Two human retinoblastoma cell lines, SO-Rb50 and Y79, were used in this study. The SO-Rb50 cell line was established in the Zhongshan Ophthalmic Center, Sun Yat-Sen University, Guangzhou, China, as previously described (15,16). The Y79 cell line was obtained from the American Type Culture Collection (ATCC; Manassas, VA, USA). The maintenance of these cell lines was carried out as previously described (17–19). In brief, the cells were cultured in RPMI-1640 medium (HyClone Co., Logan, UT, USA) supplemented with 10% fetal bovine serum, 100 U/l penicillin, and 100 U/l streptomycin at 37°C in a humidified atmosphere of 95% air/5% CO₂. The culture medium was replaced every 2 days.

A third generation of the self-inactivating lentiviral vector containing a cytomegalovirus (CMV) promoter-driven enhanced green fluorescence protein (eGFP) reporter was purchased from GeneChem Co., Ltd. (Shanghai, China). The lentiviral vector system was made from 3 types of plasmids, the pGCL-GFP vector (5′LTR, 3′LTR and woodchuck hepatitis virus post-transcriptional regulatory element), the pHelper 1.0 (gag, pol and rev element) vector and the pHelper 2.0 (VSV-G element) vectors. The pGCL-GFP vector encoding a sequence targeting the human PAX6 gene (NCBI Reference Sequence ID: NM_000280.3) was assembled by GeneChem. PCR and DNA sequencing confirmed the accurate insertion of the small interfering RNA (siRNA) sequences. For the preparation of the recombinant lentiviral vectors, the pHelper 1.0 plasmid (15 μg), the pHelper 2.0 plasmid (10 μg) and the pGCL-GFP-siRNA or the pGCL-GFP plasmid (negative control, 20 μg) were co-transfected into subconfluent 293T cells in serum-free medium using Lipofectamine 2000 reagent (Invitrogen Co., Carlsbad, CA, USA). Untransfected cells were used as blank controls. After 8 h of incubation, the medium was changed to serum-containing medium. High titers of recombinant lentiviral vectors with PAX6-RNAi were harvested after the supernatant became concentrated 48 h later. Before the experiments, 7 human PAX6-specific double-stranded, siRNA sequences were synthesized and screened (Shanghai Genepharma Co., Ltd., Shanghai, China). We selected 2 sequences for co-transfection into the cell lines in our study. The target sites were CGTCCATCTTTGCTTGGGAAA and TACCAAGCGTGTCATCAATAA.

Before the experiments, 7 human PAX6-specific double-stranded, siRNA sequences (KD1-KD7) were synthesized and screened. The details of the target sequences are presented in Table I. For transfection, the human retinoblastoma cells were seeded in 48-well plates at a concentration of 5×10⁵ cells/ml in a volume of 200 μl for each well. The cells were transfected either with PAX6-RNAi-GFP (PAX6 inhibition study group) or with GFP lentiviral vectors with scrambled siRNA (sequence: ‘TTCTCCGAACGTGTCACGT’) in serum-free medium for 20 h with a total multiplicity of infection (MOI) of 80. At 4 days after transfection, the reporter GFP gene expression was examined under a fluorescence microscopy. The transfected human retinoblastoma cell lines were screened using FACS for the following experiments.

The standard colorimetric cell counting kit-8 (CCK-8; Dojindo Laboratories, Kumamoto, Japan) was used for the determination of the number of viable cells in the cell proliferation assays. The transfected retinoblastoma cells were seeded with a volume of 90 μl cell suspension (5,000 cells/well) into 96-well plates and 10 μl CCK-8 were added to each well followed by incubation for 4 h at 37°C. Optical densities were read using a microplate reader scanning at 450 nm. This procedure was repeated every 24 h during a 6-day period. Cell survival rates were measured at 3 time points of the cell growth curve through the log phase of growth for each cell line.

