MiR-125b-5p Inhibited The Progression of Hepatoblastoma As A Shared miRNA of The lncRNA NEAT1 and YES1

Background: microRNAs have been studied widely in hepatoblastoma. However, the role of miR-125b-5p and its relationship with the lncRNA sNEAT1 and YES1 in hepatoblastoma have not been reported previously. We aimed to reveal the role of NEAT1/miR-125b-5p/YES1 in the progression of hepatoblastoma. Methods: We collected tumor tissues and their adjacent tissues from 12 hepatoblastoma patients. qRT-PCR was applied to detect the expression of miR-125b-5p, and the relationship of miR-125b-5p with clinicopathological characteristics was analyzed. Dual luciferase reporter assays and RNA pull down assays were used to identify the relationships among NEAT1, miR-125b-5p and YES1. CCK8, Transwell assays and wound healing assays were used to examine cell viability, invasion and migration. In vivo experiments were also applied to detect the effect of miR-125b-5p on hepatoblastoma. Results: miR-125b-5p was signicantly downregulated in hepatoblastoma tissue and cells. The higher the PRETEXT grade, the lower the miR-125b-5p level. NEAT1 could bind to miR-125b-5p and inhibit its expression. miR-125b-5p could target YES1 and inhibit its expression. Overexpression of miR-125b-5p decreased the proliferation, invasion, and migratory ability of hepatoblastoma cells. YES1 could rescue the above effects. At the same time, overexpression of miR-125b-5p resulted in decreased YES1 and tumor growth inhibition in vivo. Conclusion: miR-125b-5p acted as a shared miRNA NEAT1 and YES1


Background
Hepatoblastoma (HB) is the most common malignant tumor of the liver in children. It often arises from liver progenitor cells or hepatoblasts during the embryonic development of the liver 1 . Hepatoblastoma accounts for approximately 1% of childhood cancers, and its incidence is approximately 1.5 per million 2 . With the improvement of infant survival rates and advances in treatment, the survival rate of children with the disease has increased signi cantly over the past few decades 3,4 . Due to the characteristics of hepatoblastoma, it is clinically unresectable in nearly half of all cases. With the introduction of new adjuvant chemotherapy, the size of the tumor can be reduced, allowing for more cases of resection 5 .
Despite improvements in diagnosis and treatment, the prognosis of the disease is still poor 6 . At present, the role of many genes, lncRNAs and miRNAs in HB are being studied 7,8 , but the truly pathogenesis has not been revealed. Therefore, it is urgent for researchers to identify new candidate targets for the treatment of HB.
miRNAs are a type of small nonprotein-coding RNA comprised of approximately 20-25 nucleotides. They inhibit translation by binding to the 3'-UTR region of their target genes 9 . Many miRNAs have been studied in hepatoblastoma. Schmid et al. 10 revealed that miR-492 signi cantly enhances the cell proliferation, migration, and invasion of hepatoblastoma cells. They also identi ed and validated that CD44 is a direct and functional target of miR-492. High miR-492 expression is correlated with high risk or aggressive tumors and has potential for predicting reduced event-free survival. A wide detection of miRNAs was applied by Olgun 11 . They selected 4 miRNAs for that study, and they found only the level of miR-17 expression was signi cantly decreased in tumor samples. miR-125b usually acts as a tumor suppressor in many tumors, and it also protects damaged nerves [12][13][14] . Zhuo et al. 15 reported that miR-125b suppresses the proliferation, migration, and invasion of Huh6 cells by targeting DKK3. Although the inhibitory effects of miR-125b on human cancer cell proliferation and metastasis have been extensively investigated 16 , the mechanism remains unknown.
YES proto-oncogene 1 (YES1) is a prominent member of the Src family of tyrosine kinases (SFKs). It acts as a regulator of tumor growth and is involved in the development of many human cancers 17 . In pancreatic cancer, inactivation of YES1 can signi cantly inhibit cell proliferation, invasion, and migration. YES1 can be regarded as an important therapeutic target in pancreatic cancer 18 . YES1 is a direct target of miR-128-3p and TUG1 modulates YES1 expression by sponging miR-128-3p. YES1 silencing repressed prostate cancer cell progression in vitro and mitigated tumor growth in vivo 19 .
There are no previous reports about YES1 in hepatoblastoma. To better understand the mechanism of hepatoblastoma progression, this study mainly explored the effects of miR-125b-5p and its possible target genes on hepatoblastoma. We aimed to explain the interaction between miR-125b-5p and YES1 in HB and provide more possible candidate targets for the treatment of HB.

