Ara-C was obtained from Sigma (St. Louis, MO, USA). It was dissolved in sterile H₂O to a stock concentration of 0.4 g/ml. U-0126 was obtained from Cell Signaling Technology (Beverly, MA, USA) and dissolved in DMSO (Sigma) at a stock concentration of 10 mM. The stock solutions were wrapped in foil and maintained at 4°C or −20°C.

NB4 and Jurkat cell lines and the AGS and MKN-45 gastric cancer cell line were purchased from the Shanghai Institutes for Biological Sciences (Shanghai, China). The cells were maintained in humidified room air containing 5% CO₂ at 37°C, NB4 and Jurkat were cultured in RPMI-1640 medium and AGS and MKN-45 cells were cultured in DMEM medium supplemented with 10% fetal bovine serum (FBS) and 1% penicillin-streptomycin. RPMI-1640, DMEM and FBS were purchased from Gibco-BRL (Gibco, Long Island, NY, USA). Cells in the logarithmic phase of growth were used in all the experiments.

The cells were seeded at a concentration of 2×10³ cells/200 μl/well in 96-well culture plates for a cell proliferation assay with Cell Counting Kit-8 reagent (CCK-8; Dojindo Molecular Technologies, Inc., Kumamoto, Japan). Briefly, the cultured wells were treated with 20 μl/well of CCK-8 for 2 h prior to the end of incubation and the optical density of wells was measured at 450 nm using a microplate reader (Bio-Rad, Hercules, CA, USA). Results of the cell proliferation measurement were expressed as the absorbance at OD450.

Gelatinase zymography was performed in 10% Novex pre-cast SDS polyacrylamide gel in the presence of 0.1% gelatin under non-reducing conditions. Culture media (20 μl) were mixed with sample buffer and loaded with SDS-PAGE with Tris glycine SDS buffer, according to the manufacturer’s instructions (Novex). The samples were not boiled prior to electrophoresis. Following electrophoresis the gels were washed twice in 2.5% Triton X-100 for 30 min at room temperature to remove SDS. The gels were then incubated at 37°C overnight in substrate buffer containing 50 mM Tris HCl and 10 mM CaCl₂ at pH 8.0, stained with 0.5% Coomassie blue R250 in 50% methanol and 10% glacial acetic acid for 30 min, and destained. Protein standards were run concurrently and approximate molecular weights were determined by plotting the relative mobilities of known proteins.

Total RNA was extracted from the cells using RNAiso™ plus (Takara Bio, Otsu, Japan). The concentration and purity of RNA were determined by absorbance at 260/280 nm. cDNA was synthesized from 1-μg total RNA using an RNA PCR kit (both from Takara Bio, Shiga, Japan). Total RNA was reverse transcribed (10 μl of total volume) at 42°C for 30 min, 99°C for 5 min and 4°C for 5 min. The total volume of PCR reaction was 20 μl. PCR conditions were 1 cycle at 94°C for 2 min, 30 cycles at 94°C for 30 sec, 55°C for 30 sec, and 72°C for 2 min. The primers used were: GAPDH forward: 5′-TGG ACT CTG GAA TCC ATT CTG-3′ and reverse: 5′-AAA ATC CCT GTT CCC ACT CA-3′, CD-147 forward: 5′-CCA TGC TGG TCT GCA AGT CAG-3′ and reverse: 5′-CCG TTC ATG AGG GCC TTG TC-3′, MMP-2 forward: 5′-GAA GGC TGT GTT CTT TGC AG-3′ and reverse: 5′-AGG CTG GTC AGT GGC TTG-3′ and MMP-9 forward: 5′-TGC CAG TTT CCA TTC ATC TTC CAA-3′ and reverse: 5′-CTG CGG TGT GGT GGT GGT T-3′. PCR was performed on Bio-Rad MyCycler and PCR products were separated by 1.5% agarose gel electrophoresis. Images were captured using a gel imaging system (Bio-Rad, Hercules, CA, USA).

The human CD-147 siRNA and negative control siRNA were purchased from Genepharma (Shanghai, China). The siRNA sequence for CD-147 was: forward: 5′-GGU UCU UCG UGA GUU CCU CTT-3′ and reverse: 5′-GAG GAA CUC ACG AAG AAC CTG-3′. The negative control siRNA sequence was: forward: 5′-UUC UCC GAA CGU GUC ACG UTT-3′ and reverse: 5′-ACG UGA CAC GUU CGG AGA ATT-3′ for NC. DharmaFect 4 transfection reagent was purchased from Dharmacon, Inc. (Lafayette, Co, USA). AGS cells were seeded in 6-well plates at a density of 2.5×10⁵ cells/well in MEM and grown for 16 h. Transfection was performed according to the instructions provided by Dharmacon using 4 μl of DharmaFect-4 and a 100 nM/well final siRNA concentration. The cells were cultured for another 24 h. Total protein extracts were isolated and analyzed using anti-CD-147 (Invitrogen, Carlsbad, CA, USA) and GAPDH (Kangcheng, Shanghai, China) antibody immunoblotting. The cells were collected, washed twice with pre-cold PBS, and invasion was measured.

