The SK-OV-3 cell line (Passage 13) was kindly donated by Dimitra Dafou, Translational Research Laboratories, University College Hospital, London in July 2008, from an authenticated stock provided by the American Type Culture Collection (ATCC). The cell line was cultured in Dulbecco's Modified Eagle's Media (DMEM) supplemented with 10% fetal bovine serum (FBS) and antibiotics (100 u/ml penicillin and 100 μg/ml streptomycin; Invitrogen, Paisley, UK)

Rabbit anti-human Eag antibody was purchased from Alomone Labs (Israel). Rabbit anti-human HERG antibody was from Abcam laboratories (UK) while secondary anti-rabbit fluorescein isothiocyanate (FITC) conjugated antibodies were obtained from Sigma (Poole UK). The antibodies were specific for the Eag and HERG protein, as demonstrated by specific staining of Eag and HERG protein in ovarian cancer tissue and SKOV-3 cells on Western blotting (Figure 1).

The proliferation of SK-OV-3 cells was examined in the presence and absence of voltage gated K⁺ channel blockers by (3-(4, 5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium, inner salt) MTS assay using the CellTiter 96 aqueous non-radioactive cell proliferation assay kit (Promega, UK). The assay was performed according to the manufacturer's instructions to assess the effect of the voltage gated K⁺ channel blockers, 4-aminopyridine (4-AP) and tetraethylammonium (TEA) on proliferation of the SK-OV-3 cell line. 5000 cells were added to each well of a 96 well plate (Perkin Life Sciences) and incubated for 24 hours in 5% CO₂/air at 37°C. Thereafter, the media was aspirated and replaced with 100 μl fresh media containing either 4-AP or TEA at various concentrations or media only to serve as a control. The cells were incubated for 96 hours and proliferation assessed daily by the addition of MTS (3-(4, 5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium, inner salt) reagent. Plates were then read at 490 nM using a Victor 1427 multilabel counter (Wallac). All observations were performed in triplicate and experiments repeated thrice.

SK-OV-3 cells were grown on cover slips in 24 well plates and incubated for 24 hours until 80% confluent. Cells were then fixed with 4% paraformaldehyde for 20 min and treated with 0.5% Igepal before blocking with 3% BSA/5% glycine in PBS and 10% goat serum in PBS. Eag and HERG (both 1: 50 and 1:100 in10% goat serum/PBS) antibodies were added to the cover slips then incubated overnight at 4°C. Wells containing 10% goat serum without the primary antibody were used as a negative control. After incubation, FITC conjugated secondary antibody at 1 in 50 dilution in 10% goat serum/PBS appropriate to Eag and HERG was added and cells incubated at room temperature for 90 min. Images of the cells were captured and analysed using Cell F software (Olympus UK).

The ovarian cancer tissue microarray (TMA) (SDLREC Ref 0205/495) was prepared from 336 patients who underwent surgery for ovarian cancer from 1^st January 1982 to 31^st December 1998. Clinicopathological characteristics recorded included age at diagnosis, International Federation of Gynecologists and Obstetricians (FIGO) stage, grade, histological subtype, details of adjuvant treatment, extent of cytoreduction and Disease Specific Survival (DSS) which was then used for analysis. DSS was calculated from the date of the operation until 31 Dec 2005, when any remaining survivors were censored.

We calculated that a sample size of 336 patients allowed an 80% chance of detecting a hazard ratio of ≥ 1.40 and ≤0.71 (N Query Advisor software). The TMA were prepared by Ian Spendlove, University of Nottingham as described previously 20 , and were kindly donated to us. For each tumour, 5 μM section, slides were stained with H&E to locate representative areas of viable tissue. Needle core biopsies (0.6 mm) from the corresponding areas on the paraffin embedded tumour blocks were then placed in prespecified coordinates in recipient paraffin array blocks using a manual tissue arrayer (Beecher instruments). Five copies of the array were assembled using different points within the representative tumour area. 5 μM thick sections from each TMA block were placed on coated glass slides to allow for the procedure to be performed. Thus all the biopsies in the TMA were representative of the sample and were validated.

