Abstract

Accumulating evidences indicate that circular RNAs (circRNAs) play a vital role in diverse cancer biology. However, the contributions of circRNAs to hepatocellular carcinoma (HCC) and their underlying mechanism remain largely unknown. The present study aims at investigating the role of circRNA-104718 in HCC progression, which has been observed to be significantly up-regulated in HCC tissues. We found that, higher expression of circRNA-104718 also leds to a poor prognosis in HCC patients. Using luciferase binding assays and RNA immunoprecipitation studies, we identified circRNA-104718 is physically associated and co-expressed with microRNA (miR)-218-5p in HCC. Mechanistically, we demonstrated that circRNA-104718 functions as a competing endogenous RNAs (ceRNAs) and competes with thioredoxin domain-containing protein 5 (TXNDC5) mRNA and directly binds to miR-218-5p. Functionally, we found that ectopically expressed circRNA-104718 accelerated cell proliferation, migration, invasion, and inhibited apoptosis. In vivo studies on a nude mice model showed that circRNA-104718 overexpression could increase the tumor size and the rate of metastasis. Silencing of circRNA-104718 could decrease both the tumor size and metastasis significantly. Conversely, we also observed overexpression of miR-218-5p could in turn decrease the proliferation, migration, invasion, and increase apoptosis. Furthermore, circRNA-104718 could relieve the suppression of miR-218-5p target TXNDC5 and thereby cause an inhibition of miR’s functions. In summary, our results indicate that circRNA-104718 acts as a ceRNA and promotes HCC progression through the targeting of miR-218-5p/TXNDC5 signaling pathway. Thus, we propose that circRNA-104718 would be a promising target for HCC diagnosis and therapy.

Introduction

Hepatocellular carcinoma (HCC) is among one of the most prevalent causes of liver malignancy worldwide and still stands as the most leading cause for cancer related deaths [1]. Incidence of the disease and mortality due to HCC is still on rise despite recent advances in screening and treatment strategies [2]. With estimated new cases at 30,610 and estimated deaths at 20,540 in 2018, liver-associated cancers still continue to take its toll among cancer-related mortalities [3]. Critically, the incidence of HCC is predicted to be on a rise in the coming years as non-alcoholic steatohepatitis and obesity is on a rise in countries such as United States [2]. Recently, increasing number of studies are emerging to understand the molecular basis of HCC and to develop specific improved treatment strategies.

Non-coding RNA (ncRNA) is a class of RNA which was till recently considered as untranslated and non-functional. Circular RNA (circRNA) is a class of ncRNAs and is named so as both their 3′ and 5′ end are joined to one another forming a circle like structure [4]. Recently, investigations have featured the significance of circRNA in cancer and indicated it as critical for developing treatment strategies. Especially, their role in influencing tumorigenesis has been all around reported in different malignant growths [5]. Interestingly, many circRNAs have been observed to be either up-regulated or down-regulated in HCCs [6]. Previous reports have shown, circRNA-0007874 [7], circRNA-0001727 [8], and circRNA-001649 [9] to be down-regulated and which in turn seems to increase the tumor progression in HCCs. On the contrary, circRNAs such as circRNA-0005075 [10] and circRNA 100338 [11] are up-regulated and still seem to contribute to the progression of tumorigenesis in HCCs. Such variations have risen the demand for more detailed research to understand the molecular basis for these circRNA.

One role of circRNA that has been under immense focus has been its role in regulation of microRNAs (miR) [4]. miRs are 22 nucleotides long ncRNAs, with vital tasks related to regulating other genes. Remarkably, miRs have been identified to have anti-tumorigenic roles, which make them interesting candidates for development of treatment strategies [12]. In HCCs, miRs such as miR-9 [13], miR-129-5p [14], miR-124-3p [15], and miR-141-3p [11] have all been identified to be down-regulated in HCC. Most of these studies indicate that circRNA and miR could act as competitive endogeneous RNA (ceRNA) in binding and regulating each other. This in turn restricts the activity of miR in binding and activating other genes. miRNAs bind specifically to miRNA responsive elements (MREs) in specific genes and regulates them, but circRNA which share similar MREs act as intensive sponges and absorb the miRs [16]. This decreases their levels and prohibits their regulatory activity on other genes.