Cells were lysed in SDS lysis buffer on ice for 30 min. Cell debris was removed by centrifugation at 14,000 × g at 4°C for 5 min, and protein contents of the cell lysates were determined using a Bio-Rad protein assay kit (Hercules, CA, USA). Cell lysates with equal protein content were then loaded and separated by 12% SDS-PAGE. The protein bands were electrotransferred onto PVDF membrane and blocked with 5% non-fat milk in TBS/T buffer (20 mmol/l Tris base, 135 mmol/l NaCl, 0.1% Tween-20, pH 7.6) for 1 h. After incubation of the membrane with the appropriate antibodies, i.e., anti-phosphorylated or non-phosphorylated ERK, anti-CD-147, MMP-2 and MMP-9 (all from Cell Signaling Technology, MA, USA), for at least 4 h, specific protein bands were visualized using SuperSignal West Pico/Fico chemiluminescent substrate (Millipore Co., Bedford, MA, USA). As an internal control, the GAPDH contents in the samples were also immunoblotted using polyclonal anti-actin antibody as the primary antibody.

Invasion of gastric cancer cells was measured using the BD BioCoat™ BD Matrigel™ invasion chamber (BD Biosciences, San Jose, USA) according to the manufacturer’s instructions. Briefly, AGS and MKN-45 cells were treated with the inhibitor or siRNA for 24 h. The cells were seeded in the membrane of the upper chamber of the Transwell at a concentration of 3–5×10⁵ cells/ml in 2 ml of DMEM medium. The medium in the upper chamber was serum-free. The medium in the lower chamber contained 5% fetal calf serum as a source of chemoattractants. The cells that passed through the Matrigel-coated membrane were counted.

Each sample was analyzed in triplicate, and experiments were repeated three times. In all figures, error bars are standard deviations. Statistical analyses were performed using Microsoft Office Excel 2003 (Microsoft, Albuquerque, New Mexico, USA) and Statisca ver. 10 (StatSoft, Tulsa, OK, USA). Differences between mean values were evaluated by the unpaired t-test. Differences were considered statistically significant at P<0.05.

AGS and MKN-45 gastric cancer cells were treated with Ara-C (0, 1, 2, 4 μg) for 24 h (Fig. 1). The results indicated that Ara-C did not inhibit the proliferation of gastric cancer cells (Fig. 1A and B). NB4 and Jurkat Ara-C-sensitive cells were treated with Ara-C (0, 1, 2 and 4 μg) for 24 h. The data showed that Ara-C inhibited the proliferation of NB4 and Jurkat cells in a concentration gradient-dependent manner (Fig. 1C and D).

After the treatment of AGS and MKN-45 cells with Ara-C (0, 1, 2 and 4 μg) for 24 h, the invasion of these gastric cancer cells was detected. The data showed that Ara-C (0, 1, 2 and 4 μg) did not suppress AGS and MKN-45 cell invasion, but increased invasion depending on the Ara-C dose (Fig. 2A and C). The activity of MMP-2 and -9 was analyzed by gelatinase zymography experiments. The results indicated that the activity of MMP-2 and -9 was enhanced by Ara-C concentration in a gradient-dependent manner in the supernatant of AGS and MKN-45 cells (Fig. 2B and D).

Ara-C-induced MMP-2 and -9 activity allows us to determine whether CD-147 is involved in this process, as the literature supports that CD-147 is involved in the regulation of MMP-2 and -9 in tumor cell invasion (11–13). CD-147 expression was evaluated after AGS and MKN-45 cells were treated with Ara-C (0, 1, 2 and 4 μg) for 24 h. The results showed that Ara-C significantly upregulated CD-147 mRNA and protein levels (Fig. 3A–D). The expression of MMP-2 and MMP-9 was also analyzed, and the results showed that Ara-C significantly upregulated MMP-2 and -9 mRNA and protein levels (Fig. 3A–D).

To determine whether Ara-C mediates the CD-147-MMP-2/MMP-9 signaling pathway involved in gastric cancer cell invasiveness, CD-147 siRNAs were transfected into AGS cells for 24 h. The cells were treated with or without 4 μg Ara-C, and western blotting revealed that Ara-C-inducing MMP-2/MMP-9 expression was inhibited by CD-147 siRNAs (Fig. 4A). Invasive in vitro experiments showed that CD-147 siRNAs completely blocked off the Ara-C increased gastric cancer cell invasiveness. At the same time, CD-147 siRNAs also restored the growth inhibitory effect of Ara-C in gastric cancer cells (Fig. 4B and C).

The activation of ERK signaling pathway is closely associated with activation of the CD-147-MMP-2/MMP-9 signaling pathway (14,15). In order to determine the role of ERK signaling molecules in Ara-C-induced CD-147 and its downstream MMP-2 and -9, AGS cells were treated with 4 μg Ara-C at the indicated time points. MMP-2 and -9, and CD-147 protein levels were upregulated after activation of ERK signaling molecules (Fig. 5A). U-0126, the ERK signaling molecule inhibitor (16–18), was used to pre-treat gastric cancer cells was found to block Ara-C-induced ERK-CD-147 -MMP-2/MMP-9 activation (Fig. 5B). Accordingly, U-0126 allowed Ara-C to inhibit gastric cancer cell proliferation activity (Fig. 5C). In invasive in vitro experiments it was observed that 4 μg Ara-C increased the invasiveness of gastric cancer cells, while the ERK inhibitor U-0126 reduced gastric cancer cell invasion. More importantly, U-0126 completely blocked off the Ara-C increased gastric cancer cell invasiveness (Fig. 5D).