All animal experiments were performed according to the institutional guidelines approved by the Animal Care and Use Committee of Central South University, Changsha, China. We also received ethics approval from the Institutional Review Board of Central South University.

The NPC cell line, 5-8F, a subclone of SUNE-1 which was isolated from poorly differentiated squamous cell carcinoma tissue and exhibited high metastatic and tumorigenic ability, was preserved by our institute. It was cultured in RPMI-1640 supplemented with 10% newborn calf serum (NCS) at 37°C in a humidified 5% CO₂ atmosphere.

RNA from human normal spleen tissue was reverse transcribed into cDNA with M-MLV reverse transcriptase (Promega, Madison, WI, USA). The entire open reading frame (ORF) fragment (approximately 3.3 kb) of DLC-1 gene was amplified by RT-PCR with the cDNA as the template. The DLC-1 ORF forward and reverse amplification primers were 5′-GCCTGCCGTGCTTGATGTGC-3′ and 5′-TGGTGGAA GCGGTTGCGTTG-3′, respectively. The PCR product was TA-cloned into the pBS-T vector (Tiangen, Beijing, China) and sequenced. The entire DLC-1 ORF sequence was removed from pBS-T/DLC-1 following BamHI and KpnI digestion and was subcloned into pcDNA3.1(+) (Invitrogen, Carlsbad, CA, USA). The recombinant vector (pcDNA3.1/DLC-1) was transformed into DH5α Escherichia coli and the plasmid DNA was isolated using the Miniprep kit (Qiagen, Hilden, Germany) for transfection.

The 5-8F cells were seeded into six-well plates with RPMI-1640 medium at a density of 3×10⁵ cells/well. When reached 70–80% confluence, the cells were transfected with 2 μg plasmid DNA (pcDNA3.1(+) or pcDNA3.1/DLC-1) using 8 μl FuGENE 6 as described by the manufacturer (Roche, Basel, Switzerland). After 48 h, the medium was replaced with fresh RPMI-1640 medium with 10% NCS and Geneticin (G418) in order to select clones stably harboring DLC-1 or the empty vector. The selected clones were named 5-8F-DLC-1 or 5-8F-vector, respectively.

Immunoblotting experiments and ICC analysis were performed according to the procedure outlined in our previous study (11). For each cell line, 2×10⁶ cells were harvested and cell lysates were prepared using commercial cell lysis buffer (Pierce Biotechnology, Inc., Rockford, IL, USA) in western blot analysis. Equal amounts of protein (30 μg) from whole cell lysates were separated by SDS-PAGE and transferred onto polyvinylidene fluoride (PVDF) membranes (Millipore, Billerica, MA, USA). The membranes were blocked with 5% non-fat milk in TBST buffer (1× Tris-buffered saline and 0.1% Tween-20) for 2 h, and then incubated overnight at 4°C with mouse monoclonal anti-human DLC-1 antibody (1:300, BD Biosciences, Franklin Lakes, NJ, USA) and β-actin antibody (1:2000, Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA) which was used as a loading control. After washing with TBST buffer three times, the membranes were incubated with horseradish peroxidase-conjugated goat anti-mouse IgG antibody (1:2,000, Santa Cruz Biotechnology). Finally, the ECL western blotting detection system (Pierce Biotechnology, Inc.) and X-ray films were used to develop the blot images.

A 1:150 dilution of mouse monoclonal anti-human DLC-1 antibody was used in immunocytochemistry analysis. Briefly, cells seeded on slides were washed in PBS, fixed with 4% paraformaldehyde in PBS for 20 min, and treated with 0.5% Triton X-100 in PBS (PBST) for 30 min. Subsequently, endogenous peroxidase was inactivated by incubation in 3% H₂O₂ in dH₂O for 15 min at room temperature. After being rinsed with PBST, the cells were treated with 5% bovine serum albumin (BSA) in PBST (2 h at room temperature). DLC-1 antibody, the Histostain™-Plus kit (Zymed Laboratories Inc., South San Francisco, CA, USA) and the diaminobenzidine (DAB) substrate kit were then used to detect DLC-1 protein. Finally, the slides were lightly counterstained with hematoxylin, dehydrated, mounted with neutral balsam (Shanghai Specimen and Model Factory, Shanghai, China) and photographed under a microscope. For the negative control, the primary DLC-1 antibody was replaced with PBS.