Total RNA was extracted from the cells using TRIzol reagent [Tiangen Biotech (Beijing) Co., Ltd., Beijing, China; Cat. no. DP405]. The reverse transcription and the PCR amplification reactions were performed according to the M-MLV reverse transcriptase protocol (Tiangen Biotech) (Cat. no. KR104). qPCR (ABI PRISM 7500, version 1.41) was performed to detect the mRNA levels of PAX6 and cell cycle-related molecules in both retinoblastoma cell lines. The details of the primers for qPCR are presented in Table II and qPCR was carried out as previously described (20–22). The reaction system was 20 μl, including 4 μl RNase-free ddH₂O, 4 μl cDNA template, 1.7 μl mixed primers, 10 μl SuperReal PreMix and 0.3 μl ROX Reference Dye l (Tiangen Biotech) (Cat. no. FP204). The PCR running conditions were as follows: 120 sec at 95°C for the initial denaturation followed by 45 cycles of 20 sec at 95°C for denaturation, 25 sec at temperature for annealing, and 30 sec at 95°C for extension. The threshold cycle (Ct value), which is the cycle number at which the amount of the amplified gene of interest reaches a fixed threshold, was subsequently determined. Relative quantification values of the target gene mRNA levels were normalized to endogenous human β-actin gene levels and calculated using the 2^−ΔΔCt method. Each experiment was performed 3 times.

The early apoptosis of the transfected cells was detected using the PE Annexin V Apoptosis Detection kit I (BD Pharmingen, San Diego, CA, USA) (Cat. no. 559763). The 2 retinoblastoma cell lines were washed twice with cold PBS and then re-suspended in 1× binding buffer at a concentration of 1×10⁶ cells/ml. A volume of 100 μl of the solution (1×10⁵ cells) was transferred to a 5 ml culture tube and 5 μl of PE Annexin V and 5 μl 7-AAD were added. The cells were gently dispersed and then incubated for 15 min at room temperature (25°C) in the dark. Finally, 400 μl of 1× binding buffer were added to each tube. The cells were analyzed using a BD FACSCalibur flow cytometer (BD Biosciences, San Diego, CA, USA) equipped with a 540 nm exciting laser. The results are shown as percentages of the count of the early apoptotic cells to the total cell count in the SO-Rb50 and Y79 retinoblastoma cell lines. Each experiment was performed 3 times.

A cell suspension corresponding to 1×10⁶ cells was collected and centrifuged. The supernatant was discarded, 1 ml of phosphate-buffered saline (PBS) at room temperature was added and the cell pellet was re-suspended. The full volume of re-suspended cells was transferred to 3 ml of absolute ethanol pre-cooled at −20°C by pipetting the cell suspension slowly into the ethanol while swirling at top speed. The cells were left in ethanol at −20°C for 15 min. The cells were centrifuged, the ethanol was discarded, 4 ml of PBS were added at room temperature and the cells were re-suspended to allow them to rehydrate for 15 min. Again, the cell suspension was centrifuged and the supernatant was discarded, 1 ml of DNA staining mixed solution was added and the cell suspension re-suspended. Following incubation for 30 min at room temperature, the cells were analyzed by FACS in the presence of the dye. The cells were then passed through a BD FACSCalibur flow cytometer equipped with a 488 nm argon laser to measure the DNA content. The data analysis was performed using Cell Quest (BD Biosciences) and ModFit LT software (Verity Software House Inc, Topsham, ME, USA) software. The results are presented as percentages of the total cell count in different phases of the cell cycle, namely the G0/G1 phase (diploid cells), S phase (diploid and tetraploid cells), and G2/M phase (tetraploid cells). Each experiment was performed 3 times.

