Purpose: There is a lack of prognostic and predictive biomarkers in epithelial ovarian carcinoma, and the targeting of oncogenic signaling pathways has had limited impact on patient survival in this highly heterogeneous disease. The origin licensing machinery, which renders chromosomes competent for DNA replication, acts as a convergence point for upstream signaling pathways. We tested the hypothesis that Cdc7 kinase, a core component of the licensing machinery, is predictive of clinical outcome and may constitute a novel therapeutic target in epithelial ovarian carcinoma.

Experimental Design: A total of 143 cases of ovarian cancer and 5 cases of normal ovary were analyzed for Cdc7 protein expression dynamics and clinicopathologic features. To assess the therapeutic potential of Cdc7, expression was down-regulated by RNA interference in SKOV-3 and Caov-3 ovarian cancer cells.

Results: Increased Cdc7 protein levels were significantly associated with arrested tumor differentiation (P = 0.004), advanced clinical stage (P = 0.01), genomic instability (P < 0.001), and accelerated cell cycle progression. Multivariate analysis shows that Cdc7 predicts disease-free survival independent of patient age, tumor grade and stage (hazard ratio, 2.03; confidence interval, 1.53-2.68; P < 0.001), with the hazard ratio for relapse increasing to 10.90 (confidence interval, 4.07-29.17) for the stages 3 to 4/upper Cdc7 tertile group relative to stages 1 to 2/lower Cdc7 tertile tumors. In SKOV-3 and Caov-3 cells, Cdc7 siRNA knockdown triggered high levels of apoptosis, whereas untransformed cells arrest in G1 phase and remain viable.

Conclusions: Our findings show that Cdc7 kinase predicts survival and is a potent anticancer target in epithelial ovarian carcinoma, highlighting its potential as a predictor of susceptibility to small molecule kinase inhibitors currently in development.

Translational Relevance

Personalized medicine relies on biomarkers with both prognostic and predictive value. There is a paucity of such biomarkers for the management of epithelial ovarian carcinoma, and targeting of oncogenic signaling pathways in this tumor type has had limited impact on patient survival. The DNA replication licensing system is a convergence point for upstream signaling pathways. Cdc7 kinase, a core component of the licensing machinery, is therefore an attractive diagnostic and therapeutic target. Here we report Cdc7 as an independent prognostic biomarker and novel therapeutic target in epithelial ovarian carcinoma. In a cohort of 143 patients with epithelial ovarian carcinoma, increased Cdc7 expression was associated with arrested differentiation, advanced clinical stage, and genomic instability, and was found to be an independent predictor of disease-free survival. Cdc7 down-regulation by RNA interference in two ovarian cancer cell lines triggered apoptosis, suggesting Cdc7 as a potent anticancer target in epithelial ovarian carcinoma and highlighting its potential as a predictor of therapeutic response to small molecule kinase inhibitors.

Improvement in the long-term survival of individuals with epithelial ovarian carcinoma has been modest (1). The advent of molecularly targeted therapies, however, could dramatically alter clinical outcome. The process of epithelial ovarian carcinogenesis carries with it a complex array of oncogenic alterations, which give rise to a diverse and heterogeneous series of tumors, both in terms of etiology and clinical outcome. Efforts to find therapeutic targets in epithelial ovarian carcinoma have focused on deregulated proteins involved in upstream mitogenic signaling pathways (2). However, the prognostic and predictive value of these molecules remains unclear (37). For example, overexpression of epidermal growth factor receptor and human epidermal growth factor receptor 2/neu does not correlate with clinical response to agents directed against these upstream kinases (5, 6, 8). These findings highlight the complexity of targeting upstream growth signaling pathways and the difficulties encountered in exploiting their regulatory molecules as predictive biomarkers. The apparent pathogenetic variability in epithelial ovarian carcinoma, however, is underpinned by a core set of acquired capabilities, one of which is relentless cell proliferation powered by deregulation of the cell cycle machinery (9, 10). Targeting of the cell cycle machinery, which acts as an integration point for upstream signaling pathways, is therefore an attractive alternative approach to the identification of new prognostic and predictive markers (11).

The DNA replication licensing pathway has emerged as a powerful mechanism for controlling cell proliferation (1113). During late mitosis and early G1, the replication licensing factors ORC, Cdc6, Cdt1, and Mcm2-7 assemble into prereplicative complexes that render replication origins “licensed” for DNA synthesis. During S phase, cyclin-dependent kinases and the ASK-dependent Cdc7 kinase induce a conformational change in the prereplicative complexes, resulting in recruitment of Cdc45, Mcm10, and additional initiator proteins that collectively promote origin unwinding and recruitment of DNA polymerases. Expression of the licensing repressor geminin during S-G2-M phases prevents inappropriate reinitiation events at origins that have already been activated (14). Mcm2-7 have been identified as powerful biomarkers for cancer detection and prognostication (11, 15), and combined analysis of minichromosome maintenance protein expression dynamics and biomarkers of S-G2-M progression (geminin, Aurora A, Plk1, and the Aurora kinase substrate Histone H3) allows determination of tumor cell cycle kinetics (11, 16, 17). This cell cycle biomarker panel allows out-of-cycle quiescent (G0), differentiated, or senescent cells to be distinguished from those residing in-cycle, and, furthermore, can assign cells to G1, S-G2 and M phase of the cell division cycle (Fig. 1; ref. 11). This information is of value in tumor prognostication and as a predictor of therapeutic response to cell cycle phase–specific anticancer agents (11, 17).

Fig. 1.

Diagrammatic representation of the mitotic cell division cycle. Inset table shows presence of Mcm2-7, geminin, Aurora A, and Plk1 during the cell cycle, and the appearance of a phosphorylation mark on histone 3 at serine 10.

Fig. 1.

Diagrammatic representation of the mitotic cell division cycle. Inset table shows presence of Mcm2-7, geminin, Aurora A, and Plk1 during the cell cycle, and the appearance of a phosphorylation mark on histone 3 at serine 10.

