Cervical cancer-derived cell lines were obtained as previously described (16). InBl cells were derived from a patient diagnosed with epidermoid, non-keratinized, metastatic cervical cancer (FIGO, stage IVB). The CaLo cell line was derived from a tumour biopsy from a patient diagnosed with epidermoid, non-keratinized cervical cancer (FIGO, stage IIB). The two cell lines were maintained in Dulbecco’s modified Eagle’s Medium (DMEM; Invitro, Mexico City, Mexico) supplemented with 2% foetal bovine serum (FBS; Invitrogen Life Technologies, Carlsbad, CA, USA). The K562 human chronic myelogenous leukaemia cell line was used as a positive control for the expression of EpoR (21). K562 cells were maintained in RPMI-1640 (Invitrogen Life Technologies) containing 10% FBS. All cells were incubated at 37°C in a humidified atmosphere of 95% air and 5% CO₂.

The cells were resuspended in lysis buffer (50 mM Tris-HCl, pH 7.4; 150 mM NaCl; 1 mM EDTA; 1% NP40; 0.25% sodium deoxycholate), containing 100 μl/ml complete protease inhibitors cocktail (Roche Applied Science, Mannheim, Germany) and 10 μl/ml phosphatase inhibitors (Sigma-Aldrich, St. Louis, MO, USA). For the isolation of membrane proteins, cells were resuspended in a lysis buffer containing 10 mM Tris-HCl (pH 7.4), 50 mM NaCl, 5 mM EDTA, 1% Triton X-100 and 0.05% SDS. Total protein content was determined using the DC protein assay kit (BioRad Laboratories, Hercules, CA, USA). The protein (30 μg) was resolved by 10% SDS-PAGE and transferred onto polyvinylidene fluoride (PVDF) membranes (Millipore, Billerica, MA, USA). Membranes were incubated at 4°C, overnight with specific antibodies diluted 1:1,000 and then washed and incubated with the appropriate horseradish peroxidase-conjugated secondary antibodies diluted 1:5,000 (Zymed Laboratories, Invitrogen Life Technologies, Carlsbad, CA, USA). For the detection of c-Kit, membranes were incubated with a biotinylated-swine antigoat, mouse and rabbit polyclonal antibody (DakoCytomation, Glostrup, Denmark) diluted 1:2,000 for 2 h at room temperature, followed by incubation with horseradish peroxidase-conjugated streptavidin (DakoCytomation) diluted 1:3,000 for 2 h at room temperature. Proteins were detected by chemiluminescence using the Amersham ECL plus Western Blotting Detection System (GE Healthcare Bio-Sciences, Piscataway, NJ, USA). For the detection of human EpoR, a goat anti-human EpoR antibody (Sigma-Aldrich) was used, which was produced using purified recombinant human erythropoietin soluble receptor as an immunogen. For the detection of c-Kit, a mouse anti-human c-Kit from Cell Signalling Technology Inc. (Danvers, MA, USA) was used. For the study of signalling cascades, rabbit anti-STAT5, rabbit anti-STAT5 (phospho Tyr 694), rabbit anti-ERK1 and ERK2 (ERK1/2) and rabbit anti-ERK (phospho Thr185 + Thr202 + Tyr204 + Tyr187) were used, all from GeneTex Inc. (Irvine, CA, USA). As an internal control, a rabbit anti-GAPDH (GeneTex Inc.) was included.

Cells were harvested, fixed in 2% paraformaldehyde, and stained with the carboxyfluorescein-conjugated mouse monoclonal anti-human erythropoietin receptor or phycoerythrin-conjugated mouse monoclonal anti-human c-Kit/CD117 (R&D Systems, Minneapolis, MN, USA), diluted 1:4 in phosphate-buffered saline (PBS) for 1 h. Cells were washed with PBS and analysed using a FACScalibur with Cell Quest software (Beckton Dickinson, San Jose, CA, USA).

To evaluate cell proliferation, cells were incubated with different concentrations of human recombinant Epo (1, 10, 20, 30 and 50 U/ml; Sigma-Aldrich) or left untreated for 4 days. Proliferation was determined by the colorimetric MTT assay. To evaluate the effect of SCF on survival, the cells were incubated in the culture medium without FBS, treated with 15 U/ml of SCF, or left untreated for 14 days. Cell viability was evaluated by the MTT assay.

The effect of Epo and SCF on migration was evaluated using the QCM™ 24-well colorimetric cell migration assay (Chemicon International, Millipore, Temecula, CA, USA), which is based on the Boyden Chamber migration assay. The cells were starved by incubating 12 h prior to assay in FBS-free medium. Cells (1×10⁵) were then seeded onto the upper chamber in 300 μl of FBS-free medium and then supplemented with 5% bovine serum albumin (BSA). FBS-free medium (500 μl) containing 5, 10, 20 and 30 U/ml Epo or SCF, or 10/20, 10/10 and 30/30 U/ml of Epo/SCF combinations, was loaded into the lower chambers. As a negative control, FBS-free medium was used. Medium supplemented with 10% FBS was included as a positive control. The plates were incubated for 24 h. Cell migration was determined by colorimetric measurement at 560 nm.

