Introduction
Cervical cancer is the second most common malignant tumor in female in the world. The progress from low-grade squamous intraepithelial lesions (LSIL) to high-grade squamous intraepithelial lesions (HSIL) then to invasive tumors are closely related to changes in regulation of some cellular processes such as cell cycle, apoptosis and DNA repair (1–3). With the improvement in diagnostic technology and medical treatment, the outcome of patients with cervical cancer has been significantly improved, however, the prognosis of patients with distant metastasis is still poor. According to statistics, the 5-year survival rate of patients with cervical cancer at stage IV classified by the International Federation of Obstetrics and Gynecology (FIGO) is only 3–13% (4). Research has shown that the recurrence rate of cervical cancer was 35%, among which, distant metastases accounted for 11–16% (5,6). In addition, invasion and metastasis are among the most important signs of malignant tumors and the cause for ~90% of clinical patients (7). Therefore, studies on the molecular mechanisms underlying tumor invasion and metastasis are very important for understanding tumor occurrence and development. Tumor migration involves the following processes: i) tumor cell adhesion to the basement membrane of tumor adjacent tissues or extracellular matrix; ii) tumor cells degrading the basement membrane and extracellular matrix; and iii) enhancement of tumor cell motility and tumor metastasis through the vascular system or lymphatic system. Therefore, all factors that could affect the above mentioned processes are likely to be involved in tumor invasion and metastasis (8). Rho GTPases are a class of small molecular signaling proteins regulating cell adhesion, polarization, proliferation, division, invasion, migration as well as other important cellular functions (9–11). Related cellular and animal experiments have demonstrated their important roles in tumor formation (12,13). Furthermore, a number of studies indicated that expression of Rho family proteins was changed in many human tumors such as lung, prostate, breast and colon cancer (14). However, their roles in cervical cancer have not been clarified. Cdc42 is a member of Rho family proteins with GTPase activity. Recent studies have found that Cdc42 is highly expressed in many malignant tumors, and closely related to tumorigenesis, invasion and metastasis (15,16). In the present study, we explored the Cdc42 expression in cervical cancer and its effects on cervical cancer invasion and migration.
Materials and methods
Specimen
Seventy-seven biopsy samples including 15 cases of cervicitis, 18 cases of CIN I, 15 cases of CIN II, 15 cases of CIN III and 11 cases of cervical cancer were collected from outpatients who were admitted to the Peking University Third Hospital during January 2010 to June 2012 due to cervical lesions.
Immunohistochemistry
Paraffin sections of the specimens were dewaxed by submerging in xylene twice for 5 min each and dehydrated by soaking in turn in 100, 95, 90, 80 and 70% ethanol for 3 min each. After incubation in sodium citrate at 96–98°C in a water bath for 30 min, samples were air dried, incubated in 3% hydrogen peroxide for 10 min and washed with phosphate-buffered saline (PBS) 3 times for 2 min each. Samples were then incubated with Cdc42 antibody (1:200) overnight at 4°C, washed 3 times with PBS for 2 min each, and incubated with reagent I at room temperature for 20 min. After washing with PBS three times for 2 min each, samples were incubated with reagent II at room temperature for 30 min, washed with PBS three times for 2 min each, and incubated with DBA at dark for 5 sec. After washed with tap water for ~2 min, samples were stained with H&E, and dehydrated by incubating in turn in 70, 80, 90, 95 and 100% ethanol for 5 min each, then in xylene twice for 5 min each and mounted on slides.
A specimen of lung squamous cell carcinoma tissue was used as positive control and prepared as mentioned above. A cervical tumor sample prepared as indicated above except using PBS to replace Cdc42 antibody was used as a negative control.