Close modal

The ASK-dependent Cdc7 kinase is a highly conserved serine/threonine kinase that promotes origin firing and entry into S phase by phosphorylating the minichromosome maintenance complex (Fig. 1; ref. 18). Cdc7 therefore represents a potentially attractive diagnostic and therapeutic target. Cdc7 not only lies at an integration point for mitogenic signaling pathways, but also plays a key role in maintaining genomic stability through intra-S-phase checkpoint pathways in response to DNA damage and stalled replication forks (19). Furthermore, it has been suggested that inhibition of origin firing through targeting Cdc7 provokes a tumor cell–specific apoptotic response (20).

Given the heterogeneity of epithelial ovarian carcinoma, it is becoming increasingly clear that improving survival rates in this malignancy inevitably requires individualized therapeutic decisions. However, despite a large number of mechanism-based prognostic biomarkers and therapeutic agents at various stages of development, none, as yet, has a clinically proven role in the management of epithelial ovarian carcinoma (2). In light of the biological, prognostic, and therapeutic implications of Cdc7 in tumorigenesis, we report here an investigation into the putative role of this kinase as a prognostic marker and novel therapeutic target in epithelial ovarian carcinoma.

Study cohort. Patients diagnosed with epithelial ovarian carcinoma from January 1, 1999 to December 31, 2004 were identified consecutively and retrospectively from the Ovarian Carcinoma Database held in the Department of Oncology, University College London Hospital Gynaecological Cancer Centre, UCL Hospitals, London, United Kingdom. A total of 143 cases were chosen based on availability of archival human ovarian tissue blocks from diagnostic resection specimens. Specimens were retrieved from the archives of the Departments of Pathology at University College London Hospital, University College London Medical School Hospital, Whittington Hospital, and Barnet and Chase Farm Hospitals, United Kingdom. Histologic specimens were assessed for histologic subtype and nuclear grade according to WHO criteria at diagnosis.

Most patients had been reviewed after completing treatment every 3 to 6 mo for 2 y, and annually thereafter. The following clinical information was available: date of birth, date of diagnosis, operative findings including amount of residual disease, International Federation of Gynecology and Obstetrics (FIGO) stage based on findings at clinical examination and surgical exploration together with cytology results, CA125 values at diagnosis and relapse, performance status at start of chemotherapy, date of relapse, date of last follow-up, and date and cause of death. These include details regarding patient treatment, which was chosen according to recommended evidence-based practice at the time of diagnosis. Clinical and pathologic characteristics of the study cohort are summarized in Supplementary Table S1.

Primary study end points were disease-free survival (time from diagnosis to first tumor relapse) and overall survival (time from diagnosis to death or last follow-up date). Of the 143 patients, 67 (47%) relapsed within the study period with a mean time to relapse among those who relapsed of 16.9 mo (SD, 11.0 mo; range, 0-47 mo). Mean follow-up time among those who had not yet relapsed was 33.2 mo (SD, 18.5 mo; range 5-75 mo). A total of 34 patients (24%) died within the study period and 107 were alive at the last follow-up. Mean survival time was 21.9 mo (SD, 15.6 mo; range, 0-60 mo) for those who had died. Mean follow-up time among those who had not yet died was 33.3 mo (SD, 18.8 mo; range, 5-75 mo). Two patients were lost to follow-up. Ethical approval was obtained from the Joint University College London/University College London Hospital Committees on the Ethics of Human Research.

Antibodies. Cdc7 monoclonal antibody (mAb) was obtained from MBL International (Caltag-Medsystems), Mcm2 phosphorylated on serine 53 RAb from Bethyl Laboratories, caspase 3 mAb from Novus Biologicals (Stratech Scientific), caspase 8 mAb from Upstate, caspase 9 mAb from Santa Cruz Biotechnology, and poly (ADP-ribose) polymerase-1 mAb from BD Biosciences Pharmingen.

Immunoexpression profiling. Archival formalin-fixed, paraffin-embedded tissue obtained at initial diagnosis was available for all patients in the series, and for each specimen a block was chosen that contained a representative sample of invasive tumor. Sections of fallopian tube tissue were obtained from patients undergoing total abdominal hysterectomy and bilateral salpingo-oophorectomy for uterine prolapse or for the treatment of benign uterine leiomyomas. Antigen retrieval from archival tissue and immunohistochemical staining for Cdc7 were done as described (17). Primary antibody directed against Cdc7 was applied to tissue sections at a dilution of 1:100. Incubation without primary antibody was used as negative control and colonic epithelial sections as positive controls. Cdc7 expression analysis was conducted as described (17) and the labeling index of the kinase in each tumor was calculated using the following formula: LI = number of positive cells/total number of cells × 100. Reassessment of 10 randomly selected cases for each biomarker by an independent assessor showed high levels of interobserver agreement.

DNA image cytometry. Tumor DNA ploidy status was determined using the Fairfield DNA Ploidy System (Fairfield Imaging) as described (17). For statistical analysis, tetraploid and polyploid tumors were grouped together with aneuploid tumors.

Cell culture. WI-38 human diploid fibroblasts were cultured in DMEM (Invitrogen) supplemented with 10% FCS (Invitrogen) and subjected to density-dependent growth arrest (G0) as described (21). Myeloblastic HL60 cells were cultured and induced to differentiate with 1,25-dihydroxyvitamin D3 (1,25 Vit D3; Sigma) as described (22). IMR-90 cells [American Type Culture Collection (ATCC) CCL-186] were cultured in DMEM (Invitrogen) supplemented with 10% FCS (Invitrogen). HBEpC cells (ECACC 502-05a) were cultured in bronchial epithelial cell serum-free growth medium (Health Protection Agency Culture Collections). SKOV-3 cells (ATCC HTB-77) were maintained in McCoy's 5A medium (Invitrogen) supplemented with 15% FCS. Caov-3 cells (ATCC HTB-75) were grown in ATCC-formulated DMEM (ATCC) supplemented with 10% FCS. HeLa S3 cells were cultured and synchronized as described (23). Growth media for WI-38, HL60, IMR-90, SKOV-3, and HeLa S3 cell lines also contained 100 U/mL penicillin and 0.1 mg/mL streptomycin.

