miR-125a-5p inhibits tumorigenesis in hepatocellular carcinoma

Hepatocellular carcinoma (HCC) is one of the most prevalent and deadly cancers world-wide. miR-125a-5p is a tumor suppressor in HCC and other cancers, but its mechanisms of action during HCC tumorigenesis remain largely unknown. In this study, we found that miR-125a-5p expression was significantly lower in HCC tissues and cell lines than matched normal tissues and liver cells. miR-125a-5p overexpression inhibited HCC cell proliferation and induced apoptosis in vitro and in vivo, while miR-125a-5p knockdown had the opposite effects. In addition, PTPN1 and MAP3K11 were identified as targets of miR-125a-5p. Knocking down PTPN1 and MAP3K11 activated the JNK MAPK signaling pathway to suppress HCC cell proliferation and induce apoptosis. Our findings suggest that miR-125a-5p may be a useful therapeutic target for treatment of HCC patients.


INTRODUCTION
Hepatocellular carcinoma (HCC) was the sixth most common type of malignant tumor and the fourth leading cause of cancer-related death worldwide in 2018 [1]. An estimated 80-90% of HCC cases are caused by cirrhosis resulting from chronic hepatitis B (HBV) or hepatitis C (HCV) infections [2]. Chronic HBV/HCV infections, alcoholic liver disease, metabolic syndromes, and some rare autoimmune or genetic conditions may also be risk factors for HCC [3,4]. Even though treatments for HCC, including surgical resection, chemotherapeutic regimens, liver transplantation, local ablation, and targeted therapy, have greatly improved over the past few decades, the survival rates remain low for HCC patients [5]. A better understanding of the molecular mechanisms underlying the development and progression of HCC might help identify potential therapeutic targets and improve clinical outcomes. microRNAs (miRNAs), a type of non-coding RNA, are about 23 nucleotides in length and act as important gene regulators by repressing translation or degrading RNA in animals and plants [6,7]. miRNAs control the expression of their target mRNAs mainly by binding to 3'-untranslated regions (3′-UTRs) [8]. miRNAs can promote or inhibit cell proliferation, apoptosis, metastasis, and angiogenesis, and abnormal miRNA expression is associated with many cancers and other diseases [9]. Recent findings indicate that miRNAs might act as potential diagnostic and prognostic biomarkers and as therapeutic targets in HCC depending on their biological characteristics and functions [10][11][12].
Protein tyrosine phosphatase N1 (PTPN1), also known as protein tyrosine phosphatase 1B (PTP1B), is a member of the protein tyrosine phosphatase family. The role of PTPN1 in cancer is controversial, as it has been shown to act as both a tumor suppressor and a tumor promoter [25]. In B-cell lymphoma and esophageal cancer, it is a tumor suppressor [26,27]. In contrast, PTPN1 is as a tumor promoter in colorectal, non-small cell lung, breast, gastric, and colon cancer [28][29][30][31][32][33]. The role of PTPN1 in HCC remains unknown.
Mitogen activated protein kinase kinasekinase 11 (MAP3K11), also called mixed lineage kinase 3 (MLK3), is a member of the mixed lineage kinase subfamily of serine/threonine kinases. MAP3K11 is an oncogene that promotes tumorigenesis in breast and colorectal cancer [34][35][36][37]. Whether MAP3K11 also plays a role in HCC has not yet been investigated.
In this study, we investigated the role of miR-125a-5p in HCC and whether PTPN1 and MAP3K11 are involved in its effects.

miR-125a-5p is down-regulated in HCC tissues and cell lines
We used a microarray to compare miRNA expression in eight matched pairs of HCC tissues and adjacent normal tissues. Differentially expressed miRNAs, which were defined by a ≥ 2-fold change in expression between the two tissue types and associated P-values ≤ 0.05, are shown in the heat map in Figure 1A. The microarray assay indicated that miR-125a-5p expression was decreased in HCC tissues compared to adjacent normal tissues; this was also the most statistically significant difference observed ( Figure 1A, Supplementary Table  1). qRT-PCR also revealed that miR-125a-5p expression was down-regulated in the eight matched pairs of HCC tissues compared to the adjacent normal tissues ( Figure  1B). Moreover, miR-125a-5p expression was down-regulated in HCC tumor tissues compared to adjacent normal tissues in TCGA from the Starbase database [38] (http://starbase.sysu.edu.cn/) ( Figure 1C). In addition, qRT-PCR indicated that miR-125a-5p expression was lower in HCC cell lines (Bel-7404, SK-Hep1, Bel-7404, and MHCC-97H) than in a normal liver cell line (L02) ( Figure 1D). However, Kaplan-Meier survival curves indicated that overall survival in HCC patients was not associated with miR-125a-5p expression level [39] (http://www.kmplot.com/mirpower) ( Figure 1E). This may be due to the small number of samples examined. Furthermore, there is a particularly large amount of heterogeneity among HCC patients which might also have prevented detection of associations between miR-125a-5p expression and survival.

