LncRNA TMPO‐AS1 promotes hepatocellular carcinoma cell proliferation, migration and invasion through sponging miR‐329‐3p to stimulate FOXK1‐mediated AKT/mTOR signaling pathway

Abstract Purpose Hepatocellular carcinoma (HCC) is one of the leading causes of cancer‐related death worldwide. Numerous analyses have revealed the abnormal expression of long non‐coding RNAs (lncRNAs) in HCC cells. This study aims to explore biological functions of lncRNA TMPO‐AS1 (TMPO antisense RNA 1) in HCC cell proliferation, apoptosis, invasion and migration. Methods The gene expression in HCC tissues and cell lines were measured by qRT‐PCR. The role of TMPO‐AS1 in HCC was confirmed by CCK‐8, colony formation, TUNEL, transwell and western blot as well as by in vivo experiments. RNA pull down and luciferase reporter assays were utilized to prove the binding relationship between TMPO‐AS1/FOXK1 (forkhead box K1) andmiR‐329‐3p. Rescue assays elucidated the regulatory effects of TMPO‐AS1/miR‐329‐3p/FOXK1/AKT/mTOR pathway on cellular activities in HCC. Results TMPO‐AS1was upregulated in HCC tissues and cells and its depletion inhibits HCC cell proliferation, invasion, migration, and EMT process as well as tumor growth. Furthermore, TMPO‐AS1 could bind with miR‐329‐3p, which suppressed HCC cell proliferation. FOXK1 served as the target gene of miR‐329‐3p and TMPO‐AS1 upregulated FOXK1 by sponging miR‐329‐3p in HCC cells. Additionally, FOXK1 overexpression or miR‐329‐3p inhibitor neutralized the repressing effects of TMPO‐AS1 knockdown on HCC development. Finally, it verified that TMPO‐AS1 could regulate AKT/mTOR pathway via FOXK1 to promote HCC. Conclusion TMPO‐AS1 contributes to HCC progression by sponging miR‐329‐3p to activate FOXK1‐mediated AKT/mTOR signaling pathway.


| INTRODUCTION
Hepatocellular carcinoma (HCC) is identified as one of the main reasons of caner-associated death in the world and half of death cases occur in China. 1,2 People have a high risk of developing HCC. 3,4 Although the oncological and surgical treatments have been improved in recent years, the prognosis of HCC patients is still not optimistic and the postoperative recurrence rate is high, resulting in that the survival rate of HCC patients is low. 1,2,5,6 It is well-known that the molecular mechanisms underlying the biological processes of HCC are not well-researched, which requires us to find a new lncRNA that may contribute to the diagnosis and prognosis of patients.
LncRNAs are a subgroup of non-coding RNAs that have no or limited protein-coding ability, consisting of more than 200 nucleotides. 7 As lncRNAs exert their functions in many biological processes, their abnormal expressions are implicated in various cancers. 8,9 Many studies show that lncRNAs play important roles in the biological development of cancer cells, such as cell proliferation, migration, and EMT formation. [10][11][12] More importantly, aberrant expression of lnc RNA TMPO antisense RNA 1 (TMPO-AS1) promotes lung adenocarcinoma, prostate cancer. 8,13 The miR-329-3p is a critical miRNA and serves a tumor inhibitor in multiple cancers, such as suppressing cervical cancer cell proliferation. [14][15][16][17] Nevertheless, the role of TMPO-AS1 and miR-329-3p in HCC has not been investigated before. Our interest is to find whether TMPO-AS1/miR-329-3p axis modulates HCC development.
Therefore, in this study, we will be exploring the biological function of TMPO-AS1in HCC progression, which might inspire us to find an effective treatment target for HCC.

| Tissue samples
About 48 HCC tissues and normal tissues were obtained from patients with HCC undergoing surgical resection at our hospital. Patients treated with radiotherapy or chemotherapy before study were excluded. Each participant signed the written informed consent. All fresh tissues were frozen in liquid nitrogen and stored immediately at −80°C. Ethical approval was obtained from the Ethics Committee of our hospital.

