Five cell lines, including 92.1, OMM2.2, OMM2.5, Mel285 and Mel290 (provided by Dr Martine Jager of Leiden University, The Netherlands), were cultured in RPMI-1640 medium supplemented with 10% fetal calf serum (FCS), 2.5 μg/ml fungizone^®/amphotericin B, 50 μg/ml gentamicin and 2 mM L-glutamine (Gibco, Invitrogen, Carlsbad, CA, USA) as previously described (19). The cells were cultured at 37°C in a humidified air/CO₂ atmosphere. Excess tumor tissues not required for diagnosis were obtained from primary uveal melanoma tumors of 12 patients who were enrolled at the Xi’an No. 4 Hospital following approval by the Institutional Research Committee. The tissues were preserved at −80°C. Human samples were obtained after informed consent was provided by the patients, and in accordance with the Declaration of Helsinki. The present study was approved by the Ethics Committee of Xi’an No. 4 Hospital.

An MTT assay was employed for the evaluation of melanoma cell growth and proliferation. The experiments were carried out in 96-well plates according to the manufacturer’s instructions (Roche GmbH, Mannheim, Germany). In the MTT test, tetrazolium salts were transformed by active enzymes of the cells into intracellular formazan deposits and the cells were incubated for 4 h with the tetrazolium salts. After 4 h, the purple formazan salts formed became soluble. Absorbance was determined at 490 nm.

Cell migration assays were performed with a two-chamber Transwell^® device (Corning, Edison, NJ, USA). The cells were collected and suspended in RPMI-1640 medium containing 10% FCS. A Transwell apparatus with an 8-μm pore size membrane (Corning) was used to analyze the migration activity. The cells were suspended in 120 ml RPMI-1640 medium containing 0.1% FCS and then seeded in the upper chamber of the device, while 500 ml RPMI-1640 medium containing 10% FCS was added to the lower chamber. The device was incubated at 37°C for 12 h. The inner side of the upper chamber was wiped with a wet cotton swab to remove the cells, while the outer side of the chamber was gently rinsed with phosphate-buffered saline (PBS) and treated with 95% ethanol for 30 min. The membrane was then washed with PBS again and stained with a 0.1% crystal violet staining solution for 30 min. After being dried, images of the membrane were taken for >5 fields and the cells were counted. At least three experiments were performed using the Transwell assays for each cell line.

The plasmids pGEM-HA-GαqQL were constructed as described in a previous study (20,21). The 1.1 kb HindIII-NotI fragment containing the coding sequence for HA-Gαq was ligated into the blunted SalI sites of pGEM-9Zf, which was confirmed by restriction mapping and nucleotide sequencing.

Scrambled small interfering RNA (siRNA) and siRNA targeting GNAQ (sc-35429-SH) were purchased from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA, USA). The cells were transfected with scrambled or GNAQ siRNA according to the manufacturer’s instructions. Briefly, the scrambled and GNAQ siRNAs (30 pmol) were diluted in 500 μl Dulbecco’s modified Eagle’s medium (DMEM) and mixed with 5 μl Lipofectamine^® RNAiMAX (Invitrogen). After 15 min of incubation at room temperature, the complexes were added to the cells to a final volume of 3 ml medium. The cells were then harvested at the indicated time periods for subsequent analysis. The efficiency of the GNAQ siRNA was confirmed by western blot analysis of the flag expression.

The Mel285 cells were tranfected with blank vectors or HA-GαqQL. Following fixation with 4% formaldehyde in PBS for 10 min, the cells were permeabilized with 0.3% Triton X-100 in PBS for 10 min, followed by 1 h of blocking with 5% bovine serum albumin in PBS. The cells were stained with the primary antibody, rabbit anti-YAP, overnight at 4°C. Fluorescence staining was performed using Alexa 488-conjugated goat anti-rabbit (green) antibody. Confocal microscopy was performed using a multiphoton Zeiss LSM 510 laser scanning microscope (Carl Zeiss, Jena, Germany). Confocal images were obtained with LSM 510 software.

The mRNA of uveal melanoma cells or tumor tissues was extracted with TRIzol^® RNA-extraction reagent (Gibco, Rockville, MD, USA). Approximately 5 μg of total RNA for each sample were reverse-transcribed into first-strand cDNA for quantitative polymerase chain reaction (RT-qPCR) analysis. RT-qPCR was performed in a final volume of 10 μl, containing 5 μl of SsoFast™ EvaGreen^® Supermix (Bio-Rad, Hercules, CA, USA), 1 μl of cDNA (1:50 dilution) and 2 μl each of the forward and reverse primers (1 mM). The steps in the RT-qPCR were performed as follows: 94°C for 2 min for initial denaturation; 94°C for 20 sec, 58°C for 15 sec and 72°C for 15 sec; 2 sec were used for plate reading for 40 cycles; and a melt curve was generated between 65 and 95°C. β-actin was used as a quantitative and qualitative control to normalize the gene expression. Data are analyzed using the formula: R = 2^{−(ΔCt sample – ΔCt control)}. The primers used in this experiment are shown in Table I.