Samples collected
We included 12 cases of hepatoblastoma who were operated on and con rmed by pathology in 2015 to 2016. All the patients were operated on before receiving radiotherapy, chemotherapy, or immunotherapy.
The sample consisted of 7 boys and 5 girls, and their average age was 1.85 ± 1.46. The adjacent tumor tissues from these patients were collected for controls. These hepatoblastomas were classi ed using the PRETEXT system 20 . The expression level of miR-125b-5p was determined by ROC curves. Higher than the median was de ned as high expression, lower was de ned as low expression. All the patients were followed-up after treatment through December 2019. The survival time was calculated in months.

Cell proliferation assay
When the HepG2 and HuH-6 cells reached a logarithmic growth phase, we adjusted the cell concentration to approximately 1 × 10 4 /well in 96-well plates after digestion. The following day, the cells were transfected and then cultured for 24 h, 48 h, and 72 h. We added 10 µl CCK-8 solution and incubated the cells in the incubator for 2 h. The absorbance at 450 nm of each well was measured to draw the cell growth curve.

Cell invasion and wound healing assays
Cell invasion and wound healing assays were performed referenced to 8 .
Dual luciferase reporter assay Wildtype (WT) or mut sequences of the NEAT1 or YES1 3' UTR were synthesized and cloned into the pGL3-reporter vector. HepG2 or HuH-6 cells were co transfected with wild type or mutant plasmids and mir-125b-5p mimic or mimic NC vectors. After 48 h, we used a luciferase detection kit (#D0010, Solarbio, Beijing, China) and a dual uorescent enzyme reporter gene analysis system (Promega, Madison, WI, USA) to detect luciferase activity.

RNA pull down assay
The RNA pull-down experiment was performed as described previously 21 . The M-280 streptavidin beads were purchased from Thermo Fisher (#11205D, Thermo Fisher Science, Waltham, Massachusetts, USA).
Puri ed RNA was used to detect the expression of NEAT1.
In vivo experiment 24 male BALB/c nude mice (purchased from Guangdong Medical Laboratory Animal Center) aged 5-6 weeks were randomly divided into a mimic NC group and a miR-125b-5p mimic group. Then, 0.2 ml 1 × 10 6 transfected cells were injected subcutaneously into each mouse's right armpit. We observed the mental diet and the body weight of the nude mice and recorded the tumor volume. When the tumor diameter reached 1.5 cm, the nude mice were killed, and the tumor was removed and weighed.

qRT-PCR
Total RNA was extracted from tissues and cells by TRIzol (#15596906, Thermo Fisher Science) and reversed transcribed into cDNA. The cDNA was used to apply qRT-PCR and U6 was set as the internal reference. Gene mRNAs were detected according to the TaqMan Gene Expression Assays protocol (#4331182, Thermo Fisher Science). GAPDH was used as the internal reference. The primer sequences are shown in Table 1. The relative expression levels were calculated using the 2 -∆∆CT method.

Statistical analysis
Data in our study were analyzed by SPSS 21.0 (SPSS, Inc, Chicago, IL, USA). Survival curve was drawn using Kaplan-Meier method and analyzed by Log-rank test. Data was shown in mean ± standard.
Compared between two groups using t-test and multiple groups using ANOVA with Tukey's test for post hoc test. P < 0.05 indicated the difference was statistically signi cant.
Results miR-125b-5p is expressed lower in hepatoblastoma and is related to prognosis By qRT-PCR, we found that the expression of miR-125b-5p in cancer tissues was signi cantly lower than that in adjacent tissues (P < 0.01, Figure 1A). As shown in Figure 1B, we found that miR-125b-5p was expressed at lower levels in PRETEXT -grade (P < 0.01). However, the miR-125b-5p expression level had no relationship with age and gender. We analyzed the overall survival of the patients based on their expression of miR-125b-5p. The patients with high expression of miR-125b-5p had longer survival times than those with low expression of miR-125b-5p (P = 0.013, Figure 1C).