| Quantitative real-time polymerase chain reaction (qRT-PCR)
Total RNA was isolated from cells utilizing TRIzol reagent (Invitrogen) and was reverse-transcribed into cDNA by Reverse Transcription Kit (Invitrogen). The qRT-PCR was performed in the Bio-Rad CFX96 system using SYBR-Green Real-Time Kit (Takara, Tokyo, Japan). The relative expression was normalized to GAPDH or U6 and fold expression changes were calculated with 2 -ΔΔCt method.

| Colony formation assay
1 × 10 3 Huh7 and Hep3B cells were plated onto a fresh six-well plate for 2-week incubation. Formaldehyde (Sigma-Aldrich) and crystal violet (Sigma-Aldrich) were used for fixing and dying the colonies. The visible colonies were counted manually.

| Terminal-deoxynucleoitidyl transferase-mediated nick end labeling (TUNEL) assay
The apoptosis of Hep3B and Huh7cells were evaluated via TUNEL Apoptosis Kit (Invitrogen). DAPI (Koritai Biotechnology) was bought for nucleus staining. The cells were surveyed and captured through fluorescence microscopy (Olympus).

| Transwell assay
Cell invasion and migration were evaluated using transwell chambers (Millipore) with or without Matrigel (Corning Inc, USA).Hep3B and Huh7cells were cultivated for 24 hours and then nonmigrated cells were removed. About 4% paraformaldehyde and crystal violet were used for fixing and staining the migrated cells and photographed by an IX71 inverted microscope (Olympus, Tokyo, Japan).About invasion assays, cells were incubated in serum-free medium and seeded into the top compartment of the Matrigel-coated plates, and the medium containing 10% FBS was added to the bottom compartment. After 48 hours, cells were fixed in 4% paraformaldehyde and then dyed using crystal violet. Finally, invaded or migrated cells were counted under a microscope (Olympus).

| Subcellular fractionation assay
In order to determine cellular localization of TMPO-AS1, cytoplasmic and nuclear fractions were collected via a Nuclear/ Cytoplasmic Isolation Kit (Norgenbiotek Corporation, Thorold, ON, Canada). Relative expression level of TMPO-AS1 was determined by qRT-PCR. GAPDH or U6 was the cytoplasmic control or nuclear control.

| Animal studies
A total of 15 male BALB/c athymic nude mice (5-week-old, 17-20 g), were brought from the National Laboratory Animal Center (Beijing, China) and were randomly divided into three groups. HCC xenografts in the nude mice were established by subcutaneously injecting with sh-NC or sh-TMPO-AS1#1 transfected HCC cells. The untreated mice served as the mock group. Tumor volume was measured using a vernier caliper every 4 days utilizing the formula "volume = 1/2 × length ×width." The mice were sacrificed after 28 days of injection, and the tumor was also weighed. This experiment was approved by the Ethics Committee of our hospital.

| Statistical analysis
All statistical analyses were performed by SPSS (SPSS Inc, Chicago, IL, USA). Data were expressed as mean ± SD from more than three independent experiments. Student's t-test and one-way ANOVA were used to calculate the difference between two or more groups. The overall survival analysis was performed using Kaplan-Meier method and gene correlation analysis was conducted by Pearson's correlation analysis method. P value of 0.05 or less was regarded as significance.