One study showed circRNA-0007874 to be down-regulated in HCCs. With the help of bioinformatics tools, it was identified that this circRNA binds and down-regulates miR-9 [17]. In HCCs, miR-9 has been identified to be up-regulated which in turn regulates the expression of a critical cell cycle gene, p21. Hence, here silencing of miR-9 was identified to be a critical strategy to treat HCC [17,18]. Another study by Xu et al.,[11] showed ciRS-7 (circular RNA sponge for miR-7) to be up-regulated in HCC, and this increases tumorigenesis by inhibiting miR-7 which is a vital regulator of many important genes such as EGFR, PIK3CD, and CCNE1. Silencing of ciRS-7 seems to decrease the tumorigenesis by increasing the activity of miR-7 in HCCs. Such interesting contradictory roles have led to an increasing demand for more enhanced identification and analysis of circRNAs in HCCs.

The present study aims at identifying and understanding the role of circRNA-104718 in HCCs. Previously, a study by Xu et al., [10] identified in their circRNA microarray analysis, circRNA-104718 to be highly up-regulated in HCC patients. Further in our study, we identified that patients with high expression of circRNA-104718 had a poor prognosis. Using an in vitro HCC model, we identified circRNA-104718 to increase the proliferation, migration, invasion, and decrease apoptosis. In vivo model showed circRNA-104718 increased metastasis and tumor size. We further elucidated its role in sponging miR-218-5p and regulating the activity of thioredoxin domain-containing protein 5 (TXNDC5) gene, which is up-regulated in many tumors, including liver cancer and involved in the proliferation and migration of tumor cells, and thus act as a tumor-enhancing gene [19]. Hence, the present study for the first time describes the molecular mechanism behind circRNA-104718’s role in tumorigenesis, thus contributing to the development of treatment strategies for HCC.

Materials and methods

Patients and tissue samples

At Zhongshan Hospital, Shanghai, China, a total of 103 pairs of tumor biopsies and corresponding adjacent tissues were collected from HCC patients who had surgery. The present study was conducted in accordance with the Helsinki Declaration and approved by the Zhongshan Hospital Ethics Committee, and all patients provided with the signed consent forms. All experimental protocols were approved by Fudan University’s Animal Care and Use Committee and followed the National Institutes of Health’s Guide to Care and Use of Laboratory Animals. Commercial tissue microarrays (TMAs) were obtained from Shanghai bio-chip Co. Ltd. (Shanghai, China). The HCC TMA used in the present study included 68 primary HCC and 60 adjacent non-cancerous tissues of the liver, with donor age ranging from 18 to 73 years (mean age, 48.31 years). The detailed information regarding the clinical features of the patients is presented in Table 1.

Table 1
Association of circRNA-104718 expression with the clinical characteristics of patients with HCC
Characteristics Biopsies (n=103) TMAs (n=68) circRNA-104718 low expression (n=79) circRNA-104718 high expression (n=92) P-value 
Gender      
  Female, n (%) 59 (57.3) 58 (85.3) 50 (63.3) 67 (72.8) 0.12058 
  Male, n (%) 44 (42.7) 10 (14.7) 29 (36.7) 25(27.2)  
Age, y      
  ≥60, n (%) 64 (62.1) 12 (17.6) 42 (53.2) 34 (37.0) 0.02420 
  <60, n (%) 39 (37.9) 56 (82.4) 37 (46.8) 58 (63.0)  
TNM stage      
  I, II 68 (66.0) 26 (38.2) 46 (58.2) 48 (52.2) 0.26151 
  III, IV 35 (34.0) 42 (61.8) 33 (41.8) 44 (47.8)  
pT stage      
  T1–T2 56 (54.4) 34 (50.0) 47 (59.5) 43 (46.7) 0.06514 
  T3–T4 47 (45.6) 34 (50.0) 32 (40.5) 49 (53.3)  
Serum AFP/ng/ml      
  <25 53 (51.5)  29 (61.7) 29 (51.8) 0.20878 
  ≥25 50 (48.5)  18 (38.3) 27 (48.2)  
Tumor size, cm      
  ≥5, n (%) 41 (39.8)  15 (31.9) 24 (42.9) 0.17462 
  <5, n (%) 62 (60.2)  32 (68.1) 32 (57.1)  
Vascular invasion      
  Yes 45 (43.7) 16 (23.5) 40 (50.6) 21 (22.8) 0.00014 
  No 58 (56.3) 52 (76.5) 39 (49.4) 71 (77.2)  
Characteristics Biopsies (n=103) TMAs (n=68) circRNA-104718 low expression (n=79) circRNA-104718 high expression (n=92) P-value 
Gender      
  Female, n (%) 59 (57.3) 58 (85.3) 50 (63.3) 67 (72.8) 0.12058 
  Male, n (%) 44 (42.7) 10 (14.7) 29 (36.7) 25(27.2)  
Age, y      
  ≥60, n (%) 64 (62.1) 12 (17.6) 42 (53.2) 34 (37.0) 0.02420 
  <60, n (%) 39 (37.9) 56 (82.4) 37 (46.8) 58 (63.0)  
TNM stage      
  I, II 68 (66.0) 26 (38.2) 46 (58.2) 48 (52.2) 0.26151 
  III, IV 35 (34.0) 42 (61.8) 33 (41.8) 44 (47.8)  
pT stage      
  T1–T2 56 (54.4) 34 (50.0) 47 (59.5) 43 (46.7) 0.06514 
  T3–T4 47 (45.6) 34 (50.0) 32 (40.5) 49 (53.3)  
Serum AFP/ng/ml      
  <25 53 (51.5)  29 (61.7) 29 (51.8) 0.20878 
  ≥25 50 (48.5)  18 (38.3) 27 (48.2)  
Tumor size, cm      
  ≥5, n (%) 41 (39.8)  15 (31.9) 24 (42.9) 0.17462 
  <5, n (%) 62 (60.2)  32 (68.1) 32 (57.1)  
Vascular invasion      
  Yes 45 (43.7) 16 (23.5) 40 (50.6) 21 (22.8) 0.00014 
  No 58 (56.3) 52 (76.5) 39 (49.4) 71 (77.2)  

