MicroRNA-146a-5p enhances radiosensitivity in hepatocellular carcinoma through replication protein A3-induced activation of the DNA repair pathway

MicroRNA-146a-5p enhances radiosensitivity in hepatocellular carcinoma through replication protein A3-induced activation of the DNA repair pathway. Am J Physiol Cell Physiol 316: C299–C311, 2019. doi:10.1152/ajpcell. 00189.2018.—Hepatocellular carcinoma (HCC) is known for its high mortality rate worldwide. Based on intensive studies, microRNA (miRNA) expression functions in tumor suppression. Therefore, we aimed to evaluate the contribution of miR-146a-5p to radiosensitivity in HCC through the activation of the DNA damage repair pathway by binding to replication protein A3 (RPA3). First, the limma package of R was performed to differentially analyze HCC expression chip, and regulative miRNA of RPA3 was predicted. Expression of miR-146a- 5p, RPA3, and DNA damage repair pathway-related factors in tissues and cells was determined. The effects of radiotherapy on the expres- sion of miR-146a-5p and RPA3 as well as on cell radiosensitivity, proliferation, cell cycle, and apoptosis were also assessed. The results showed that there exists a close correlation between miR-146a and the radiotherapy effect on HCC progression through regulation of RPA3 and the DNA repair pathway. The positive rate of ATM, pCHK2, and Rad51 in HCC tissues was higher when compared with that of the paracancerous tissues. SMMC-7721 and HepG2 cell proliferation were signiﬁcantly inhibited following 8 Gy 6Mv dose. MiR-146a-5p restrained the expression of RPA3 and promoted the expression of relative genes associated with the DNA repair pathway. In addition, miR-146a-5p overexpression suppresses cell proliferation and enhances radiosensitivity and cell apoptosis in HCC cells. In conclusion, the present study revealed that miR-146a-5p could lead to the restriction of proliferation and the promotion of radiosensitivity and apo- ptosis in HCC cells through activation of DNA repair pathway and inhibition of RPA3. 3-phosphate dehydrogenase (GAPDH) served as the internal reference of RPA3, ATM, and pCHK2. The reliability of the PCR results was evaluated using the solubility curve. Relative expression of the target gene was calculated using the equation 2 (cid:3)(cid:4) C T . Western blot analysis. The radio-immunoprecipitation assay (RIPA) buffer (3 ml/g) was added into transfected SMMC-7721 and HepG2 cells. The cells were then lysed to obtain the protein samples of each group, and protein concentration was determined afterward. Proteins were separated by electrophoresis and subsequently transferred onto a nitrocellulose membrane and then blocked by 5% skim milk powder overnight at 4°C. Following the addition of diluted primary antibodies rabbit-anti-human monoclonal antibody RPA3 (AB_10860648, 1: 2,000, ab109394; Abcam), ACOT12 (AB_ 10860648, 1: 2,000, ab83326; Abcam), ATM (AB_1640207, 1: 5,000, ab81292; Abcam), pCHK2 (AB_10863751, 1: 1,000, ab109413; Ab-cam), Rad51 (AB_2722613, 1: 1,000, ab133534; Abcam), and rabbit polyclonal antibody GAPDH (AB_307275, 1: 2,500, ab9485; Ab-cam), the mixture was incubated overnight. The secondary antibody of horseradish peroxidase-labeled goat anti-rabbit IgG (1: 1,000, Boster Biological Technology, Wuhan, Hubei, China) was added and incubated at 37°C for 1 h. Afterward, the membrane was immersed into an enhanced chemiluminescence reaction solution (Pierce, Waltham, MA). Following the development and ﬁxing, the results were observed. With GAPDH used as the internal control, the gray value ratio of target bands to internal control band was considered as the relative protein expression level. This method was also applicable to tissue experiments. data with normal distribution and inhomogeneous variance were determined with Welch’s t -test. The repeated-measures analysis of variance was employed for cell proliferation. Comparisons among multiple groups were performed with one-way analysis of variance with the Tukey’s post hoc tests for multiple pairwise comparisons. If the data did not conform to the normality of data or the homogeneity of variance, the rank-sum test was used. When P (cid:3) 0.05, the difference was considered statistically signiﬁcant. proliferation. With radiation treatment, miR-146a-5p was upregulated, RPA3 expression was inhibited, radio- sensitivity and apoptosis were elevated, and proliferation was suppressed via activation of the DNA repair pathway.