Cells (1×10³ ) were seeded in 0.3% agar in culture medium supplemented with 5% FBS and 30 U/ml Epo or 15 U/ml SCF, or a combination of 30 U/ml Epo and 15 U/ml SCF, over a layer of 0.6% agar in 24-well plates. Following 7 days of incubation, colonies were stained with 1% crystal violet in 20% methanol, then imaged and counted.

To inhibit the expression of EpoR in the cell membrane, the cells were incubated with 20 μM lovastatin (Sigma-Aldrich), which is a selective inhibitor of the 3-hydroxy-3-methylglutaryl (HMG)-CoA reductase, for 20 h at 37°C. Inhibition of surface expression was evaluated by western blotting of membrane protein extracts. For transient inhibition of c-Kit gene expression, a commercial small interfering RNA (siRNA) system was used (Santa Cruz Biotechnology Inc., CA, USA). Pools of target-specific 19–25 nucleotide siRNA (Santa Cruz catalogue no. 29225) were used to transfect cells according to the instructions of the manufacturer. As a negative control, cells were transfected with scrambled RNA. Cells were pre-incubated in antibiotic-free culture medium for 24 h. siRNAs were diluted in transfection medium, mixed with the transfection reagent and incubated at room temperature for 30 min. The mixture was overlaid onto the cells and then incubated for 7 h at 37°C. Culture medium containing a double concentration of FBS and antibiotics was added without removing the transfection mixture and further incubated for 24 h. The medium was replaced with fresh culture medium supplemented with 10% FBS and the cells were assayed 48 h later. Expression of c-Kit was evaluated by western blotting. To inhibit the JAK2 phosphorylation, the cells were incubated with 10 μM Tyrphostin AG490 (Sigma-Aldrich) diluted in ethanol for 24 h. PI3K was inhibited by incubating the cells with 100 nM Wortmannin (Sigma-Aldrich) diluted in DMSO for 24 h. To inhibit ERK1/2 kinases, the cells were incubated with 20 μM U0126 ethanolate (Sigma-Aldrich) diluted in DMSO for 24 h.

Columns and reagents were purchased from Miltenyi Biotec (Teterow, Germany). For the isolation of c-Kit (also known as CD117) positive cells, the CD117 MicroBead kit, including paramagnetic microbeads conjugated to monoclonal mouse anti-human CD117 antibody, were used according to the instructions of the manufacturer. Briefly, 1×10⁸ cells were resuspended in 300 μl of column buffer, then 100 μl of FcR blocking reagent and 100 μl of CD117 MicroBeads were added, the cells were incubated for 15 min at 4°C. Following washing of the cells, they were resuspended in 500 μl of buffer. LS⁺ columns were attached to the magnet and washed with 3 ml of buffer, then the cell suspension was applied onto the column. The column was washed three times and the unlabeled cells flowing through were collected for analysis. The column was removed from the separator magnet, 5 ml of buffer was added onto the column, and magnetically labelled cells were flushed out by pushing the plunger into the column.

Results are presented as the mean ± standard error (SEM). The t-test was used for the comparison between treatment groups and between cell lines. Confidence intervals (CI; 95%) and P values were calculated. The test was two-tailed and P<0.05 was considered to indicate a statistically significant result.

Expression of EpoR and c-Kit was investigated in CaLo and InBl cells. Western blot analysis demonstrated the presence of a band of the predicted molecular weight for EpoR (65 kDa) and c-Kit (145 kDa) in CaLo and InBl cells (Fig. 1A). For EpoR to be functional, it must be translocated to the cell surface. Therefore, the expression of EpoR in the cell membrane was analysed by flow cytometry. As demonstrated in Fig. 1B, membrane expression of EpoR was detected in the CaLo and InBl cell lines. By contrast, detection of c-Kit by western blotting was problematic and required a biotin-mediated amplification step, suggesting that the receptor was either expressed at low levels or by a low number of cells. Therefore, the proportion of c-Kit expressing cells were evaluated by flow cytometry. As observed in Fig. 1B, a population of 12.8% of CaLo cells and 11.4% of InBl cells revealed a positive membrane expression of c-Kit.

In a previous study, we identified that exogenous Epo induces proliferation of cervical cancer cell lines. Therefore, to investigate whether stimulation with Epo induces proliferation of CaLo and InBl, the cell lines were incubated in the presence of increasing concentrations of Epo. As expected, the cell lines demonstrated a dose-dependent increase in cell proliferation (Fig. 2A). However, proliferation of CaLo cells was significantly augmented from doses >30 U/ml, whereas enhancement of InBl proliferation was evident from doses >1 U/ml. In fact, throughout the various concentrations of Epo tested, proliferation of InBl was constitutively higher than that of CaLo (Fig. 2A). To verify that cell proliferation was mediated by EpoR, the cells were pre-incubated with lovastatin, which reduced the cell membrane expression of EpoR (Fig. 2B). As summarized in Fig. 2C, incubation with lovastatin reverted the proliferation effect induced by 30 U/ml of Epo, demonstrating that proliferation was mediated by EpoR.