| TMPO-AS1-regulated HCC cell proliferation, apoptosis, and invasion
To explore the role of TMPO-AS1 in HCC, firstly, qRT-PCR measured the expression of TMPO-AS1 in HCC tissues and cells. TMPO-AS1was upregulated in HCC tissues and cells (Huh7, Hep3B, and LM3) compared with normal liver tissues and human liver epithelial cell (THLE-3) ( Figure 1A,B). Additionally, Kaplan-Meier analysis illustrated that the highTMPO-AS1 expression was closely associated with poor prognosis of HCC patients ( Figure 1C). Then the associations between TMPO-AS1 and clinicopathological characteristics in patients with HCC were investigated. High TMPO-AS1 expression was related to large tumor size, lymphatic metastasis, and advanced stage of cancer patients ( Table 1).The knockdown efficiency of TMPO-AS1 was measured in Huh7 orHep3B cells which contained higher expression of TMPO-AS1than LM3 cells. TMPO-AS1 expression was decreased by transfecting sh-TMPO-AS1#1/2 plasmids ( Figure 1D). Next, several loss-of-function assays were performed. Results from colony formation and CCK-8assays indicated that colony formation ability and cell viability in sh-TMPO-AS1#1/2 transfected Huh7 and Hep3B cells were decreased ( Figure 1E,F). TUNEL assay detected that TUNEL-positive cells were increased by knocking down TMPO-AS1, indicating that cell apoptosis was stimulated. Furthermore, results from transwell assay showed that TMPO-AS1downregulation decreased number of invaded and migrated cells ( Figure 1H, Figure S1A). The western blot assay examined the expression of migration-related proteins. When downregulating TMPO-AS1, the expression of E-cadherin was increased, and that of N-cadherin, MMP2 as well as MMP7 was decreased, indicating that EMT process of HCC cells was obstructed by knocking downTMPO-AS1 ( Figure S1B). Additionally, animal studies were also carried out. Compared with control groups, the tumors obtained from mice injected with sh-TMPO-AS1#1 transfected cells were obviously smaller ( Figure S1C). Besides, tumor volume and weight were also smaller and lighter than control groups ( Figure S1D,E). The above results demonstrated that TMPO-AS1 is involved in the development of HCC both in vitro and in vivo.

HCC cells
According to the result that TMPO-AS1 mainly distributed in cytoplasm, confirmed by subcellular fractionation assay (Figure 2A).We speculated that TMPO-AS1 possibly affects cancer progression via ceRNA regulation pattern. Subsequently, starBase was utilized to screen the miRNAs that might bind with TMPO-AS1. After qRT-PCR analyses, 29 miRNAs were downregulated in HCC cells and 51 miR-NAs were negatively regulated by TMPO-AS1.Finally two prequalified miRNAs (miR-329-3p and miR-4500) were obtained by taking the intersection of two groups of miR-NAs ( Figure S2A). Then, the expression of miR-329-3p and miR-4500 was tested in HCC tissues and normal tissues. The results of qRT-PCR assay showed that the expression of miR-4500 presented no obvious differences in HCC tissues and normal tissues, while miR-329-3p was significantly downregulated in HCC tissues ( Figure S2BB, Figure 2B). Besides, Pearson's correlation analysis validated that miR-329-3p was negatively correlated with TMPO-AS1 in HCC ( Figure 2C). Through starBase, the binding site between miR-329-3p and TMPO-AS1 was obtained ( Figure 2D). Therefore, miR-329-3p was chosen as the object of the studies below. MiR-329-3p was less expressed in HCC cells compared with THLE-3 cells ( Figure 2E). Similarly, to explore the biological function of miR-329-3p in HCC,miR-329-3p overexpression efficiency was evaluated by transfecting miR-329-3p mimics into Huh7 and Hep3B cells, and the expression of miR-329-3p was increased ( Figure 2F). Luciferase reporter assay detected that the luciferase activity of TMPO-AS1-WT was decreased by miR-329-3p mimics, and that of TMPO-AS1-Mut was not affected ( Figure 2G). RNA pull down assay showed that expression of TMPO-AS1was enriched in the Bio-miR-329-3p-WT group rather than the mock, Bio-NC, or Bio-miR-329-3p-Mut group ( Figure 2H). Additionally, CCK-8 and colony formation assays demonstrated that overexpressing miR-329-3p inhibited cell proliferation ( Figure 2I,J). In conclusion, TMPO-AS1 could bind with miR-329-3p and miR-329-3p overexpression suppressed HCC cell proliferation.

