METTL3‐mediated maturation of miR‐589‐5p promotes the malignant development of liver cancer

Abstract MiR‐589‐5p could promote liver cancer, but the specific mechanisms are largely unknown. This study examined the role and mechanisms of miR‐589‐5p in liver cancer. The expressions of miR‐589‐5p, METTL3 and m6A in liver cancers were determined by RT‐qPCR. The relationship between miR‐589‐5p and METTL3‐mediated m6A methylation was examined by m6A RNA immunoprecipitation. After transfection, the viability, migration, invasion and expressions of METTL3 and miR‐589‐5p in liver cancer cells were detected by CCK‐8, wound‐healing, transwell and RT‐qPCR. After the xenograft tumour was established in mice, the tumour volume was determined and the expressions of METTL3, miR‐589‐5p, MMP‐2, TIMP‐2, E‐cadherin, N‐cadherin and Vimentin in tumour tissue were detected by RT‐qPCR and Western blotting. In vitro study showed that miR‐589‐5p and METTL3 were highly expressed in liver cancer. METTL3 was positively correlated with miR‐589‐5p. METTL3 up‐regulated the expression of miR‐589‐5p and promoted the maturation of miR‐589‐5p. Overexpressed miR‐589‐5p and METTL3 promoted the viability, migration and invasion of liver cancer cells, while the effects of silencing miR‐589‐5p and METTL3 on the cells were the opposite. The effects of METTL3 overexpression and silencing were reversed by miR‐589‐5p inhibitor and mimic, respectively. In vivo study showed that METLL3 silencing inhibited the growth of xenograft tumour and the expressions of METTL3, MMP‐2, N‐cadherin and Vimentin, promoted the expressions of TIMP‐2 and E‐cadherin, while miR‐589‐5p mimic caused the opposite results and further reversed the effects of METLL3 silencing. In summary, this study found that METTL3‐mediated maturation of miR‐589‐5p promoted the malignant development of liver cancer.

Therefore, it is of great significance to develop novel methods for treating liver cancer.
MicroRNAs (miRNAs) are small non-coding RNAs that play crucial roles in several physiological and pathological processes in different diseases and cancers, including in liver cancer. 5,6 For instance, serum miR-885-5p in liver cancer patients is a marker for prediction and prognosis of advanced liver cancer 7 ; miR-155-5p enhances the sensitivity of liver cancer cells to Adriamycin through modulating autophagy and can function as a target to treat liver cancer 8 ; miR-328-3p inhibits the malignant biological activities of liver cancer cells through regulating ERMP1 and may be a novel marker for liver cancer prognosis. 9 MiR-589-5p is a less studied miRNA, but also has a regulatory effect on liver cancer. 10 MiR-589-5p is highly expressed in liver cancer tissue and facilitates the growth of liver cancer cells. 10 A study reported that miR-589-5p could enhance the maintenance of stemness and chemoresistance of liver cancer. 11 However, the regulatory mechanisms involved in the effect of miR-589-5p in liver cancer are still largely unknown.
N6-methyladenosine (m6A) RNA methylation, which is a dynamic and reversible modification, is widely discovered in many coding and non-coding RNAs. 12 m6A methylation mediates over 80% of RNA methylation and is considered to be a common post-transcriptional modification of messenger RNA in eukaryotes, and it is also a common post-transcriptional modification in non-coding RNAs including in miRNAs. 13,14 m6A methylation is involved in various aspects of RNA metabolism, such as the regulation of mRNA stability, splicing, transport, localization and translation, RNA-protein interactions. 15 m6A on pri-miRNAs can stimulate miRNA maturation. 16 m6A participates in various biological processes, including embryogenesis, cancer formation, DNA damage response and spermatogenesis. 15,[17][18][19][20] m6A RNA methylation is conducted by its writers, erasers and readers to add, remove or recognize m6A, respectively. YTHDC1, YTHDC2, YTHDF1, YTHDF2 and YTHDF3 are identified as readers, and readers are proteins, which recognize m6A sites and perform multiple functions in nucleus or cytoplasm. 21,22 The functional interaction between m6A 'writers' and 'erasers' determines the dynamic and reversible regulation of m6A modification. 23 METTL3 is a m6A writer and is identified as the major methyltransferase in the methylation to regulate the development of different types of cancers by modulating the maturation of miR-NAs. 12,[24][25][26] For example, METTL3 can modulate the maturation of miR-126-5p to promote the progression of ovarian cancer, 27 and METTL3 enhances the metastasis of colorectal cancer by promoting the maturation of primary miR-1246 (pri-miR-1246). 28 However, whether the effect of miR-589-5p on liver cancer was associated with the METTL3-mediated m6A methylation has not been researched.
In the present research, we performed in vitro and in vivo experiments to examine whether the effect and mechanism of miR-589-5p on liver cancer was associated with the METTL3-mediated m6A methylation.