Proteins were extracted from uveal melanoma cells and the protein concentration was quantified by the Bradford assay. A total of 20 μg of protein was separated by 12% sodium dodecyl sulfate polyacrylamide gel electrophoresis and transferred to a nitrocellulose membrane (BioTeke, Beijing, China), which was incubated with 2% non-fat dry milk in Tris-buffered saline (TBS) to block non-specific binding at room temperature for 1 h. After being blocked, the membrane was incubated in blocking buffer containing primary antibodies overnight at 4°C. Antibodies including anti-Jag-1, anti-NICD, anti-Hes-1 and anti-β-actin were purchased from Santa Cruz Biotechnology, Inc.. Subsequently, the membrane was washed with TBS-Tween for 10 min and incubated with horseradish peroxidase-conjugated secondary antibody (Tiangen Corporation, Beijing, China) diluted in blocking buffer for 1 h at room temperature. After being washed again with TBS-Tween buffer for 10 min, proteins were detected using enhanced chemiluminescence (Pierce, Rockford, IL, USA).

The statistical significance of differences was assessed using SPSS 15.0 (SPSS Science, Chicago, IL, USA). Differences between the two groups were analyzed by the Student’s t-test. Correlations were assessed using the Spearman’s rank correlation coefficient. Differences between multiple groups were analyzed by the ANOVA. P<0.05 was considered to indicate a statistically significant result.

Mutant GNAQ has been identified in certain uveal melanoma cell lines including 92.1, OMM2.2 and OMM2.5 (22). To determine the effect of GNAQ on uveal melanoma cells, the cultured 92.1, OMM2.2 and OMM2.5 cells were transfected with scrambled or GNAQ siRNA. Transfection of GNAQ siRNA resulted in a 67, 70 and 74% inhibition of GNAQ mRNA expression in 92.1, OMM2.2 and OMM2.5 cells, respectively (Fig. 1A). The viability of these transfected cells was evaluated by MTT over 24–96 h. The results showed that GNAQ knockdown significantly inhibited the viability of 92.1, OMM2.2 and OMM2.5 cells (Fig. 1B–D). GNAQ knockdown also reduced the migratory rates of 92.1, OMM2.2 and OMM2.5 cells compared with the respective controls (Fig. 1E). Moreover, cells such as Mel285 and Mel290 were absent from the mutant GNAQ. Therefore, we examined the effect of the GNAQ expression on these cells through transfection with human influenza hemagglutinin A epitope (HA)-tagged GαqQL. Consequently, HA-GαqQL transfection resulted in a significant increase in the GNAQ mRNA expression (Fig. 1F), and increased the viability and migration of the Mel285 and Mel290 cells (Fig. 1G–I). These results indicated that mutant GNAQ may promote uveal melanoma cell viability and migration.

Activation of Notch signaling is involved in the carcinogenesis of various tissues including ocular tissues (23–25). As mutant GNAQ was highly expressed in uveal melanoma cells (26), we hypothesized that GNAQ causes Notch signal activation and contributed to cell growth and migration. To validate our hypothesis, GNAQ expression in 92.1 cells was knocked down by GNAQ siRNA, and the expression of Jag-1 (Notch ligand), NICD and Hes-1 (Notch target gene) was analyzed by western blotting. A decreased expression of Jag-1, NCID and Hes-1 was observed in the GNAQ knockdown cells (Fig. 2A). Furthermore, HA-GαqQL transfection in the Mel285 cells notably upregulated the expression of Jag-1, NCID and Hes-1 (Fig. 2B). These results revealed that the Notch signaling may be triggered by mutant GNAQ in uveal melanoma cells.

To confirm the effect of Notch activation induced by GNAQ on uveal melanoma cells, the Mel285 cells were transfected with HA-GαqQL and treated with or without 5 μmol/l MRK003 (a potent γ-secretase inhibitor). Compared with the HA-GαqQL transfection group, treatment with MRK003 significantly inhibited cell viability and migration (Fig. 3A and B). Furthermore, Hes-1 expression in these cells was decreased by MRK003 (Fig. 3C). These results indicated that inactive Notch impeded the effect of GNAQ on uveal melanoma cells.

It has been shown that nuclear YAP induces Notch activation (27). Therefore, we investigated whether YAP is involved in GNAQ-induced Notch activation. Mel285 cells were transfected with HA-GαqQL, and the YAP phosphorylation levels and nuclear translocation were detected. Compared with the Mel285 cells transfected with empty vectors, those transfected with HA-GαqQL showed decreased YAP phosphorylation (Fig. 4A) and increased YAP nuclear translocation (Fig. 4B). Furthermore, YAP expression in the 92.1 cells was inhibited, and Jag-1 and Hes-1 expression was reduced by YAP siRNA (Fig. 4C). Additionally, the viability and migration of the 92.1 cells was inhibited by the YAP siRNA transfection (Fig. 4D and E). These results demonstrated that GNAQ inhibits YAP phosphorylation and promotes YAP translocation, which mediated Notch activation in uveal melanoma cells.

To confirm the connection between the GNAQ, YAP and Notch signaling in human uveal melanoma, total RNA was isolated from 12 fresh-frozen human uveal melanoma samples, and the mRNA levels of GNAQ, YAP, Jag-1 and Hes-1 were analyzed. Positive correlations between the GNAQ and Jag-1 mRNA levels and between the GNAQ and Hes-1 mRNA levels were observed (Fig. 5A and B). However, no correlation was found between the GNAQ and YAP mRNA levels, which suggested that GNAQ regulated YAP protein phosphorylation but not its transcript level.