RNA interference. For small interfering RNA (siRNA) experiments the following double-strand RNA duplex specific for the Cdc7 coding region was used: sense strand 5′-GCTCAGCAGGAAAGGTGTTTT-3′ and antisense strand 5′-AACACCTTTCCTGCTGAGCTT-3′. Non-targeting siRNA was used as a negative control. siRNA duplexes were synthesized by Ambion (Warrington) and transfected using Lipofectamine 2000 (Invitrogen) according to the manufacturer's recommendations. One day prior to transfection cells were plated in medium containing FCS and antibiotics, such that they would be 30% confluent at the time of transfection. One hour prior to transfection, medium was changed to medium containing FCS but without antibiotics. For Cdc7 RNA interference experiments, cells were plated in 6-well format and transfected with either control siRNA or Cdc7 siRNA or left untreated. Cells were harvested at 24-h time intervals posttransfection, by washing with PBS and treating with trypsin for 3 min at 37°C.

Cell population growth assessment. Cell population growth was assessed using serial measurement of cell suspensions on the Z2 series Coulter Counter (Beckman Coulter) following the manufacturer's instructions. Phase contrast microscopy was done with an inverted Axiovert 200M (Cell Observer; Zeiss), equipped with an AxioCam CCD camera (Zeiss) and Axiovision software.

Flow cytometric analysis. Bivariate flow cytometric analysis was done as described (24). Apoptosis was quantified using the Annexin V-FITC Apoptosis Detection Kit (BioVision). Cells were stained with propidium iodide and Annexin-V FITC following the manufacturer's instructions and quantified by flow cytometry.

Preparation of total cell protein extracts and immunoblotting. Cells harvested by treatment with trypsin were washed in PBS and resuspended in radio immunoprecipitation assay buffer [Tris-Cl (pH 7.4) 50 mmol/L, NaCl 150 mmol/L, NP40 1%, sodium deoxycholate 1%, SDS 0.1%, EDTA 1 mmol/L] at 5 × 106 cells/mL. After incubation on ice for 30 min the lysate was sonicated for 10 s, clarified by high speed centrifugation (13,000 g, 15 min, 4°C) and stored at −80°C. For immunoblotting, 50 μg protein extract were loaded in each lane and separated by 4–20% SDS-PAGE. Protein was transferred from polyacrylamide gels onto nitrocellulose membranes (Amersham Biosciences) by semidry electroblotting in transfer buffer. Blocking, antibody incubations, and washing steps were done as described (17) using the following conditions: PBS/10% milk for Cdc7 and PBS/5% milk for all other antibodies.

Statistical methods. Spearman's rank correlation coefficient was used to examine associations between Cdc7 and the combinatorial ratios, geminin/Ki67, and Mcm2/Ki67. Relationships between Cdc7 expression and tumor grade, stage and ploidy status were assessed using nonparametric Jonckheere-Terpstra and Mann-Whitney U tests as appropriate. Analysis of disease-free and overall survival data was carried out using Kaplan-Meier plot (using tertiles for Cdc7), log-rank test, and Cox regression (treating Cdc7 expression level as a continuous variable). Hazard ratios (HR) with 95% confidence intervals (95% CI) for Cdc7 were estimated firstly unadjusted (univariate analysis), and secondly adjusted for age, grade, and stage (multivariate analysis). Patients with incomplete data were excluded from the multivariate analysis. All tests were two-sided and used a significance level of 0.05. Analysis was carried out using SPSS 12.0 for Windows (SPSS).

Cdc7 expression in normal ovarian tissue and epithelial ovarian cancer. Monospecificity of a commercially available antibody against Cdc7 was confirmed in total cell extracts from asynchronous WI-38, SKOV-3, HeLa S3, and HL60 cells by detection of a single protein with a molecular mass consistent with the reported electrophoretic mobility of the corresponding human antigen (Fig. 2A). Next we studied Cdc7 kinase expression in proliferating human cells and out-of-cycle states. Cdc7 is present throughout the mitotic cell division cycle in synchronized HeLa cells but is tightly down-regulated in quiescent (G0) human diploid fibroblasts (0 hours) and during differentiation of myeloblastic HL60 cells (Fig. 2B). Following release from G0, protein levels of Cdc7 accumulate by 12 hours, prior to reentry into S phase. This indicates that Cdc7 constitutes a proliferation marker and that down-regulation of this kinase contributes to the replication arrest and loss of proliferative capacity that characterize the quiescent (G0) and differentiated out-of-cycle states (12). Next, we analyzed the expression dynamics of Cdc7 in diagnostic tumor specimens from a published cohort of 143 epithelial ovarian carcinoma cases (17) and in 5 cases of normal ovarian tissue. Only a very small percentage of cells (0.2%) expressed Cdc7 in normal ovarian surface epithelium (Fig. 2C), consistent with reports showing repression of replication licensing in quiescent and differentiated out-of-cycle states (21, 22, 24). Similar results were seen in normal fallopian tube epithelium (Supplementary Fig. S1), a site which has recently been implicated as a source of some ovarian carcinomas (25). On the contrary, in invasive carcinoma we found high levels of Cdc7 expression (Fig. 2C), indicative of a hyperproliferative state and supporting the hypothesis that this kinase is deregulated during the malignant transformation of epithelial ovarian cells.

Fig. 2.