INTRODUCTION
Hepatocellular carcinoma (HCC) falls under the category of heterogeneous diseases, and its development is often caused by various etiologies, including hepatitis B virus (HBV), metabolic syndrome, and chronic alcohol abuse (25). At present, HCC is the fifth most frequently occurring type of cancer, which accounts for ϳ5% of all cancers worldwide, and its incidence is on the rise, with more than 500,000 new cases reported each year (10). The treatment for HCC patients depends on the different stages of the disease, which includes ablation, transplantation or resection at an early stage, chemoembolization at an intermediate stage, kinase inhibitor sorafenib at advanced stage, and transplantation at end stage (3). In the advanced stages of HCC, radiation therapy is considered to be an effective therapeutic method for tumor control, which includes internal radiation therapy with radioisotopes (15). Although radiotherapy is considered to be a major therapeutic option for HCC patients, its efficacy is under the limitation of intrinsic radio resistance of the tumor, which is influenced mainly by the activity of the DNA damage repair pathway (5). MicroRNAs (miRNAs) can efficiently attain tumor radiosensitivity control during the different stages of the disease through its effects on the radio-related signal transduction pathways, DNA damage repair, tumor microenvironment, apoptosis, and cell cycle checkpoint (36). miRNAs are short noncoding RNAs, and miR-146a-5p has been found to be the most downregulated miRNA during liver fibrosis (8). In Wharton's jelly mesenchymal stem cells (WJ-MSCs), a decrease in miR-146a-5p could inhibit the proliferation and enhance the migration of these stem cells (12). Meanwhile, the carcinogenesis and deterioration of HCC are influenced by the downregulation of miR-146a-5p expression, which indicates that the upregulation of miR-146a-5p could lead to HCC suppression (35). Replication protein A (RPA) is a single-stranded (ssDNA)-binding protein that can affect DNA recombination, DNA repair, and replication, and cell cycle checkpoint and RPA3 is its subunit (29). An increase in the expression of RPA3 plays an important role in nasopharyngeal carcinoma (NPC) radio resistance and is expected to play a role as a biomarker in predicting radiosensitivity and prognosis in NPC patients (24). In addition, the upregulation of RPA3 expression could promote tumor progression in HCC cells, which leads to high patient mortality (29). DNA repair pathways can be improperly activated by cancer cells to overcome standard anticancer treatments with preserved genome integrity (21). If repair is impossible, normal cells will promote DNA damage repair and cause cell cycle arrest and apoptosis through DNA repair pathways (2). In this study, RPA3 was identified as the target gene of miR-146a-5p based on the microRNA.org biology prediction website. In the present  ATM, ataxia-telangiectasia mutated; pCHK2, phosphorylation of checkpoint kinase 2; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; miR-146a-5p, microRNA-146a-5p; RT-qPCR, reverse transcription quantitative polymerase chain reaction; RPA3, replication protein A3.
Radiotherapy regimen and efficacy determination. Patients were asked to lie in a supine position on the treatment table. Enhanced CT scanning was performed on the patients with an interval of 3 mm. The two ends of the lesion were scanned by continuous CT with an interval of 5 mm. The imaging data were input into the UNICORN 3DTPS treatment planning system through the network, and the organs at risk were delineated by a physician. The radiation dose used on patient was not higher than the respective tolerance dose. The dose-volume histogram was used to evaluate and optimize the radiotherapy scheme. The patients were treated with fractionated irradia-tion, and different sizes of radiation fields were adopted according to different pathological properties. The whole liver was divided into several longitudinal segments with a width of 2.5 cm and was treated by 6 Mv-X-ray radiations. Radiation therapy started from the first segment on the right side, and the ventral surface was radiated in the same way as the dorsal surface. The radiation used on patients ranged from 250 to 300 cGy every day. The radiation field was narrowed, and 10 -20 Gy was added depending on the decrease in tumor size. The 6 Mv-X-ray produced by the linear accelerator (Varian 600; Varian Medical Systems, Palo Alto, CA) was used to radiate patients with a dose rate of 200 cGy/minute. The tumor tissues were extracted following radiotherapy.
