Introduction
Retinoblastoma (RB), is the most common type of primary ocular malignant tumor in infants, with a worldwide incidence of ~1/20,000 cases (1). Up to 1,000 new RB cases are diagnosed each year in China, which accounts for ~20% of cases worldwide (2). Notably, pediatric patients are likely to succumb within the 1–2 years following RB onset if no treatments are received (3). To relieve the symptoms of RB, various therapies have been developed, in particular, chemotherapy followed by adjuvant therapy has been gradually prioritized considering the severe complications caused by other treatment strategies (4,5). However, as drug resistance is a major reason for treatment failure or RB recurrence, numerous studies have attempted to identify novel biomarkers for RB (6–8).
It has been documented that the expression of various microRNAs (miRNAs/miRs) are abnormally expressed within neoplastic tissues and cells (9). Among them, miR-34a was demonstrated to induce cell apoptosis, delay cell cycle progression and encourage cell senescence, which are considered important characteristics of tumor onset and aggravation (10–13). Furthermore, miR-34 was demonstrated to facilitate the onset and development of colorectal cancer through regulating DNA methylation (14). This miRNA also blocks epidermal growth factor receptor (EGFR) signaling by modifying phosphoinositide 3-kinase, ultimately contributing to the inhibition of gastric cancer progression (15). Notably, downregulated miR-34a expression induces the multiplication of RB cells (16), though the underlying mechanisms remain unclear.
Numerous studies have confirmed that miR-34a serves an important role in regulating tumor cell sensitivity to chemotherapeutic drugs, and this effect may be primarily associated with the post-transcriptional regulation of oncogenes by miR-34a (17–19). For instance, Vinall et al (20) reported that miR-34a increases the sensitivity of bladder cancer cells to cisplatin by inhibiting the expression levels of cyclin dependent kinase 6 and sirtuin 1. It has also suggested that miR-34a may affect the expression of various drug-resistant proteins within tumor cells by altering the Notch1 signaling pathway (21) Notably, the combined effect of miR-34a and melanoma antigen-A (MAGE-A) on the chemosensitivity of tumor cells has been confirmed (22). Furthermore, the MAGE-A family proteins have been demonstrated to be involved in guide cell carcinogenesis by interfering with the p53 signaling pathway and thereby reduce the sensitivity of tumor cells to chemotherapeutic drugs (23,24). Nevertheless, few studies have focused on the associations among miR-34a, MAGE-A and p53 underlying the chemosensitivity of RB (1).
The present study aimed to investigate the involvement of the miR-34a/MAGE-A/p53 signaling pathway on the sensitivity of RB to chemotherapeutic drugs using tissue samples and in vitro experiments, which may provide a basis for clinically diagnosing and treating RB in the future.
Materials and methods
Collection of tissues
A total of 293 RB and adjacent tumor tissue samples were collected from the Department of Ophthalmology, The First Affiliated Hospital of Zhengzhou University between April 2015 and January 2017. The mean age of patients was 23.64±17.04 months (11–61 months), with 155 females and 138 males. Notably, only infants <6 years old were included, and patients with family heredity were excluded. All the tissues were fixed with 4% paraformaldehyde at room temperature for 30 min, and were surgically diagnosed as RB by ≥2 senior pathologists who were blinded to the clinicopathological features of patients. The immediate relatives of all patients with RB provided written informed consent, and the present study was approved by the Ethics Committee of The First Affiliated Hospital of Zhengzhou University.