Cdc7 is deregulated in epithelial ovarian carcinoma. A, immunoblots of asynchronously proliferating SKOV-3, HeLa S3, WI-38, and HL60 total cell lysates with a mouse monoclonal antibody against Cdc7. B, immunoblots of Cdc7 and actin (loading control) in total cell lysates from synchronized HeLa S3 cells, WI-38 released from density-dependent growth arrest in G0, and HL60 cells following induction of the monocyte/macrophage differentiation program (tables show the percentage of cells in G1, S, and G2/M phases at the indicated time points). C, photomicrographs of normal ovarian surface epithelium and high-grade epithelial ovarian carcinoma stained with a Cdc7 antibody (original magnification 400×; inset 1000×).

Fig. 2.

Cdc7 is deregulated in epithelial ovarian carcinoma. A, immunoblots of asynchronously proliferating SKOV-3, HeLa S3, WI-38, and HL60 total cell lysates with a mouse monoclonal antibody against Cdc7. B, immunoblots of Cdc7 and actin (loading control) in total cell lysates from synchronized HeLa S3 cells, WI-38 released from density-dependent growth arrest in G0, and HL60 cells following induction of the monocyte/macrophage differentiation program (tables show the percentage of cells in G1, S, and G2/M phases at the indicated time points). C, photomicrographs of normal ovarian surface epithelium and high-grade epithelial ovarian carcinoma stained with a Cdc7 antibody (original magnification 400×; inset 1000×).

Close modal

Increased Cdc7 expression is associated with arrested tumor differentiation, advanced clinical stage, and genomic instability. Next we tested for an association between Cdc7 expression and conventional predictors of clinical outcome across the series of 143 cases. Cdc7 expression levels were highly significantly associated with tumor grade of differentiation (P = 0.004), increasing with the degree of tumor anaplasia or arrested differentiation (Fig. 3A). There was also a highly significant association between tumor DNA ploidy status and Cdc7 expression levels (P < 0.001), reflecting an increased proportion of cycling cells in aneuploid tumors as compared with diploid tumors (Fig. 3B). In line with these data, we found a highly significant association between tumor grade and DNA ploidy status (P < 0.001), highlighting the intricate link among cell cycle deregulation, tumor anaplasia, and genomic instability in this tumor type (10, 17).

Fig. 3.

Relationship between Cdc7 and clinicopathologic parameters. Box-whisker plots showing labeling indices for Cdc7 in relation to (A) tumor grade of differentiation (grade 1, n = 16; grade 2, n = 33; grade 3, n = 75), (B) tumor DNA ploidy status (aneuploid, n = 91; diploid, n = 52) and (C) tumor stage (early, n = 48; advanced, n = 94). The median (solid black line), interquartile range (boxed), and range (enclosed by lines) of Cdc7 expression are shown (outlying cases depicted by isolated points).

Fig. 3.

Relationship between Cdc7 and clinicopathologic parameters. Box-whisker plots showing labeling indices for Cdc7 in relation to (A) tumor grade of differentiation (grade 1, n = 16; grade 2, n = 33; grade 3, n = 75), (B) tumor DNA ploidy status (aneuploid, n = 91; diploid, n = 52) and (C) tumor stage (early, n = 48; advanced, n = 94). The median (solid black line), interquartile range (boxed), and range (enclosed by lines) of Cdc7 expression are shown (outlying cases depicted by isolated points).

Close modal

Increased Cdc7 expression is associated with accelerated cell cycle progression in epithelial ovarian carcinoma. Cdc7 expression was significantly associated with tumor stage (P = 0.01), suggesting that its deregulation occurs during progression from early- to advanced-stage disease (Fig. 3C). We have recently shown that advancing tumor stage is linked to accelerated cell cycle progression in epithelial ovarian carcinoma, as indicated by an increase in the geminin/Ki67 ratio and a decrease in the Mcm2/Ki67 ratio, the latter finding reflecting a switch from a replication-licensed but nonproliferating state to an actively cycling state (11, 17). In the same cohort, we found a significant positive correlation between Cdc7 expression and the geminin/Ki67 ratio (Spearman correlation coefficient, 0.23; 95% CI, 0.07-0.38; P = 0.006), indicating that increased Cdc7 expression is coupled to accelerated cell cycle progression observed in advanced-stage epithelial ovarian carcinoma, consistent with the rate-limiting role of Cdc7 for S phase entry and progression. Increased Cdc7 expression was also found to strongly correlate with the expression of other markers of S-G2-M phase progression including geminin (Spearman correlation coefficient, 0.43; 95% CI, 0.29-0.55; P < 0.001), Aurora A (0.50; 95% CI, 0.37-0.61; P < 0.001), Aurora B (0.53; 95% CI, 0.40-0.64; P < 0.001), and phospho-histone H3 (H3S10ph; 0.48; 95% CI, 0.34-0.60; P < 0.001; Fig. 1; ref. 11). Furthermore, there was a significant negative correlation between Cdc7 expression and the Mcm2/Ki67 ratio (Spearman correlation coefficient, −0.24; 95% CI, -0.39 to -0.08; P = 0.004), indicating that the switch from a replication-licensed but nonproliferating state to an actively cycling state is linked to an increase in Cdc7 expression (11). These findings suggest that deregulation of Cdc7 expression may facilitate the accelerated cell cycle transit times characteristic of advanced-stage epithelial ovarian carcinoma.