| Oncogenic function of FOXK1 in HCC cells is dependent on miR-329-3p
To find out the target gene of miR-329-3p, firstly, five possible downstream mRNAs of miR-329-3p were screened out

| TMPO-AS1 promotes HCC by regulating miR-329-3p/FOXK1 axis
Previous assays showed that TMPO-AS1, miR-329-3p, and FOXK1constituted a ceRNA network in HCC cells, so we proposed whetherTMPO-AS1 affected the development of HCC via regulating miR-329-3p and FOXK1. Therefore, on one hand,miR-329-3p inhibitor was used in the rescue experiments. The knockdown efficiency of miR-329-3p was detected in miR-329-3p inhibitor transfected cells, and miR-329-3p expression was remarkably decreased ( Figure S3A). Afterward, the results of the CCK-8 and colony formation assays proved that miR-329-3p downregulation reversed the repressing effects of TMPO-AS1 silence on cell proliferation ( Figure S3B,C). The facilitating role of TMPO-AS1 depletion on cell apoptosis was also neutralized by the miR-329-3p inhibitor ( Figure S3D). In addition, when suppressing miR-329-3p expression, the impeditive effects of TMPO-AS1 insufficiency on cell invasion, migration, and EMT were also counteracted ( Figure S3E,F,G). On the other hand, we measured the overexpression efficiency of FOXK1 in HCC cells. qRT-PCR and western blot assays showed that FOXK1 overexpression neutralized the inhibitory effect of TMPO-AS1 depletion on mRNA and protein expressions of FOXK1 ( Figure 4A,B). Proliferation assays detected that FOXK1 upregulation reversed the restraining influence of TMPO-AS1 knockdown on cell proliferation ( Figure 4C,D). TUNEL assay showed that FOXK1 upregulation offsets the accelerating effects of TMPO-AS1 knockdown on cell apoptosis ( Figure 4E). Furthermore, results from transwell and western blot assays elucidated that FOXK1 offsets the obstructive effects of TMPO-AS1 deficiency on invasion, migration, and EMT process of HCC cells( Figure 4F, Figure S4E-G).

| TMPO-AS1 promoted AKT/mTOR pathway in HCC cells by modulating FOXK1
Several papers have reported that FOXK1 could regulate the AKT/mTOR pathway. [18][19][20] Therefore, we tried to investigate whetherTMPO-AS1 could modulate AKT/mTOR pathway via FOXK1in HCC cells. Firstly, western blot assay examined the expression of AKT/mTOR pathway-related proteins. TMPO-AS1 knockdown markedly decreased the p-AKT and p-mTOR expressions in Huh7 and Hep3Bcells. Moreover, this inhibitory effect of sh-TMPO-AS1was then offset by FOXK1 upregulation, indicating that TMPO-AS1 exerted the underlying regulatory function on AKT/mTOR pathway ( Figure 5A). Subsequently, rescue assays were performed to verify our assumptions with using IGF-1, the AKT/mTOR pathway activator. Results of CCK-8, colony formation and TUNEL assay suggested that IGF-1 treatment reversed the suppressing effects of TMPO-AS1depletion on cell proliferation as well as the encouraging effects of that on cell apoptosis ( Figure 5B-D). Transwell and western blot assays showed that the interruptive effects of TMPO-AS1depletion on cell invasion, migration and EMT were remedied by IGF-1 treatment ( Figure 5E-G). These results showed that TMPO-AS1 facilitates HCC development through upregulating FOXK1 expression to stimulate AKT/mTOR pathway.

