Search

Background Ovarian cancers are hallmarked by chromosomal instability. New therapies deliver improved patient outcomes in relevant phenotypes, however therapy resistance and poor long-term survival signal requirements for better patient preselection. An impaired DNA damage response (DDR) is a major chemosensitivity determinant. Comprising five pathways, DDR redundancy is complex and rarely studied alongside chemoresistance influence from mitochondrial dysfunction. We developed functional assays to monitor DDR and mitochondrial states and trialled this suite on patient explants. Methods We profiled DDR and mitochondrial signatures in cultures from 16 primary-setting ovarian cancer patients receiving platinum chemotherapy. Explant signature relationships to patient progression-free (PFS) and overall survival (OS) were assessed by multiple statistical and machine-learning methods. Results DR dysregulation was wide-ranging. Defective HR (HRD) and NHEJ were near-mutually exclusive. HRD patients (44%) had increased SSB abrogation. HR competence was associated with perturbed mitochondria (78% vs 57% HRD) while every relapse patient harboured dysfunctional mitochondria. DDR signatures classified explant platinum cytotoxicity and mitochondrial dysregulation. Importantly, explant signatures classified patient PFS and OS. Conclusions Whilst individual pathway scores are mechanistically insufficient to describe resistance, holistic DDR and mitochondrial states accurately predict patient survival. Our assay suite demonstrates promise for translational chemosensitivity prediction.

Purpose: DNA damage defects are common in ovarian cancer and can be used to stratify treatment. Although most work has focused on homologous recombination (HR), DNA double-strand breaks are repaired primarily by nonhomologous end joining (NHEJ). Defects in NHEJ have been shown to contribute to genomic instability and have been associated with the development of chemoresistance.Experimental Design: NHEJ was assessed in a panel of ovarian cancer cell lines and 47 primary ascetic-derived ovarian cancer cultures, by measuring the ability of cell extracts to end-join linearized plasmid monomers into multimers. mRNA and protein expression of components of NHEJ was determined using RT-qPCR and Western blotting. Cytotoxicities of cisplatin and the PARP inhibitor rucaparib were assessed using sulforhodamine B (SRB) assays. HR function was assessed using γH2AX/RAD51 foci assay.Results: NHEJ was defective (D) in four of six cell lines and 20 of 47 primary cultures. NHEJ function was independent of HR competence (C). NHEJD cultures were resistant to rucaparib (P = 0.0022). When HR and NHEJ functions were taken into account, only NHEJC/HRD cultures were sensitive to rucaparib (compared with NHEJC/HRC P = 0.034, NHEJD/HRC P = 0.0002, and NHEJD/HRD P = 0.0045). The DNA-PK inhibitor, NU7441, induced resistance to rucaparib (P = 0.014) and HR function recovery in a BRCA1-defective cell line.Conclusions: This study has shown that NHEJ is defective in 40% of ovarian cancers, which is independent of HR function and associated with resistance to PARP inhibitors in ex vivo primary cultures. Clin Cancer Res; 23(8); 2050–60. ©2016 AACR.

Supplementary Table ST1. Patient characteristics for primary culture samples divided by NHEJ status. Supplementary figure S1. Predictors of NHEJ function A. pDNA-PKcs foci 1 hour after 2 Gy IR normalised to untreated controls. Foci were counted across 3 fields of view in > 50 cells for each condition.