Immunohistochemistry. Small cubes of HCC tissues and paracancerous tissues were fixed by 10% formaldehyde solution for Ն24 h. Paraffin-embedded tissues were cut into 4-m serial sections and then incubated at 60°C for 1 h. Tissue slices were then dewaxed with conventional xylene, dehydrated with gradient ethanol, and placed in 3% H 2O2 for 10 min. Following high antigen pressure repair for 1-3 min, 10% normal goat serum blocking solution (Beijing Kangwei Century Biotechnology, Beijing, China) was added to the slices. The slices were then incubated overnight at 4°C with appropriate rabbit anti-human monoclonal antibodies ATM (AB_1640207, 1: 50,000, ab81292; Abcam), pCHK2 (AB_10863751, 1: 200,000, ab109413, Abcam), and Rad51 (AB_2722613, 1: 10,000; ab133534; Abcam). Next, the slices were added with biotin-labeled secondary antibody and incubated at 37°C for 30 min, followed by treatment with streptavidin-peroxidase solution (Beijing Zhongshan Biotechnology, Beijing, China). Following staining with diaminobenzidine for 5-10 min, the slices were washed with distilled water for 10 min, immersed in hematoxylin for 4 min, washed with running water, and rinsed with ethanol for 10 s. After bathing in running water, ammonia was added to obtain a blue color. The slices were dehydrated, cleared, and mounted with neutral balsam. The positively stained area and its percentage in the total area were measured in five randomly selected high-power fields. Radiation treatment. HepG2, Hep-3B, and SMMC-7721 cells and the QGY-7703 and Bel-7402 cell lines were cultured separately. Afterward, the cells in each medium were separately irradiated by different radioactive rays (0, 2, 4, 6, and 8) Gy from 6Mv-X-ray produced by a linear accelerator (Varian 600; Varian Medical Systems) simultaneously. The following day after irradiation, HCC cells in each medium were detected by 3-(4,5-dimethylthiazol-2-yl)-2,5diphenyltetrazolium bromide (MTT) assay to determine cell proliferation status. HCC cells with high radiosensitivity were selected for further experiments to determine the best radiation treatment plan, and then the transfected cells in different groups were treated with the most appropriate and effective radiation plan.
Cell transfection and grouping. HCC cells (SMMC-7721 and HepG2) in logarithmic growth phase were transiently transfected using Lifopectamine 2000. The day before transfection, the cells were detached and inoculated into a six-well plate. When the cell confluence reached 90 -95%, the transfection reagent, 1 l of Lipofectamine 2000, was diluted with 50 l of serum-free culture medium. Following total mixture, the diluted transfection reagent was mixed with diluted DNA within 5 min and placed at room temperature for 20 min and cultured at 37°C and 5% CO 2. Once the culture medium was replaced after 4 h, the cells were further cultured for 24 -72 h for subsequent experiments. The cells were then divided into the following groups: blank group (without any treatment), negative control (NC) group, miR-146a-5p mimic group (transfected with miR-146a-5p mimic), miR-146a-5p inhibitor group (transfected with miR-146a-5p inhibitor), X-ray group (pure X-ray irradiation, irradiated using linear accelerator 6 Mv, radiation field 15 ϫ 15 cm, source-skin distance 100 cm, 8 Gy dose, culture for 24 h), miR-146a-5p mimic ϩ X-ray group (transfected with miR-146a-5p mimic, irradiated with 6 Mv and 8 Gy dose, culture for 24 h), and miR-146a-5p inhibitor ϩ X-ray group (transfected with miR-146a-5p inhibitor, irradiated with 6 Mv and 8 Gy dose, culture for 24 h).