Cdc7 is an independent predictor of patient survival in epithelial ovarian cancer. Univariate analysis showed that Cdc7 kinase levels predict both disease-free survival (HR, 1.96; CI, 1.53-2.52; P < 0.001) and overall survival (HR, 1.68; CI, 1.16-2.44; P = 0.007; Fig. 4A) in this cohort, further implicating Cdc7 deregulation in the development of aggressive phenotypes in epithelial ovarian carcinoma. Importantly, Cdc7 retained its independent prognostic value in predicting disease-free survival (HR, 2.03; CI, 1.53-2.68; P < 0.001; Fig. 4B) and overall survival (HR, 1.69; CI, 1.16-2.46; P = 0.006) in multivariate analysis. Notably, Cdc7 showed prognostic utility over and above tumor FIGO stage (Fig. 4C), the only conventional prognostic indicator with independent value in predicting relapse (P < 0.001) in this cohort. Relative to stage 1 to 2 patients, the HR for relapse for patients with stage 3 to 4 tumors was 3.72 (95% CI, 2.11-6.55). When the tertiles of Cdc7 (lower tertile, <6.5%; middle tertile, 6.5-12.5%; upper tertile, >12.5%) were incorporated within stage, then relative to those patients with stage 1 to 2 and in the lower tertile of Cdc7 (n = 5/23 events/patients), the HR for relapse was 10.90 (CI, 4.07-29.17) for the stage 3 to 4/upper Cdc7 tertile (n = 26/30 events/patients) group. Thus, Cdc7 expression analysis can improve the prognostic value of tumor stage by approximately 3-fold. Notably, the greatest discriminatory effect of Cdc7 expression was observed for stage 3 and 4 tumors, which make up the vast majority of epithelial ovarian carcinoma cases (26). At present there are few prognostic indicators available to guide the clinical management of these tumors (27). Therefore, the independent prognostic information obtained from Cdc7 could be of major benefit for this group of patients.

Fig. 4.

Cdc7 is an independent predictor of survival. Survival curves showing association between Cdc7 and patient survival. A, Kaplan-Meier survival plot shows that Cdc7 expression predicts overall survival in this study cohort (HR, 1.68; 1.16-2.44; P = 0.007). The cohort was split into tertiles of Cdc7 expression (lower tertile, <6.5% of cells positive for Cdc7, n = 8/47 events/patients; middle tertile, 6.5-12.5%, n = 8/48 events/patients; upper tertile >12.5%, n = 12/48 events/patients). B, multivariate survival plot shows that Cdc7 predicts disease-free survival independent of patient age, tumor grade and stage (HR, 2.03; 1.53-2.68; P < 0.001). The cohort was split into tertiles of Cdc7 expression (lower tertile, <6.5% of cells positive for Cdc7, n = 17/47 events/patients; middle tertile, 6.5-12.5%, n = 18/48 events/patients; upper tertile >12.5%, n = 32/48 events/patients). C, Kaplan-Meier survival plot shows that Cdc7 has independent prognostic value in predicting disease-free survival over and above tumor stage, the hazard ratio for relapse increasing to 10.90 (CI, 4.07-29.17) for the stage 3 to 4/upper Cdc7 tertile group relative to those patients with stage 1 to 2 and in the lower tertile of Cdc7 (stage 1 to 2/lower tertile, n = 5/23 events/patients; stage 1 to 2/middle tertile, n = 6/23 events/patients; stage 1 to 2/upper tertile, n = 5/16 events/patients; stage 3 to 4/lower tertile, n = 12/24 events/patients; stage 3 to 4/middle tertile, n = 12/24 events/patients; stage 3 to 4/upper tertile, n = 26/30 events/patients).

Fig. 4.

Cdc7 is an independent predictor of survival. Survival curves showing association between Cdc7 and patient survival. A, Kaplan-Meier survival plot shows that Cdc7 expression predicts overall survival in this study cohort (HR, 1.68; 1.16-2.44; P = 0.007). The cohort was split into tertiles of Cdc7 expression (lower tertile, <6.5% of cells positive for Cdc7, n = 8/47 events/patients; middle tertile, 6.5-12.5%, n = 8/48 events/patients; upper tertile >12.5%, n = 12/48 events/patients). B, multivariate survival plot shows that Cdc7 predicts disease-free survival independent of patient age, tumor grade and stage (HR, 2.03; 1.53-2.68; P < 0.001). The cohort was split into tertiles of Cdc7 expression (lower tertile, <6.5% of cells positive for Cdc7, n = 17/47 events/patients; middle tertile, 6.5-12.5%, n = 18/48 events/patients; upper tertile >12.5%, n = 32/48 events/patients). C, Kaplan-Meier survival plot shows that Cdc7 has independent prognostic value in predicting disease-free survival over and above tumor stage, the hazard ratio for relapse increasing to 10.90 (CI, 4.07-29.17) for the stage 3 to 4/upper Cdc7 tertile group relative to those patients with stage 1 to 2 and in the lower tertile of Cdc7 (stage 1 to 2/lower tertile, n = 5/23 events/patients; stage 1 to 2/middle tertile, n = 6/23 events/patients; stage 1 to 2/upper tertile, n = 5/16 events/patients; stage 3 to 4/lower tertile, n = 12/24 events/patients; stage 3 to 4/middle tertile, n = 12/24 events/patients; stage 3 to 4/upper tertile, n = 26/30 events/patients).

Close modal

Cdc7 depletion triggers high levels of apoptosis in SKOV-3 and Caov-3 ovarian cancer cells. To assess the therapeutic potential of Cdc7 in epithelial ovarian carcinoma, we used RNA interference to down-regulate Cdc7 expression in SKOV-3 (p53-null/Rb-negative) and Caov-3 (p53-negative/Rb-negative) ovarian cancer cells (2831). Transfection of SKOV-3 and Caov-3 cells, and HBEpC human bronchial epithelial cells and IMR-90 human diploid fibroblasts (untransformed control cells) with Cdc7 siRNAs, but not control siRNAs, reduced Cdc7 mRNA levels by ∼90% in all cell lines by 72 hours posttransfection. Correspondingly, Cdc7 protein levels became almost undetectable and the phosphorylation of Mcm2 serine 53 (Mcm2S53ph), a known target residue for Cdc7 kinase, was absent (Fig. 5A, Supplementary Figs. S2A and S3A, and data not shown). Down-regulation of Cdc7 expression resulted in a decline in cell numbers in the population of Cdc7-depleted SKOV-3 cells by 96 hours, whereas 8.2- and 6.7-fold increases were noted for untreated and control siRNA-transfected cells over the same time period (Fig. 5B). Similarly, in Caov-3 cells, which have a lower transfection efficiency compared with SKOV-3 cells, a 0.3-fold increase in cell numbers was measured by 96 hours posttransfection with Cdc7 siRNA, significantly lower than the 1.7-fold increase observed for both untreated and control siRNA–transfected cells (Supplementary Fig. S2B). HBEpC cells transfected with Cdc7 siRNA showed a small population increase (0.7-fold) by 96 hours posttransfection, compared with the 3.9- and 3.4-fold increases seen in untreated cells and cells transfected with control siRNAs (Supplementary Fig. S3B). Similarly, Cdc7 siRNA–transfected IMR-90 cells grew at half the growth rate of untreated or control siRNA–transfected cells over a period of 72 hours posttransfection, after which cell proliferation ceased completely with no apparent cell loss (data not shown). The decrease in cell numbers after Cdc7 knockdown was observed with two different Cdc7 siRNAs, suggesting that this phenotype directly correlates with Cdc7 depletion and is unlikely to be due to concomitant nonspecific down-regulation of an unknown gene.