Cell lines and culture

All the human HCC cell lines used in the present study (SMMC-7721, HepG2, MHCC-LM3, and SK-Hep1), control hepatic normal cell (L02) and HEK293T cells were all purchased from Shanghai Institute of Cell Biology, Chinese Academy of Sciences (Shanghai, China). Cells were maintained in Dulbecco’s Modified Eagle’s medium (DMEM; Gibco, Carlsbad, CA, U.S.A.) medium with 10% heat-inactivated fetal bovine serum (FBS, Gibco), 1% penicillin (100 U/ml) and 0.1 mg/ml streptomycin (Solarbio, Beijing, China) in a humidified chamber with 5% CO2 and 95% air at 37°C.

RNA extraction and quantitative RT-PCR

Total RNA from tissues or cells was extracted using Trizol reagent (Invitrogen, Grand Island, NY, U.S.A.). For miRNA analysis, complementary deoxyribonucleic (cDNA) was obtained using the TaqMan MicroRNA Reverse Transcription Kit (Applied Biosystems, Foster City, CA, U.S.A.) and quantitave real-time polymerase chain reaction (qRT-PCR) was performed using TaqMan miRNA assay kit (Applied Biosystems). U6 small nuclear RNA (U6 snRNA) was used as an endogenous control for normalization. For mRNA analysis, cDNA was synthesized by using M-MLV reverse transcriptase (Invitrogen) and reverse transcription primers Oligo(dT). qRT-PCR was then performed with SYBR Green Real-Time PCR Master Mixes (ThermoFisher, Waltham, MA, U.S.A.) on a 7900HT Fast RealTime PCR machine (Applied Biosystems). Primer sequences used in the present study as follows:

  • circRNA-104718: forward, 5′-GAGGCCCAGCTCAGTGTG-3′; reverse, 5′-GGCTGCAATGGTCTTGTTGG-3′

  • TXNDC5: forward, 5′-TCACTGAGGGAGTACGTGGA-3′; reverse, 5′-AGCAGTGCAGTCTACTTCGG-3′

  • Glyceraldehyde-3-phosphate dehydrogenase (GAPDH): forward, 5′-GGAAGCTTGTCATCAATGGAAATC-3′; reverse, 5′-TGATGACCCTTTTGGCTCCC-3′

  • miR-218-5p loop primer: 5′-CGACTCGATCCAGTCTCAGGGTCCGAGGTATTCGATCGAGTCGCACT-3′

  • miR-218-5p: forward, 5′-AACACGAACTAGATTGGTACA-3′; reverse, 5′-AGTCTCAGGGTCCGAGGTATTC-3′

  • U6: forward, 5′-CTCGCTTCGGCAGCACA-3′; reverse, 5′-AACGCTTCACGAATTTGCGT-3′

Western blot analysis

Proteins isolation, quantification, and Western blotting were performed as previously mentioned [20]. Cells were washed and proteins were extracted with lysis buffer (100 μl/50 ml). Equal amounts of protein sample were separated by SDS/PAGE and transferred to PVDF membranes (Millipore, Bedford, Mass). Membranes were incubated overnight at 4°C with primary antibodies anti-TXNDC5 and GAPDH (Abcam, Cambridge, MA, U.S.A.). Membranes were then thoroughly washed with Tris-buffered saline-Tween 20 (TBST) and incubated for 1 h at room temperature with horseradish peroxidase-labeled secondary antibody (Santa Cruz Technology, Santa Cruz, CA, U.S.A.). After washing with TBST, the membranes were visualized with an enhanced chemiluminescence (ECL) system. The mean values from three experiments were obtained.