MTT assay was performed as previously described (12). The 5-8F-DLC-1 and 5-8F-vector cells were seeded in 96-well plates at a density of 2×10³ cells/well in 200 μl culture medium (RPMI-1640 with serum). Three wells had no cells and were used as the control for the minimum absorbance. Cells were propagated at 37°C in an incubator with humidified 5% CO₂ atmosphere for one to seven days. The medium was discarded and the cells were incubated with 20 μl/well MTT solution (5 mg/ml) for 4 h at 37°C followed by the addition of 150 μl/well DMSO. Ten minutes later, the absorbance value was measured with an ELISA plate reader (ELx800, BioTek Instruments, Inc., Winooski, VT, USA) at a test wavelength of 490 nm. The growth curves were drawn by EXCEL software.

The 5-8F-vector and 5-8F-DLC-1 cells were seeded in six-well plates at 1×10³ cells/well and each cell line was seeded in triplicate. Cells were cultured in RPMI-1640 supplemented with 10% NCS at 37°C in a humidified 5% CO₂ atmosphere for eight days. After discarding the culture medium and washing with D-Hank’s solution (three times), the cells were fixed with methanol for 15 min and stained with 0.4% crystal violet for 10–30 min. After washing with water and drying in the air, clones containing >50 cells were counted under an inverse microscope (TE2000U; Nikon, Osaka, Japan).

As has been described in our previous study (11), the 5-8F-vector and 5-8F-DLC-1 cells (1×10⁶ cells per sample) were collected and washed with PBS, fixed in 70% ethanol and stored at 4°C. All samples were resuspended in PBS containing RNase A (100 U/ml, Sigma, St. Louis, MO, USA) and incubated at 37°C for 30 min, stained with propidium iodide (PI, 50 μg/ml, Sigma), and analyzed by FCM (BD FACSCalibur™; BD Biosciences). The cell cycle distribution was calculated from the resultant DNA histogram using Mod Fit LT software.

A total of 12 (six male and six female) four to six-week-old BALB/c-nu/nu nude mice were purchased from Shanghai SLAC Laboratory Animal Co., Ltd. The 5-8F-DLC-1 and 5-8F-vector cells (5×10⁶ cells/each type) were injected subcutaneously into the proximal dorsal midline of three four to six-week-old male nude mice and three four to six-week-old female nude mice. Tumor size was subsequently measured in two dimensions twice a week. All the mice were sacrificed 45 days after injection and were examined for the presence of tumors. Tumors were removed from the nude mice, measured and weighed for tumorigenicity analysis.

The 5-8F-vector and 5-8F-DLC-1 cells were seeded into six-well plates at a density of 1×10⁶ cells/well. The following day, the confluent cell monolayer was wounded with a sterile 200 μl tip and plates were returned to cell culture incubator. Images were captured at the beginning and after 24 h of cultivation, and the migration ability of the cells was evaluated by measuring the width of the wounds.

Transwell inserts were loaded into 24-well plates, and 200 μl migration buffer (0.1% BSA in RPMI-1640 medium) were added to the top chamber at 37°C for 1 h. A total of 5×10⁵ cells in 200 μl migration buffer were seeded into the top chamber pre-treated with migration buffer and incubated for 18 h. The lower chamber was supplemented with 600 μl complete medium with 15% fetal calf serum (FCS). After an 18-h incubation, the cells on the upper side of the membranes were removed using a cotton swab and cells on the lower side were fixed in methanol and stained with 0.1% crystal violet and counted in five independent microscopic fields at ×200 magnification.

The protocol for invasion assay was similar to that for cell migration assay except that Transwell chambers were coated with Matrigel.

