The human Ishikawa endometrial cancer cell line was kindly provided by Professor Marcello Maggiolini (University of Calabria, Italy). The human HeLa cervical cancer cell line was kindly provided by Professor Roberto Maggio (University of L’Aquila, Italy), while the human CASKI cervical- and SiHa vulva-cancer cell line were kindly provided by Professor P.J.F. Snijders (VU Medisch Centrum, Germany). Tumor cell lines were cultured in the appropriate medium supplemented with 10% fetal calf serum (FCS). Treatments with 10 μmol/l MEK/ERK inhibitor U0126 (1, 4-diamino-2, 3-dicyano-1, 4-bis[2-aminophenylthio] butadiene; Promega) or with AZD1152 (60 nM), an Aurora-B kinase inhibitor, was done as shown in the figures. Radiation was delivered at room temperature using an X-ray LINAC at the dose rate of 2.5 Gy/min. For clonogenic survival assay, exponentially growing cells, diluted to appropriate densities, were plated with complete medium in presence of U0126, AZD1152 or vehicle/control [dimethyl sulfoxide (DMSO) 0.1%] and then irradiated with graded doses (0–2-4–6 Gy). Cells were then cultured in drug-free medium for 14 days, fixed with methanol/acetic acid (10:1, v/v) and stained with crystal violet. Colonies containing >50 cells were counted. The plating efficiency (PE) was calculated as the number of colonies observed/the number of cell plated; the surviving fraction (SF) was calculated as: colonies counted/cells seeded (PE/100).

Cells from adherent culture were counted using hemocytometer and tested for exclusion of trypan blue. Results represent the average of triplicate experiments, including standard error.

Proteins of whole cell lysates were assessed using the Lowry method (40) and equal amounts of proteins were separated on SDS-PAGE. The proteins were transferred to a nitrocellulose membrane (Schleicher & Schell Bioscience GmbH, Germany) by electroblotting. Immunoblottings were performed with the following antibodies directed against c-Myc, ERK1/2, phospho-ERK1/2, p27, cyclin D1, CDK2, Aurora-B, DNAPKcs and α-tubulin (B-7) (all from Santa Cruz Biotechnology, Santa Cruz, CA, USA). Peroxidase-conjugate anti-mouse or anti-rabbit IgG (Amersham-Pharmacia Biotech, UK or Santa Cruz) were used for enhanced chemiluminescence (ECL) detection.

After the appropriate treatments, cells (1×10⁶) were fixed for 30 min in 70% ethanol and pelleted by centrifugation (720 g; 5 min). After removal of ethanol, cells were incubated and resuspended in 1 ml of DNA staining solution (PBS containing 200 mg/ml RNase A, 20 mg/ml propidium iodide plus 0.1% Triton X-100) and left at room temperature for 60 min. Ten thousand events per sample were acquired using a FACScan flow cytometer (Becton-Dickinson, San Jose, CA, USA) and the percentage of cells in G₁/S, G₂/M and Sub-G₂/M phases of the cell cycle were determined using CellQuest software (Becton-Dickinson). Apoptosis was analyzed by using Annexin V staining (GenScript, Piscataway, NJ, USA). All cells were then measured on a FACScan flow cytometer with an argon laser at 488 nm for excitation and analyzed using CellQuest software. Apoptotic cells were detected by the percentage of Annexin V stained cells. The results were expressed as the percentage of death by apoptosis induced by a specific treatment.

Soft agar assays were performed as previously described (34). Briefly, 2×10³ cells were suspended in 0.3% Bacto-agar (Life Technologies) containing the appropriate medium supplemented with 10% FCS and/or 10 μM U0126 and seeded in 60-mm cell culture plates. The plates were incubated at 37°C for 14 days. The numbers of colonies containing ≥6 cells were counted.