Reverse transcription-quantitative polymerase chain reaction. Total RNA was extracted using TRIzol (Invitrogen). Reverse transcription was performed according to the kit's instructions (Shanghai Jikai Gene Chemical Technology) to obtain cDNA. Primers and probes of miR-146a-5p, RPA3, ATM, pCHK2, and Rad51 were designed and synthesized by Shanghai Yihe Applied Biotechnology (Shanghai, China) ( Table 1). Quantitative PCR ABI7500 (ABI, Oyster Bay, NY) was applied in RT-quantitative (q)PCR experiment. U6 was considered as the internal control of miR-146a-5p, whereas glyceraldehyde- . Western blot analysis. The radio-immunoprecipitation assay (RIPA) buffer (3 ml/g) was added into transfected SMMC-7721 and HepG2 cells. The cells were then lysed to obtain the protein samples of each group, and protein concentration was determined afterward. Proteins were separated by electrophoresis and subsequently transferred onto a nitrocellulose membrane and then blocked by 5% skim milk powder overnight at 4°C. Following the addition of diluted primary antibodies rabbit-anti-human monoclonal antibody RPA3 (AB_10860648, 1: 2,000, ab109394; Abcam), ACOT12 (AB_ 10860648, 1: 2,000, ab83326; Abcam), ATM (AB_1640207, 1: 5,000, ab81292; Abcam), pCHK2 (AB_10863751, 1: 1,000, ab109413; Abcam), Rad51 (AB_2722613, 1: 1,000, ab133534; Abcam), and rabbit polyclonal antibody GAPDH (AB_307275, 1: 2,500, ab9485; Abcam), the mixture was incubated overnight. The secondary antibody of horseradish peroxidase-labeled goat anti-rabbit IgG (1: 1,000, Boster Biological Technology, Wuhan, Hubei, China) was added and incubated at 37°C for 1 h. Afterward, the membrane was immersed into an enhanced chemiluminescence reaction solution (Pierce, Waltham, MA). Following the development and fixing, the results were observed. With GAPDH used as the internal control, the gray value ratio of target bands to internal control band was considered as the relative protein expression level. This method was also applicable to tissue experiments.
Clonogenic assay. Trypsin was used to detach transfected cells into a single-cell suspension that was diluted 10 times (final cell concentration was 10 3 ). Each type of cell was inoculated into three wells with 100/well in the 0 Gy group, 200/well in the 2 Gy group, 400/well in the 4 Gy group, 800/well in the 6 Gy group, 1,000/well in the 8 Gy group, and 2,000/well in 10 Gy group, respectively. Following the addition of cell suspension and 2 ml of culture medium in each well of a six-well plate, cells were further cultured at 37°C with 5% CO2 for 14 days. After clone formation in the culture plate, the culture was terminated, followed by the addition of crystal violet staining for 30 min. After crystal violet was washed out, the plate was dried with air. The plate was then reversed, and clones with more than 50 cells were visually counted. The survival fraction was worked out using the following formula: survival fraction (counted at last) ϭ cloning efficiency of irradiation cell/cloning efficiency of the control cell.
MTT assay. When the growth density of transfected cells reached ϳ80%, transfected SMMC-7721 and HepG2 cells in each group were seeded in a 96-well plate at a density of 5 ϫ 10 7 cells/l, followed by the addition of 20 l of MTT solution (5 g/l) into each well after being cultured for 24, 48, and 72 h. The supernatant culture medium was discarded after 4 h, and 150 l of dimethyl sulfoxide (DMSO) was added into each well and gently shaken for 10 min. The optical density  (OD) was measured at 490 nm, using a microplate reader for each well to reflect cell growth activity, and the cell growth curve was plotted. Flow cytometry. Propidium iodide (PI) method was used to detect cell cycle. After transfection for 48 h, the cells were collected and detached using 0.25% trypsin. The adjusted sample cell number was 1 ϫ 10 6 cells/ml. Following centrifugation of the cells (1 ml) at 402 g for 10 min, the supernatant was discarded. Cells were fixed by pr-cooled ethanol overnight at 4°C. PI staining solution (1 ml, 50 mg/l) containing RNAase was added into 100 l of cell suspension (10 6 cells/ml). The cells were kept with the avoidance of light for 30 min and filtered with nylon net with 300 meshes. The samples were analyzed using a flow cytometer to detect cell cycle by red fluorescence at the excitation wavelength of 488 nm.