Cultivation of RB cells and cell transfection
Human RB cell lines, HXO-Rb44, SO-Rb50, Y79 and WERI-Rb-1, were purchased from the American Type Culture Collection (Manassas, VA, USA). The cells were cultured with RMPI-1640 supplemented with 10% fetal calf serum (both from Invitrogen; Thermo Fisher Scientific, Inc., Waltham, MA, USA) in a humidified atmosphere with 5% CO2. The culture medium was replaced every 3–4 days, and all cell lines were maintained within 20 generations. When the confluence of RB cells reached ~50%, the Lipofectamine™ 2000 Liposome Transfection kit (Invitrogen; Thermo Fisher Scientific, Inc.) was used to transfect cells. The following oligonucleotides were used: miR-34a mimics (40 nM) forward, 5′-UGGCAGUGUCUUAGCUGGUUGU-3′ and reverse, 5′-AACCAGCUAAGACACUGCCAUU-3′; miR-34a inhibitor (40 nM) forward, 5′-ACAACCAGCUAAGACACUGCCA-3′ and reverse, 5′-UGGCAGUGUCUUAGCUGGUUGU-3′; miR-negative control (NC) (40 nM) forward, 5′-CAGUACUUUUGUGUAGUA CAA-3′ and reverse, 5′-UUGUACUACACAAAAGUACUG-3′; pcDNA3.1-MAGE-A (30 nM); small interfering RNA (si)-MAGE-A 3/6/12 (30 nM) forward, 5′-CUACCUGGAGUACCGGCAG-3′ and reverse, 5′-UGGCAGUGUCUUAGCUGGUUGU-3′; pcDNA3.1-p53 (30 nM) or si-p53 (30 nM) forward, 5′-GAAGAAAAUUUCCGCAAAA-3′ and reverse, 5′-CUU UUGCGGAAAUUUUCUUC-3′ (all from Guangzhou RiboBio Co., Ltd., Guangzhou, China).
Reverse transcription-quantitative polymerase chain reaction (RT-qPCR)
Tissues and cells were lysed with addition of 1 ml TRIzol solution (Invitrogen; Thermo Fisher Scientific, Inc.), and total RNA was extracted according to the manufacturer’s protocol. The purity and concentration of RNA were determined using spectrophotometry. Subsequently, the extracted RNA was reverse transcribed into cDNA, and amplified using PCR according to the PrimeScript one-step RT-qPCR kit (Takara Bio, Inc., Otsu, Japan) protocol. The temperature protocol for RT was as follows: 16°C for 30 min; 42°C for 30 min; and 85°C for 5 min. ABI Primer Express software (version 2.0; Guangzhou RiboBio Co., Ltd.) was used to design the primers (Table I). The primers were added to the 25 µl PCR reaction system, and the reaction conditions were as follows: Pre-denaturation at 95°C for 2 min; and 30 cycles of denaturation at 94°C for 45 sec, renaturation at 59°C for 45 sec and extension at 72°C for 30 sec. The expression levels of miR-34a, MAGE-A and p53 were determined using a CFX96™ thermocycler (Bio-Rad Laboratories, Inc., Hercules, CA, USA), and quantified according to the comparative Cq method (2−∆∆Cq) (4). U6 was designated as the internal reference for miR-34a, and β-actin was adopted as the internal reference for MAGE-A and p53.
Luciferase reporter gene assay
The MAGE-A fragments that contained binding sites of miR-34a were amplified by PCR, and were then cloned into pmirGLO double luciferase expression vector (Promega Corporation, Madison, WI, USA) to form MAGE-A-wt. Through a similar approach, MAGE-A-mut was constructed, except that the binding sites of miR-34a within MAGE-A were mutated. Subsequently, RB cells that were transfected with MAGE-A-wt, MAGE-A-mut or pRL-TK reporter gene vector were transfected with miR-34a mimic, miR-34a inhibitor or miR-NC using Lipofectamine 2000. The resultant firefly luciferase and Renilla luciferase activity were measured using a Double Luciferase Reporter assay system (Promega Corporation) following 24 h of transfection.
Western blotting
Total protein was extracted using immunoprecipitation assay lysis buffer (Beyotime Institute of Biotechnology, Haimen, China) and quantified using the Bradford method. Each sample (40 µg) was separated using 8% SDS-PAGE and then transferred onto polyvinyli-dene fluoride membranes. Following blocking with 50 g/l skim milk powder for 1.5 h at room temperature, bovine serum albumin was added to diluted primary antibodies against MAGE-A (mouse anti-human; 1:500; Santa Cruz Biotechnology, Inc., Dallas, TX, USA; cat. no. sc-20034) and p53 (mouse anti-human; 1:500; Sigma-Aldrich; Merck KGaA, Darmstadt, Germany; cat. no. P5813), E-cadherin, N-cadherin, vimentin, snail (mouse anti-human; 1:1,000; Abcam, Cambridge, UK; cat. nos. ab76055, ab98952, ab8978 and ab82846, respectively) and GAPDH (rabbit anti-human; 1:1,000; Santa Cruz Biotechnology, Inc.; cat. no. sc-47724). After the samples were incubated at 4°C overnight, the membranes were washed with 1 ml/l TBS-Tween-20 three times. Then, the rabbit anti-mouse IgG H&L (DyLight® 550) preadsorbed secondary antibody (1:5,000; Abcam; cat. no. ab98786) was added, and the membranes were incubated at room temperature for 2 h. Finally, GelDoc2000 Enhanced Chemiluminescence reagent (Bio-Rad Laboratories, Inc.) was used for development in the darkroom, and the expression levels of the aforementioned proteins were assessed using ImageJ software (version 1.46; National Institutes of Health, Bethesda, MD, USA) was employed to analyze the protein bands.