Luciferase reporter assay

Luciferase assay was performed as previously mentioned [21]. HEK-293T cells were seeded in 24-well plates in triplicate and transfected with corresponding plasmids and miR-218-5p mimics or inhibitors. The luciferase reporters including circRNA-104718-WT, circRNA-104718-MUT, TXNDC5-WT, and TXNDC5-MUT were constructed by Genechem (Shanghai, China). Site-directed mutagenesis was performed using the QuickChange Lightning kit (Stratagene, La Jolla, CA, U.S.A.). Sequencing was used to confirm the correct mutations had been generated. Luciferase activity was measured using Dual-luciferase reporter assay system (Promega, Madison, WI) according to the manufacturer’s instructions. Relative luciferase activity was normalized to the Renilla luciferase internal control.

Fluorescence in situ hybridization

The hybridization experiment was as previously mentioned [22]. The incubations were performed overnight with circRNA-104718 and hsa-miR-218-5p probes. Specimens were analyzed on a Nikon inverted fluorescence microscope. The circRNA-104718 probe for fluorescence in situ hybridization (FISH) was 5′-GGT ACC GCG TGT GGA GCC GCA TCC TGC AGG CCG TGA ATG CGC-3′ and the miR-218-5p probe for FISH was 5′-TTG TGC TTG ATC TAA CCA TGT-3′.

Wound-healing assay

24 h after transfection, a straight scratch was created in the center of each well using a micropipette tip. Cell migration was assessed by measuring the movement of the cells into the scratch in the well. The wound closure speed after 24 and 48 h was determined and normalized to the length at 0 h. Each experiment was performed in triplicate.

Cell invasion assays

Cell invasion assay was performed using a 24-well Transwell chamber (BD, U.S.A.). At 48 h after transfection, SK-Hep1 and MHCC-LM3 cells were trypsinized and transferred to the Matrigel (BD, U.S.A.)-coated upper chamber containing 100 μl of serum-free medium. FBS was added to the lower chamber as chemoattractant. After 24 h, cells on the bottom of the chambers were fixed in 4% paraformaldehyde and stained with 0.1% crystal violet. Cells that invaded into the lower surface were counted in at least five random fields (magnification: ×200, Nikon). Each experiment was performed in triplicate.

RNA immunoprecipitation assay

HEK-293T cells (∼1 × 107) were washed with cold PBS after transfection with miR-218-5p mimics or mimic-NC for 48 h, and lysed with RNA immunoprecipitation (RIP) lysis buffer (EMD Millipore, Billerica, MA, U.S.A.) according to manufacturer’s instructions. Then cell lysates were incubated with RIP immunoprecipitation buffer containing magnetic beads conjugated with human anti-Argonaute2 (AGO2) antibody (Millipore, Billerica, MA, U.S.A.) or negative control mouse IgG (Millipore, Billerica, MA, U.S.A.). Samples were incubated with Proteinase K and then immunoprecipitated RNA was isolated. Extracted RNAs were analyzed by RT-PCR or qRT-PCR to identify the presence of circRNA-104718.

Plasmids construction and cell transfection

For circRNA-104718 overexpression, the full-length circRNA-104718 cDNA product was amplified in 293 T cells and then cloned into over expression vector pLCDH-ciR (Genechem, Shanghai, China), which contained a front and back circular frame; while the mock vector with no circRNA-104718 sequence served as a control. The sequence of circRNA-104718 siRNA 5′-ATACTCGGTCCAACATCTAGC-3′ was used as negative control were synthesized at Genechem (Shanghai, China). All plasmids were isolated using the DNA Midiprep kit (Qiagen, Germany). For miR-218-5p overexpression and knockdown, miR-218-5p mimics, miR-218-5p inhibitor, and two scrambled miRNAs used as negative controls (mimic-NC for miR-218-5p mimics and inhibitor-NC for miR-218-5p inhibitor, respectively) were purchased from Genechem. Transfections were performed by using Lipofectamine 2000 reagent (Invitrogen, Carlsbad, CA, U.S.A.) following manufacturer’s instructions.