Cell apoptosis was detected by Annexin V-FITC/PI staining. Cell treatment was the same with cell cycle analysis. Cells were cultured at 37°C and 5% CO 2 for 48 h, collected, and washed with PBS twice. Following centrifugation, the cells were resuspended in 200 l of binding buffer and uniformly mixed with 10 l of Annexin V-FITC and 5 l of PI, and a reaction took place with the avoidance of light at room temperature for 15 min before adding 300 l of binding buffer. Cell apoptosis was detected by flow cytometry (6HT, Wuhan Cellwar Bio-technology, Wuhan, Hubei, China) at the excitation wavelength of 488 nm.
Statistical analysis. All data were analyzed by SPSS 21.0 (IBM, Armonk, NY). Measurement data were expressed as means Ϯ SD. First, the normality test and variance homogeneity test were carried out. Unpaired t-test (independent-samples t-test) was performed to test data with normal distribution and homogeneous variance, whereas data with normal distribution and inhomogeneous variance were determined with Welch's t-test. The repeated-measures analysis of variance was employed for cell proliferation. Comparisons among multiple groups were performed with one-way analysis of variance with the Tukey's post hoc tests for multiple pairwise comparisons. If the data did not conform to the normality of data or the homogeneity of variance, the rank-sum test was used. When P Ͻ 0.05, the difference was considered statistically significant.

RESULTS
MiR-146a improves the effect of radiotherapy on HCC by binding to RPA3 and the DNA repair pathway. HCC expression chip GSE40367 was obtained, and only 26 differentially expressed genes were obtained by differential expression analysis of the sample. The heat map (Fig. 1A) was constructed on 26 differentially expressed genes. The results showed that compared with the control group, 18 genes had markedly elevated levels and eight genes presented with decreased levels. STAT4, ACOT12, and RPA3 were the three genes with the most significant difference between the minimum P value. Levels of these three genes in TCGA of the HCC samples and normal sample were further retrieved (Fig. 1, B-D), and the findings revealed significantly increased levels in RAP3 and decreased levels in ACOT12, whereas there were no significant changes observed in STAT4 levels. Based on further information retrieval on the study of RPA3 and ACOT12 on HCC, there were no related reports that found the effect of ACOT12 on HCC, whereas there were a few reports regarding RPA3 on HCC (26,29). In addition, among the study on RPA3, researchers pointed out that RPA3 showed prominently upregulated levels in HCC and promoted cell proliferation and invasion of HCC. These results were in line with what we have obtained through chip analysis and TCGA expression data analysis. Although the related function of RPA3 on HCC has been reported, there is very little known regarding the effect of RPA3 on radiotherapy of HCC. Further information retrieved on RPA3-related signaling pathway has demonstrated that RPA3 is closely associated with the DNA repair pathway (11,24), and DNA repair has been confirmed to play a role in tumor radiotherapy (9,23). The above analysis and related reports revealed that RPA3 may affect radiotherapy of HCC through the DNA repair pathway. To get more information about the  function mechanism of RPA3 on HCC, databases like DIANA were used to predict regulative miRNA of RPA3. The findings signified 29 regulative miRNAs in the DIANA database, and seven regulative miRNAs were predicted in the mirDIP database, whereas 5 miRNAs of conserved sites were predicted in the TargetScan database. To improve the accuracy, the top 15 miRNAs in the DIANA database and prediction results of mirDIP and TargetScan databases were analyzed through Venn histogram (Fig. 1E). The results showed that only two miR-NAs, miR-146a and miR-146b, were located in the prediction result intersection of three databases. The further information retrieval about the study of these 2 miRNAs on HCC showed that a few reports have revealed that miR-146a is closely related to HCC progression (18,34,37). These findings dem-   Fig. 2). HCC tissues have higher positive expression rate of ATM, pCHK2, and Rad51. After immunohistochemistry was performed to test the positive expression rate of ATM, pCHK2, and Rad51, the results showed that ATM, pCHK2, and Rad51 were expressed in cytoplasm and nuclei, and the positive rate of ATM, pCHK2, and Rad51 in HCC tissues was higher than in paracancerous tissues (P Ͻ 0.05; Fig. 3, A and B).