Evaluation of drug susceptibility
A 100 µl cell suspension at the density of 4×103/ml was added into each well of the 96-well plates. When cells adhered to the wall, the following drug treatments were performed: Vincristine (0.05, 0.10, 0.50, 1.00 and 5.00 µg/ml), etoposide (0.5, 1.0, 5.0, 10.0 and 50.0 µg/ml), carboplatin (1.5, 7.5, 15.0, 75.0 and 150.0 µg/ml), and Adriamycin (1.0, 5.0, 10.0, 50.0 and 100.0 µg/ml) (all from Sigma-Aldrich; Merck KGaA). Subsequently, the samples were agitated and cultured at 37°C for 48 h in 5% CO2. Then, 0.5% MTT (10 µl) was added to each well and incubated at 37°C for 4–6 h in 5% CO2. Following the removal of the supernatant from the wells, each well was supplemented with 100 µl acidulated isopropanol. Following culturing at room temperature for 15–30 min, the samples were detected based on the optical density (OD) values at 570 nm using a Freedom Evolyzer-2200 Enzyme-Linked Immunometric meter (Tecan Group Ltd., Männedorf, Switzerland). Finally, the inhibition rate (IR, %) was calculated according to the formula as follows: [1− (ODmedicine / ODcontrol)] × 100. The drug concentration that caused death of half of the cells also known as the half maximal inhibitory concentration (IC50), was also calculated.
Cell apoptosis analysis
The cells were washed with 0.01 mol/l PBS, and were centrifuged at 1,500 × g for 5 min at 4°C. The supernatant was discarded, and 500 µl 1X binding buffer (Thermo Fisher Scientific, Inc.) was added to adjust the concentration of cells to 1×106/ml. Subsequently, the cells in each tube were mixed with 500 µl cell suspension, 5 µl Annexin V-fluorescein isothiocyanate and 10 µl propidium iodide (Invitrogen; Thermo Fisher Scientific, Inc.). Following the incubation of the cells at room temperature for 10 min, the samples were analyzed using a FACSCalibur flow cytometer with CellQuest software (version 3.3; both from BD Biosciences, Franklin Lakes, NJ, USA).
Statistical analysis
All data were analyzed using SPSS 13.0 statistical software (SPSS, Inc., Chicago, IL, USA). The results are expressed as the mean ± standard deviation. All the experiments were repeated for at least three times. Differences between two groups were analyzed using the Student’s t-test, and those of ≥3 groups were compared using one-way analysis of variance with the Student-Newman-Keuls post hoc test. As for the categorical data (n, %), the χ2 test was used for analysis. Spearman’s correlation analysis was also performed to assess the associations among miR-34a, MAGE-A and p53 expression. The Kaplan-Meier estimator method was applied to calculate overall survival of the studied population, and the log-rank test was used to evaluate the significance of differences. Cox regression model was used to perform univariate and multivariate analyses. Furthermore, in order to investigate the role of miR-34a and MAGE-A underlying RB etiology, their binding sites were predicted using the Targetscan database (http://www.targetscan.org/). P<0.05 was considered to indicate a statistically significant difference.
Results
miR-34a, MAGE-A and p53 expression levels in human RB tissues and cell lines
RT-qPCR was used to detect the expression levels of MAGE-A, p53 and miR-34a within human RB tissues and paracarcinoma tissues. MAGE-A and p53 expression levels within RB tissues were significantly upregulated compared with paracarcinoma tissues (P<0.01); however, the expression of miR-34a was significantly downregulated in RB tissues (P<0.01) (Fig. 1A). Further analyses demonstrated that miR-34a expression was negatively correlated with MAGE-A expression in the investigated RB tissues (rs −0.51; P<0.001), and positively correlated with p53 expression (rs, 0.61; P<0.001) (Fig. 1B). In addition, a negative correlation was identified between MAGE-A and p53 expression levels among the collected RB samples (rs, −0.33; P<0.001).