Purpose: DNA damage defects are common in ovarian cancer and can be used to stratify treatment. Although most work has focused on homologous recombination (HR), DNA double-strand breaks are repaired primarily by nonhomologous end joining (NHEJ). Defects in NHEJ have been shown to contribute to genomic instability and have been associated with the development of chemoresistance. Experimental Design: NHEJ was assessed in a panel of ovarian cancer cell lines and 47 primary ascetic-derived ovarian cancer cultures, by measuring the ability of cell extracts to end-join linearized plasmid monomers into multimers. mRNA and protein expression of components of NHEJ was determined using RT-qPCR and Western blotting. Cytotoxicities of cisplatin and the PARP inhibitor rucaparib were assessed using sulforhodamine B (SRB) assays. HR function was assessed using γH2AX/RAD51 foci assay. Results: NHEJ was defective (D) in four of six cell lines and 20 of 47 primary cultures. NHEJ function was independent of HR competence (C). NHEJD cultures were resistant to rucaparib (P = 0.0022). When HR and NHEJ functions were taken into account, only NHEJC/HRD cultures were sensitive to rucaparib (compared with NHEJC/HRC P = 0.034, NHEJD/HRC P = 0.0002, and NHEJD/HRD P = 0.0045). The DNA-PK inhibitor, NU7441, induced resistance to rucaparib (P = 0.014) and HR function recovery in a BRCA1-defective cell line. Conclusions: This study has shown that NHEJ is defective in 40% of ovarian cancers, which is independent of HR function and associated with resistance to PARP inhibitors in ex vivo primary cultures. Clin Cancer Res; 23(8); 2050–60. ©2016 AACR.

IntroductionEpithelial ovarian cancer is recognized to be heterogeneous but is currently treated with a single treatment strategy. Successful patient stratification of emerging chemotherapy agents is dependent upon the availability of reliable biomarkers indicative of the entire tumor.AimThe aim of this study was to evaluate intertumor and intratumor heterogeneity within a series of epithelial ovarian cancer using homologous recombination (HR) DNA repair status.MethodsPrimary cultures generated from ascites and solid tumor from multiple intra-abdominal sites were characterized by their morphology and expression of protein markers. Results were compared with Formalin fixed paraffin embedded tissue pathology.Homologous recombination function was determined by quantification of nuclear Rad51 foci. Growth inhibition (sulforhodamine B) assays were used to calculate the GI50 for cisplatin and rucaparib.ResultsAscites with matched solid tumor were cultured from 25 patients.Concordance in functional HR status between ascites and solid tumor subcultures was seen in only 13 (52%) of 25 patients. Heterogeneity in HR status was seen even in patients with homogeneous histological subtype. Homologous recombination defective cultures were significantly more sensitive to cisplatin and rucaparib.Additionally, intertumor and intratumor heterogeneity was seen between the expression of epithelial and ovarian markers (EpCAM, cytokeratin, CA125, MOC-31, and vimentin). There was no relationship between heterogeneity of HR functional status and antigen expression.ConclusionsIntertumor and intratumor functional HR heterogeneity exists that cannot be detected using histological classification. This has implications for biomarker-directed treatment.

Background Cell lines provide a powerful model to study cancer and here we describe a new spontaneously immortalised epithelial ovarian cancer cell line (NUOC-1) derived from the ascites collected at a time of primary debulking surgery for a mixed endometrioid / clear cell / High Grade Serous (HGS) histology. Results This spontaneously immortalised cell line was found to maintain morphology and epithelial markers throughout long-term culture. NUOC-1 cells grow as an adherent monolayer with a doubling time of 58 hours. The cells are TP53 wildtype, positive for PTEN, HER2 and HER3 expression but negative for oestrogen, progesterone and androgen receptor expression. NUOC-1 cells are competent in homologous recombination and non-homologous end joining, but base excision repair defective. Karyotype analysis demonstrated a complex tetraploid karyotype. SNP array analysis of parent and derived subpopulations (NUOC-1-A1 and NUOC-1-A2) cells demonstrated heterogeneous cell populations with numerous copy number alterations and a pro-amplification phenotype. The characteristics of this new cell line lends it to be an excellent model for investigation of a number of the identified targets. Materials and Methods The cell line has been characterised for growth, drug sensitivity, expression of common ovarian markers and mutations, clonogenic potential and ability to form xenografts in SCID mice. Copy number changes and clonal evolution were assessed by SNP arrays.

