Human apurinic/apyrimidinic endonuclease (APE1) is a prognostic factor in ovarian, gastro-oesophageal and pancreatico-biliary cancers

Background: Altered DNA repair may be associated with aggressive tumour biology and impact upon response to chemotherapy and radiotherapy. We investigated whether expression of human AP endonuclease (APE1), a key multifunctional protein involved in DNA BER, would impact on clinicopathological outcomes in ovarian, gastro-oesophageal, and pancreatico-biliary cancer. Methods: Formalin-fixed human ovarian, gastro-oesophageal, and pancreatico-biliary cancers were constructed into TMAs. Expression of APE1 was analysed by IHC and correlated to clinicopathological variables. Results: In ovarian cancer, nuclear APE1 expression was seen in 71.9% (97 out of 135) of tumours and correlated with tumour type (P=0.006), optimal debulking (P=0.009), and overall survival (P=0.05). In gastro-oesophageal cancers previously exposed to neoadjuvant chemotherapy, 34.8% (16 out of 46) of tumours were positive in the nucleus and this correlated with shorter overall survival (P=0.005), whereas cytoplasmic localisation correlated with tumour dedifferentiation (P=0.034). In pancreatico-biliary cancer, nuclear staining was seen in 44% (32 out of 72) of tumours. Absence of cytoplasmic staining was associated with perineural invasion (P=0.007), vascular invasion (P=0.05), and poorly differentiated tumours (P=0.068). A trend was noticed with advanced stage (P=0.077). Conclusions: Positive clinicopathological correlations of APE1 expression suggest that APE1 is a potential drug target in ovarian, gastro-oesophageal, and pancreatico-biliary cancers.

DNA base excision repair (BER) is involved in the repair of bases that have been damaged by alkylation, oxidation, or ringsaturation in addition to processing deaminated bases and DNA single-strand breaks. Although there is more than one subpathway of BER, in most cases excision of a damaged base by a DNA glycosylase enzyme leads to the formation of a potentially cytotoxic apurinic/apyrimidinic (AP) site intermediate (Hickson et al, 2000). This is a target for an AP endonuclease (APE1), which cleaves the phosphodiester backbone on the 5 0 side of the AP site via a hydrolytic mechanism. The major AP endonuclease in human cells, APE1 (also called previously HAP1, Ref-1, and APEX1), accounts for over 95% of the total AP endonuclease activity in most cultured human cell lines (Demple et al, 1991;Robson and Hickson, 1991). In addition to its DNA-repair activity, APE1 also regulates redox function (Robson and Hickson, 1991) and transcription (Bhakat et al, 2003;Fuchs et al, 2003).
Using either antisense oligonucleotides or siRNA approaches, several groups have reported that depletion of intracellular APE1 sensitises mammalian cells to a variety of DNA-damaging agents (Chen and Olkowski, 1994;Ono et al, 1994;Walker et al, 1994;Robertson et al, 1997;Evans et al, 2000;Silber et al, 2002). In pancreatic cancer cell lines for example, downregulation of APE1 potentiated the cytotoxicity of gemcitabine (Lau et al, 2004). APE1 downregulation has also been shown to block ovarian cancer cell growth (Fishel et al, 2008). In melanoma cell lines, APE1 downregulation led to increased apoptosis, whereas APE1 overexpression conferred protection from cisplatin-or H 2 O 2induced apoptosis. (Yang et al, 2005).
We have undertaken a drug discovery programme to isolate small-molecule inhibitors of APE1 (Madhusudan et al, 2005). In the previously published study, we provided the first preclinical evidence that APE1 inhibition by a small-molecule inhibitor potentiated the cytotoxicity of a panel of base-targeting agents thought to be repaired through BER (Madhusudan et al, 2005). To identify drug-like leads, we recently adopted an industry-standard, high-throughput, virtual screening strategy using first-generation inhibitors as templates and screened a chemical library of 2.6 million compounds. Several novel inhibitors of APE1 were identified with the ability to potentiate the cytoxicity of alkylating agents (unpublished data).
In the current study, we provide evidence that APE1 is a prognostic marker in ovarian, gastric, and pancreatico-biliary cancer. We demonstrate positive clinicopathological correlations of APE1 nuclear expression in ovarian, gastric, and pancreaticobiliary cancers, suggesting that APE1 is also a potential drug target in these tumours.