The patients with RB were divided into the high miR-34a expression (≥0.48) and low miR-34a expression groups (<0.48) (Table II). Patients were also divided into high and low MAGE-A expression groups using the mean overall expression (1.97) as a threshold. It was indicated that low miR-34a and high MAGE-A expression were significantly associated with higher stages (D-E) of the International Classification of Retinoblastoma (ICRB) system (5) (χ2, 4.67; P=0.031; χ2, 58.92; P<0.001), poor differentiation (χ2, 7.76; P=0.005; χ2, 8.87; P=0.003), positive choroidal invasion (χ2, 6.21; P=0.013; χ2, 4.53; P=0.033) and positive optic nerve involvement (χ2, 7.92; P=0.005; χ2, 9.52; P=0.002), when compared with the high miR-34a and low MAGE-A expression groups (Table II). However, no significant associations were identified between the two genes, and age, sex, family history or eye effected. In addition, Kaplan-Meier estimator analyses demonstrated that patients with RB with low miR-or high MAGE-A expression were associated with significantly lower overall survival rates compared with the high miR-34a or low MAGE-A expression groups (P<0.05; Fig. 1C). Multivariate Cox regression analysis demonstrated that low miR-34a expression [hazard ratio (HR), 2.10; 95% confidence interval (CI), 1.10–4.02; P=0.025], high MAGE-A expression (HR, 2.13; 95% CI, 1.04–4.17; P=0.037), stage D-E of the ICRB system (HR, 1.89; 95% CI, 1.04–3.45; P=0.037), poor differentiation (HR, 1.75; 95% CI, 1.02–2.94; P=0.040) and optic nerve involvement (HR, 1.92; 95% CI, 1.11–3.23; P=0.019) were independent prognostic factors for RB (Table III).
Sensitivities to carboplatin, etoposide, Adriamycin and vincristine in HXO-Rb44, SO-Rb50, Y79 and WERI-RB1 cell lines
The results of the MTT assay revealed that SO-Rb50 cells exhibited a maximal sensitivity to carboplatin (IC50, 17.43 µg/ml) and Adriamycin (IC50, 5.49 µg/ml) (Fig. 2A-C). Furthermore, the inhibitory response to etoposide were ranked as follows: Y79 (IC50, 20.96 µg/ml) > SO-Rb50 (IC50, 20.84 µg/ml) > WERI-RB1 (IC50, 9.73 µg/ml) > HXO-Rb44 (IC50, 4.92 µg/ml) (Fig. 2B). Y79 (IC50, 0.07 µg/ml) and SO-Rb50 (IC50, 0.0.11 µg/ml) exhibited higher sensitivity to vincristine compared with WERI-RBI (IC50, 0.46 µg/ml) and HXO-Rb44 (IC50, 0.51 µg/ml) (Fig. 2D). Furthermore, miR-34a and p53 expression levels were highest in SO-Rb50 cells out of all four cell lines (P<0.05; Fig. 2E). MAGE-A expression in HXO-Rb50 and SO-Rb50 were the lowest among all the cell lines (P<0.05). Thus, SO-Rb50 was selected for the following experiments.
miR-34a and MAGE-A regulate the viability and apoptosis of RB cells
The SO-Rb50 cells were treated with 17.43 µg/ml carboplatin, 20.84 µg/ml etoposide, 5.49 µg/ml Adriamycin or 0.11 µg/ml vincristine. It was demonstrated that the survival rates of SO-Rb50 were significantly inhibited following miR-34a mimics transfection (P<0.05), while silencing of miR-34a appeared to significantly increase the survival rate of SO-Rb50 cells compared with the NC group (P<0.05) (Fig. 3A). Contrary to the results of miR-34a, upregulation of MAGE-A expression was associated with significantly increased survival rates of SO-Rb50 cells in response to carboplatin, etoposide, Adriamycin and vincristine treatment, when compared with the NC group (all P<0.05; Fig. 3A). Concurrently, transfection of si-MAGE-A significantly suppressed the survival rate of SO-Rb50 cells under all four treatment conditions, when compared with the NC group (all P<0.05). In addition, the viability of SO-Rb50 cells was significantly attenuated in the miR-34a mimics and si-MAGE-A groups (P<0.05), while it was significantly improved in the miR-34a inhibitor and pcDNA3-MAGE-A groups compared with the NC group (all P<0.05; Fig. 3B).