Fig. 5.

Cdc7 knockdown in SKOV-3 cells induces an abortive S phase and apoptotic cell death. A, Western blot analysis of untreated, control siRNA- and Cdc7 siRNA–transfected SKOV-3 cell lysates probed with antibodies against Cdc7; Mcm2S53ph; pro-caspases 3, 8, and 9; PARP-1; and actin (loading control). B, cell growth rate of untreated, control siRNA- and Cdc7 siRNA–transfected SKOV-3 cells shown as fold increase in cell numbers relative to starting population. C, DNA content of SKOV-3 cells transfected with Cdc7 siRNA at 0 h and 72 h posttransfection. D, phase contrast micrographs of SKOV-3 cells transfected with control siRNA or Cdc7 siRNA at 0 h and 72 h posttransfection (original magnification 200×; inset 1,000×) and confocal fluorescence microscopy images of propidium iodide–stained SKOV-3 cells transfected with control siRNA or Cdc7 siRNA 72 h posttransfection. E, flow cytometric monitoring of apoptosis by Annexin-V assay of untreated SKOV-3 cells or cells treated with control or Cdc7 siRNA. Panels show the quantification of propidium iodide and Annexin-V staining at 72 h posttransfection and show the percentage of cells in each quadrant.

Fig. 5.

Cdc7 knockdown in SKOV-3 cells induces an abortive S phase and apoptotic cell death. A, Western blot analysis of untreated, control siRNA- and Cdc7 siRNA–transfected SKOV-3 cell lysates probed with antibodies against Cdc7; Mcm2S53ph; pro-caspases 3, 8, and 9; PARP-1; and actin (loading control). B, cell growth rate of untreated, control siRNA- and Cdc7 siRNA–transfected SKOV-3 cells shown as fold increase in cell numbers relative to starting population. C, DNA content of SKOV-3 cells transfected with Cdc7 siRNA at 0 h and 72 h posttransfection. D, phase contrast micrographs of SKOV-3 cells transfected with control siRNA or Cdc7 siRNA at 0 h and 72 h posttransfection (original magnification 200×; inset 1,000×) and confocal fluorescence microscopy images of propidium iodide–stained SKOV-3 cells transfected with control siRNA or Cdc7 siRNA 72 h posttransfection. E, flow cytometric monitoring of apoptosis by Annexin-V assay of untreated SKOV-3 cells or cells treated with control or Cdc7 siRNA. Panels show the quantification of propidium iodide and Annexin-V staining at 72 h posttransfection and show the percentage of cells in each quadrant.

Close modal

Fluorescence-activated cell sorting analysis of Cdc7-depleted SKOV-3 cells (Fig. 5C) showed that by 24 hours posttransfection cells had accumulated with G1 DNA content, with only a small fraction of S phase cells (8% of cell population compared with 20% and 17% in untreated and control siRNA–transfected cells, respectively). A fraction of cells with less than 2C DNA content (8%) was noted in Cdc7 siRNA–transfected cells, indicative of cell death, which was not present in untreated or control siRNA–transfected cells. By 48 hours, the proportion of cells with sub-G1 DNA content had doubled to 17% in Cdc7-depleted cells, with a concomitant decrease in cells with G2-M DNA content (10% compared with 20% and 18% in untreated and control siRNA–transfected cells, respectively). By 72 hours posttransfection, the majority of Cdc7-depleted cells had either less than 2C DNA content (37%) or G1 DNA content (39%), with a small fraction of the cell population in S-G2-M phase. A similar trend was observed in the second ovarian cancer cell line, Caov-3 (Supplementary Fig. S2C). Down-regulation of Cdc7 expression in the HBEpC and IMR-90 lines caused an accumulation of cells with G1 DNA content (70% and 84%, respectively, of the population by 96 hours) and a reduction in S (10% and 5%) and G2-M fractions (19% and 12%). The fraction of cells with less than 2C DNA content 96 hours posttransfection was low (6%; Supplementary Fig. S3C) for HBEpC and negligible (∼1%; data not shown) for IMR-90.

Phase contrast microscopy confirmed that the progressive decline in SKOV-3 and Caov-3 cell numbers following Cdc7 depletion was due to cell death, whereas control cells and Cdc7 siRNA–transfected HBEpC and IMR-90 cells remained viable (Fig. 5D, Supplementary Figs. S2D and S3D, and data not shown). Moreover, confocal fluorescence microscopy of Cdc7-depleted SKOV-3 cells revealed characteristic morphologic changes associated with apoptosis, including cell shrinkage and nuclear blebbing, which were not observed in control cells or Cdc7 siRNA–transfected HBEpC or IMR-90 cells. Morphologic changes such as cell shrinkage and nuclear blebbing together with detection of a sub-G1 population by fluorescence-activated cell sorting suggests that apoptosis was activated in Cdc7-depleted SKOV-3 cells. Further evidence for apoptosis was obtained by Annexin V-FITC detection of phosphatidylserine on the cell surface. Annexin V-FITC staining was detected in 32% of Cdc7 siRNA–transfected SKOV-3 cells. This was significantly higher than observed in cells not treated (1%), or treated with control siRNA (19%; Fig. 5E). Indeed, in SKOV-3 and Caov-3 cell extracts prepared 72 hours posttransfection there was a marked decrease in levels of the pro-caspases 3, 8, and 9 and full-length poly (ADP-ribose) polymerase-1 (Fig. 5A, Supplementary Fig. S2A), and the active fragment of poly (ADP-ribose) polymerase-1 was detectable, confirming activation of the apoptotic pathway (Fig. 5A). Taken together, these data suggest that Cdc7 depletion in SKOV-3 and Caov-3 cells triggers a block or strong delay in S phase progression leading to cell death in a subpopulation of cells, whereas untransformed cells enter a G1-arrested state.