| Ethics statement
The collection and use of clinical tissues was approved by the Ethics Committee of Zhejiang Provincial People's Hospital (Z2020042309G), and informed consent was provided by all the patients whose tissues were used in this research. Animal trial in this study was approved by the Committee of Experimental Animals of Zhejiang Provincial People's Hospital (Z2020101006G).

| Bioinformatics analysis
The expression of miR-589-5p in LIHC was predicted on StarBase

| Clinical samples
Liver cancer tissues and adjacent tissues were collected from 70 liver cancer patients who received operation treatment in Zhejiang Provincial People's Hospital. The clinicopathological characteristics (age, sex, hepatitis B virus (HBV), serum alpha-fetoprotein (AFP) level, tumour size, number of tumour nodules, cirrhosis, venous infiltration, Edmondson-Steiner grading and tumour-node-metastasis (TNM), tumour stage of all the patients with highly expressed or lowexpressed miR-589-5p and METTL3) were recorded (Tables 1 and 2).

| RNA m6A quantification
The experiment was performed using the m6A RNA Methylation Assay Kit (ab185912) according to a previous publication. 28 In brief, three pairs of clinical samples were randomly chosen, and the total RNA in the tissue was isolated by TRIzol (15596018, Invitrogen).
After the concentration of the RNA was determined by the Ultra-micro UV-Visible Spectrophotometer (NanoDrop, Thermo Scientific), 300 ng RNA was added into the detection well, which was pre-added with 80 μl binding buffer. After incubation at 37°C for 90 min, the detection well was washed with 150 μl 1 × wash buffer three times. Then, 50 μl capture antibody was added into the well and incubated for 60 min at room temperature, followed by washing with 150 μl 1 × wash buffer for three times. Then, 50 μl detection antibody was added into the well and incubated for 30 min, followed by washing with 150 μl 1 × wash buffer for four times.
Then, 50 μl enhancer solution was added into the well and incubated for 30 min, followed by washing with 150 μl 1 × wash buffer for five times. After that, the well was added with 150 μl developer solution and incubated for 5 min. After the well was added with 100 μl stop solution, the absorbance of the well was detected under an Imark microplate reader (Bio-Rad) under a wavelength of 450 nm.

| Mutagenesis assay
This experiment was conducted following a previous publication. 28 In brief, after the m6A point mutations in pri-miR-589 were achieved using the QuikChange Lightning Multi Site-Directed Mutagenesis Kit (210513; Agilent), the mutant m6A in pri-miR-589 (MUT-pri-miR-589) and the wide type pri-miR-589 (WT-pri-miR-589) plasmids were constructed. The plasmids were then used to transfect the HEK293T cells.

| m6A RNA immunoprecipitation (MeRIP) assay
This experiment was conducted using the Magna MeRIP m6A Kit (17-10499, Merck) following a previous publication. 28 In brief, transfected HEK293T cells were collected and lysed by the RIP lysis buffer. 3 μg of m6A antibody (ab208577, Abcam) was conjugated to protein A/G magnetic beads at 4°C for 18 h. Then, the antibodyconjugated beads were incubated with the antibody in IP buffer with RNase inhibitor and protease inhibitor. The interacting RNAs were isolated and detected by qRT-PCR.

| Wound-healing assay
After the transfection, 1 × 10 5 Hep3B and SK-Hep1 cells were added into a 6-well plate to the confluence of about 100%. Then, a straight wound was created in each well with a 20μl pipette tip. After that, the previous medium was replaced with an FBS-free medium and   normally fed for 30 days, the mice were anaesthetized with 50 mg/ kg sodium pentobarbital (B005; Jiancheng) and further sacrificed by cervical dislocation. Subsequently, the tumour was taken out, photographed, and the volume was measured according to the formula that major diameter (L) × (minor diameter) 2 (W 2 )/2 mm. 3 Finally, the tumour was used in Western blotting and RT-qPCR assays.