Study design and setting
Expression of APE1 in ovarian cancer was investigated using a tissue microarray (TMA) of 157 ovarian cancer cases. Tissue was obtained from patients with primary ovarian cancer treated at Nottingham University Hospitals (NUH) between 2000 and 2007. Survival was calculated from the operation date until 1 April 2009 when any remaining survivors were censored. During the study period, patients were treated with either single agent carboplatin (65 patients (41.4%)) or platinum-based combination chemotherapy (89 patients (56.7%): 79 out of 89 patients received carboplatin and paclitaxel; 4 out of 89 received cyclophosphamide/doxorubicin/cis-platinum (CAP); 3 out of 89 were treated in the ICON-5 trial using paclitaxel, carboplatin plus gemcitabine, or topotecan (Bookman et al, 2009); and 3 out of 89 were treated according to the SCOTROC trial protocol by carboplatin, docetaxel ± topotecan; Kaye et al, 2009). Platinum resistance was defined as patients who had progression during first-line platinum chemotherapy or relapse within 6 months after treatment.
Expression of APE1 in gastro-oesophageal cancers was investigated using two TMA sets. The first set consisted of 142 gastric/ gastro-oesophageal cancer cases not exposed to neoadjuvant chemotherapy. With recent incorporation of neoadjuvant chemotherapy as a standard treatment option for operable gastro-oesophageal tumours (Cunningham et al, 2006), we also established a second TMA of 103 gastric/gastro-oesophageal cancer cases exposed to preoperative platinum-based chemotherapy. Tissue was obtained from patients treated at Nottingham University Hospitals (NUH) between 2001 and 2008. Survival was calculated from the date of diagnosis until 13 January 2009 when any remaining survivors were censored. During the study period, patients in the neoadjuvant arm with adenocarcinomas were treated with either neoadjuvant ECF (epirubicin (50 mg m À2 ), cisplatin (60 mg m À2 ), and continuous infusional 5-FU (200 mg m À2 per day)) or ECX (epirubicin (50 mg m À2 ), cisplatin (60 mg m À2 ), and capecetabine (625 mg m À2 p.o. b.d continuously)) chemotherapy up to three cycles prior to surgery. Patients with squamous cell carcinoma were treated with CF (cisplatin (80 mg m À2 ) and infusional 5-FU (1000 mg m À2 daily for 4 days); Allum et al, 2009) chemotherapy up to two cycles prior to surgery.
Expression of APE1 in pancreatico-biliary cancer was investigated using a TMA of 120 cases. All the participants were patients who underwent pancreatic resection (for benign or malignant pancreatic disease) between June 2001 and June 2006 at NUH Queen's Medical Centre Campus. Cores from 72 malignant tumours were selected for analysis within this study, of which 34 had pancreatic adenocarcinoma (PAC), 20 had ampullary adenocarcinoma (AAC), and 18 had cholangiocarcinoma. Only 15 patients had received adjuvant 5-FU/folinic acid chemotherapy in this cohort.
The conduct of this study was approved by the Ethics Committee of Nottingham University Hospitals.

Construction of TMA
Tissue microarrays were constructed as described previously (Kononen et al, 1998). In short, area-specialised histopathologists identified and marked formalin-fixed, paraffin-embedded tissue blocks containing tumour tissue on haematoxylin-and eosinstained slides. The marked areas in these donor paraffin blocks were used to construct the TMA. Triplicate tissue cores with a diameter of 0.6 mm were taken from the marked areas and arrayed into a recipient paraffin block using a tissue puncher/arrayer (Beecher Instruments, Silver Spring, MD, USA) as previously described (Kononen et al, 1998). Five-micrometre sections of the tissue array block were cut and placed on Fisherbrand Colorfrost/ Plus microscope slides (Fisher Scientific, Pittsburgh, PA, USA) for immunohistochemical staining.

Immunohistochemistry
A standard streptavidin -biotin complex method was used. Negative controls were obtained by omitting the primary antibody in each case. The tissue slides were deparaffinised with xylene and then rehydrated through five decreasing concentrations of alcohol (100, 90, 70, 50, and 30%) for 2 min each. Endogenous peroxidise activity was blocked by incubation in a 1% hydrogen peroxide/ methanol buffer. Antigen retrieval was performed by microwave treatment of the slides in sodium citrate buffer (pH 6.0) for 10 min. The slides were rinsed in phosphate buffer solution (PBS) and incubated with blocking serum diluted in PBS to block nonspecific staining. The slides were incubated for 1 h with the primary rabbit polyclonal anti-APE-1 antibody clone NB100-101 (Novus Biologicals Inc., Littleton, CO, USA) at a dilution of 1 : 500. After washing with PBS, sections were incubated with the secondary antibody (Vector Laboratories, Burlingame, CA, USA) for 30 min followed by the avidin -biotin complex for a further 30 min. 3 -3 0 -Diaminobenzidine tetrahydochloride was used as chromogen. All sections were counterstained with Gill's haematoxylin.