The ultimate goal in the management of epithelial ovarian carcinoma is tumor cell–specific killing with negligible effects on normal cells. The pursuit of this goal over recent years has seen a dramatic shift in anticancer drug development, from conventional platinum-based chemotherapy to targeted agents (32). Targeted agents, for the most part, spare normal cells and have the potential to be well-tolerated therapies that increase survival and improve quality of life. However, despite a rapid increase in the identification of such molecular targets, at present relatively few predict response to mechanism-based therapeutic agents (2). Purely prognostic markers have an important role in the management of epithelial ovarian carcinoma (1), but the aim of personalized cancer medicine can only be achieved through the development of biomarkers with both prognostic and predictive value.

The success of this approach has long been recognized in the treatment of breast cancer, firstly with the use of tamoxifen and aromatase inhibitors in hormone receptor–expressing tumors, and more recently with the advent of trastuzumab in Her2-positive tumors. For a biomarker to predict a therapeutic response in epithelial ovarian carcinoma, it must be linked to its pathobiology. For example, it should be deregulated during epithelial ovarian tumorigenesis and therefore reflect the malignant potential of the tumor. Secondly, it must be readily and accurately detectable in diagnostic tissue specimens or ascitic fluid, and its expression should provide prognostic information. Finally, its targeted inhibition or restoration of function should show a tumor-specific killing effect, the intensity of which may be predicted from the molecule's expression level. The findings we report here indicate that Cdc7 kinase has the potential to fulfill all of these criteria.

Our data show that analysis of Cdc7 expression in diagnostic tissue specimens can distinguish between indolent and aggressive epithelial ovarian tumors. In our cohort, there were highly significant associations between Cdc7 expression and conventional biological indicators of poor clinical outcome, namely tumor anaplasia, aneuploidy, and advanced FIGO stage. Increased Cdc7 expression was also linked to accelerated cell cycle progression and reduction in the length of G1 phase, indicative of increasingly aggressive tumor behavior. These data are supported by the finding that Cdc7 is an independent prognostic indicator of both relapse and overall survival. Importantly, Cdc7 provides additional discriminatory prognostic information in advanced-stage disease, which may be invaluable in this patient group. The association between increased Cdc7 expression levels and the development of an aggressive tumor phenotype further reinforces the hypothesis that Cdc7 is an attractive anticancer target in epithelial ovarian carcinoma.

Depletion of Cdc7 by RNA interference has identified this essential kinase as a potent anticancer target in epithelial ovarian carcinoma. Cdc7 knockdown in SKOV-3 and Caov-3 cells causes an abortive S phase leading to p53-independent apoptotic cell death. By contrast, untransformed cells reversibly arrest in G1 following Cdc7 knockdown and remain in a viable nonproliferative state. Our findings are in keeping with data reported for other transformed cell lines (20, 33). It has been postulated that in untransformed cells inhibition of origin firing may trigger a “licensing checkpoint,” leading to a block to DNA replication initiation and stable G1 arrest (20). Cancer cells seem to have lost this checkpoint function and initiate DNA synthesis with only a subset of replication origins primed, which may result in the stalling and collapse of replication forks and double-strand breaks, triggering intra-S phase and/or G2-M checkpoints and apoptosis.

Taken together, our data show that Cdc7 is a strong independent prognostic marker in epithelial ovarian carcinoma and that targeted Cdc7 inhibition can lead to tumor cell–specific killing. The main aim of future studies with emerging small-molecule Cdc7 inhibitors (33, 34) will be to assess translational mechanism-based end points in parallel with survival and quality of life outcomes in this highly heterogeneous malignancy.

No potential conflicts of interest were disclosed.

Grant support: Cancer Research UK Programme Grant C428/A6263.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Note: Supplementary data for this article are available at Clinical Cancer Research Online (http://clincancerres.aacrjournals.org/).

A.A. Kulkarni, S.R. Kingsbury, and S. Tudzarova contributed equally to this work.

We are grateful to Dr Elisabetta Leo for her assistance in studying protein expression profiles in cell lysates.