miR-125b-5p inhibited the progression of hepatoblastoma
To investigate the role of miR-125b-5p in hepatoblastoma, we compared the expression levels of miR-125b-5p in HepG2, HuH-6, and HL-7702 cells. Figure 2A shows that miR-125b-5p was expressed lower in HepG2 and HuH-6 cells compared with HL-7702 cells (P<0.05). This result was consistent with the results from tissues. Then, we transfected miR-125b-5p mimic and mimic NC into the hepatoblastoma cells to observe the effect of miR-125b-5p. We found that the expression of miR-125b-5p was signi cantly higher in the miR-125b-5p mimic group (P<0.01, Figure 2B). This indicated the transfection was successful.
Through the cell proliferation assays, we found cells treated with the miR-125b-5p mimic had lower vitality (P<0.05, Figure 2C). Their cell invasion and migration ability were also decreased (all P < 0.05, Figure 2D, E). These results suggested that miR-125b-5p signi cantly inhibited the progression of hepatoblastoma in vitro.
NEAT1 regulated the expression of YES1 by miR-125b-5p Through the starBase website (http://starbase.sysu.edu.cn/), we found miR-125b-5p was a shared microRNA of NEAT1 and YES1. NEAT1 was overexpressed or knocked down in hepatoblastoma cells and then we detected the expression of miR-125b-5p by qRT-PCR. The results shown that compared to the Blank group, miR-125b-5p expression was increased in the si-NEAT1 group and decreased in the oe-NEAT1 group (all P < 0.05, Figure 3A).
Then, we explored the relationship between NEAT1 and miR-125b-5p in HepG2 cells using dual luciferase reporter gene assays. This suggested that the miR-125b-5p mimic signi cantly inhibited the relative luciferase activity of NEAT1-WT but had no effect on NEAT1-Mut (P < 0.05, Figure 3B). The results from RNA pull-down assays also indicated NEAT1 could combine with miR-125b-5p ( Figure 3C). We cotransfected si-NEAT1 and a miR-125b-5p inhibitor to clarify the relationship between NEAT1 and YES1.
qRT-PCR was applied to detect the expression of miR-125b-5p and YES1. When NEAT1 was knocked down, the expression of miR-125b-5p was upregulated but YES1 was downregulated. Silencing miR-125b-5p could reverse the interference of NEAT1 ( Figure 3D). The above results implied NEAT1 inhibited the expression of miR-125b-5p by binding with miR-125b-5p, thus promoting the expression of YES1. miR-125b-5p inhibited the growth of hepatoblastoma by targeting YES1 miR-125b-5p targeted to YES1 was predicted by TargetScan and we proved it through dual luciferase reporter gene assays ( Figure 4A). When miR-125b-5p was overexpressed, the expression level of miR-125b-5p was increased more in the miR-125b-5p mimic group than in the mimic NC group. However, the expression level of YES1 was signi cantly decreased in the miR-125b-5p mimic group relative to the mimic NC group (P < 0.05, Figure 4B). miR-125b-5p mimic or Lv-YES1 were transfected into HepG2 and HuH-6 cells. Through qRT-PCR detection, we found that compared to the mimic NC+ Lv-NC group, miR-125b-5p was upregulated and YES1 was downregulated in the miR-125b-5p mimic+Lv-NC group (P < 0.05). The expression of YES1 was increased in the mimic NC+ Lv-YES1 group and this was reversed when cotransfected with miR-125b-5p mimic ( Figure 4C). Next, we examined their cell proliferation, invasion, and migration ability separately ( Figure 4D-F). According to these results, we found that the miR-125b-5p mimic could inhibit cell proliferation, invasion, and migration ability. However, YES1 enhanced these abilities and reversed the effect of miR-125b-5p.
In vivo, we found the tumor volume and tumor weight were signi cantly decreased in the miR-125b-5p mimic + Lv-NC group compared to the miR-125b-5p NC group. However, YES1 increased the tumor volume and tumor weight, and reversed the effect of miR-125b-5p (P < 0.05, Figure 4G, H). qRT-PCR was used to detect the expression of miR-125b-5p and YES1. This showed that compared with miR-125b-5p NC, the expression of miR-125b-5p was increased and that of YES1 was decreased in the miR-125b-5p mimic group (P < 0.05, Figure 4I). This suggested miR-125b-5p inhibited the expression of YES1 and played a role in inhibiting the tumor progression of hepatoblastoma in vivo.