PTPN1 and MAP3K11 are up-regulated and negatively correlated with miR-125a-5p expression in HCC
Next, we analyzed PTPN1 and MAP3K11 expression in HCC. Both genes were up-regulated in HCC according to the Starbase database [38] (http://starbase. sysu.edu.cn/) ( Figure 4A and 4B), and high PTPN1 and MAP3K11 expression was associated with poorer overall survival in the GEPIA database [46] (http://gepia.cancer-pku.cn/) ( Figure 4C and 4D). qRT-PCR revealed that PTPN1 and MAP3K11 were upregulated in the eight HCC tissue samples compared to the adjacent normal tissues ( Figure 4E and 4F). qRT-PCR and Western blot indicated that PTPN1 and MAP3K11 also were up-regulated in HCC cell lines ( Figure 4G-4I, Supplementary Figure 3A). Moreover, Spearman correlation analysis indicated that miR-125a-5p expression was negatively correlated with PTPN1 and MAP3K11 expression ( Figure 4J and 4K). Finally, knockdown of PTPN1 and MAP3K11 resulted in reduced PTPN1 and MAP3K11 protein expression (Supplementary Figure 3B and 3C). Together, these results suggest that PTPN1 and MAP3K11 are upregulated and negatively correlated with miR-125a-5p expression in HCC.

miR-125a-5p suppresses PTPN1 and MAP3K11 expression via the MAPK signaling pathway in HCC
In mammals, there are four major subgroups of the MAPK family: the extracellular signal-regulated kinases 1 and 2 (ERK1/2), JNK, p38, and ERK5 pathways [47]. As mentioned above, "activation of JUN kinase activity" was one of the biological functions associated with the 131 common elements identified using target prediction databases, implying that the JNK MAPK signaling pathway might be essential for the biological function of miR-125a-5p in HCC. Next, we confirmed that miR-125a-5p overexpression reduced, while miR-125a-5p knockdown increased, the expression of JNK1/2/3 and c-Jun in Bel-7404 and SK-Hep1 cells     Figure 4B and 4C). These results indicate that miR-125a-5p suppressed PTPN1 and MAP3K11 expression via the MAPK signaling pathway in HCC.

miR-125a-5p suppresses cell proliferation and induces apoptosis in HCC by targeting PTPN1 and MAP3K11 via the MAPK signaling pathway
Loss of function experiments were performed using PTPN1 siRNA 3 and MAP3K11 siRNA 1, which had the highest knockdown efficiencies for their targets. The loss-of-function experiments included four groups: a control group, a group transfected with miR-125a-5p inhibitors, a group transfected with PTPN1 siRNA 3 or MAP3K11 siRNA 1, and a group co-transfected with miR-125a-5p inhibitors and PTPN1 siRNA 3 or MAP3K11 siRNA 1. CCK8 and colony formation assays showed that cell proliferation was increased in the miR-125a-5p inhibitors group and suppressed in the PTPN1 or MAP3K11 group. Immunofluorescence assays revealed that cell apoptosis was suppressed in the miR-125a-5p inhibitors group and upregulated in the PTPN1 or MAP3K11 group. Cell proliferation and apoptosis rates were similar in the co-transfected and control groups ( Figure 6A-6F, Supplementary Figure  5A and 5B). In addition, co-transfection restored PTPN1, MAP3K11, JNK1/2/3, and c-Jun protein expression to control levels ( Figure 6G and 6H, Supplementary Figure 5C). In addition, we performed a gain of function assay by inserting PTPN1 and MAP3K11 CDS into pCDNA3.1 (+) vectors. Western blot assays confirmed acceptable rescue efficiency and ensured that the siRNAs had no off-target effects (Supplementary Figure 5D). Together, these results indicate that miR-125a-5p suppresses cell proliferation and induces apoptosis in HCC by targeting PTPN1 and MAP3K11 via the MAPK signaling pathway.