Approximately 1×10⁷ transfected human retinoblastoma cells of each of the 2 cell lines was collected into a 15 ml centrifuge tube and re-suspended in 600 μl cell lysis buffer [20 mM Tris (pH 7.5), 150 mM NaCl, 1 mM EDTA, 1% Triton X-100, 2.5 mM sodium pyrophosphate, 1 mM β-glycerophosphate, 1 mM Na₃VO₄, 1 μg/ml leupeptin, 1 mM phenylmethanesulfonyl fluoride (PMSF)]. The cells remained in that medium on ice for 30 min with re-dispersion every 5 min. The lysates were then centrifuged at 12,000 rpm for 5 min at 4°C. The protein concentrations were determined using a NanoPhotometer (Implen GmbH, München, Germany). The proteins were denatured in 5× loading buffer for 10 min and were separated by sodium dodecyl sulfate polyacrylamide gel electropheresis (SDS-PAGE) using 5% stacking and 12% separating gels. They were then electroplotted onto polyvinylidene difluoride (PVDF) membranes (0.2 μm, Immobilon-P; Millipore, Billerica, MA, USA) in the way of a semi dry process with 20 V for 30 min. After being blocked in a blocking solution (Beyotime Institute of Biotechnology, Beijing, China; Cat. no. P0023B) for 30–60 min, the membranes were incubated overnight at 4°C with primary antibodies, including mouse anti-human β-actin (diluted at 2,000; Beyotime Institute of Biotechnology; Cat. no. AA128), mouse anti-human PAX6 (diluted 500; Abcam Co., Hong Kong, China; Cat. no. Ab78545), rabbit anti-human cyclin-dependent protein kinase 2 (CDK2) (diluted at 1,000; Cat. no. 2546), mouse anti-human proliferating cell nuclear antigen (PCNA) (diluted at 1,000; Cat. no. 2586), mouse anti-human cyclin-dependent kinase 1 (CDK1) (diluted at 1,000; Cat. no. 9116), rabbit anti-human Bcl-2 (diluted at 1,000; Cat. no. 2870) and rabbit anti-human BAX (diluted at 1,000; Cat. no. 5023) (all from Cell Signaling Technology, Inc., Beverly, MA, USA). After being rinsed in PBS solution with 0.5%v/v Tween-20 (PBST) thrice for 10 min, the PVDF membranes were incubated at 37°C for 1 h with the secondary goat anti-mouse IgG(H+L) antibody (diluted at 2,000) (Cat. no. A0216) or goat anti-rabbit IgG(H+L) antibody (diluted at 2,000) (both from Beyotime Institute of Biotechnology) (Cat. no. A0208) conjugated with horseradish peroxidase (HRP). Using PBST, the rinsed PVDF membranes were subjected to enhanced chemiluminescence (ECL) using an ECL detection kit (Beyotime Institute of Biotechnology) (Cat. no. P0018) and quantified using Quantity One software (Bio-Rad Laboratories Inc., Hercules, CA, USA). Each western blot analysis was performed 3 times. For further analysis, the means were calculated and the data were normalized.

Statistical analysis was performed using a commercially available software package, SPSS (version 20.0 for Windows IBM; SPSS, Inc., Chicago, IL, USA). The distributions of the parameters were evaluated using the Levene test. The data are presented as the means ± standard deviation. Statistical analysis of the differences was carried out using an independent samples t-test and paired-samples t-test. A P-value of <0.05 was considered to indicate a statistically significant difference.

To investigate the function of PAX6 in both retinoblastoma cell lines, the gene was silenced by transfecting the retinoblastoma cells with lentiviral vectors carrying GFP-PAX6-RNAi sequences (PAX6 inhibition study group). The SO-Rb50 and Y79 cell lines were transfected with the lentiviral vectors at an MOI of 80. The successfully transfected cells expressed GFP and were examined under a fluorescence microscope at 4 days after transfection (Fig. 1). The efficiency of the infection was approximately 80% at an MOI of 80. The stably transfected cells were screened by FACS for the next step of the study.

Seven human PAX6-specific small interfering RNA sequences (KD1-KD7) were primarily screened using RT-qPCR in the SO-Rb50 cell line. The highest inhibition rate was 45% in KD4, indicating that the siRNA failed to have sufficient inhibitory effects (Fig. 2A). To obtain a higher inhibition rate, 2 combined target sites (KD1 + KD4, KD3 + KD4 and KD4 + KD5) were used to inhibit PAX6 in the 2 human retinoblastoma cell lines. The inhibition rate of the KD4 + KD5 group reached 70%, providing sufficeint inhibitory effects for the following experiments (Fig. 2B, left panel). The inhibitory effects on endogenous PAX6 expression were confirmed by western blot analysis. The protein levels of PAX6 in the SO-Rb50 and Y79 retinoblastoma cell lines were significantly lower in the knockdown groups than in the control groups (Fig. 2B).