| Western blotting assay
Total proteins from the mice tumour tissues were isolated by NP-

| RT-qPCR
The  (Table S2), the cDNA was amplified in the QuantStudio6 system (Applied Biosystems) and the gene expressions were quantified using the 2 −ΔΔCt method.

| Statistical analysis
GraphPad 8.0 software was used to analyse the data in this study.
The data were expressed as mean ± SD. The data of the two groups were compared using independent sample t test (paired sample t test between cancerous tissues and adjacent tissues), and one-way ANOVA was used for the analysis of variance between multiple groups. p < 0.05 was considered as statistically significant.

| MiR-589-5p was highly expressed in liver cancer tissues and cells
MiR-589-5p was predicted to be highly expressed in LIHC through analysis in starBase (p = 8.4e-23, Figure 1A), and this was further confirmed using clinical liver cancer tissues (p < 0.001, Figure 1B) and cultured liver cancer cell lines (p < 0.01, Figure 1C) by RT-qPCR analysis. Furthermore, the clinicopathological characteristics of all patients with high or low expression of miR-589-5p were analysed, as exhibited in Table 1; a high level of miR-589-5p was associated with the tumour size (p = 0.029), Edmondson-Steiner grading (p = 0.016) and TNM tumour stage (p = 0.004), indicating that miR-589-5p might have a regulatory effect on liver cancer. Given that the level of miR-589-5p in Hep3B and SK-Hep1 was relatively higher than other cancer cells among the 6 liver cancer lines ( Figure 1C), the two cells were chosen for later use.

| MiR-589-5p mimic promoted the viability, migration and invasion of liver cancer cells, but the effects of miR-589-5p inhibitor were the opposite
To investigate the effect of miR-589-5p on liver cancer cells, the mimic and inhibitor of miR-589-5p were transfected into the Hep3B and SK-Hep1 cells, as shown in Figure 1D, the expression of miR-589-5p was up-regulated by miR-589-5p mimic (p < 0.001), but down-regulated by a miR-589-5p inhibitor (p < 0.05). After cell trans-

fection, cell biological activities including cell viability and abilities
to migrate and invade were evaluated. As shown in Figure 1E, the viability of Hep3B and SK-Hep1 cells was increased by miR-589-5p mimic (p < 0.05) but reduced by miR-589-5p inhibitor (p < 0.05). In addition, miR-589-5p further promoted the migration (Figure 2A

| METTL3 was highly expressed in liver cancer, and it was positively correlated to the expression of miR-589-5p
StarBase was used to analyse the correlation between METTL3 and miRNAs, as shown in Figure 3A, METTL3 and miR-589-5p were positively correlated. In addition, highly expressed METTL3 in clinical liver cancer samples was further verified (p < 0.001, Figure 3B); moreover, the positive correlation of METTL3 and miR-589-5p expressions was also confirmed in liver cancer samples ( Figure 3C).
Meanwhile, as compared with the normal cells, liver cancer cells showed a high level of METTL3 (p < 0.001, Figure 3D), indicating that METTL3 had a regulatory effect on liver cancer. To further verify the effect of METTL3 on miR-589-5p expression, two siRNAs for METTL3 (METTL3 siRNA1 and METTL3 siRNA2) were transfected into Hep3B and SK-Hep1 cells for later use. As shown in Figure 3E, the expression of METTL3 in the two cells were both down-regulated by METTL3 siRNA1 and METTL3 siRNA2 (p < 0.001, Figure 3E). In addition, after silencing METTL3 in the two liver cancer cells, the expression of miR-589-5p was downregulated (p < 0.001, Figure 3F).  Table 2, it could be observed that a high level of METTL3 was associated with the tumour size (p = 0.008), Edmondson-Steiner grading (p = 0.029) and TNM tumour stage (p = 0.003). Therefore, we then focussed on examining the effect and mechanism of METTL3 in liver cancer.