| DISCUSSION
Many papers have documented that lncRNAs play crucial roles in HCC progression. For illustration, lncRNA n339260 facilitates vasculogenic mimicry and strengthens cancer stem cell in HCC. 21 The lncRNA SchLAH represses HCC metastasis via interacting with FUSin sarcoma. 22 Also, lncRNAs can serve as therapeutic targets for HCC patients. For instance, lncRNA AWPPH contributes to the development of HCC progression by YBX1 and acts as a prognostic marker. 23 The lncRNA CCHE1 suggests a bad prognosis of HCC and promotes HCC progression by activating the ERK/MAPK pathway. 24 In our research, TMPO-AS1 was upregulated in HCC tissues and cell lines. Besides, the oncogenic role of TMPO-AS1 has been confirmed in cancers, such as cervical cancer 25 and nonsmall cell lung cancer. 11 Furthermore, TMPO-AS1 is also identified as an underlying therapeutic target for patients diagnosed with cancers, including colorectal cancer, 26 prostate cancer, 10 and lung adenocarcinoma. 27 Therefore, we conjectured whether TMPO-AS1 also contributed to the development of HCC, and inhibitingTMPO-AS1 benefited the improvement of the prognosis of HCC patients. Our research found that high TMPO-AS1 expression was closely associated with unsatisfactory prognosis of HCC patients. The results of the functional assays demonstrated that TMPO-AS1 depletion suppressed cell proliferation, invasion, migration, and EMT process, meanwhile, inhibited tumor growth, suggesting the tumor promoter role of TMPO-AS1 in HCC.
Moreover, to figure out whether TMPO-AS1 regulated HCC via ceRNA pattern, the downstream miR-NAs of TMPO-AS1 were explored. In the present study, F I G U R E 4 TMPO-AS1 upregulated FOXK1 to promote HCC. A, B, qRT-PCR and western blot assays were conducted to detect FOXK1mRNA and protein expressions in differently transfected groups. C, D, The capacity of cell proliferation was detected by CCK-8 and colony formation assays. E, Cell apoptosis was detected by TUNEL assay. F, Cell invasion was tested by transwell assay. * P < .05, ** P < .01 miR-329-3p was screened by bioinformatics tool and mechanism experiments. Besides, miR-329-3p was certified as a tumor suppressor in cervical cancer, 14,15 and the number of related researches about its role in cancers is limited. In this study, miR-329-3p overexpression offsets the repressing influence of TMPO-AS1 depletion on HCC cell growth. Subsequently, the decisive mRNA in ceRNA network, FOXK1 was explored. According to previous researches, FOXK1 promotes liver cancer, 19 breast cancer, 28 esophageal cancer, 29 colorectal cancer, 30 prostate cancer 31 etc. Whether TMPO-AS1 could modulate HCC progression by targeting FOXK1/miR-329-3p axis was worth investigating. Additionally, the regulatory effects of FOXK1on AKT/mTOR pathway have been elucidated by F I G U R E 5 AKT/mTOR pathway was promoted by TMPO-AS1 in HCC cells. A, Western blot assay assessed the level of AKT/mTOR pathway-associated proteins in Huh7 and Hep3B cells. B, C, CCK-8 and colony formation assays detected cell proliferation in different treatment groups. D-F, TUNEL and transwell assays detected cell apoptosis and invasion/migration abilities respectively in different treatment groups. G, Western blot assay detected the level of EMT-and migration-related proteins. * P < .05, ** P < .01 other investigations. [18][19][20] Hence, whetherTMPO-AS1 could stimulate AKT/mTOR pathway via regulating FOXK1 to promote HCC further interested us. Through western blot assays, TMPO-AS1 knockdown obviously inhibited p-Akt and p-mTOR expressions, and this inhibitory influence was counteracted by FOXK1 upregulation. In addition, IGF-1 (an activator of AKT/mTOR pathway) treatment could reverse the obstructive effects of TMPO-AS1 knockdown on cell proliferation, invasion, migration, and EMT process, indicating the significant influence of AKT/mTOR pathway stimulation on the regulatory role of TMPO-AS1 in HCC.
In a summary, TMPO-AS1 promotes HCC through sponging miR-329-3p to upregulate FOXK1 and activate AKT/mTOR signaling pathway, which was the first to illustrate the potential of TMPO-AS1 as a therapeutic target in HCC. In addition, the upstream molecular mechanism of TMPO-AS1 in HCC cells waits to be explored in our further studies.