Inhibition of endogenous SCF expression induces apoptosis in CaLo and InBl cells (16). In the present study, it was investigated whether exogenous SCF would protect cells from starvation-induced death. As observed in Fig. 2D, a small but significant proportion of CaLo cells survived FBS withdrawal when cultured in the presence of SCF. Of note, InBl cells were able not only to survive but also to proliferate as a response to SCF. Inhibition of c-Kit expression using siRNA (Fig. 2E) completely eliminated the effect of SCF (Fig. 2F). These observations suggest that SCF protects cells from starvation-induced death by activating its corresponding receptor, c-Kit.

Anchorage independent cell growth has been associated with metastatic potential. Therefore, it was next evaluated if Epo and/or SCF would be able to stimulate the formation of cell colonies in soft agar. The results are demonstrated in Fig. 3A. Administration of Epo or SCF alone produced a modest number of colonies in the cell lines. By contrast, the combination of Epo and SCF promoted a significant increase of cell colonies in CaLo and a highly significant increment of cell colony numbers in InBl cells. Cell migration is fundamental for tumour dissemination. Therefore, to investigate whether activation of EpoR and c-Kit induces cell migration, the cell lines were seeded on the upper insert of Boyden chambers, migration was tested in the presence of 5, 10, 20 and 30 U/ml of either Epo or SCF. As demonstrated in Fig. 3B, CaLo cells were stimulated to migrate by Epo and SCF at the various different doses tested. The response of InBl cells was significantly higher at 5 U/ml of SCF, while only the higher concentration of Epo (30 U/ml) induced a significant number of migrating cells (Fig. 3C). Our results appear to indicate that Epo and SCF alone are associated with migration but not with clonal expansion of cervical tumours. This was in contrast with the combination of Epo/SCF, that was able to stimulate migration and colony formation.

Following this, we investigated the effect of a combination of Epo and SCF on the migratory behaviour of InBl cells. Our results demonstrate that the cell lines respond to Epo and SCF, thus for the following experiments, only InBl cells were tested. Three different combinations (10/20, 10/10 and 30/30 U/ml) of Epo/SCF were used. As observed in Fig. 4A, all of the combinations induced the migration of InBl, being 10/20 significantly higher. This particular combination was subsequently compared with the effect of the growth factors alone. The results are demonstrated in Fig. 4B. The proportion of cells migrating as a response to Epo alone was significantly lower than that induced by the combination of cytokine factors. By contrast, the proportion of migrating cells in the presence of SCF was comparable to that induced by the combination of growth factors. Our results appear to suggest that SCF is a strong inducer of migration. However, as described above, we identified that only 11.4% of InBl cells express c-Kit. Therefore, in order to corroborate that cells expressing c-Kit would in turn migrate as a response to SCF, we isolated c-Kit⁺ cells by utilising paramagnetic microbeads conjugated to monoclonal anti-human CD117 antibody (Fig. 4C). The c-Kit⁺ and c-Kit⁻ cells were tested in a migration assay in the presence of Epo, SCF or Epo/SCF (Fig. 4D). In all of the conditions the proportion of migrating c-Kit⁺ cells was significantly higher than the proportion of c-Kit⁻ cells. Additionally, migration induced by the combination of Epo/SCF was superior to that produced by SCF alone in the c-Kit⁺ cells.

Our observations suggest that SCF, Epo and Epo/SCF induce the migration of cervical cancer cells. Binding of soluble SCF to c-Kit and Epo to its receptor activates three main signalling pathways. Thus, in an attempt to ascertain the participation of each cascade in migration, the cells were pre-incubated with Tyrphostin AG490, Wortmannin and U0126 ethanolate to inhibit JAK2, PI3K and ERK1/2, respectively and then tested in a migration assay. As depicted in Fig. 5A, migration induced by Epo, SCF and the combination of Epo/SCF was significantly reduced by inhibiting JAK2 phosphorylation. Conversely, inhibition of PI3K resulted in a modest decrement of migrating cells. By contrast, migration induced by Epo, SCF and Epo/SCF was abrogated by the inhibition of ERK1/ERK2. This observation strongly suggests that migration is regulated by JAK2-mediated signalling and the MAPK/ERK pathway. To corroborate the activation of these signalling pathways, we evaluated the Epo-, SCF- and Epo/SCF-mediated phosphorylation of STAT5 and ERK1/2 in a time-course experiment. As demonstrated in Fig. 5B, activation of STAT5 was evident only 1 min following stimulation with Epo, showing maximum phosphorylation at 30 min. Stimulation with SCF induced a rapid activation of STAT5, presenting a maximum at 1 min following stimulation, with this response decaying after 30 min. Activation of STAT5 was only detected at 3, 5 and 10 min following treatment with the combination of Epo/SCF. Unlike STAT5, phosphorylation of ERK1/2 was weakly induced by Epo. By marked contrast, SCF, as well as the combination of Epo/SCF, promoted a strong phosphorylation of ERK1/2 from 1 min following treatment. Notably, ERK1/2 activation persisted at high levels until the end of the time-course experiment (60 min).