Tumor cells (1.5×10⁵ cells/well) were plated in 6-well plates and grown in the appropriate medium supplemented with 10% FCS for 2 days. These cells were transfected with ERK1siRNA and ERK2 siRNA (sc-29308 and sc-44224 respectively, Santa Cruz Biotechnology). All siRNA duplexes were transfected using Oligofectamine reagent (Invitrogen) according to the manufacturer’s instructions. After ERK transfection, cells were grown in the appropriate medium for 72 h.

Transwell membrane (Corning Costar Corp.) was used. Cancer cells were trypsinized, washed and kept suspended in the appropriate medium without FCS. To the lower wells of the chambers, migration-inducing medium (with 10% FCS) was added. Upper wells were filled with serum-free medium with cells (20,000 cells per well) in the absence or in the presence of the appropriate treatments. After 8 h, filters were removed and fixed with methanol and subsequently the cells on the upper side were wiped off using a Q-tip. Filters were stained with 20% Giemsa solution. Evaluation of complete transmigration was performed under the microscope and random fields were scanned (four fields per filter) for the presence of cells at the lower membrane side only. Invasion assays were done in a similar manner as the migration assays described above, unless the inserts were pre-coated with Matrigel (BD Biosciences).

Continuous variables were summarized as the mean and standard deviation (SD) and the statistical comparisons between control and treatments were established by carrying out the ANOVA test or the t-test when appropriate. Dichotomous variables were summarized by absolute and/or relative frequencies and statistical comparisons between control and treated groups were established by carrying out the Fisher’s exact test. For multiple comparisons the level of significance was corrected according to Bonferroni correction. All tests were two-sided and were determined by Monte Carlo significance. P<0.05 was considered statistically significant.

In order to verify the effects of MEK-inhibitor, U0126, a time course experiment with or without U0126 treatment (10 μM) was performed. Cells were treated with U0126 either for 3 h or for 4 days and subsequently processed for immunoblotting, cell count and FACS analysis. As shown in Fig. 1A and B, U0126 induced a rapid (3 h) and persistent (4 days) decrease in phospho-active ERKs (Fig. 1A) concomitant with a decrease in the proliferation rate ranging from 65.5 to 74.9% (Fig. 1B). As shown in Fig. 1C, treatment with U0126 resulted in preferential accumulation of tumor cells in the G₁-S phase of cell cycle, with the percentage of Ishikawa, SiHa, Caski and HeLa cells in thiS phase of 61, 69, 74 and 77%, respectively. This phenomenon was consistent with cyclin D1 and CDK2 downregulation and p27^Cip/Kip upregulation (Fig. 1D). c-Myc, one of the most tumorigenic transcription factors and downstream target of ERKs, while increasing in the expression levels comparing 3 h and 4 days untreated cultured cells, was early (3 h) and persistently (≤4 days) inhibited by the U0126 treatment as shown in Fig. 1D.

To verify whether this inhibition affects anchorage-independent growth of cancer cells, we performed soft agar and cell growth assay in modified layer. Soft agar assay showed that tumor cells cultured with vehicle only formed cell aggregates of different sizes depending on the cell line used (Fig. 2A–D, upper panel). U0126 (10 μM) greatly affected the propensity of tumor cells to grow in modified layer (polyHEMA coated dishes) compared to controls (Fig. 2A–D, middle panel) with a decrease in the proliferation rate of 86, 84, 93 and 89% in Ishikawa, SiHa, Caski and HeLa cell lines, respectively (ANOVA test; P<0.01). The decrease of phospho-active-ERK levels in tumor cells cultured on poly-HEMA-coated dishes was still present after U0126 treatments (Fig. 2A–D, lower panel). Finally, U0126 (10 μM) significantly inhibited Ishikawa, SiHa, Caski and HeLa invasion and migration both at 3 h and 4 days after treatment, with the most evident effect at the longest time (Fig. 3). All together, these results indicate that U0126 induces growth arrest by blocking the molecular mechanism responsible of G₁/S cell cycle phase progression and reduces the signals enabling tumorigenic and metastatic potential of tumor cells.