To examined the proliferation of the 2 retinoblastoma cell lines following the inhibition of PAX6, cell proliferation assay was performed using the standard colorimetric CCK-8. OD values were measured at 5 time points of the cell growth curve for each cell line. A significant increase in the cell survival rates was observed in the knockdown group in these 2 cell lines. The OD values of the 2 cell lines increased sharply at 1–5 days after cell plating in the knockdown group compared with the control groups (Fig. 3).

The effects of endogenous PAX6 inhibition on cell apoptosis were examined using the PE Annexin V Apoptosis Detection kit I. Flow cytometric analysis revealed a reduced early apoptotic rate in the PAX6-knockdown groups compared to the control groups. The percentage of apoptotic cells was significantly lower in the PAX6-knockdown group than the corresponding negative GFP-control groups (t=4.036, P>0.05, n=3) and the negative control group without transfection (t=7.948, P<0.05, n=3) (Fig. 4).

We then determined the effects of the inhibition of endogenous PAX6 on the cell cycle by FACS. For the SO-Rb50 cell line, the percentage cell count in the S phase was significantly higher in the PAX6-knockdown group than in the negative GFP-control group (40.00±1.10 vs. 34.69 ± 3.17%; t=−4.44; P<0.05, n=3). For the the Y79 cell line, the percentage cell count in the G2/M phase was significantly higher in the PAX6-knockdown group than in the negative GFP-control group (16.92±1.89 vs. 10.78±2.23%; t=−8.31, P<0.05, n=3) (Fig. 5).

To determine how the suppression of PAX6 affects cell cycle distribution, we measured the mRNA levels of cell cycle-related genes by RT-qPCR. In the SO-Rb50 retinoblastoma cell line, the mRNA levels of cdc25A, CDK2, PCNA and CDK1 were significantly higher in the PAX6 inhibition study group than in negative GFP-control group (cdc25A, 2.93±0.16 vs. 1.00±0.00; t=−21.47; P<0.01, n=3; CDK2, 1.75±0.11 vs. 1.00±0.00; t=−11.56; P<0.01, n=3; PCNA, 3.16±0.35 vs. 1.00±0.00; t=−10.86; P<0.01, n=3; CDK1, 2.11±0.21 vs. 1.00±0.00; t=−9.22; P<0.05, n=3) (Fig. 6A). However, the mRNA level of p21 was significantly lower in the PAX6 inhibition study group than in negative GFP-control group (0.35±0.03 vs. 1.00±0.00; t=37.25; P<0.01, n=3) (Fig. 6A). In the Y79 retinoblastoma cell line, the mRNA levels of cdc25A, CDK2 and p21 were significantly lower in the PAX6 inhibition study group than in negative GFP-control group (cdc25A, 0.36±0.07 vs. 1.00±0.00; t=15.91; P<0.01, n=3; CDK2, 0.28±0.03 vs. 1.00±0.00; t=46.77; P<0.01, n=3; p21, 0.37±0.08 vs. 1.00±0.00; t=14.37; P<0.01, n=3) (Fig. 6B). The mRNA levels of PCNA and CDK1 were significantly higher in the PAX6 inhibition study group than in negative GFP-control group (PCNA, 2.79±0.16 vs. 1.00±0.00; t=−19.47; P<0.01, n=3; CDK1, 1.56±0.31 vs. 1.00±0.00; t=−3.15; P<0.05, n=3) (Fig. 6B).

To determine the molecular mechanisms involved in the inhibition of PAX6 expression, as well as the signaling pathways associated with the cell cycle and apoptosis, we measured the expression of apoptosis-related Bcl-2 and BAX and the cell cycle regulatory proteins, CDK1 and PCNA, by western blot analysis. In both retinoblastoma cell lines, the levels of Bcl-2, CDK1 and PCNA were significantly upregulated in the PAX6 inhibition study groups than in negative GFP-control groups (Bcl-2, t=−5.90, P<0.05 and t=−4.86, P<0.05, resp. n=3; CDK1, t=−5.27, P<0.05; and t=−7.60, P<0.05, resp. n=3; PCNA, t=−11.49, P<0.01; and t=−4.32, P=0.05, resp. n=3) (Fig. 7). However, the level of BAX was downregulated in both retinoblastoma cell lines (t=4.51, P<0.05; and t=67.96, P<0.01, resp. n=3) (Fig. 7).