HCC tissues have higher miR-146a-5p but lower RPA3 mRNA expression. The miR-146a-5p and RPA3 expression in HCC tissues was detected, and the results showed that 69 HCC patients treated with the same dose (8 Gy and 6 Mv) of radioactive rays had significantly increased miR-146a-5p but markedly decreased RPA3 mRNA expression in HCC tissues when compared with preradiation (P Ͻ 0.05; Fig. 4).
MiR-146a-5p is upregulate whereas RPA3 is downregulated in postradiotherapy effective cases. Following radiotherapy, the expression of miR-146a-5p and RPA3 in HCC tissues was further assessed. Among the 69 patients with HCC treated with an 8-Gy, 6-Mv dose of radiation, 52 cases presented with radiotherapy sensitivity, with one case of CR and 52 cases of PR, and 17 cases were found with radiotherapy insensitivity, with 15 cases of SC and two cases of PD. Compared with preradiation, the radiotherapy sensitivity group demonstrated increased expression of miR-146a-5p (P Ͻ 0.05), whereas there was a decline in mRNA expression of RPA3 (P Ͻ 0.05). In the radiotherapy insensitivity group, the expression of miR-146a-5p showed no significant difference (P Ͼ 0.05), whereas there was an only slight enhancement in mRNA expression of RPA3 (P Ͼ 0.05) (Fig. 5, A and B).

Proliferation of SMMC-7721 and HepG2 cells is markedly inhibited following radiation with 8-Gy, 6-Mv doses.
The effect of radiotherapy on proliferation of SMMC-7721, HepG2, Hep-3B, QGY-7703, and Bel-7402 cells was analyzed by MTT assay, and the results indicated that cell proliferation rates of SMMC-7721, HepG2, Hep-3B, QGY-7703, and Bel-7402 were reduced as the dose was increased (P Ͻ 0.05). SMMC-7721 and HepG2 presented with the highest radiosensitivity based on the falling range, which had the highest ranking (P Ͻ 0.05). The decrease in cell proliferation rate with the 8-Gy radiation dose was faster in SMMC-7721 and HepG2 (Fig. 7). These results indicated that proliferation of SMMC-7721 and HepG2 cells was significantly inhibited with the 8-Gy and 6-Mv radiation dose, and as a result they were selected for subsequent experiments.
MiR-146a-5p activates the DNA damage repair pathway by downregulation of RPA3. The results from RT-qPCR and Western blot analysis that were conducted to determine the expression of RPA3 and the DNA damage repair pathwayrelated factors (ATM, pCHK2, and Rad51) showed that SMMC-7721 and HepG2 cells had the same trend (Figs. 8 and 9); the mRNA and protein expression of RPA3, ATM, pCHK2, and Rad51 showed no significant difference between the blank group and the NC group (P Ͼ 0.05). Compared with the blank group and the NC group, the miR-146a-5p mimic group and the miR-146a-5p mimic ϩ X-ray group had significantly elevated expression of miR-146a-5p, and the miR-146a-5p mimic, X-ray, and miR-146a-5p mimic ϩ X-ray groups demonstrated significantly increased mRNA and protein expression of ATM, pCHK2, and Rad51, whereas mRNA and protein expression of RPA3 was markedly inhibited (all P Ͻ 0.05); in the miR-146a-5p inhibitor group, the expression of miR-146a-5p and mRNA and protein expression of ATM, pCHK2, and Rad51 were decreased, and the mRNA and protein expression of RPA3 was increased (P Ͻ 0.05); in the miR-146a-5p inhibitor ϩ X-ray group, the expression of miR-146a-5p was Blank NC miR-146a-5p mimic x-ray miR-146a-5p mimic + x-ray Blank NC miR-146a-5p mimic Blank NC miR-146a-5p mimic miR-146a-5p inhibitor x-ray miR-146a-5p mimic + x-ray miR-146a-5p inhibitor + x-ray Blank NC miR-146a-5p mimic miR-146a-5p inhibitor x-ray miR-146a-5p mimic + x-ray inhibitor, X-ray, miR-146a-5p mimic ϩ X-ray, and miR-146a-5p inhibitor ϩ X-ray. C: flow cytometry cycle graph of HepG2 cells. D: cell cycle percentage of HepG2 cells in response to the treatment of miR-146a-5p mimic, miR-146a-5p inhibitor, X-ray, miR-146a-5p mimic ϩ X-ray, miR-146a-5p inhibitor ϩ X-ray. The experiment subjects were SMMC-7721 and HepG2 cells; the experiment was repeated 3 times. Results are measurement data expressed as means Ϯ SD and analyzed by repeated-measures analysis of variance. *P Ͻ 0.05 vs. blank and negative control (NC) groups; #P Ͻ 0.05 vs. the X-ray group.
