miR-34a inhibits proliferation, migration and invasion of paediatric neuroblastoma cells via targeting HNF4α

Abstract Objective To investigate the potential mechanism of microRNA-34a (miR-34a) on proliferation, migration and invasion of paediatric neuroblastoma cells. Methods The expression of miR-34a and hepatocyte nuclear factor 4α (HNF4α) in paediatric neuroblastoma tissues were detected by RT-q PCR and Western blot, respectively. Cell proliferation, migration, invasion and the expression of matrix metalloproteinase 2 (MMP-2) and matrix metalloproteinase 14 (MMP-14) after transfection of miR-34a mimics or HNF4α siRNA into SH-SY5Y cells were detected by MTT assay, Transwell assay and Western blot assay, respectively. The target relationship between miR-34a and HNF4α was verified by TargetScan online prediction and dual-luciferase assay. Cell proliferation, migration and invasion of SH-SY5Y cells after overexpression of miR-34a and HNF4α were detected. Results The expression level of miR-34a was decreased (p < .05) while the expression level of HNF4α was increased (p < .05) in paediatric neuroblastoma tissues. Over- expression of Mi-34a or knockdown of HNF4α in SH-SY5Y cells could lead to a decreased of cell proliferation, migration, invasion and the expression of MMP-2 and MMP-14 (p < .05). The results of TargetScan online prediction and dual-luciferase assay indicted that HNF4α was a potential target gene for miR-34a. Overexpression of HNF4α could reverse the inhibition of miR-34a on proliferation, migration and invasion of SH-SY5Y cells. Conclusion The expression of miR-34a was down-regulated in paediatric neuroblastoma tissues, and overexpression of miR-34a could inhibit proliferation, migration and invasion of SH-SY5Y cells by targeting HNF4α.


Introduction
Neuroblastoma (NB), originating from sympathetic neurons, is a neoplasm caused by the deterioration of undifferentiated neuroblasts. It occurs mainly in the abdomen, mediastinum and neck of children and is the most common extracranial solid tumour in children [1]. The incidence of neuroblastoma in children under 15 years old was 10.54/100,000, ranking the first in peripheral nervous system malignancies in children [2]. Neuroblastoma is characterized by high degree of malignancy, rapid development, strong invasion and great differences in clinical manifestations. Although surgery, chemotherapy, radiotherapy, immunotherapy and other multidisciplinary comprehensive treatment can alleviate its clinical symptoms, the overall survival rate of NB in China is still difficult to improve. Immunotherapy and molecular targeted therapy are new directions in the clinical treatment of NB [3,4].
MicroRNA (miRNA) is a kind of endogenous non-coding RNA molecules that regulate gene expression by degrading target mRNA or inhibiting its translation. It has attracted much attention because of its important regulatory role in the development of the body and the development of diseases. At present, some miRNA molecules as tumour suppressors have shown a broad prospect in the clinical treatment of cancer [5]. Previous studies have found that a variety of miRNAs are abnormally expressed in NB tissues and are associated with factors such as NB metastatic, tumour size and lymph node metastasis, including up-regulation of miR-181, miR-21, and miR-15a and downregulation of miR-34a, miR-101 and miR-145 [6][7][8]. In this study, the effect of miR-34a expression and overexpression of miR-34a in NB tissues on proliferation, migration and invasion of NB cells was examined to explore its mechanism in the development of NB.

Clinical specimens
Thirty-five cases of paediatric neuroblastoma diagnosed by the department of pathology in the Henan Xinxiang Central hospital from July 2016 to April 2018 were selected as subjects. There were 19 males and 16 females aged from 3 months to 8 years, who underwent surgical resection of cancer tissues and were quickly put into liquid nitrogen. All the patients did not receive chemotherapy and radiotherapy before surgery. In addition, 15 cases of normal children's adrenal tissue preserved by the pathology department of our hospital were selected as the normal control. The tissues were collected and stored in the refrigerator at À80 C. The neuroblastoma tissues and adrenal tissues were labelled as the non-tumour tissues group and the cancer tissues group, respectively. This study was approved by Henan Xinxiang Central Hospital, and prior written and informed consent was obtained from each patient.