Climbing slices of 5-8F-DLC-1 and 5-8F-vector cells were prepared and washed with ice-cold PBS buffer on the following day. The cells were then fixed with ice-cold 4% paraformaldhyde for 20 min, incubated in 0.2% (v/v) Triton X-100 for 20 min and stained with 1 μg/ml phalloidin-TRITC for 40 min at room temperature. The slices were washed with PBS three times (10 min per time) after each of the above steps. The cytoskeletal distribution was observed under a ZEISS LSM 5 confocal laser microscope (Carl Zeiss Inc., Oberkochen, Germany).

Total RNA from the 5-8F-vector and 5-8FDLC-1 cells was extracted using TRIzol (Invitrogen) according to the manufacturer’s instructions. Microarray analysis was performed by CapitalBio Corp. (Beijing, P.R. China) using Human Genome U133 Plus2.0 GeneChip Arrays (Affymetrix, Sunnyvale, CA, USA) which covers >54,000 probe sets representing 47,000 transcripts and variants and including 38,500 human genes as previously described (13). The raw data were normalized with Robust Multichip Analysis (RMA). The differential expression of genes was detected with Significance Analysis of Microarrays (SAM). A two-fold or greater change in intensity was used as the criterion for inclusion in our filtered data set and the P-value was set ≤0.05.

To detect the expression of DLC-1 in 5-8F cells transfected with DLC-1 or the pcDNA3.1(+) vector and to confirm the microarray results, RT-PCR analysis was performed as previously described (11). Total cellular RNA was extracted using TRIzol (Invitrogen) and reverse transcribed using a Reverse Transcription kit (Fermentas, Hanover, MD, USA) according to the manufacturer’s instructions. The cDNAs obtained were used as the template for amplification with Taq enzyme mixture (Fermentas). The primer sets, the length of PCR products, annealing temperature and cycles for amplification are listed in Table I. DLC-1 primers were taken from a previous study (10).

MTT assay, colony formation assay, FCM, and cell migration and invasion assay were carried out in triplicate. The data are presented as the means ± SD and processed with SPSS for Windows 11.5 by using an independent-sample t-test. A P-value <0.05 was considered to indicate a statistically significant difference.

The recombinant vector was sequenced and the results showed that the coding sequence was correct except for a single nucleotide polymorphism (SNP) site at +1380 (G→A, V354M) (Fig. 1A). This vector was designated as pcDLC-1. Plasmid, either pcDLC-1 or pcDNA3.1(+) was transfected into 5-8F cells using FuGENE 6 and the cell lines stably expressing pcDLC-1 or pcDNA3.1(+) plasmids were named 5-8F-DLC-1 or 5-8F-vector, respectively. DLC-1 was highly expressed in the 5-8F-DLC-1 cells while it was absent in the 5-8F-vector cells when detected by RT-PCR (Fig. 1B) and western blot analysis (Fig. 1C). Immunocytochemical detection showed that the DLC-1 protein was mainly present in the cytoplasm of the 5-8F-DLC-1 cells (Fig. 1D).

We analyzed changes in the biological characteristics of 5-8F-DLC-1 cells. Compared with the 5-8F-vector cells, the 5-8F-DLC-1 cells showed significant growth inhibition (P<0.05, Fig. 2A) and a reduction in cloning efficiency as measured by MTT analysis and colony formation assay (23.1 vs. 52.5%, respectively) (Fig. 2C). FCM showed that the 5-8F-DLC-1 cells were arrested at the G0/G1 phase (Fig. 2B). Compared to the 5-8F-vector cells, the ratio of 5-8F-DLC-1 cells at the G0/G1 phase increased (67.25 vs. 45.39%), while the ratio of cells at the S phase (21.24 vs. 31.26%) and the G2/M phase (11.51 vs. 23.35%) decreased significantly. In vivo tumorigenicity analysis confirmed that the size of the tumors formed by 5-8F-DLC-1 cells in nude mice was much smaller than that of the tumors formed by 5-8F-vector cells (P<0.05, Fig. 2D). These results indicated that the stable expression of DLC-1 blocked 5-8F cells at the G0/G1 phase and resulted in attenuated proliferation and colony formation ability in vitro and a lower tumorigenicity potential in vivo.