miR-125a-5p suppresses cell proliferation and induces apoptosis in HCC by targeting PTPN1 and MAP3K11 via the MAPK signaling pathway in vivo
We established a subcutaneous tumor model in mice and removed the xenografted tissues after images were taken. Representative images of the mice and xenografted tissues are shown in Figure 7A and 7B. Analysis of the xenografted tissue volumes and weights revealed that miR-125a-5p overexpression suppressed, while miR-125a-5p knockdown promoted, tumor growth ( Figure 7C and 7D). Moreover, PTPN1 and MAP3K11 expression was higher in HCC tumor tissues than in normal liver tissue in the Human Protein Atlas database [48] (http://www.proteinatlas. org) ( Figure 7E). A Chi-square test indicated that PTPN1 and MAP3K11 expression was also higher in HCC tissues than normal liver tissues in human liver cancer tissue microarray(TMA) slides purchased from U.S. Biomax (NO: LV1501) containing 10 normal liver and 120 HCC tissues (Supplementary Tables 3-5). Brdu and Ki67 staining of the xenograft tissues indicated that miR-125a-5p overexpression decreased the number of proliferating cells and increased the number of apoptotic cells, while miR-125a-5p knockdown had the opposite effects ( Figure 7F). In addition, miR-125a-5p overexpression decreased PTPN1, MAP3K11, JNK1/2/3, and c-Jun expression, while miR-125a-5p knockdown had the opposite effects ( Figure 7F). Taken together, these results demonstrate that miR-125a-5p may suppress cell proliferation and induce apoptosis in HCC by targeting PTPN1 and MAP3K11 via the MAPK signaling pathway.

DISCUSSION
In this study, we found that miR-125a-5p was downregulated in HCC tissues and cell lines and that it inhibited HCC cell proliferation and induced cell apoptosis both in vitro and in vivo. In addition, PTPN1 and MAP3K11 were identified as targets of miR-125a-5p. Furthermore, miR-125a-5p inhibited HCC cell proliferation and induced cell apoptosis by targeting PTPN1 and MAP3K11 via the MAPK signaling pathway.
Protein tyrosine phosphatase and tyrosine kinases modulate cellular levels of tyrosine phosphorylation and regulate cellular events such as proliferation, differentiation, and motility [49]. It is well established that the protein tyrosine phosphatase PTPN1 directly regulates the insulin and the leptin signaling pathway, making it a promising therapeutic target for type II diabetes and obesity [50]. However, PTPN1 plays conflicting roles in different types of tumors [26][27][28][29][30][31][32][33]. Here, we found that PTPN1 was a tumor promoter, and its expression increased in HCC tissues and cell lines. Moreover, high PTPN1 expression was associated with poorer overall survival in the GEPIA database.
MAP3K11 phosphorylates several MAP2Ks to activate c-Jun-N-terminal kinase (JNK) and downstream c-Jun transcription factors [51]. MAP3K11 can also activate p38 and help facilitate the activation of B-Raf, which stimulates the ERK 1/2 signaling pathway [52,53]. Previous studies reported that MAP3K11 acts as an oncogene and participates in the tumorigenesis process in breast and colorectal cancer [34][35][36][37]. Here, we found that MAP3K11 expression was increased in HCC tissues and cell lines, and higher MAP3K11 was associated with poorer overall survival in the GEPIA database.
We also found that miR-125a-5p suppressed HCC cell proliferation and induced cell apoptosis by directly targeting PTPN1 and MAP3K11 via the JNK MAPK signaling pathway in HCC. The JNK pathway regulates many physiological processes, including inflammatory responses and cell proliferation, differentiation, and death [54]. The JNK pathway is activated by two upstream mitogen activated protein kinase kinases (MAP2Ks) that directly phosphorylate JNKs on threonine (Thr183) and tyrosine (Tyr185) residues [55]. In turn, MAP2Ks are activated by upstream pathways through mitogen activated protein kinase kinase kinases (MAP3Ks), such as MEKK-4, MLK2, and MLK3 [56]. Upon activation by the upstream MAP2Ks, JNKs phosphorylate and activate several nuclear and nonnuclear proteins downstream of JNKs, such as c-Jun, Jun B, Jun D, c-Myc, p53, and the Bcl-2 family [57].
These proteins control cellular proliferation, differentiation, and apoptosis [58]. MAP3K11, which is a MAP3K, is an upstream activator of JNKs, which in turn activate the downstream proteins c-Jun, p53, and the Bcl-2 family and ultimately result in phenotypic changes in HCC cells. We found that miR-125a-5p overexpression increased p53 and Bax expression and reduced Bcl-2 expression, while miR-125a-5p knockdown led had the opposite effects. PTPN1 also activated JNKs and had the same effects as MAP3K11. Once c-Jun is activated, it increases the activity of transcription factor activator protein-1 (AP-1) upon dimerization with Fosproteins [57]. This increases expression of apoptotic precursor factors such as NF-κB, CD95L, p53, and TNF-α, leading to cell apoptosis [59,60]. Additionally, activated JNKs can promote apoptosis by directly inducing p53 activity [61][62][63]. JNKs also promote the activation of the Bcl-2 family in mitochondria; this induces the release of cytochrome C, which interacts with caspase-3 to cause apoptosis, into the cytoplasm [64,65]. Here, we also found that miR-125a-5p overexpression increased, while miR-125a-5p knockdown reduced, cleaved caspase substrate and caspase-3 expression, implying that JNKs also directly act on mitochondria to activate caspase-3 and induce cell apoptosis.
In summary, we found in this study that miR-125a-5p expression was significantly decreased in HCC tissues and cell lines compared to matched normal tissues and normal liver cells. miR-125a-5p overexpression inhibited HCC cell proliferation and induced apoptosis in vitro and in vivo, while miR-125a-5p knockdown had the opposite effects. Furthermore, PTPN1 and MAP3K11 were identified as targets of miR-125a-5p, and knockdown of PTPN1 and MAP3K11 inhibited the JNK signaling pathway to suppress HCC cell proliferation and induce apoptosis. These results indicate that miR-125a-5p inhibits HCC cell proliferation and induces cell apoptosis by directly targeting PTPN1 and MAP3K11 via the MAPK signaling pathway (Figure 8). miR-125a-5p mimics or JNK signaling inhibitors alone or in combination might therefore be potential treatments for HCC.