Materials
Human neuroblastoma cell line SH-SY5Y was purchased from the Cell Bank of the Chinese Academy of Sciences (Shanghai, China). Foetal bovine serum, DMEM culture medium, trypsin and pcDNA 3.1 vector were purchased from Thermo Fisher Scientific (Waltham, MA). Trizol Plus RNA purification kit and Lipofectamine 2000 were purchased from Invitrogen (Carlsbad, CA). TaKaRa reverse transcription kit and real-time fluorescence quantitative kit were purchased from Bao Bioengineering Co., Ltd. (Dalian, China). PVDF membrane, ECL luminescent solution, RIPA lysate, crystal purple 4% and paraformaldehyde were purchased from Solarbio Company (Beijng, China). MTT, dimethyl sulphoxide were purchased from Sigma-Aldrich (St. Louis, MO). The double luciferase assay kit was purchased from Promega (Madison, WI). HNF4a 3'-UTR wild and mutant type plasmids were constructed by Han Heng biological Co., Ltd (Shanghai, China). Transwell was purchased from Corning Inc (Corning, NY). Control mimics and miR-34a mimics were designed and synthesized by Gemma Company (Shanghai, China). Rabbit anti-human hepatocyte nuclear factor 4a (HNF4a) polyclonal antibody, rabbit anti-human MMP-2 polyclonal antibody, rabbit antihuman MMP-14 polyclonal antibody and rabbit anti-human GAPDH polyclonal antibody were purchased from Emerald Technology Co., Ltd (Wuhan, China). Control siRNA and HNF4a siRNA were designed and synthesized by Wei Zhen Biotechnology Co., Ltd (Shandong, China). The primers were synthesized by Liuhe Huada Gene Company (Beijing, China). Cell incubator was purchased from Thermo Fisher Scientific (Waltham, MA). Enzyme labelling instrument of HBS-1096B was purchased from de tie Experimental Equipment Co., Ltd (Nanjing, China). Microscope was purchased from Bingyu Optical Instrument Co., Ltd (Shanghai, China).

Cell culture and grouping
To resuscitate these frozen cells, frozen SH-SY5Y cells were taken out and melted in 37 C water bath. The cell suspension was transferred to a sterile centrifuge tube, and then washed with the serum-free DMEM medium. Resuscitated SH-SY5Y cells were recultured with DMEM containing 10% foetal bovine serum at 37 C and 5% CO 2 saturated humidity. The cells were subcultured at 80% confluence. SH-SY5Y cells were collected and suspended after trypsin digestion. The cell suspension concentration was adjusted to 1 Â 10 5 /mL and inoculated into 6-well plate. After the cells grew to about 45% confluence degree and cells transformation were performed showing as control mimics, miR-34a mimics, control siRNA, HNF4a siRNA, miR-34a mimics and pcDNA 3.1 empty vector, miR-34a mimics and HNF4a overexpression vectors were transfected into SH-SY5Y cells according to Lipofectamine 2000 transfection reagent instructions, generating the miR-NC group, miR-34a mimics group, si-NC group, si-HNF4a group, miR-34a mimics þ pcDNA3.1 group and miR-34a mimics þ pcDNA3.1-HNF4a group, respectively.

RT-qPCR
Total RNA was extracted from primary tissues and cell lines by the Trizol reagent according to the producer's instructions. TaKaRa reverse transcription kit and fluorescence quantitative kit were used for reverse transcription and system preparation. GAPDH and U6 was used as internal reference for PCR amplification. Each RNA sample was repeated three times. The 2 À᭝᭝Ct method was used to calculate the relative expression of genes. Primers and sequences are shown in Table 1.

Western blot
Western blot assays were performed to detect the protein expression of HNF4a, MMP-2 and MMP-14 in SH-SY5Y cells as well as neuroblastoma and adrenal tissues with various treatments. In brief, RIPA cell lysis solution containing protease inhibitor was used to lyse cells for 30 min. After 4 C, 12,000 g centrifugation for 15 min, the supernatant was obtained. The protein sample was transferred to PVDF membrane after SDS-PAGE electrophoresis, and 5% skim milk powder was sealed at room temperature for 1 h. Rabbit anti-human HNF4a polyclonal antibody (1:1000), rabbit anti-human MMP-2 polyclonal antibody (1:1000), rabbit anti-human MMP-14 polyclonal antibody (1:1000) and rabbit anti-human GAPDH polyclonal antibody (1:1000) were added respectively and incubated at 4 C overnight. After washed with TBST, horseradish peroxidase (HRP) labelled secondary antibody was added and incubated at room temperature for 2 h. TBST was used to wash the membrane three times, 10 min each time. Bands were detected by ECL Kit. Repeat the experiment three times.