The results were evaluated using two methods. The first one was performed by the pathologists of the Peking University Third Hospital based on the percentage of positive cells and staining intensity. In detail, 5 randomly selected fields of each immunohistochemically stained section were observed under light microscope at ×200 magnification and photographed. Based on the percentage of positive cells, samples were classified into five different grades: grade 0, I, II, III and IV and scored correspondingly as 0, 1, 2, 3 and 4 points; if <5%, 5–25%, 26–50%, 51–75%, and >75% were positive, respectively. Brown staining was considered as positive. Staining intensity was divided into four levels: no color was defined as grade 0 and scored as 0 point, light brown was defined as grade I and scored as 1 point, brown was defined as grade II and scored as 2 points, and dark brown was defined as grade III and scored as 3 points. Based on the above two indicators, samples were classified as negative, weakly positive (+), positive (++) and strongly positive (+++) if their overall score was ≤1, 2–3, 4–5 and >5, respectively. The second method was based on analysis using Lecia Q550 CW image analysis software. Five randomly selected fields of each immunohistochemically stained section were observed under light microscope at ×200 magnification and photographed. The OD value of positive regions was determined using the software and used to calculate the mean of each slide. The differences in OD values among samples were statistically analyzed.
Western blot analysis
Confluent cells were washed with PBS and incubated for 10 min in lysis buffer (50 mM Tris-HCl, pH 7.4, 150 mM NaCl, 1% Nonidet P-40, 0.25% sodium deoxycholate, 1 mM EDTA, 2 mM Na3VO4, 1 mM NaF, 2 μg/ml leupeptin, 2 μg/ml antipain, 2 μg/ml soybean trypsin inhibitor, and 2 μg/ml lima trypsin inhibitor). Cells were harvested by scraping and then centrifuged for 5 min at 4°C. For immunoblot analyses, 50 μg of cellular protein was resolved by 10% SDS-PAGE, transferred to nitrocellulose membranes, and probed with specific antibodies directed against Cdc42 (1:200) and GAPDH (1:200), respectively, using protocols provided by the suppliers. Densitometric analyses of the western blots were performed using a ChemiImager 4000 (Alpha Innotech).
Transfection
Cells in 6-well plates were maintained in 1 ml of serum-containing medium and transfected with 1.5 μg of plasmid pGFP, pGFP-Cdc42 CA, pGFP-Cdc42 WT or pGFP-Cdc42 DN using Lipofectamine 2000 (0.15%) following the protocol provided by the manufacturer 24 h after cells were split at 1:5 ratio. Lipofectamine 2000 was removed by changing into fresh medium containing 10% FBS 5 h post-transfection, and cells were analyzed 36 h following transfection.
Cell invasion assay
Cell invasion was assayed using a Transwell. In brief, cell culture chambers containing polycarbonate membrane inserts with 8-μm pore size (Corning Costar Corp.) were coated with Matrigel and dried at 4°C. Transfected cells were briefly incubated with trypsin to obtain a single-cell suspension, and 1×105 cells in serum-free medium were added to the upper chamber. The bottom chamber was filled with 600 μl of complete medium, and the assembly was incubated at 37°C for 48 h to allow cell invasion. Membranes were washed with PBS, and cells that did not pass through the membrane and were gently removed from the upper surface using cotton swabs, and cells on the lower surfaces of the membranes were stained with 1% crystal violet and counted under a microscope.
Cell migration assay
Cell migration assay was performed using a Transwell following the procedure similar the cell invasion assay except that the membrane was coated with 2% gelatin.
Immunofluorescence assay
Cells were fixed with paraformaldehyde for 20 min, washed twice with PBS at room temperature, permeablized with 1 ml of 0.2% Triton-100 for 30 min, washed again with PBS twice at room temperature, incubated with 4% BSA at room temperature for 30 min, washed again with PBS twice at room temperature, and incubated with anti-phalloidin (1:50) antibody for 1 h at room temperature. After being washed with PBS twice, the stained cytoskeleton was observed using a confocal microscope (Lecia SP5).
Statistical analysis
SPSS 17.0 statistical software was used for statistical analysis. The difference of Cdc42 expression between HeLa cells and Crl-2614 cells was analyzed using t-test and among normal cervical tissue, CIN samples and cervical tumor tissues was compared using the χ2 test. Differences in cell invasion and migration among different cells were analyzed using ANOVA. P<0.05 was considered statistically significant.