ATR is an attractive target in cancer therapy because it signals replication stress and DNA lesions for repair and to S/G2 checkpoints. Cancer-specific defects in the DNA damage response (DDR) may render cancer cells vulnerable to ATR inhibition alone. We determined the cytotoxicity of the ATR inhibitor VE-821 in isogenically matched cells with DDR imbalance. Cell cycle arrest, DNA damage accumulation and repair were determined following VE-821 exposure. Defects in homologous recombination repair (HRR: ATM, BRCA2 and XRCC3) and base excision repair (BER: XRCC1) conferred sensitivity to VE-821. Surprisingly, the loss of different components of the trimeric non-homologous end-joining (NHEJ) protein DNA-PK had opposing effects. Loss of the DNA-binding component, Ku80, caused hypersensitivity to VE-821, but loss of its partner catalytic subunit, DNA-PKcs, did not. Unexpectedly, VE-821 was particularly cytotoxic to human and hamster cells expressing high levels of DNA-PKcs. High DNA-PKcs was associated with replicative stress and activation of the DDR. VE-821 suppressed HRR, determined by RAD51 focus formation, to a greater extent in cells with high DNA-PKcs. Defects in HRR and BER and high DNA-PKcs expression, that are common in cancer, confer sensitivity to ATR inhibitor monotherapy and may be developed as predictive biomarkers for personalised medicine.

Introduction: Ovarian cancer is recognized as a heterogeneous disease but is currently treated with a single treatment strategy. Improved outcomes will require stratified use of novel cytotoxic agents exploiting the molecular pathology of the tumor but this is dependent upon the availability of reliable biomarkers. Most research testing the efficacy of biomarker stratified treatment use a single tumor sample, assuming the biomarker profile to be homogeneous. The reliability of such biomarkers may however be complicated by intra-tumor heterogeneity (ITH). The aim of this study was to evaluate the extent of ITH within a series of epithelial ovarian cancers using homologous recombination (HR) DNA repair status as a marker, given that this has been shown to predict sensitivity to PARP inhibitors ex vivo and cisplatin in vivo. Methods: Primary cultures (PCO) were generated from ascitic fluid and solid tumor taken from a variety of intra-abdominal sites. Cultures were characterized based upon their morphology and using an established panel of immunofluorescent-labeled antibodies. Results were cross-referenced with formal pathological examination of FFPE tissue and cytology of ascites. The functional status of HR was determined for each culture by quantification of nuclear Rad51 focus formation following induction of DNA DSB. Standard growth inhibition (SRB) assays were used to calculate the GI50 for cisplatin. Results: PCO cultures from ascites with matched solid tumor were cultured from 19 patients. Solid tumor was sampled from multiple intra-abdominal regions in ten patients. In the remaining nine patients solid tumor was taken from a single site from the bulk of the pelvic tumor. Inter- and intra-tumor heterogeneity was seen between the expression of epithelial and ovarian markers (EpCAM, cytokeratin, CA125, MOC-31 and vimentin). PCO samples taken from eight patients showed homogeneous staining patterns between all subcultures. Eleven patients however demonstrated variation in the staining patterns between subcultures from ascites and solid tumor or between solid tumors from varying sites. This heterogeneity was not reflected in the histological classification of the tumors with only 2/ 11 heterogeneous patients being classified as mixed histological subtype upon formal pathological examination. Variability in the HR status between ascites and solid tumor subcultures was seen in 9/19 (47%) patients. Functional heterogeneity was seen even in patients with homogeneous expression of markers and homogeneous histological subtype. The remaining 10 patients showed agreement in HR status in ascitic and all solid subcultures. When two or more solid tumor samples from anatomically distinct sites were assessed in the same patient there was variability between cultures in 7/9 patients. The sensitivity of PCO subcultures to cisplatin, as hypothesized, correlated with HR status with a GI50 of 265.6μM in HR competent cultures and 6.27μM in defective cultures, p Conclusions: Inter- and intra-tumor heterogeneity exists that cannot be detected using traditional histological classification. The success of stratification for novel therapeutic agents on the basis of biomarker testing is dependent upon the sample analyzed being representative of the tumor as a whole. This study suggests that a single sample cannot be assumed to be representative in terms of morphology, expression of markers or more importantly, in its DNA repair functional status and therefore subsequent sensitivity to therapy. Accepting that functional ITH exists and tumor subpopulations are variably responsive to both therapeutic agents suggests that despite response to therapy in some subpopulations, resistant subpopulations may survive, proliferate, and result in disease relapse. This is not necessarily a failure of treatment but it does suggests that alternative treatment strategies in heterogeneous tumors may be more successful. This abstract is also presented as Poster B53. Citation Format: Rachel L. O'Donnell, Laura C. Woodhouse, Aiste McCormick, Angelika Kaufmann, Michelle Dixon, Asima Mukhopadhyay, Nicola J. Curtin, Richard J. Edmondson. Intra- and intertumor heterogeneity in epithelial ovarian cancer: Consequences for biomarker-dependent stratification of therapies. [abstract]. In: Proceedings of the AACR Special Conference on Advances in Ovarian Cancer Research: From Concept to Clinic; Sep 18-21, 2013; Miami, FL. Philadelphia (PA): AACR; Clin Cancer Res 2013;19(19 Suppl):Abstract nr PR10.