Evaluation of immune staining
The tumour cores were evaluated by specialist pathologists and oncologists blinded to the clinicopathological characteristics of patients. Whole-field inspection of the core was included in the assessment and intensities of staining were grouped as follows: 0 ¼ no staining; 1 ¼ weak staining; 2 ¼ moderate staining; and 3 ¼ strong staining. Cytoplasmic and nuclear staining were assessed separately for each core. Only stained malignant cells (epithelial cancer cells in ovarian cancer, ductal and acinar exocrine epithelial cells in pancreatico-billiary cancer) were included in the evaluation of staining. Stained stromal cells, inflammatory cells, and endocrine cells within islets of Langerhans were disregarded. Nuclear and cytoplasmic score were assigned separately for each core, and the mean of three cores was calculated for each sample. The results shown represent the mean score from the three cores available for each assessable patient sample. Not all cores within the TMA were suitable for immunohistochemistry (IHC) analyses due to small technical problems such that some cores were missing or lacked tumours.

Statistical analysis
Statistical analysis of data was performed using SPSS version 15.0 for Windows (SPSS Inc., Chicago, IL, USA). Univariate analysis of associations was determined using the Pearson w 2 -test. Survival rates were calculated from the time of resection until the end of the follow-up period and Kaplan -Meier curves were plotted. The statistical significance of differences between survival rates was determined using the log-rank test. Survival was censored if the patient was still alive or died of other causes. P-values o0.05 were identified as statistically significant. The threshold for significance was adjusted to o0.01 for subset analyses.

Patient demographics
Ovarian cancer This cohort comprised female patients with a median age of 61 years (range 33 -87, mean 60 years) ( Table 1a). Out of the 157 patients included in the study, 54 (34%) were dead before the end of the follow-up period. Histologically, most patients were found to have a serous cystadenocarcinoma (56%), followed by endometrioid (21%), clear-cell carcinoma (13%), mucinous cystadenocarcinoma (8%), or other types (2%). Tumours displayed poor histological differentiation in 73% of cases. The majority of tumours were classified as FIGO stage III (45%).
Gastro-oesophageal cancer There were two groups of patients: those who received at least one cycle of neoadjuvant chemotherapy (neoadjuvant group) and those who underwent primary surgery (primary group) (Table 1b). There were 103 patients in the neoadjuvant group, with a median age of 63 years and 81% were male. T3 tumours were the majority, constituting 62% of cases.
A total of 142 cases were in the primary group, with a median age of 74 years: 74% were male, 53% had T3 tumours.
Pancreatico-biliary cancer Of the 72 patients with pancreaticobilairy tumours, 44 patients were male and 28 were female (Table 1c). Thirty-four had PAC, 20 had AAC, and 18 had cholangiocarcinoma. The median follow-up time was 20.4 months; 35 patients (38%) died of their cancer at the time of the observation period, while only 12 patients (13%) were still alive at 5 years following the operation. The median survival was 37.8 months.

APE1 expression and clinicopathological correlations
Ovarian cancer Apyrimidinic endonuclease-1 expression was noted both in the cytoplasmic and nuclear compartments in ovarian cancer; however, nuclear staining was more prevalent (Table 2). Of the 135 evaluable cores, 97 had nuclear expression (71.9%), but only 40 had cytoplasmic expression (29.6%). It was seen more often localised to the nucleus only (45.9%) ( Figure 1A) as compared with 3.7% showing exclusively cytoplasmic staining. There was a significant difference in the level of APE1 expression among the different histological subtypes of ovarian cancer, with serous and mucinous tumours displaying nuclear APE1 more frequently than endometrioid and clear-cell carcinomas (P ¼ 0.006) ( Table 3). Nuclear localisation of APE1 correlated with a lower likelihood of achieving optimal debulking after initial surgery (P ¼ 0.009), suggesting a more aggressive phenotype. This was also reflected on survival; APE1-positive cases had worse   (Figure 2A). No significant correlations were seen between APE1 and FIGO stage or with histological grade.
Platinum resistance (defined as patients who had progression during first-line platinum chemotherapy or relapse within 6 months after treatment) was more frequently seen in APE1positive cases (35.8%) as compared with APE1-negative patients (21.1%); however, this did not reach statistical significance (P ¼ 0.09).
Gastro-oesophageal cancer In the primary-group TMAs, 41 out of 93 tumours that were stained for APE1 showed specific nuclear staining (44.1%) and 48 out of 93 (51.6%) were positive in the cytoplasm compartment as well (Table 2). In the neoadjuvantgroup TMAs, 16 out of 46 tumours that were stained for APE1 showed specific nuclear staining (34.8%) ( Figure 1B) and 32 out of 46 (69.6%) were positive in the cytoplasm. In the neoadjuvant group, nuclear APE1 correlated with worse overall survival in    Abbreviation: APE1 ¼ apurinic/apyrimidinic endonuclease-1. A strong correlation was found (P ¼ 0.006).  Figure 2B). Cytoplasmic APE1 correlated with tumour differentiation (P ¼ 0.034). No significant correlations were seen in the primary surgery group.