C307 miR-146a-5p TARGETING RPA3 IN HCC lower, whereas mRNA and protein expression of other genes showed no significant change (P Ͼ 0.05). Compared with the X-ray group, the miR-146a-5p mimic ϩ X-ray group illustrated significantly increased miR-146a-5p expression and mRNA and protein expression of ATM, pCHK2, and Rad51 and significantly decreased mRNA and protein expression of RPA3, whereas the miR-146a-5p inhibitor ϩ X-ray group showed the opposite trends (all P Ͻ 0. 05).
Overexpression of miR-146a-5p reduces the proliferation of HCC cells. MTT assay was introduced to detect the proliferation of HCC cells. Based on the results, SMMC-7721 and HepG2 cells presented the same trend. There was no significant difference observed among the blank, NC, and miR-146a-5p inhibitor ϩ X-ray groups (P Ͼ 0.05). Compared with the blank group and the NC group, the miR-146a-5p mimic, X-ray, and miR-146a-5p mimic ϩ X-ray groups had decreased proliferation rate of HCC cells (all P Ͻ 0.05). Proliferation rate of HCC cells in the miR-146a-5p inhibitor group was increased (P Ͻ 0.05; Fig. 11). In comparison with the X-ray group, the miR-146a-5p mimic ϩ X-ray group displayed decreased an proliferation rate, whereas the miR-146a-5p inhibitor ϩ X-ray group had an increased proliferation rate (all P Ͻ 0.05).
Overexpressed miR-146a-5p inhibits cell cycle progression and promotes the apoptosis of HCC cells. The cell apoptosis and cell cycle were determined. The results (Figs. 12 and 13) showed that SMMC-7721 and HepG2 cells had the same trend. There was no significant difference observed among the blank, NC, and miR-146a-5p inhibitor ϩ X-ray groups (P Ͼ 0.05). Compared with the blank group and the NC group, the miR-146a-5p mimic, X-ray, and miR-146a-5p mimic ϩ X-ray groups had a prolonged G0/G1 phrase and shortened S phase with higher apoptotic rate (P Ͻ 0.05) in HCC cells, and all index changes in the miR-146a-5p mimic ϩ X-ray group were more obvious (all P Ͻ 0.05), whereas the miR-146a-5p inhibitor group showed shortened G0/G1 phase and prolonged S phase with lower apoptotic rate (P Ͻ 0.05). Compared with the X-ray group, the miR-146a-5p mimic ϩ X-ray group demonstrated prolonged G0/G1 phase and shortened S phase with higher apoptotic rate (P Ͻ 0. 05); the miR-146a-5p inhibitor ϩ X-ray group displayed shortened G0/G1 phase and prolonged S phase with lower apoptotic rate (P Ͻ 0.05).

DISCUSSION
Despite the progress made in the treatment and the prognosis of HCC, it remains to be one of the most commonly occurring and major causes of mortality worldwide (16). The previous study has demonstrated that miR-26b contributes to the enhancement of radiosensitivity by binding to erythropoietin, which produces human hepatocellular A2 in HCC (14). Furthermore, a recent study has illustrated that miRNA regulation can affect tumor radiosensitivity through various perspectives, including cell apoptosis and the DNA damage repair, making miRNA regulation an important factor that could be beneficial in tumor diagnosis and treatment, such as radiotherapy (36). In the present study, we aim to examine the regulatory mecha-nism by which the biological function of miR-146a-5p plays a role in the enhancement of radiosensitivity in HCC, which in turn helps improve HCC treatment.