Cell migration and invasion detection
The migration and invasion abilities of NB cells were detected by Transwell assay. Generally, SH-SY5Y cells with miR-NC, miR-34a mimics, si-NC, si-HNF4a, miR-34a mimics þ pcDNA 3.1 and miR-34a mimics þ pcDNA 3.1-HNF4a were digested by trypsin and resuspended with the serum free DMEM medium overnight. For migration assays, 3 Â 10 4 cells were inoculated into the upper chamber of transwell uncoated matrix glue. For invasion assay, 3 Â 10 4 cells were seeded into the upper chamber coated with matrix glue. Cells in the upper chamber were cultured in 600 ll DMEM medium and the medium containing 10% foetal bovine serum as chemoattractant was added into the lower compartment with six duplicated pores in each group. After 24 h of conventional culture, the cells in the upper chamber were gently wiped off with cotton swab. Cells attached to the lower surface were fixed with 4% paraformaldehyde for 30 min, stained with 0.1% crystal violet for 10 min, washed with double steaming water until the water was clear, followed by natural air-drying and then observed under microscope and photographed in five random fields. The number of penetrating cells was counted.

Statistical analysis
The experiment was repeated for 3 times in each group. The obtained experimental data were statistically analyzed by SPSS17.0 (SPSS, Chicago, IL), and the results were expressed by mean ± standard deviation ( x ± s). The T test was used for data comparison between two groups, and one-way ANOVA was used for data comparison among multiple groups. p < .05 was considered statistically significant.

Expression of miR-34a and HNF4a in neuroblastoma tissue
Expression of miR-34a and HNF4a were detected by RT-qPCR and Western blot. Results as shown in Figure 1: compared with normal adjacent tissues, miR-34a expression was significantly down-regulated in human neuroblastoma tissues (p < .05), while the expression of HNF4a was significantly upregulated in mRNA and protein levels (p < .05).

Overexpression of miR-34a inhibited the proliferation, migration and invasion of SH-SY5Y cells
The proliferation, migration and invasion were detected after transfection of miR-34a mimics to SH-SY5Y cells. The results ( Figure 2) show: the proliferation capacity of the miR-34a mimics group was significantly lower than that of the miR-NC group at 48 h and 72 h (p < .05). The number of migration and invasion cells in the field of the miR-34a mimics group was significantly lower than that of the miR-NC group (p < .05).

Knocking down HNF4a inhibited proliferation, migration and invasion of SH-SY5Y cells
After transfection of HNF4a to SH-SY5Y cells, the expression of the HNF4a was down-regulated (p < .05) (Figure 3(A,B)). The proliferation capacity of the si-HNF4a group cell was significantly lower than that of the si-NX group at 48 h and 72 h (p < .05) (Figure 3(C)). The number of migratory and invasive cells decreased in the visual field of si-HNF4a group (p < .05) (Figure 3(D)).
Effects of overexpression of miR-34a or knockdown of HNF4a on the expression of MMP-2 and MMP-14 in SH-SY5Y cells MMP-2 and MMP-14 expression were detected after transfection of miR-34a mimics or HNF4a siRNA into SH-SY5Y cells.
The result was shown in Figure 4: mRNA and protein levels of MMP-2 and MMP-14 decreased significantly (p < .05), the difference was statistically significant.