Background: Sapacitabine is an oral nucleoside analogue that is metabolised in-vivo to CNDAC (2′-C-cyano-2′-deoxy-1-β-D-arabino-pentofuranosylcytosine). It has a novel mechanism of action by causing single-strand breaks and replication-associated double-strand breaks (DSBs) after its incorporation into DNA. DSB, if unrepaired, are cytotoxic and are predominantly repaired by homologous recombination (HR). Over 50% of high-grade serous ovarian cancers are estimated to be defective in HR, only a small proportion of which are accounted for by germline BRCA mutations. HR-defective ovarian cancers have been shown to be sensitive to PARP inhibitors ex-vivo and platinum in-vivo and we therefore hypothesised that HR defective ovarian cancers would be sensitive to CNDAC and aimed to explore the potential use of CNDAC in ovarian cancer stratified for HR function. Methods: Primary cultures (PCO) were generated from ascitic fluid collected from patients undergoing ovarian cancer surgery, characterised and confirmed to be ovarian and epithelial in origin based upon morphology, expression of immunofluorescent-labelled markers and cross-referencing with formal examination of FFPE tissue and cytology of ascites. The functional status of HR was determined for each cell line and PCO culture by quantification of nuclear Rad51 focus formation following induction of DNA DSB. SRB assays were used to calculate the GI50 for cisplatin and CNDAC, (Cyclacel, UK), in cell lines nad PCO cultures. Results: As hypothesised, HR defective cell lines were highly sensitive to CNDAC (VC8 BRCA2 mutant: GI50=366 nM) in comparison to their matched HR competent cell lines (VC8-BRCA2 reconstructed: GI50=18 μM; VC8-PARPi-resistant BRCA revertant: GI50=3 μM). Sensitivity to CNDAC in patient PCO cultures (n=39) was heterogeneous with a mean GI50 of 297 nM (95% CI 136-646 nM). Stratifying PCO cultures according to their HR status demonstrated greater sensitivity in HR defective cultures with a mean GI50 of 135 nM (46-394; n=12) in defective cultures, compared to 477 nM in HR competent cultures (164-1386; n=27). Of the 34 cisplatin resistant PCO cultures (GI50 >5 μM), 20 (59%) remained sensitive to CNDAC with a mean GI50 of 426 nM (148-1223; n=13) in HR competent cultures and a GI50 of 230 nM (70-759; n=7) in HR defective cultures. CNDAC sensitivity was evaluated in base excision repair (BER) defective EM9 cell line (GI50=374 nM) compared with parental wild type AA8 cells (GI50=6.9 μM), p Conclusions: Cell lines defective in HR or BER are highly sensitive to CNDAC suggesting a possible use of HR or BER status as biomarkers for the stratified use of CNDAC in the treatment of ovarian cancer. When applied to ex-vivo studies HR status as a single marker cannot reliably predict sensitivity and a panel of pathways may therefore be needed to enable reliable treatment stratification. Nevertheless, stratification based on these markers is likely to permit enrichment of patients more likely to respond to CNDAC. Additionally CNDAC may have a role in the treatment of platinum resistant disease as approximately 60% of patients with resistant disease may benefit. Further work is required to explore the possibility of CNDAC as a chemosensitiser in platinum sensitive disease and as a second line agent in platinum resistant disease. Citation Format: Rachel L. O'Donnell, James CC Murray, Adriana GB Buskin, Sheelagh Frame, David G. Blake, Michelle Dixon, Aiste McCormick, Angelika Kaufmann, E Ruth Plummer, Richard J. Edmondson, Nicola J. Curtin. Therapeutic potential of sapacitabine in ovarian cancers defective in homologous recombination. [abstract]. In: Proceedings of the AACR Special Conference on Advances in Ovarian Cancer Research: From Concept to Clinic; Sep 18-21, 2013; Miami, FL. Philadelphia (PA): AACR; Clin Cancer Res 2013;19(19 Suppl):Abstract nr A33.

