in CCNE1-amplified NIHOVCAR3 xenograft tumors with the indicated treatments described in A. F-H. Quantification of pCHEK2 staining of (F) Total number of tumor cells, (G) Average area of pCHEK2-positive tumor cells, and (H). Number of PGCCs. Ten high-power fields (HPF) were analyzed in each group of tumor samples using ImageJ. p value was calculated using Student t test (between groups) and ANOVA (among groups).

Figure 5.. Aberrant mitotic spindle and fragmented DNA in CCNE1-amplified tumors upon mTOR inhibitor treatment.

A. Immunofluorescence of mitotic spindles in CCNE1-amplified NIHOVCAR3 tumor xenografts treated with mTORi or vehicle control as described in Fig. 4A. B. Percentage of tumor cells with deranged mitotic spindles in mitotic phase, with fragmented DNA in each of the indicated tumor treatment groups, as indicated. Tumor cells with unorganized/multi-polar spindles and fragmented chromosome DNA were counted as abnormal. Percentage of abnormal mitotic cells of total mitotic cells from 10 high-power fields/5 tumors/treatment group. p values were calculated using ANOVA. C. Percent of PGCCs in control and mTORi treated (Torin1 (10nM) or Rad001 (10nM)) or control vehicle for 3 days CCNE1-amplified (OVCAR3) and control HGSOC lines. Equal number of cells from each sample were fixed and DNA content assessed by Hoechst staining followed by FACS analysis. Data represented 3 independent experiments with duplicates each. Cells with DNA content>4n were scored as PGCCs. p value was calculated using Student t test (between groups) and ANOVA (among groups).

Figure 6.. Impact of mTOR or PARP inhibition on ALDH+ PGCC population in CCNE1-amplified ovarian tumor cells in vivo.

A. Representative IF images of ALDH1A1 in CCNE-amplified NIHOVCAR3 tumors from the indicated treatment groups as described in Fig 4. B. Quantification of ALDH1A1+ (i) PGCC and (ii) Total ALDH1A1+ tumor cells from the indicated tumor treatment groups. Ten high-power fields (HPF) were analyzed in each group of tumor samples representing 5 tumors each. p value was calculated using unpaired t-test C. Percentage of ALDH-positive tumor cells in NIHOVCAR3 tumors from the indicated treatment groups. Tumor cells isolated from each tumor were used for Aldefluor assay followed by FACS analysis. Each data point represents data of an individual tumor. p value was calculated using Student t-test (between groups) and ANOVA (among groups). D. Impact of Rad001 and bevacizumab therapy on patients with CCNE1-amplified Tumors. Swimmer plots of patient responses for four patients with CCNE1-amplified tumors treated with Rad001 and bevacizumab. Note patient 27 withdrew from the trial prior to completing 1 cycle of therapy. E. Blood CA-125 levels of CCNE1-amplifed patients.