Calpain system protein expression and activity in ovarian cancer

The ovarian tissue microarray was composed of tumour cores from 575 ovarian cancer cases; with 448 chemo-naïve samples and 22 cases sampled post-chemotherapy (no chemotherapy details were available on the remainder). Staging was according to the International Federation of Obstetrics and Gynecology (FIGO) staging system. Grading was performed using the Shimizu–Silverberg grading system. The clinicopathological variables of the cohort are listed in Table 1. Patients were diagnosed with ovarian cancer and received treatment at Nottingham University Hospitals between 1991 and 2011. The median follow-up period was 8 years: ranging between 3 years and 20 years. The median overall survival time was 44 months: ranging from 0 to 223 months. The majority of the patients (n = 357) received platinum-based chemotherapy among which 168 patients were treated with chemotherapy containing taxanes (Table S1). Progression-free survival was defined as the length of time between start of treatment and clinical identification of recurrence. Data on the resistance to chemotherapy were recorded, classified by the Gynaecologic Oncology Group (GOG) as either refractory (not responding to chemotherapy), resistant (an initial response to chemotherapy with recurrence within 6 months) or sensitive (when there was either no recurrence or recurred after 6 months). Ethical approval was obtained from Derbyshire Ethics Committee (07/H0401/156). This study is reported in accordance with REMARK (reporting recommendations for tumour marker prognostic studies) criteria (McShane et al. 2005).

Protein expression was investigated using sections taken from tissue microarray blocks, which were constructed as follows. Tumour samples had been immediately fixed in 10% formalin and embedded into paraffin blocks. Donor blocks were sectioned and stained with haematoxylin and eosin with a specialist gynaecological pathologist reviewing and marking the area of interest to core. A Beecher instrument manual tissue microarrayer was used to produce the TMA paraffin blocks by extracting cores of 0.6 mm from the donor blocks and placing them in pre-punched holes in the TMA blocks, each with approximately 160 cores. For the majority of the cases, a single tissue core was used for each patient. There were 48 cases that had two cores per case and 3 that had three cores per case. A small number of cores from other tissues were also placed in each TMA block to act as potential subsequent positive controls and to aid orientation. Fresh sections (4 µm) were cut from each block and placed on coated glass slides for the immunohistochemical assessment of protein expression.

Immunohistochemistry was performed as described previously (Storr et al. 2012a), following re-optimisations. Briefly, slides were heated at 60 °C for 10 min then dewaxed in xylene and rehydrated in industrial methylated spirits. Sections were pre-treated by heat-induced epitope retrieval in 0.01 mol/L sodium citrate buffer, pH 6, for 10 min (750W) + 10 min (450W) in a microwave oven. Staining was achieved using a Novolink Polymer Detection System (Leica, Denmark) following the manufacturers’ instructions. Briefly, endogenous peroxidase activity was neutralised with peroxidase block reagent for 5 min at room temperature; followed by application of protein block reagent for 5 min at room temperature, to minimise non-specific interactions of the subsequent detection reagents after washing with Tris buffered saline. Primary antibody was diluted in bond primary antibody diluent (Leica, Denmark) and applied to the tissue overnight at 4 °C. The primary antibodies were diluted in bond primary antibody diluent (Leica, Denmark), calpain-1 (1:1000, Santa Cruz Biotechnology, INC.), calpain-2 (1:2500, Chemicon^® International Millipore), calpain-4 (1:100,000, Chemicon^® International Millipore) and calpastatin antibody (1:50,000, Chemicon^® International Millipore), and applied to the tissue overnight at 4 °C. The specificity of these antibodies was initially confirmed, by western blotting, before use. Post-primary reagent, a polymer penetration enhancer, was applied on the slides for 30 min; followed by NovoLink Polymer [anti-mouse/rabbit IgG-Poly- horseradish peroxidase (HRP)] for another 30 min, with each step followed by a Tris buffered saline wash. Immunohistochemical reactions were visualised with 3, 3′-diaminobenzidine (DAB) for 5 min and counterstained with haematoxylin for 6 min. Slides were then dehydrated with industrial methylated spirits and xylene and mounted with DPX (SIGMA). Negative controls had primary antibody omitted.