HCC tissue specimens
Eight pairs of fresh HCC tissue and adjacent normal tissue samples were collected from patients with HCC in our previous study [66]. Patients with other comorbidities, such as liver cirrhosis or hepatitis, were excluded from that study; only those with primary liver cancer alone were included. The specimens were frozen immediately after surgery in liquid nitrogen. None of the patients received radiotherapy, chemotherapy, or other treatments before surgery. All patients provided written informed consent for use of the samples during their hospitalization. The informed consent documents were reviewed and approved by the Ethics Committee of Shanghai Municipal Hospital of Traditional Chinese Medicine. miRNA microarray miRNA expression profiles in the HCC and matched control tissues were examined in a miRNA array (Shanghai Wcgene Biotech, Shanghai, China) using highthroughput quantitative RT-PCR (qRT-PCR). Total RNA was isolated from tissue samples, purified, polyadenylated in a poly (A) polymerase reaction, and then reversedtranscribed into cDNA. Individual miRNAs were quantified with specific MystiCq microRNA qPCR assay primers (Sigma, St. Louis, USA).

RNA extraction, reverse transcription, and quantitative real-time polymerase chain reaction (qRT-PCR)
Total RNA was extracted from cultured cells and tissues using TRIzol Reagent (Sigma, MKCB9720, St. Louis, USA) according to the manufacturer's instructions. Complementary DNA (cDNA) was reverse transcribed using an RT Reagent Kit (Takara, RR036A, Dalian, China) and RT Master Mix (Takara, RR037A, Dalian, China). qPCR was performed using Premix EX Taq (Takara, RR420A, Dalian, China). All primers used are listed in Supplementary Table 6 and were synthesized by Generay (Shanghai, China). U6 or GAPDH was used as an endogenous control.

Cell counting kit-8 (CCK8) cell proliferation assays
10,000 cells were cultivated in a 96-well plate in a 37°C incubator with 5% CO2. 10% CCK8 (Bimake, B34302, TX, USA) was added to each well 24 h, 48 h, 72 h, and 96 h after the cells were transfected. Absorbance optical density (OD) values were measured 1 h later at 450 nm wavelength. Each assay was performed three times.

Colony formation assay
For the colony formation assay, 300 cells were cultivated in 12-well plates in a 37°C incubator with 5% CO2 24 h after transfection. Ten days later, colonies resulting from the surviving cells were fixed with 95% alcohol (Sangon Biotech, A507050, Shanghai, China), stained with 0.1% crystal violet (Sigma, #SLBP6921V, St. Louis, USA), counted using Image J, and photographed. Each assay was performed three times.
Finally, luciferase activity was measured using a Dual Luciferase Reporter Assay Kit (Promega, E1910, Madison, USA) according to the manufacturer's instructions.

Western blot analysis
Total protein was extracted from cells after transfection using lysis buffer for western and IP (KeyGen, KGP701, Nanjing, China), and protein concentration was measured using a BCA protein assay kit (Beyotime, P0010, Shanghai, China

Tumor xenografts in animals
A total of 15 4-week-old athymic BALB/c nude mice (SLAC, Shanghai, China) were divided among three groups with five mice in each group. hsa-miR-125a-5p overexpression, knockdown, or control Bel-7404 cells (5 × 10 6 in 200 μL PBS) were subcutaneously injected into the mice. All cells expressed EGFP. Tumor volumes for each mouse were calculated as follows: tumor volume = length × width × width/2. After 3 weeks, the IVIS Imaging system (Spectral Instruments, USA) was used to observe the tumor xenografts in the mice. The mice were then sacrificed by cervical dislocation under anesthesia with ether, and xenograft tumor tissues were removed for examination.