Results
Cdc42 expression in normal cervical tissue, CIN I or below, CIN II or above, and cervical tumor tissues
Pathological results indicated that Cdc42 expression rate was 100% (11/11) in cervical cancer tissues, 100% (35/35) in CIN II or above tissues and 6.5% (2/31) in CIN I or below tissues (Fig. 1A–D). Further analysis showed that Cdc42+++, the highest expression grade, and Cdc42++ were found in 63.6% (7/11) and the remaining 36.4% (4/11) in cervical cancer tissues, respectively, while Cdc42++ and Cdc42+ were found in 65.6% (23/35) and 20% (7/35) of CIN II or above tissues, respectively (Table I). Image analysis showed that the expression value of Cdc42 was 0.1933±0.0091, 0.2135±0.0192 and 0.2516±0.0135 in CIN I or below, CIN II or above, as well as in cervical cancer tissues, respectively. The expression of Cdc42 was significantly different among these three groups (F=96.94, P<0.05) and increased with elevation of cervical lesions (Fig. 1F). Overall, the results of these two analysis were in a good agreement, both indicating that Cdc42 expression was significantly different among CIN I or below, CIN II or above, as well as in cervical cancer tissues.
Cdc42 expression in normal cervical cells and HeLa cells
Western blot analysis of Cdc42 (Fig. 2A) showed that expression of Cdc42 was significantly higher in cervical cancer cell line HeLa cells than in in the normal cervical cell line Crl-2614 (t=20.33, P<0.05) (Fig. 2B).
Effects of Cdc42 on invasion and migration of HeLa cells Transfection of Cdc42 in HeLa cells
To examine the role of Cdc42 in HeLa cells, we transfected either constitutively active Cdc42 plasmid pGFP-Cdc42 CA, dominantly negative Cdc42 plasmid pGFP-Cdc42 DN or wild-type Cdc42 plasmid pGFP-Cdc42 WT as well as control plasmid pGFP. Fig. 3 showed the fluorescent images of transfected HeLa cells, indicating the successful transfection of each plasmid in HeLa cells. In addition, we also examined the transfection results using western blot analysis (Fig. 4), and confirmed the expression of the transfected Cdc42 in HeLa cells.
Effects of Cdc42 on invasiveness of HeLa cells
Invasion assay using Transwell showed that the invasiveness of HeLa cells transfected with Cdc42 CA and Cdc42 WT was significantly higher than that of non-transfected HeLa cells while that of HeLa cells transfected with Cdc42 DN was significantly lower than that of the non-transfected HeLa cells (F=684.7, P<0.01) (Figs. 5 and 6).
Effects of Cdc42 on migration of HeLa cells
Migration assay using Transwell showed that the migration ability of HeLa cells transfected with Cdc42 CA and Cdc42 WT was significantly higher than that of non-transfected HeLa cells while that of HeLa cells transfected with Cdc42 DN was significantly lower than that of the non-transfected HeLa cells (F=545.8, P<0.01) (Figs. 7 and 8).
Impact of Cdc42 on pseudopodia formation in HeLa cells
Confocal microscopy analysis of immunofluorescence stained cytoskeleton of non-transfected HeLa cells and Cdc42 CA-, Cdc42 WT- and Cdc42 DN-transfected HeLa cells showed that compared with non-transfected HeLa cells, Cdc42 CA transfected HeLa cells had more, larger and thicker pseudopodia, Cdc42 WT-transfected HeLa cells showed more, but similar-shaped pseudopodia, while Cdc42 DN transfected HeLa cells showed no obviously changed pseudopodia (Fig. 9). Table II showed the number of pseudopodia in Cdc42 CA-, Cdc42 WT- and Cdc42 DN-transfected and non-transfected HeLa cells. The results indicated that 1amellipodia only appeared in Cdc42 CA-transfected HeLa cells and the number of total pseudopodia Cdc42 CA-transfected HeLa cells and filopodia in Cdc42 WT-transfected HeLa cells were significantly higher than that of Cdc42 DN-transfected and non-transfected HeLa (Fig. 10, Table II).