Initially, based on the target prediction program and the dual-luciferase activity determination, we found that RPA3 was a target gene of miR-146a-5p. The results from our study found that there was a decline in the expression of RPA3, whereas the expression of miR-146a-5p was elevated in HCC tissues and cell proliferation was restrained following radiotherapy. A previous study has demonstrated the effects of  Map of molecular mechanisms involved in microNRA (miR)-146a-5p binding to replication protein A3 regulation (RPA3) in hepatocellular carcinoma radiosensitivity via the DNA damage repair pathway. miR-146a-5p was downregulated with weakened inhibitory effect, resulting in upregulated RPA3 expression and impaired DNA repair pathway as well as decreased apoptosis and increased hepatocellular carcinoma cell proliferation. With radiation treatment, miR-146a-5p was upregulated, RPA3 expression was inhibited, radiosensitivity and apoptosis were elevated, and proliferation was suppressed via activation of the DNA repair pathway.
miR-146a-5p downregulation in HCC and its potential to suppress tumors (35). Another study has demonstrated that the expression of RPA3 was upregulated in HCC cells, which accelerated the progression of HCC and resulted in increased patient mortality, whereas cell invasion, colony formation, proliferation, and soft agar growth were inhibited by the downregulation of RPA3 in HepG2 cells (29). It has previously been demonstrated that radiotherapy, which is an important therapeutic pathway for HCC, was able to achieve a local control rate of Ͼ90% (6). By means of irradiation, the apoptosis of SK-MES-1 cells was improved, and its growth was restrained by overexpressed miRNA-126 (27). A study has also demonstrated that migration and invasion capacity in pancreatic cancer and breast cancer was restrained by miR-146a levels, which have the potential to be applied in therapeutic treatment, and upregulation of miR-146a has been found in multiple types of cancer, such as thyroid cancer and cervical cancer (33). Another important finding from our study was that miR-146a-5p promotes an increase in the expression of ATM, pCHK2, and Rad51 associated with the DNA repair pathway by downregulating RPA3. The previous study has found that the DNA repair pathway fell into the category of DNA damage response mechanisms, and various DNA repair pathways were activated by subnuclear DNA damage induction strategies (7). The DNA repair pathway exerted great influence on cell survival and played a role in current cancer therapeutic strategies, such as radiotherapy and cytotoxic chemotherapy (20). It has been demonstrated that the ATM kinase was activated by DNA damage, and then cellular signaling pathways of great importance were initiated (17). In addition, the previous study has shown that the Wnt/␤-catenin pathway was under the regulation of miR-146a-5p through binding to the tumor suppressor Numb (13). At the same time, lines of evidence have suggested that hepatocarcinogenesis was triggered by accumulated lesions such as chromosomal aberrations and DNA damages through the DNA damage response, dysregulated DNA damage repair, and signaling-to-cell cycle checkpoints (31).
Another finding from our study revealed that the overexpression of miR-146a-5p improves radiosensitivity in HCC cells and inhibits cell proliferation. There has also been a recent report suggesting that the overexpression of miR-146b-5p resulted in the enhancement of radiosensitivity and cell apoptosis, differentiation of glioma stem cells, and reduction in the expression of stem cell marker, neurosphere formation capacity, and cell viability (32). In addition, another study revealed that the overexpression of miR-33a-5p could lead to decreased cell proliferation in WM451 cells, whereas it promoted radiosensitivity in melanoma as a result of glycolysis inhibition (4). Proliferation and activation of the hepatic stellate cell were also suppressed by overexpressed miR-146a-5p (8). Cell apoptosis was enhanced, whereas cell migration, invasion, and proliferation were restrained and cell viability reduced in HCC through miR-490-5p regulation (30). In addition, the overexpression of miR-142-5p in both SMMC-7721 and HepG2 cells resulted in the significant enhancement of cell apoptosis, whereas cell cycle arrest at the G0/G1 phase and the relative cell viability were remarkably decreased (19).
In conclusion, the findings from our study demonstrated that the overexpression of miR-146a-5p contributes to the inhibition of cell proliferation and enhancement of radiosensitivity and apoptosis in HCC by activating the DNA repair pathway and downregulation of RPA3 gene expression (Fig. 14). Further large-scale studies on miR-146a-5p and its effects on radiosensitivity in HCC would be beneficial in finding new and effective therapeutic pathways for HCC.

DISCLOSURES
No conflicts of interest, financial or otherwise, are declared by the authors.