| METTL3 regulated the maturation of miR-589-5p by changing the m6A modification status of pri-miR-589
Previous research reported that METTL3 regulates the maturation of certain miRNA. 28 To determine whether METTL3 regulated the maturation of pri-miR-589 to miR-589-5p, the expressions of the pri-miR-589 and miR-589-5p in Hep3B and SK-Hep1 cells after silencing METTL3 were evaluated. As shown in Figure 4A,B, METTL3 silencing down-regulated the expression of miR-589-5p but up-regulated the expression of pri-miR-589 (p < 0.05). METTL3 is associated with the function of m6A modification, 28-31 we therefore detected the m6A level in liver cancer samples and discovered that the level of m6A in cancer tissues was significantly up-regulated as compared with the normal tissues (p < 0.001, Figure 4C). Furthermore, the m6A modification was reduced after METTL3 silencing but increased after overexpression METTL3 as compared with the Vector group (p < 0.01, Figure 4D). For further research, we obtained the sequences of pri-miR-589, pre-miR-589-5p and miR-589-5p from

University of California Santa Cruz Genomics Institute (UCSC;
http://genome.ucsc.edu/), and there were 299 nucleotides of pri-miR-589. Then, SRAMP (http://www.cuilab.cn/sramp/) was used to predict the potential modification site of m6A upstream of pri-miR-589, and the GGACA m6A motifs were determined to be the m6A deposition site in the upstream of pri-miR-589 ( Figure 4E, left). To verify the functions of these sites in the m6A modification of pri-miR-589, the expression vector containing mutant and wildtype m6A motif (MUT-pri-miR-589 and WT-pri-miR-589) was constructed. In the MUT-pri-miR-589, guanine replaced the adenine in the m6A motif ( Figure 4E, right). Then, the MUT-pri-miR-589 and WT-pri-miR-589 plasmids were transfected into the HEK293T cells ( Figure 4F), and we found that the transcript level of pri-miR-589 remained unchanged in MUT-pri-miR-589 and WT-pri-miR-589 group, while the transcript level of miR-589-5p was down-regulated in WTpri-miR-589 group when compared with theWT-pri-miR-589 group (p < 0.001). In addition, the down-regulated m6A level was also discovered in MUT-pri-miR-589 group when compared with the WTpri-miR-589 group (p < 0.001, Figure 4G). These results confirmed that the sites predicted were responsible for m6A modification of pri-miR-589, and we further verified that METTL3 regulated the maturation of miR-589-5p by changing the m6A modification status of pri-miR-589.

| METTL3 overexpression promoted the viability, migration and invasion of liver cancer cells by up-regulating miR-589-5p, but the effects of METTL3 silencing were the opposite
To further verify that the regulatory effect of METTL3 in liver cancer was realized by regulating miR-589-5p, the expression of miR-589-5p in the Hep3B and SK-Hep1 cells was detected after the transfection. As exhibited in Figure 5A,B, the expression of miR-589-5p was inhibited by METTL3 silencing and miR-589-5p inhibitor (p < 0.05), while the expression of miR-589-5p was promoted by METTL3 overexpression and miR-589-5p mimic (p < 0.001, Figure 5A,B); moreover, the effect of METTL3 silencing and overexpression on miR-589-5p expression was reversed by miR-589-5p mimic and inhibitor, respectively. The cell viability ( Figure 5C,D), migration ( Figure 6A,B,E-F) and invasion (Figure 6C,D,G,H) of the two cells were reduced by METTL3 silencing and miR-589-5p inhibitor (p < 0.05), but those cell progresses were increased by METTL3 overexpression and miR-589-5p mimic (p < 0.05). Also, the effects of METTL3 silencing and overexpression on these processes were reversed by miR-589-5p mimic and inhibitor, respectively. These results indicated that METTL3 regulated the viability, migration and invasion of liver cancer cells by regulating miR-589-5p.

| METTL3 regulated the growth of xenograft tumour by regulating miR-589-5p
To confirm the above finding in vivo, animal experiments were performed. After the xenograft tumour was collected after growing in mice for 30 days (Figure 7A), we discovered that the tumour volume was inhibited by METTL3 silencing but promoted by miR-589-5p mimic as compared with the siNC+MC group (p < 0.001, Figure 7B). Also, the effects of METTL3 silencing and miR-589-5p