DISCUSSION
Overall prognosis of advanced melanoma, ovarian, gastro-oesophageal, and pancreatico-biliary cancer remains poor. New treatment strategies are urgently required to address this largely unmet need. Cytotoxic agents (such as alkylating agents and platinum compounds) and ionising radiation used in the treatment of these tumours induce DNA damage (such as base damage and others) in cells. Cancer cells proficient in DNA repair are able to repair such DNA damage and continue to survive. This is a significant cause for therapeutic resistance and impacts negatively on patient outcomes. Pharmacological inhibition of DNA repair is likely to enhance the cytotoxicity in cancer cells and improve tumour response in patients. APE1 is a key multifunctional protein essential for DNA BER, in addition to redox regulation of transcription factors and transcriptional regulation. It is widely accepted that nuclear expression of APE1 is involved in DNA repair, whereas cytoplasmic expression may be involved in redox/ transcriptional regulation.
In the current study we have demonstrated the potential APE1 expression has as a negative prognostic factor in cancer patients.
Nuclear expression of APE1 was frequently seen in ovarian, gastrooesophageal, and pancreatico-biliary cancers. In ovarian cancer, nuclear APE1 expression was seen commonly in serous and mucinous tumours (P ¼ 0.006). In addition, sub-optimal debulking and residual tumour was seen more frequently in tumours with nuclear expression of APE1, implying aggressive tumour biology (P ¼ 0.009). In fact, this was reflected in poor overall survival in patients (P ¼ 0.05). Interestingly, we also found evidence that APE1 expression may predict platinum resistance in patients, although this did not reach statistical significance (P ¼ 0.09). The negative prognostic impact was also seen in gastro-oesophageal cancer where patients who received neoadjuvant platinum-based chemotherapy and whose tumours showed APE1 nuclear expression did significantly worse as compared with non-expressors (P ¼ 0.005). Our study is consistent with recent preclinical observations using melanoma cell lines where APE1 overexpression conferred protection from cisplatin (Yang et al, 2005). Moreover, in lung cancer, blocking APE1 function potentiated cisplatin cytotoxicity (Wang et al, 2009). These studies imply that APE1 is involved in the repair of damage induced by platinating agents, and current evidence suggests that the protective effect may be contributed by the DNA repair as well as the redoxregulatory activity of APE1. The negative prognostic significance of APE1 has also been recently demonstrated in osteosarcoma (Wang et al, 2004) and head and neck cancer (Koukourakis et al, 2001).
Apyrimidinic endonuclease-1 exhibits complex and heterogeneous staining patterns. Although the biological relevance of the compartmentalisation of APE1 is not understood, its complexity suggests that APE1 localisation in cells is tightly regulated. In addition, as the nuclear expression of APE1 is involved in DNA repair and the cytoplasmic expression may be involved in redox/ transcriptional regulation, it is speculated that the compartmentalistion may reflect distinct functions of APE1 in cells. The altered subcellular localisation of APE1 in pancreatico-biliary cancers observed in our study is distinct to observations in other tumours. Nuclear staining only was seen in 44% of all periampullary tumours. Combined nuclear and cytoplasmic expression was seen in 25% of tumours. Cytoplasmic expression of APE1 was never seen in the absence of nuclear expression. Absence of cytoplasmic expression correlated to adverse prognostic features. This is in contrast to tumours such as breast cancer. In normal breast tissue, In both cases, APE1 was associated with worse survival (log-rank test P-value ¼ 0.05 and 0.005 for ovarian and gastric cancers, respectively).
APE1 localisation is predominantly nuclear, whereas in breast carcinomas, nuclear, cytoplasmic, and nuclear/cytoplasmic stainings were observed (Kakolyris et al, 1998;Puglisi et al, 2002). Furthermore, while nuclear localisation was associated with good prognostic features (being related to better differentiation, low angiogenesis, and negative lymph node status), cytoplasmic and combined nuclear/cytoplasmic localisation were associated with poor prognostic factors, such as angiogenesis together with node and p53 positivity (Kakolyris et al, 1998;Puglisi et al, 2002). A dysregulation in nuclear vs cytoplasmic ratio towards increased cytoplasmic staining was also observed in thyroid carcinomas (Tell et al, 2000) and prostate cancer (Kelley et al, 2001).
In conclusion, we have shown prognostic implications of APE1 in ovarian, gastro-oesophageal, and pancreatico-biliary cancer. Moreover our study suggests that APE1 is a potential drug target in these tumours. The recent evidence of BER modulation using PARP inhibitors that have shown promise in recent clinical trials in ovarian and breast cancer (Lord and Ashworth, 2008), implies that APE1 inhibitors may have similar clinical application in patients.