miR-34a and MAGE-A regulate epithelial mesenchymal transition (EMT)-specific proteins within RB cells
Flow cytometry was performed to assess the percentage of apoptotic cells and explore the effect of miR-34a and MAGE-A on the apoptotic conditions of SO-Rb50 cells. The results revealed that miR-34a inhibitor and pcDNA3.1-MAGE-A groups significantly reduced the apoptotic rate of SO-Rb50 cells (P<0.05), and the apoptotic rate of miR-34a mimics and si-MAGE-A groups was significantly higher compared with that of the NC group (all P<0.05; Fig. 4A). In addition, as demonstrated in Fig. 4B, miR-34a mimics transfection resulted in increased E-cadherin expression, along with decreased N-cadherin, vimentin and snail expression (all P<0.05), whereas miR-34a inhibitor resulted in the opposite observations (all P<0.05). In addition, treatment with pcDNA3.1-MAGE-A reduced E-cadherin, and increased N-cadherin, vimentin and snail expression levels in SO-Rb50 cells compared with the NC group (all P<0.05). Concurrently, silencing of MAGE-A caused an increase in E-cadherin expression, and a reduction in N-cadherin, vimentin and snail expression levels (all P<0.05).
miR-34a targets MAGE-A to affect the expression of MAGE-A and p53
Based on the TargetScan database, three members of the MAGE-A family (MAGE-A3, MAGE-A6 and MAGE-A12) were identified as the potential targets of miR-34a (Fig. 5A). The double luciferase reporter gene assay was used to further explore whether miR-34a directly binds to the 3′-untranslated region of MAGE-A. miR-34a mimics significantly reduced the luciferase activity in the MAGE-A3, MAGE-A6 and MAGE-A12 groups (all P<0.05), but no significant regulatory effect was observed when MAGE-A3, MAGE-A6 and MAGE-A12 were mutated compared with the NC group (Fig. 5A). Furthermore, transfection with miR-34a mimics significantly decreased MAGE-A and increased p53 expression levels, while miR-34a inhibitor promoted MAGE-A and reduced p53 expression levels (all P<0.05; Fig. 5B). Additionally, overexpression of MAGE-A significantly decreased the mRNA and protein expression of p53 (P<0.05); however, no significant differences were observed regarding the effect of MAGE-A on miR-34a (Fig. 5C and D).
P53 alters RB cell chemosensitivity
p53 expression was significantly upregulated following transfection with pcDNA3.1-p53 and downregulated following si-p53 transfection compared with the NC control (both P<0.05; Fig. 5E). Furthermore, over-expression of p53 was associated with significantly reduced the survival rates of SO-Rb50 cells following treatment with carboplatin, etoposide, Adriamycin and vincristine compared with the NC group (all P<0.05; Fig. 6A). The apoptotic rate of cells in the pcDNA3.1-p53 group was significantly increased compared with the si-p53 group (P<0.05; Fig. 6B), and cell viability in the si-p53 group was increased compared with that in the pcDNA3.1-p53 group (P<0.05; Fig. 6C). Finally, reduced p53 expression resulted in the downregulation of E-cadherin expression, and upregulation of N-cadherin, vimentin and snail expression levels (all P<0.05; Fig. 6D).