Discussion
As a common malignant tumor among tumors of children, the prognosis of hepatoblastoma is often related to the type of disease differentiation. Although many genes and pathways have been studied in HB, the speci c molecular mechanism has not been revealed 22 . It has been demonstrated that the miR-100/let-7a-2/miR-125b-1 cluster works with miR-371-3 in HB. Overexpression of the rst cluster delays tumorigenesis and inhibits miR-371 while preventing tumorigenesis 23 .
In this study, HB cancer tissues and cancer cells were used to examine the role of miR-125b-5p. It was found that compared with normal tissues and cells, miR-125b-5p was expressed at low levels in cancer tissues and cancer cells. Its low expression was related to the patient's PRETEXT grade and overall survival time. The lower its expression level, the higher the PRETEXT grade and the shorter the patient's overall survival time.
Previous studies of miR-125b-5p in tumor research have only reported limited results. miR-125b-5p has been studied the most in breast cancer. It can inhibit its target gene and its downstream pathways to inhibit the proliferation, invasion and migration of breast cancer cells, and improve their sensitivity to chemotherapy [24][25][26] . In ovarian cancer, the median serum levels of miR-125b in patients with epithelial ovarian cancer were significantly lower than that of controls. Serum miR-125b levels were found to be a useful diagnostic biomarker and a biomarker to predict the responses to chemotherapy in patients with epithelial ovarian cancer 27 . In cervical cancer, it was found that LncRNA CAR10 upregulated PDPK1 to promote cancer development by sponging miR-125b-5p 28 .
To better study the mechanism of miR-125b-5p in HB, we rst explored it in HB cells. It was found that miR-125b-5p could signi cantly inhibit cell proliferation, invasion, and migration. Next, we used a bioinformatics website to predict that miR-125b-5p is a shared miRNA of the LncRNA NEAT1 and YES1 genes. LncRNAs regulate gene expression mainly at the epigenetic level, transcription level and posttranscription level 28 . Transcription level and posttranscription level regulation are mainly mediated as molecular sponges, which inhibit the expression of target genes through binding and sponging miRNA 29 . Through an RNA pull down experiment, we found that NEAT1 can bind to miR-125b-5p. YES1, as a target gene of miR-125b-5p, has reduced expression because of the interference of NEAT1. We speculate that NEAT1/miR-125b-5p/YES1 is a ceRNA mechanism that regulates HB development. YES1, as a member of the SRC family, acts as an oncogene in most tumors 30 and plays an important role in the acquired drug resistance of breast and lung cancer 31, 32 . This study found that YES1 is signi cantly overexpressed in HB. Inactivation of YES1 can signi cantly inhibit the proliferation, invasion, and migration of HB cells. YES1, as the target gene of miR-125b-5p, can reverse the role of miR-125b-5p in HB. In vivo experiments also con rmed that miR-125b-5p can signi cantly inhibit the expression of YES1 and inhibit the growth of tumors. Yes-associated protein (YAP)1 has been reported to be related to tumor development and it cooperates with other signaling pathways. It has been reported that Wnt/βcatenin works in association with Hippo/YAP to induce the development of hepatoblastoma 33, 34 . We suspect that miR-125b-5p targets and regulates YES1, causing changes in related signaling pathways that are involved in the progression of HB.
Overall, this study con rmed that miR-125b-5p is signi cantly under expressed in HB. The expression level of miR-125b-5p is related to the PRETEXT stage and the patient's overall survival time. miR-125b-5p is a shared miRNA of NEAT1 and YES1. NEAT1 binding to miR-125b-5p targets the regulation of YES1. The NEAT1/miR-125b-5p/YES1 axis is involved in HB disease progression. miR-125b-5p and YES1 may provide new targets for the treatment of HB.   D, the cell invasion ability as determined by Transwell assays (×200). E, Cell migration distance detected by a scratch test. Every experiment repeated three times and expressed as mean ± standard deviation; comparisons between two groups were analyzed using independent t test, and comparisons among multiple groups were analyzed using one-way ANOVA. * indicates P < 0.05. ** indicates P < 0.01.

Figure 3
NEAT1 regulated the expression of YES1 through miR-125b-5p A, expression of miR-125b-5p when cells were treated with overexpressed NEAT1 or knocked down NEAT1. B, relationship between NEAT1 and miR-125b-5p detected by dual luciferase reporter gene assays. C, RNA pull down was applied to identify the relationship between NEAT1 and miR-125b-5p. D, expression of miR-125b-5p and YES1. Every experiment repeated three times and expressed as mean ± standard deviation; comparisons among multiple groups were analyzed using one-way ANOVA. * indicates P < 0.05. ** indicates P < 0.01. Figure 4 miR-125b-5p inhibited tumor growth of hepatoblastoma by targeting YES1 in vivo A, TargetScan predicted and dual luciferase veri ed the relationship of miR-125b-5p and YES1. B, expression of miR-125b-5p and YES1 when miR-125b-5p was overexpressed. C, expression of YES1 in different transfected groups. D, cell proliferation detected by CCK8. E, the cell invasion ability determined by Transwell assays (×200). F, cell migration distance detected by a scratch test. G, tumor volumes in different groups. H, tumor weight in different groups. I, expression of YES1 in vivo. Every experiment repeated three times and expressed as mean ± standard deviation; comparisons between two groups were analyzed using independent t test, and comparisons among multiple groups were analyzed using one-way ANOVA. * indicates P < 0.05 compared with mimic NC+Lv-NC. # indicates P < 0.05 compared with miR-125b-5p mimic+Lv-NC. & indicates P < 0.05 compared with mimic NC+Lv-YES1.