Slides were scanned at high resolution using a Nanozoomer Digital Pathology Scanner (Hamamatsu Photonics) with 200× magnification. Staining intensity was semi-quantitatively assessed using an immunohistochemical H-score with ranking from none (0), weak (1), medium (2) to strong (3) with the percentage area of each staining intensity being multiplied by the intensity rank (H-score range: 0–300). Greater than 25% of the slides were examined by a second independent assessor blinded to scores and clinicopathologic criteria. Single measure intraclass correlation coefficient (ICC) analysis was used to determine the level of agreement between independent scorers. The single measure ICCs between scores were 0.8, 0.814, 0.733 and 0.796 for calpain-1, calpain-2, calpain-4 and calpastatin protein expression, respectively. A non-biased cut-point of the immunohistochemical scores, to dichotomise data, was determined using X-tile software using patient OS (Camp et al. 2004; Storr et al. 2012a).

Ovarian cancer cell lines PEO1 and PEO4 cells were obtained from the American Type Culture Collection. A2780, A2780-cis and SKOV-3 cells were obtained from the European Collection of Authenticated Cell Cultures. Cells were cultured in Roswell Park Memorial Institute (RPMI)-1640 medium (SIGMA) (for A2780 and A2780-cis), RPMI medium (SIGMA,) containing Sodium Pyruvate (SIGMA) 2 mM (for PEO1 and PEO4) or McCoy’s 5A Modified medium (SIGMA) (for SKOV-3) supplemented with 10% heat-inactivated iron-supplemented donor bovine serum (Gibco, Life Technologies) and penicillin/streptomycin (SIGMA) (with 10,000 units penicillin and 10 mg streptomycin/mL) in 37 °C and 5% CO₂ atmosphere. Cisplatin (1 µM) was added to A2780-cis culture media every two to three passages. Cell line authentication was conducted using the Promega Powerplex^® 16-short-tandem-repeat system. All cell lines were regularly screened for mycoplasma infection using a MycoProbe^® Mycoplasma Detection Kit (R&D Systems). Cells in logarithmic growth phase were used for experiments.

Stock solution of cisplatin (p4394, SIGMA) was made at 1 mg/ml in sodium chloride solution (0.9%) (S8776, SIGMA). Carboplatin (c2538 SIGMA) was initially diluted in sterile distilled water (dH₂O) to reach 20 mM. Both stock solutions were protected from light, stored at room temperature and used within 2 months. Calpeptin (03-34-0051, Merck Millipore) was initially diluted to 100 mM in dimethyl sulfoxide (DMSO, D2650, SIGMA), and then the aliquoted stock solution was stored at − 20 °C.

Subconfluent cells were trypsinised, washed, collected and lysed at 5 × 10⁶ cells per ml in RIPA buffer (R0278, SIGMA) supplemented with 1 × Halt Protease Inhibitor Cocktail containing protease inhibitors, phosphatase inhibitors and EDTA (ThermoFisher Scientific) at 4 °C for 10 min and then lysates were frozen at − 20 °C or − 70 °C for long-term storage.

Cell lysate was loaded into SDS-PAGE gel (Novex® ready-made NuPAGE^® 4–12% Bis-Tis Protein Gels), after which proteins were separated by gel electrophoresis and transferred onto a 0.2-µm nitrocellulose membrane (Millipore). Membranes were then blocked with 3% non-fat milk PBS containing 0.1% Tween 20 prior to incubation with primary antibody overnight at 4 °C. Antibodies were used at the following dilutions: anti-calpastatin 1:1000, anti-calpain-1 1:1000, anti-calpain-2 1:2500 and anti-calpain-4 1:1000. Membranes were washed and incubated with mouse/rabbit anti-β-actin antibody [1 in 1000 dilution, ab8226 or ab8227 (Abcam)], for 1 h at room temperature. Secondary antibodies [i.e. 680 Donkey anti-Mouse IgG (H + L) (926-32222, IRDye^®, LI-COR) and 800CW Donkey anti-Rabbit IgG (H + L) (926-32213, IRDye^®, LI-COR)] were incubated on the membrane for 1 h. Membranes were visualised using an Odyssey FC Imager (LI-COR Biosciences). The fluorescence intensity was quantified using Image Studio Version 4.0. (LI-COR Biosciences) and normalised against β-actin signals.