1
Aletti GD, Gallenberg MM, Cliby WA, Jatoi A, Hartmann LC. Current management strategies for ovarian cancer.
Mayo Clin Proc
2007
;
82
:
751
–70.
2
Crijns AP, Duiker EW, de Jong S, Willemse PH, van der Zee AG, de Vries EG. Molecular prognostic markers in ovarian cancer: toward patient-tailored therapy.
Int J Gynecol Cancer
2006
;
16
:
152
–65.
3
Psyrri A, Kassar M, Yu Z, et al. Effect of epidermal growth factor receptor expression level on survival in patients with epithelial ovarian cancer.
Clin Cancer Res
2005
;
11
:
8637
–43.
4
Mustea A, Sehouli J, Fabjani G, et al. Epidermal growth factor receptor (EGFR) mutation does not correlate with platinum resistance in ovarian carcinoma. Results of a prospective pilot study.
Anticancer Res
2007
;
27
:
1527
–30.
5
Posadas EM, Liel MS, Kwitkowski V, et al. A phase II and pharmacodynamic study of gefitinib in patients with refractory or recurrent epithelial ovarian cancer.
Cancer
2007
;
109
:
1323
–30.
6
Seiden MV, Burris HA, Matulonis U, et al. A phase II trial of EMD72000 (matuzumab), a humanized anti-EGFR monoclonal antibody, in patients with platinum-resistant ovarian and primary peritoneal malignancies.
Gynecol Oncol
2007
;
104
:
727
–31.
7
Tuefferd M, Couturier J, Penault-Llorca F, et al. HER2 status in ovarian carcinomas: a multicenter GINECO study of 320 patients.
PLoS ONE
2007
;
2
:
e1138
.
8
Serrano-Olvera A, Duenas-Gonzalez A, Gallardo-Rincon D, Candelaria M, De la Garza S. Prognostic, predictive and therapeutic implications of HER2 in invasive epithelial ovarian cancer.
Cancer Treat Rev
2006
;
32
:
180
–90.
9
Hanahan D, Weinberg RA. The hallmarks of cancer.
Cell
2000
;
100
:
57
–70.
10
Evan GI, Vousden KH. Proliferation, cell cycle and apoptosis in cancer.
Nature
2001
;
411
:
342
–8.
11
Williams GH, Stoeber K. Cell cycle markers in clinical oncology.
Curr Opin Cell Biol
2007
;
19
:
672
–9.
12
Blow JJ, Hodgson B. Replication licensing-defining the proliferative state.
Trends Cell Biol
2002
;
12
:
72
–8.
13
Coller HA. What's taking so long? S-phase entry from quiescence versus proliferation.
Nat Rev Mol Cell Biol
2007
;
8
:
667
–70.
14
Machida YJ, Hamlin JL, Dutta A. Right place, right time, and only once: replication initiation in metazoans.
Cell
2005
;
123
:
13
–24.
15
Shreeram S, Blow JJ. The role of the replication licensing system in cell proliferation and cancer.
Prog Cell Cycle Res
2003
;
5
:
287
–93.
16
Wohlschlegel JA, Kutok JL, Weng AP, Dutta A. Expression of geminin as a marker of cell proliferation in normal tissues and malignancies.
Am J Pathol
2002
;
161
:
267
–73.
17
Kulkarni AA, Loddo M, Leo E, et al. DNA replication licensing factors and Aurora kinases are linked to aneuploidy and clinical outcome in epithelial ovarian carcinoma.
Clin Cancer Res
2007
;
13
:
6153
–61.
18
Masai H, You Z, Arai K. Control of DNA replication: regulation and activation of eukaryotic replicative helicase, MCM.
IUBMB Life
2005
;
57
:
323
–35.
19
Costanzo V, Gautier J. Single-strand DNA gaps trigger an ATR- and Cdc7-dependent checkpoint.
Cell Cycle
2003
;
2
:
17
.
20
Montagnoli A, Tenca P, Sola F, et al. Cdc7 inhibition reveals a p53-dependent replication checkpoint that is defective in cancer cells.
Cancer Res
2004
;
64
:
7110
–6.
21
Kingsbury SR, Loddo M, Fanshawe T, et al. Repression of DNA replication licensing in quiescence is independent of geminin and may define the cell cycle state of progenitor cells.
Exp Cell Res
2005
;
309
:
56
–67.
22
Barkley LR, Hong HK, Kingsbury SR, James M, Stoeber K, Williams GH. Cdc6 is a rate-limiting factor for proliferative capacity during HL60 cell differentiation.
Exp Cell Res
2007
;
313
:
3789
–99.
23
Stoeber K, Tlsty TD, Happerfield L, et al. DNA replication licensing and human cell proliferation.
J Cell Sci
2001
;
114
:
2027
–41.
24
Eward KL, Obermann EC, Shreeram S, et al. DNA replication licensing in somatic and germ cells.
J Cell Sci
2004
;
117
:
5875
–86.
25
Lee Y, Miron A, Drapkin R, et al. A candidate precursor to serous carcinoma that originates in the distal fallopian tube.
J Pathol
2007
;
211
:
26
–35.
26
Holschneider CH, Berek JS. Ovarian cancer: epidemiology, biology, and prognostic factors.
Semin Surg Oncol
2000
;
19
:
3
–10.
27
Canevari S, Gariboldi M, Reid JF, Bongarzone I, Pierotti MA. Molecular predictors of response and outcome in ovarian cancer.
Crit Rev Oncol Hematol
2006
;
60
:
19
–37.
28
Mujoo K, Maneval DC, Anderson SC, Gutterman JU. Adenoviral-mediated p53 tumor suppressor gene therapy of human ovarian carcinoma.
Oncogene
1996
;
12
:
1617
–23.
29
Xie Y, Gilbert JD, Kim JH, Freytag SO. Efficacy of adenovirus-mediated CD/5-FC and HSV-1 thymidine kinase/ganciclovir suicide gene therapies concomitant with p53 gene therapy.
Clin Cancer Res
1999
;
5
:
4224
–32.
30
Warnick CT, Dabbas B, Ford CD, Strait KA. Identification of a p53 response element in the promoter region of the hMSH2 gene required for expression in A2780 ovarian cancer cells.
J Biol Chem
2001
;
276
:
27363
–70.
31
Saadatmandi N, Tyler T, Huang Y, et al. Growth suppression by a p14(ARF) exon 1β adenovirus in human tumor cell lines of varying p53 and Rb status.
Cancer Gene Ther
2002
;
9
:
830
–9.
32
DiSaia PJ, Bloss JD. Treatment of ovarian cancer: new strategies.
Gynecol Oncol
2003
;
90
:
S24
–32.
33
Montagnoli A, Valsasina B, Croci V, et al. A Cdc7 kinase inhibitor restricts initiation of DNA replication and has antitumor activity.
Nat Chem Biol
2008
;
4
:
357
–65.
34
Vanotti E, Amici R, Bargiotti A, et al. Cdc7 kinase inhibitors: pyrrolopyridinones as potential antitumor agents. 1. Synthesis and structure-activity relationships.
J Med Chem
2008
;
51
:
487
–501.

Supplementary data