To determine whether DLC-1 affects the motility of NPC cells, in vitro wound healing assay, migration assay and invasion assay were performed. Compared with the 5-8F-vector cells, the 5-8F-DLC-1 cells closed the wound more slowly (Fig. 3A) and the number of cells penetrating the Transwell membrane was significantly lower (18±4.0 vs. 42±5.6; P<0.00) (Fig. 3B), indicating that the motility of DLC-1-expressing cells was significantly impaired. When observed after a 24 h-cultivation, the number of 5-8F-DLC-1 and 5-8F-vector cells penetrating the Transwell membrane coated with Matrigel was 191.40±16.6 and 278±20.7, respectively (n=5, P<0.05) (Fig. 3C), indicating a significant reduction in the invasive ability of 5-8F-DLC-1 cells.

Phalloidin is a virulent alkaloid extracted from a toxic mushroom and can solidly bind with the cell membrane and fibrous actin (F-actin). Fluorescently-labeled phalloidin was used to specifically mark cell surface morphology and changes in F-actin distribution. In the 5-8F-vector cells, abundant F-actin was observed at the cell periphery, a prominent site of the cytoplasm, and throughout the cytoplasm (Fig. 4); while in the 5-8F-DLC-1 cells, microfilaments exhibited polar and circular distribution, focused along the cell periphery and were significantly decreased in the cytoplasm as compared with those in the 5-8F-vector cells (Fig. 4). This is an indication that DLC-1 plays a crucial role in cytoskeletal formation, which may be one of the mechanisms behind its effect on NPC cell motility.

To investigate the molecular mechanism behind the inhibitory effect of DLC-1 expression on the biological characteristics of tumor cells, microarray analysis was performed to compare gene expression profiles in 5-8F-vector and 5-8F-DLC-1 cell groups. The gene expression levels were compared between the two groups. The expression of 840 genes and 151 expressed sequence tags (ESTs) was significantly altered in the 5-8F-DLC-1 group as compared to that in the 5-8F-vector group. Among these genes, 454 were upregulated and 386 were downregulated. The most differentially expressed or altered critical tumor-related genes are listed in Tables II and III. As shown in Table II, genes that act as tumor suppressors, such as IGFBP7, TNS1, TP53 and TP63, were significantly upregulated by the DLC-1 gene. Those acting as oncogenes, such as EGFR, KRAS and TGFβ2, were significantly downregulated (Table III).

To validate our array expression findings, 17 of the differentially expressed genes, such as WNT5A, TNS1, FHL1, S100A2, RECK, DUSP2, CASP9, IGFBP7, EGFR, CDCP1, KRAS, TGFβ2, AKT3, MMP7, MUC4, BCL10 and PTK6 were selected for further verification with RT-PCR analysis. As shown in Fig. 5, eight of these genes, namely WNT5A, TNS1, FHL1, S100A2, RECK, DUSP2, CASP9 and IGFBP7 were found to be upregulated, while nine genes, including EGFR, CDCP1, KRAS, TGFβ2, AKT3, MMP7, MUC4, BCL10 and PTK6 were downregulated in the 5-8F-DLC-1 cells as compared to the control 5-8F-vector cells. The trends for either the up- or downregulation of mRNA expression obtained by RT-PCR were consistent with the microarray results.

Furthermore, using Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) software, we analyzed the microarray dataset to identify whether specific biological pathways were differentially affected by the DLC-1 gene. As shown in Fig. 6, proteins encoded by these genes were mainly observed in the cytoplasm and nucleus. Their molecular functions include protein binding, metal ion binding and nucleotide binding and they participate in some important biological process mainly associated with cell adhesion, negative regulation of cell proliferation, cell cycle arrest and the inhibition of apoptosis. Pathway analysis indicated that the altered genes were associated with a number of essential biological processes, such as tumor-related pathways including focal adhesion, MAPK signaling, VEGF and TGFβ signaling and apoptotic pathways along with several specific cancer-related and metabolism-related pathways (Fig. 7).