Introduction: The phosphatase and tensin homolog deleted on chromosome 10 (PTEN) tumour suppressor protein is a central negative regulator of the PI3K/AKT signalling cascade and induces apoptosis and suppresses cell proliferation. PTEN is found to be either inactivated or mutated in various human malignancies. The role of PTEN in cancer however is still not fully understood. PTEN loss has been suggested to impair double strand break repair by homologous recombination and has therefore been suggested to improve sensitivity to PARP inhibitors. However, conflicting results suggest that PTEN loss is associated with cisplatin resistance development. In this study we assess the role of PTEN in ovarian cancer using experimental knockdown models and primary ovarian cancer cultures. Methods: PTEN knockdown models were created in a normal epithelial and clear cell carcinoma cell lines using MISSION®shRNA lentiviral transduction particles and grown in purimycin selection media. Sensitivity of the knockdown models to common therapies were assessed using SRB assay (these included cisplatin, paplitaxol, camptothecin, doxorubicin and irradiation). We screened 28 unselected primary epithelial ovarian cancer cultures derived from ascitic fluid collected at the time of surgery and matched genomic DNA for PTEN mutations in exons 2-5, 6-7 and 9-10 using PCR amplification and Sanger sequencing. We also assessed mRNA expression of PTEN in this panel of ovarian cancer cultures using qRT-PCR. Sensitivity to cisplatin was assessed in the primary ovarian cultures using SRB assay and HR status was determined using γH2AX/RAD51 immunofluorescence assay. Results: In normal epithelial cells, PTEN knock-down had no effect on cytotoxic agent sensitivity whilst in clear cell carcinoma cells, PTEN knockdown enhanced sensitivity to cisplatin, doxorubicin and irradiation. PTEN knockdown produced G1 arrest in both normal epithelial and clear cell carcinoma cells. The primary ovarian cancer cultures included 20 high grade serous, 6 mixed endometroid / clear cell, 1 clear cell and 1 mucinous carcinoma. Whilst two point mutations were detected in primary cultures (1105T>TG, 25L>L in 6 cultures and 1508G>GA, 159R>R in 4 cultures), no somatic mutations were observed in any of the PTEN coding sequences. The mutations found did not correlate with histological type or the level of PTEN mRNA expression. PTEN mRNA expression varied over 40 fold between the cultures. PTEN expression did not correlate with HR status or cisplatin sensitivity in primary ovarian cancer cultures. Discussion: These data indicate that PTEN knockdown results in therapeutic sensitisation. However, we have also shown that in primary ovarian cancers mutations in PTEN gene are rare and that PTEN mRNA expression does correlate with cisplatin sensitivity or HR status. Therefore the results of experimental PTEN knockdown models need to be interpreted with caution and PTEN inhibition needs to be further assessed in primary cultures to assess its role ovarian cancer. Citation Format: Aiste McCormick, Eleanor Earp, Charlotte Leeson, Michelle Dixon, Rachel ODonnell, Richard J. Edmondson. PTEN expression level is important in determining sensitivity to cytotoxic agents in vitro, but is not representative of PTEN function observed in primary ovarian cancer cultures. [abstract]. In: Proceedings of the AACR Special Conference on Advances in Ovarian Cancer Research: From Concept to Clinic; Sep 18-21, 2013; Miami, FL. Philadelphia (PA): AACR; Clin Cancer Res 2013;19(19 Suppl):Abstract nr B9.