SKOV-3, PEO1 and PEO4 cells were used to assess effects of calpain inhibition on cell proliferation and chemosensitivity. To ensure cells were in log phase during assessment, 2 ml of cell suspension (8 × 10⁴/ml SKOV3 cells, 1 × 10⁵/ml PEO1 cells and 1 × 10⁵/ml PEO4 cells) were seeded into each well of six-well plates. Cells were plated for variable times before treatment, according to particular doubling times. SKOV3 and PEO1 cells were cultured for 1 day whilst PEO4 cells for 3 days before treatment. Doubling time for SKOV3’s was 32.3 ± 4.0 h, for PEO1 57.4 ± 5.7 h and for PEO4 102.7 ± 8.5 h (data not shown). Cells were then pre-treated for 1 h (SKOV3) or 90 min (PEO-1 and -4) with 50 µM calpeptin followed by an additional 48-h treatment with cisplatin (SKOV3: 5 µM, PEO1: 3 µM, PEO4: 19 µM) or carboplatin (SKOV3: 11 µM, PEO1: 14 µM, PEO4: 26 µM) (IC50 concentrations) in the presence of 50 µM calpeptin. Cells in each well were trypsinised and counted at various times thereafter.

PEO-1/-4 cells were treated with DMSO or 50 µM calpeptin for 4 h or 24 h based on calpain activity results from fluorescent plate reader assays using a fluorescent substrate CMAC, t-BOC-Leu-Met, which indicated that 50 µM calpeptin induced maximal inhibition over a 90 min–48 h treatment time (approximately 30% for PEO1 and 40% for PEO4) (data not shown). Total RNA was extracted from treated cells with RNA protect Cell Reagent (Qiagen) and purified with the RNeasy Plus Mini Kit (Qiagen) according to the manufacturer’s instructions. RNA samples from three independent experiments were pooled together. After quantification using a Nanodrop spectrophotometer (calculation of 260/280 nm ratio), 0.5 µg total RNA was reversed transcribed into cDNA using RT² First Strand Kit (Qiagen) at 42 °C for 15 min following by 95 °C for 5 min to stop the reaction. Real-time PCR was performed using the Human Epithelial to Mesenchymal Transition (EMT) RT² Profiler™ PCR Array (PAHS-090Z, Qiagen) in combination with RT² SYBR Green Mastermixes (Qiagen). The PCR array profiles the expression of 84 key genes involved in EMT process. The PCR cycling condition was set as follows: 95 °C for 10 min, 40 cycles of 95 °C for 15 s and 60 °C for 1 min and run on ABI-7500 (Applied Biosystems). Data were analysed using the ΔΔCt method.

The relationships between categorised protein expression and clinicopathologic factors were examined using Pearson’s chi-square test of association (χ²) or Fisher’s exact test if a cell count was less than 5 in a 2 × 2 table. To assess the relationship between protein expression and survival outcomes, survival curves were generated using the Kaplan–Meier method and statistical significance determined by the Log-rank test. Multivariate survival analysis was performed using a proportional hazards model by Cox regression analysis to estimate hazard ratios and 95% confidence intervals for overall survival. Spearman’s rank correlation coefficient (Spearman’s rho) test was performed to assess the correlation between the expression levels of different proteins. The correlation strength was interpreted as follows: Spearman rho (rs) less than 0.16 is too weak to be meaningful, ranging from 0.16 to 0.19 as very weak correlation, 0.20 to 0.39 as a weak correlation, 0.40 to 0.59 as a moderate correlation, and 0.60 to 0.79 as strong correlation, and 0.80 or greater as very strong correlation (Divaris et al. 2012). Results from in vitro experiments were represented as average ± standard deviation from a minimum of three independent experiments. For comparing two variables, the Student’s t test was used and one-way ANOVA was used to compare three or more groups. Statistical analyses were carried out using SPSS 22.0 software. P values < 0.05 were considered statistically significant. Data analysis of the PCR array was carried out using the online GeneGlobe Data Analysis Centre. Fold-change of target genes against the reference gene was calculated from 2^−△△Ct values. The CT cut-off was set at 35. Gene expression differences greater than twofold were considered as differentially transcribed.