HNF4 is a potential target gene of miR-34a
By online analysis of TargetScan, it was found that HNF4a 3'-UTR contained miR-34a binding sites. Further detection of luciferase activity was found: The miR-34a mimics and HNF4a 3'-UTR wild plasmids were co-transfected into SH-SY5Y cells, the activity of cell luciferase was significantly reduced (p < .05), while the activity of cell luciferase was not significantly changed when co-transfected with HNF4a 3'-UTR mutant plasmids ( Figure 5).
Overexpression of HNF4 can reverse the inhibitory effect of miR-34a on proliferation, migration and invasion of SH-SY5Y cells As shown in Figure 6: after the transfection of miR-34a mimics to SH-SY5Y cells, the cell proliferation capacity was significantly lower than that of the miR-NC group (p < .05), and the number of migration and invasion cells in the field was significantly reduced (p < .05). However, the proliferation, migration and invasion of SH-SY5Y cells were basically restored to normal level when the miR-34a mimics and pcDNA 3.1-HNF4a overexpression vectors were co-transfected into SH-SY5Y cells.

Discussion
According to reports [9,10], the treatment based on NB risk grouping can improve the overall survival rate, and the survival rate of low-risk and medium-risk NB patients can be increased to more than 90% by surgical treatment, chemotherapy, immunotherapy, etc. However, high-risk NB patients are prone to complications after surgical treatment, and toxic side effects of chemotherapy and radiotherapy are likely to cause endocrine disorders and growth retardation, and the overall survival rate is only 40-50%. The abnormal expression of miRNA in various human tumour tissues can regulate the proliferation, differentiation, apoptosis, angiogenesis and other biological processes of cancer cells, thus play an inhibitory or promoting role in tumour development [11,12]. For example, miR-499 plays a role of tumour inhibition in liver cancer by inhibiting cell migration and inducing apoptosis [13]. MiR-195 inhibits cervical cancer cell proliferation by blocking cell cycle [14]. MiR-34 was first found in nematodes in 2001. It is widely distributed in mammals, nematodes and arthropods. It is highly conserved in evolution. The miR-34 gene family is formed by local and tandem repeats in animal development and evolution [15]. Members of the miR-34 family, including miR-34a, miR-34b and miR-34c, play an important role in inhibiting the development of tumours such as liver cancer, breast cancer and lung cancer [16,17]. In addition, there is evidence that the miR-34 family is involved in neuronal differentiation. Overexpression of miR-34a can inhibit p53induced apoptosis of neural stem cells and arrest cell cycle at G1 phase to promote differentiation of neural stem cells into neurons [18]. Multiple studies showed that miR-34a can inhibit the proliferation of various tumour cells, such as osteosarcoma of prostate cancer, by blocking cell cycle and playing a role in tumour inhibition [19,20]. This study found that miR-34a was low expressed in paediatric neuroblastoma tissues, and overexpression of miR-34a could inhibit proliferation and invasion of SH-SY5Y cells. Moreover, it was found that HNF4a was the potential target of miR-34a through online prediction by TargetScan and biluciferase reporter gene experiments.
Hepatocyte nuclear factor 4a (HNF4a) belongs to the steroid hormone receptor superfamily, which can regulate cell differentiation and metabolism at the transcriptional level. Human HNF4a is located on 20q13.12 chromosome. It is a highly conserved zinc finger protein, which can be activated by various pathways such as acetylated phosphorylation. It binds with the target gene promoter sequence to form a dimer and regulates chromosomal structure, hepatocyte polarization and cholesterol metabolism [21]. Some studies found that HNF4a can be used as a marker for the diagnosis of invasive mucinous adenocarcinoma, and has inhibitory  effect on colon cancer. Knocking out HNF4a can promote the metastasis of colon cancer and contribute to the occurrence and development of colon cancer [22,23]. In neuroblastoma cells, HNF4a can promote cell invasion, metastasis and angiogenesis by targeting MMP-14. The lower differentiation of neuroblastoma and higher expression level of HNF4a, the shorter survival time of patients [24]. In this study, HNF4awas knocked down by RNA interference in SH-SY5Y cells, the ability of proliferation, migration and invasion was decreased (p < .05), and the expression of MMP-2 and MMP-14 was down-regulated at mRNA and protein levels (p < .05). Overexpression of HNF4a reversed the inhibitory effect of miR34a on proliferation, migration and invasion of SH-SY5Y cells.
In conclusion, in children with neuroblastoma, the expression of microRNA-34a was low while that of HNF4 alpha was high. Overexpression of miR-34a could inhibit the proliferation, migration and invasion of neuroblastoma cells by targeting HNF4a. This study may provide a new target for the clinical treatment of neuroblastoma in children.