After dewaxing in xylene, tissue sections were rehydrated in a graded alcohol series and endogenous peroxidase activity blocked by immersing the slides in 2% hydrogen peroxide in PBS for 20 min. Antigen retrieval was performed by microwaving the slides in 0.1 M sodium citrate buffer at 800 W for 15 min followed by 400 W for 10 min. After blocking with horse serum, the slides were incubated with either HERG antibody (1:200) for 1 hour at room temperature or Eag antibody (1:600) overnight at 4°C. Staining was developed using the Vectastain Elite kit and di-aminobezidine (DAB) chromagen (Dako, Hamburg, Germany) as per manufacturer's instructions. Slides were counterstained with Harris's hamatoxylin and dehydrated in alcohol before mounting in Depex for viewing.

The staining intensity was graded as: 0-no staining, 1-low, 2-intermediate and 3-high by three independent observers (VA, HS and GVS) who had no knowledge of the pathological variables or the disease outcome for the dataset.

Analysis was performed using SPSS (version 17.0). Continuous data were analysed using median, interquartile range (IQR) and 95% confidence intervals (CI). Fisher's exact tests were used for comparative analysis of categorical data. Kaplan Meier survival estimates using log-rank testing were determined for all baseline factors specifically stage, grade, residual disease, adjuvant chemotherapy, histological type, and Eag and HERG immunostaining. In all cases, a P value of <0.05 was considered statistically significant. The cell proliferation assay data was analysed using one way analysis of variance with Dunnett's multiple comparison test (Graphpad Prism 5).

There is demarcated staining of the plasma membrane of cells incubated with Eag antibody while nuclear staining was seen when the cells were incubated with HERG antibody (n = 3; Figure 2A and 2B, higher magnification is shown in inset). Diffuse staining of the cytoplasm and nuclei was seen with both Eag (Figure 2A) and HERG antibody (Figure 2B). No immunoreactivity was observed in negative control experiments where primary antibody was replaced with goat serum (Figure 2C).

4-aminopyridine (4-AP) inhibited cell proliferation significantly (P < 0.001) at concentrations of 1, 2 and 5 mM compared to control cells at 96 hours with maximum inhibition noted at 5 mM (n = 3). Tetraethylammonium (TEA) inhibition of proliferating cells was also noted at all concentrations at 96 hours (n = 3; P < 0.001) (Figure 2C and 2D) compared to untreated cells.

Clinicopathological characteristics along with Eag and HERG staining of 336 patients with ovarian cancer are summarised in Table 1. The median age at diagnosis and overall survival was 62 years (IQR 24-90) and 19.5 months (IQR 0-271) respectively. Nearly half of the patients (49.4%) had FIGO stage 3 disease and two thirds (65.8%) had grade 3 tumour. The predominant histological type observed is serous (49.4%) followed by endometriod (11.7%) and mucinous (9.7%). The intensity of Eag and HERG staining was stratified into two groups of high and low/intermediate staining (Figure 3A and 3B). The survival curves for these groups are shown in Figure 3C and 3D. Biopsies with an insufficient number of cancer cells to produce a reliable result were not included in the final analysis. Patients with high Eag staining had significantly poor survival of 13.8 months (CI 7.2-20.3) compared with 24.1 months (18.9-29.2); (P = 0.016) for the low/intermediate group (Table 2). HERG staining did not show significant correlation with survival (P = 0.586). Other prognostic markers for ovarian cancer such as stage, grade, residual disease, histological type and adjuvant therapy also showed significant differences in median survival (Table 2).

Following identification of a significant association of Eag staining with survival, further evaluation was carried out between the two groups of patients with Eag staining and other prognostic makers. There was no difference in the proportion of patients with stage of disease (P = 0.587), histological type (P = 0.622) or receiving adjuvant chemotherapy (P = 0.594), but there was significant correlation between high Eag staining, grade of tumour (P = 0.014) and residual disease left after the operation (P = 0.011) (Table 3). On Cox multivariate analysis (Table 4) patients with high Eag staining did not show a significant effect (P = 0.215) on overall survival compared to other prognostic markers like FIGO stage and residual disease. Staining of normal ovary (n = 6) demonstrated low level expression of both Eag and HERG on the surface epithelium compared to ovarian cancer (shown by arrow in Figure 3C).