Introduction: The most toxic forms of DNA damage, double strand breaks (DSBs), are repaired by two main pathways - homologous recombination repair (HRR) and non-homologous end joining (NHEJ). Fifty percent of ovarian cancers have defects in HRR resulting in hypersensitivity to platinum and PARP inhibitors. When NHEJ is defective, DSBs with mismatched ends can be joined by an error-prone mechanism, involving joining between regions of nucleotide microhomology, but this is highly mutagenic. Defects in the NHEJ pathway may therefore contribute to genomic instability and have been associated with resistance development. In this study we demonstrate that NHEJ is defective, resulting in error prone repair, in ~50% of ovarian cancers. Methods: In a panel of ovarian cancer cell lines (OVCAR3, SKOV3, IGROV-1, A2780, CP70 and MDAH) and 30 primary ovarian cancer cultures generated from ascites collected at the time of surgery, we assessed NHEJ using an assay measuring the ability of cell extracts to end-join linearized plasmid monomers into multimers. We determined the effect of inhibition of DNA-PK by concurrent treatment with a DNA-PK inhibitor (NU7441) on the cytotoxicity of the PARP inhibitor, rucaparib, using the sulforhodamine B (SRB) assay. mRNA expression of DNA-PK, XRCC4, LIG4, KU70 and KU80 was determined using qRT-PCR normalised to GAPDH expression. DNA-PK function was determined by immunofluorescence assessment of auto-phosphorylation of DNA-PK one hour after 2Gy irradiation. Furthermore, we assessed the effect of NU7441 on the homologous recombination function using the γH2AX / RAD51 foci formation assay in a BRCA1 defective cell line. Results: We demonstrated that many cell lines and primary ovarian cancer cultures were able to rejoin compatible DNA-ends accurately, however the amount of rejoining of mismatched ends varied significantly between cultures and four out of six cell lines and 14 of 30 primary cultures assessed were unable to accurately rejoin DSBs with mismatched ends resulting in large deletions and error prone repair. This process was not dependent on expression of DNA-PK, XRCC4 or LIG4. Also, whilst in cell lines the autophosphorylation of DNA-PK correlated with error free rejoining, accurate rejoining in primary cultures was independent of DNA-PK auto-phosphorylation. Error-prone repair in ovarian cancer extracts correlated with reduced expression of KU70 (p=0.006) and KU80 (p=0.009). We found that defects in the NHEJ pathway are associated with resistance to rucaparib in primary ovarian cultures (p=0.02). Furthermore concurrent treatment with the DNA-PK inhibitor was cytostatic and induced resistance to rucaparib in the majority of ovarian cancer cultures. These findings are supported by the finding of HR function recovery in a BRCA1 defective cell line treated with DNA-PK inhibitor. Conclusion: This study has shown, for the first time, that some ovarian cancers are unable to perform accurate end joining, such that error-prone joining predominates. Error prone repair is associated with resistance to PARP inhibitors in vitro. Error prone rejoining is highly mutagenic and may contribute to the genomic instability and development of resistance that is observed in ovarian cancer. This abstract is also presented as Poster A31. Citation Format: Aiste McCormick, Michelle Dixon, Rachel ODonnell, Nicola J. Curtin, Richard J. Edmondson. Ovarian cancers harbor defects in nonhomologous end joining resulting in error prone repair and resistance to rucaparib. [abstract]. In: Proceedings of the AACR Special Conference on Advances in Ovarian Cancer Research: From Concept to Clinic; Sep 18-21, 2013; Miami, FL. Philadelphia (PA): AACR; Clin Cancer Res 2013;19(19 Suppl):Abstract nr PR06.

Highlights • Treatment of stage 1A1 cancer of the cervix often involves preservation of the corpus. • Rarely metastasis to the corpus can occur in these cases