Tumor cells were treated with U0126 (10 μM) 24 h before the delivery of increasing doses of ionizing radiation (0–600 cGy) (Fig. 4). All cell lines were basically radioresistant at all doses tested (Fig. 4) and U0126 treatments increased the radiosensitivity with effects already evident at lowest radiation doses (Fig. 4). We further investigated whether suppressing ERKs function would influence the repair machinery of DNA double-strand breaks (DSBs) induced by irradiation. Tumor cells treated with U0126 were irradiated and the number of DNA foci representing the amount of unrepaired DSBs was counted in 4 different cell lines: i) untreated, ii) U0126-treated, iii) U0126-pretreated and iv) treated with RT (Fig. 5A–D, upper panel). For these experiments, a single dose of 4-Gy radiation was used. We have found that this dose is the most suitable to effectively distinguish the individual from combined treatments. Radiation-induced γ-H2AX foci were readily detectable at 1 h of irradiation in all treated cells (Fig. 5A–D, upper panel). Interestingly, adding U0126 together with radiation led to a substantially increased number of cells retaining these foci for ≥14 h with respect to cells treated with RT alone. The difference in the percentage of cells retaining DNA foci between the two groups was significant (P<0.01). Due to the fact that DNA-PKcs is usually activated in malignant tumor cells and its inactivation impairs DNA repair following ionizing irradiation we examined whether U0126 would affect DNA-PKcs expression. Of note, DNA-PKcs expression measured by western blotting at 24 h after treatments significantly decreased in all cell lines tested in the present study (Fig. 5A–D, lower panel).

We tested if the U0126-mediated effect on radiosesitization involved Aurora-B, known to play a central role in the control of DNA repair machinery. To this purpose cancer cells were cultured with U0126 24 h before the delivery of ionizing radiation (4 Gy) (Fig. 6) and expression levels of Aurora-B kinases was evaluated in these cell lines. As shown in Fig. 6A, MEK/ERK inhibition reduced Aurora-B expression in all cell lines used (Fig. 6A). To verify whether in these cell lines the Aurora-B played a role of increased radioresponse, experiments with AZD1152, a selective Aurora-B inhibitor, were performed. The functional inhibition of Aurora-B was assessed by quantifying p-H3, the active phosphorylated form of histone H3 required for normal chromosomal condensation. As shown in Fig. 6B, 24 h after AZD1152 (60 nM) treatment a dramatic decrease in p-H3 expression, consistent with inhibition of Aurora-B H3-phosphorylating activity, was observed in all cell lines. The functional inactivation of Aurora-B by AZD1152 increased the radiosensitivity of all cellular models (Fig. 6C) with a kinetic of radiosensitization similar to that observed during U0126 treatment (compare Fig. 4, with Fig. 6C). The radiosensitizing effect of AZD1152 was partially due to increased apoptotic stimulus that was greatly potentiated in association with radiotherapy (Fig. 6D) as measured by Annexin V assay. This evidence suggests a possible role of Aurora-B in the induction of pro-apoptotic stimulus upon radiation treatment.

In order to verify the existence of a direct functional link between ERK1/2 and Aurora-B, ERK1/2 RNA interference experiments were performed. Upon 72 h of treatment with ERK1/ERK2 siRNA, downregulation of total ERKs and Aurora-B protein levels was observed (Fig. 7A). Since MEK/ERK silencing caused downregulation of Aurora-B expression we further studied AZD1152 effects in cellular growth, growth in modified monolayer and in migration/invasion assays. The Aurora-B inactivation by AZ1152 significantly decreased growth rate and growth in modified monolayer (Fig. 7B, left and middle panels) as well as migration/invasion (Fig. 7B, right panels) potential of all tumor cell models. This body of evidence suggests that Aurora-B mediates tumorigenic potential of gynecological cancer cell lines controlling their invasion and metastatic potential.