Discussion
Currently, immunohistochemical results are mainly analyzed by pathologists or image analysis software. The former has advantages such as enabling multi-field observation, but also has limitations such as subjectivity. The latter has advantages of being more objective, and is a reproducible semi-quantitative method. However, it is mainly used in scientific research fields. Studies have shown that the former is more suitable for analyzing positive results of cytoplasm and nucleus, while the latter is more accurate for analyzing positive results of membrane staining (17). Cdc42 is expressed in both cytoplasm and membrane, to more accurately and reliably analyze the results, we utilized both methods and obtained similar results. Our findings indicated that like in other tumors, Cdc42 was also overexpressed in cervical cancer and its cell line.
To explore the roles of Cdc42 in cervical cancer, we investigated its effects on invasiveness and migration of HeLa cells using Transwells after transfecting either Cdc42 CA, Cdc42 WT or Cdc42 DN. Our results showed that overexpression of Cdc42 CA significantly improved the invasiveness and migration of HeLa cells. Tumor metastasis is an extremely complex process, in which, tumor cells must first migrate away from the primary tumor, infiltrate the surrounding tissue, invade and and survive in the bloodstream or lymphatic system, and finally escape the blood or lymph circulation to reach distant organs (18). Numerous studies have shown that Cdc42 is involved in multiple steps of this complex process, and is one of the key genes regulating tumor metastasis (19,20). The above findings are consistent with our conclusion that Cdc42 can significantly improve the capability of tumor cell invasion and migration.
To further understand the underlying mechanisms by which Cdc42 improving tumor cell invasion and migration, we performed immunofluorescence staining for F-actin in non-transfected and Cdc42 transfected HeLa cells. The results indicated that overexpression of Cdc42 CA promoted the formation of pesudopodia including 1amellipodia. The latest in vivo imaging observations and invasiveness studies showed that pseudopodia played a major role in tumor invasion and metastasis (21). Pseudopodia in cells include filopodia, lamellipodia, and invadopodia. Filopodia are highly dynamic structure and play major roles in the process of tumor cell invasion, adhesion, supporting migration and nutrition. Lamellipodia are principally involved in the adhesion of tumor cells to extracellular matrix in the initial migration stage. Invadopodia are high dynamic, actin-rich membrane structures and closely related to cell migration. The formation of invadopodia is mainly based on N-WASP-dependent branch-like structure on the actin network. Cdc42 can regulate N-WASP- and Arp2/3-mediated cytoskeleton polymerization, leading to the invasion of invadopodia to the matrix structure, thus participating in the formation and stabilization of invadopodia (20,22–24). Invadopodia are highly active in the highly invasive breast cancer cell line MDA-MB-231, where Cdc42 could activate Cdc42 binding protein CIP4 and mediate N-WASP activation (25). Our results suggested that overexpression of Cdc42 in cervical cancer has the potential to enhance tumor cell invasiveness and migration by promoting the formation of invadopodia, filopodia and lamellipodia. With the in-depth understanding of invadopodia-mediated tumor invasion and migration, the structure has become a target in cancer therapy (26). However, the mechanisms of tumor cell invasion and migration are very complex. In addition to the formation of invadopodia, the degradation of extracellular matrix due to secretion and activation of matrix metalloproteinase, tumor cell phenotype conversion, i.e. epithelial to mesenchymal transition, as well as amoeba movement phenotype, all could enhance tumor cell invasion and migration (18,27). Thus, whether the enhanced invasiveness and migration due to overexpression of Cdc42 is also mediated by other pathways has yet to be further explored.
In conclusion, expression of Cdc42 was positively correlated with the grade of cervical lesions (P<0.05) and overexpression of dominantly active Cdc42 can significantly improve the migration and invasiveness of tumor cells possibly through promoting pseudopodia formation.