Discussion
Accumulating evidence has demonstrated that aberrant expression of miR-34a regulates the progression of various neoplasms (25), including glioma (26), breast cancer (27,28), cervical cancer (29), cholangiocarcinoma (30) and multiple myeloma (31). The results of the present study demonstrated that the expression level of miR-34a was reduced in RB tissues compared with that in paracarcinoma tissues (P<0.05). Generally, miRNAs modulate diverse biological processes by targeting downstream mRNA (32–34). For instance, miR-34a has been reported to regulate carcinogenesis by targeting several mRNAs, including MAGE-A in medulloblastoma (22), programmed cell death 1 ligand 1 or lactate dehydrogenase A in breast cancer (28), and B cell lymphoma-2 in cervical cancer (29). The expression of MAGE-A has been documented to be upregulated in gastric cancer (35), epithelial ovarian cancer (36), and lung adenocarcinoma tissues and cells (37). In agreement with the aforementioned studies, the results of the RT-qPCR analysis in the present study revealed that the mRNA level of MAGE-A was significantly upregulated in RB tissues compared with paracarcinoma tissues (P<0.05). The clinical studies also indicated that high MAGE-A expression and low miR-34a expression were significantly associated with poor prognosis of patients with RB (P<0.05). Furthermore, in order to investigate the role of miR-34a and MAGE-A underlying RB etiology, their binding sites were predicted using the Targetscan database, and validated their targeted association by performing a dual luciferase reporter gene assay. The results of the current study also suggested that MAGE-A expression in SO-Rb50 cells was significantly reduced following transfection with miR-34a mimics and increased following miR-34a inhibitor transfection. Collectively, these results suggested that the downregulation of MAGE-A by miR-34a partially mediated the pathogenesis of RB.
Furthermore, the aforementioned carcinogenic mechanism of miR-34a and MAGE-A in RB was hypothesized to be the result of their contribution to enhancing cell viability and modulating cell apoptosis (38–41). In the present study, miR-34a mimics, miR-34a inhibitor or pCDNA3-MAGE-A were transfected into SO-Rb50 cells, and their effects on viability, survival rate and apoptosis of SO-Rb50 were analyzed. The results revealed that the addition of miR-34a mimics and silencing of MAGE-A significantly suppressed cell viability and cell survival, and induced apoptosis. Furthermore, significantly increased E-cadherin expression, as well as decreased N-cadherin, vimentin and snail expression levels were observed in SO-Rb50 cells. Additionally, the transfection of miR-34a inhibitor and pcDNA3.1-MAGE-A were able to impede the apoptosis of SO-Rb50 cells, and encourage its viability and survival. It may be concluded that miR-34a, through targeting MAGE-A, is involved in RB pathogenesis by altering cell survival, cell viability, cell apoptosis and expression levels of EMT-associated proteins.
Furthermore, miR-34a and MAGE-A were demonstrated to regulate p53 expression, SO-Rb50 cell viability, survival and apoptotic rates, as well as E-cadherin, N-cadherin, vimentin and snail expression levels. p53 serves as a transcription factor that is largely involved in cell growth, cell differentiation, cell senescence and cell apoptosis (38,42–44). In addition, it has been previously suggested that the inhibitory effects of MAGE-A on p53 transactivation may lead to tumor cell resistance to chemotherapeutic drugs (etoposide) (45). miR-34a confers chemosensitivity of medulloblastoma cells through modulation of MAGE-A and p53 expression levels (22). Similar to the aforementioned publications, the present study demonstrated that miR-34a, MAGE-A and p53 were able to significantly alter the resistance of SO-Rb50 cells to vincristine, etoposide, carboplatin and Adriamycin. Taken together, the results suggest that miR-34a targeted MAGE-A and thereby regulated MAGE-A to alter the chemosensitivity of RB cells, and this phenomenon was achieved possibly through alterations to the viability, survival, apoptosis and EMT of RB cells.
In conclusion, miR-34a may function as a tumor suppressor for RB by targeting MAGE-A and altering p53 expression, indicating that the miR-34a/MAGE-A/p53 axis may serve as a therapeutic target or diagnostic biomarker for RB. However, a number of limitations must be addressed. Firstly, the results may not well be generalized to other ethnicities or a larger population, due to the limited sample size incorporated in the present study. Secondly, only the SO-Rb50 cell line was examined, and more cell lines should be focused on to verify the mechanisms. Thirdly, no animal models were established, which may provide a direct understanding regarding the influence of the miR-34a/MAGE-A/p53 axis on RB development. Lastly, long non-coding RNAs and circular RNAs are situated upstream of miRNAs, and may also participate in the mechanism underlying RB carcinogenesis. As a result, further studies are required to explore the molecular mechanism of the miR-34a/MAGE-A/p53 axis on RB progression.