| DISCUSS ION
In this study, we discovered that miR-589-5p was highly expressed MiR-589-5p is a less studied miRNA and has been reported to have a regulatory effect on liver cancer. 10 Previous research discovered that miR-589-5p is promoted in liver cancer. 11 Consistently, we also proved that miR-589-5p was highly expressed in liver cancer tissues and cells. MiR-589-5p could promote liver cancer cell viability, migration and invasion. 32 Similarly, in this study, the viability, migration and invasion of liver cancer cells were promoted by miR-589-5p mimic but inhibited by miR-589-5p inhibitor, verifying the role of miR-589-5p in promoting liver cancer development. However, the mechanism of miR-589-5p in affecting liver cancer still needed to be clarified.  F I G U R E 4 METTL3 regulated the maturation of miR-589-5p by changing the m6A modification status of pri-miR-589. (A-B) The expression level of pri-miR-589 and miR-589-5p in Hep3B (A) and SK-Hep1 (B) cells after silencing METTL3 was evaluated by RT-qPCR, and U6 was an internal control. (C) The level of m6A in clinical liver cancer samples was quantified by an m6A RNA Methylation Assay Kit. (D) The m6A modification level in the presence and absence of METTL3 was evaluated by MeRIP assays. (E) GGACA m6A motifs sites were in the upstream of pri-miR-589 (Right panel); motif mutation of m6A in pri-miR-589 (Left panel). (F) The level of pri-miR-589 and miR-589-5p after transfected with MUT-pri-miR-589 and WT-pri-miR-589 plasmids in HEK293T cells was detected by RT-qPCR, and U6 was an internal control. (G) The m6A modification level in the HEK293T cells transfected with MUT-pri-miR-589 and WT-pri-miR-589 plasmids was detected by MeRIP assays. (^p < 0.05, ^^p < 0.01, ^^^p < 0.001, vs. siNC; *** p < 0.001, vs. Normal; † † † p < 0.001, vs. Anti-IgG; &&& p < 0.001, vs. Mock; ## p < 0.001, vs. Anti-m6A in Vector group; § § § p < 0.001, vs. WT-pri-miR-589; △△△ p < 0.001, vs. Anti-m6A in WT-pri-miR-589 group).  processes, such as T cell homeostasis, embryonic thousand cell growth and differentiation and cancer cell growth. [12][13][14]24 Rich m6A modification has been discovered in ovarian cancer, colorectal cancer, lung cancer and liver cancer. 27,28,33,34 Similarly, a high level of m6A was also detected in liver cancer in this study. The most recognized underlying mechanism of m6A modification in cancers is that the methylation process of m6A modification is regulated by methyltransferase. 13 METTL3, as the first reported m6A writer, is a major methyltransferase, and its expression is up-regulated in different cancers such as gastric cancer, nasopharyngeal carcinoma, cervical cancer and liver cancer. [35][36][37][38] In the current research, highly expressed METTL3 was further confirmed in DLBC, CHOL, THYM and liver cancer, and the level of METTL3 was determined to be negatively correlated with the overall survival. However, whether m6A modification in liver cancer was mediated by METTL3 remained unknown. Therefore, the MeRIP assay was performed, and for the first time, we further confirmed the association between m6A modification and METTL3 in liver cancer. m6A methylation is also a common post-transcriptional modification in non-coding RNAs including in miRNAs. Researches proved that METTL3-mediated m6A modification promotes the maturation of miRNAs, including miR-126-5p, miR-1246, miR-221 and miR-222. 27,28,39 After we discovered that METTL3 promoted the maturation of miR-589-5p, we constructed the mutant and wild-type m6A motif (in the upstream of pri-miR-589) plasmids to further verify whether the effect of METTL3 on the maturation of miR-589-5p was associated with the m6A modification of pri-miR-589. After performing MeRIP assay, we confirmed that METTL3 promoted the maturation of miR-589-5p via the m6A modification of pri-miR-589.
To further verify the regulatory effect of METTL3/miR-589-5p on liver cancer, the biological activities changes of cancer cells were determined by performing cell experiments, and the growth changes of liver cancer were determined by animal experiments.
We found that METTL3 overexpression promoted the viability, migration and invasion of liver cancer cells, and that METTL3 silencing further inhibited the growth of xenografted-tumour in nude mice but the effects of miR-589-5p mimic were the opposite.
Moreover, miR-589-5p mimic and inhibitor reversed the effects of METTL3 silencing and overexpression on liver cancer, indicating that METTL3/miR-589-5p enhanced the malignant development of liver cancer.
Our research, for the first time, verified that METTL3 promoted the maturation of miR-589-5p via enhancing m6A modification of pri-miR-589 and further enhanced the malignant development of liver cancer. The current findings provide a new diagnostic and therapeutic marker for liver cancer.

ACK N OWLED G EM ENT
Not applicable.

CO N FLI C T O F I NTE R E S T
The authors declare no conflicts of interest.