Detection of apoptosis

Cell apoptosis and cell cycle phase distribution with cellular DNA contents were carried out using flow cytometry. HCC cells were seeded into six-well plate at density 1 × 106 cells/ml and treated with 100, 300, and 500 µM of naringenin for 24 h in 5% CO2 incubator and at 37°C temperature. After 24-h incubation, the cultured cells were harvested, washed with cold PBS, fixed in 70% ethanol and treated with RNase A (10 mg/ml). Fixed cells were stained with propidium iodide (PI) dye followed by incubation for 30 min at room temperature in dark. The PI fluorescence of individual nuclei was measured using flow cytometer (BD FACS Calibur, Becton Dickinson, U.S.A.). Data were analyzed with the Cell Quest Pro V 3.2.1 software (Becton Dickinson, U.S.A.).

Animal studies

The BALB/c (nu/nu) nude mice used in the present study were all male, 6–8 weeks old, and purchased from the Chinese Science Academy (Shanghai, China). All the animal experiments were made in the animal lab in Zhangshan Hospital and the animal studies were approved by the Animal Care and Use Committee. Subcutaneous tumor growth assays, a lung metastasis model, and a tail vein injection model were implemented as described [21–23]. Metastases were detected using the IVIS@ Lumina II system (Caliper Life Sciences, Hopkinton, MA) 10 min after intraperitoneal injection of 4.0 mg luciferin (Gold Biotech) in 50 µl of saline. The xenograft tumor size was monitored every 3 days (volume = width2 × length × 1/2). Mice were euthanized at the end of the experiment and the tumors were excised. Tumors were fixed in 10% formalin, embedded in paraffin, and cut into 4-μm thick slices. The slides of paraffin-embedded xenograft tissues were probed with primary antibodies anti-ki67 (Abcam). The staining processes were performed as previously described [23] and quantified with Image ProPlus (IPP) software (Media Cybernetics, Rockville, MD, U.S.A.).

Statistical analysis

Data were analyzed using SPSS software (version 16). Each experiment was performed in triplicate and values are presented as mean ± SEM. The two-tailed Student’s t-test was used to analyze statistical differences between groups. P-values <0.05 were considered statistically significant.

Results

circRNA-104718 is up-regulated in HCC patients and is associated with poor prognosis in patients

To identify the expression levels of circRNA-104718 in HCC, we initially checked the differential expression levels in 103 HCC tissues and compared it with matched non-tumor tissues (Figure 1A). It was evident that circRNA-104718 levels increased significantly in the tumor samples. Further, we checked the tissue samples for circRNA-104718 using FISH and also observed a very high expression of circRNA-104718 in the HCC tissue samples when compared with relative controls (Figure 1B). Additionally, when we quantified the expression in 68 tumorous tissue samples, we observed that the expression in tumor tissues were highly up-regulated (Figure 1C). To understand the role of circRNA-104718 in the prognosis of HCC patients, we performed the Kaplan–Meier survival curve analysis. It was evident that patients with higher circRNA-104718 expression levels had lower survival rate when compared with the patients with lower expression of circRNA-104718 (Figure 1D). As we had established that there was a strong association between circRNA-104718 and HCC, we further wanted to establish its correlation with other clinic pathological features (age, sex, stage, tumor size, AFP etc.). Even though we could observe certain trends such as increased expression in females than in males, none of these were significantly different. We propose increasing the number of patients in each group could potentially address this difference (Table 1). Clinicopathological analysis showed that high expression of circRNA-104718 was associated with vascular invasion (P=0.00014), which is an important clinical parameter to predict survival, but there was not statistically significant difference in TNM stage (P=0.26150>0.05), serum AFP (P=0.20878>0.05), tumor size (P=0.17462>0.05) and pT stage (P=0.06514>0.05). Thus, our analysis indicates that the high circRNA-104718 is associated with poor prognosis and tumor metastasis.

Increase of circRNA-104718 expression in HCC tissues and its correlation with prognosis of patients

Figure 1
Increase of circRNA-104718 expression in HCC tissues and its correlation with prognosis of patients

(A) The differential expression of circRNA-104718 in HCC tissues and matched non-tumor tissues of 103 patients as indicated. The median expression level of each circRNA-104718 level was indicated by horizontal line in scatter plot figure (T: tumor tissues, N: adjacent non-tumor tissues). (B) FISH was performed to detect the expression of circRNA-104718 in HCC tissue chip (60 pairs of tissue samples and 68 tumor tissues). (C) The intensity of staining was scored in five separate fields for each sample. (D) Kaplan–Meier survival curve indicated the high circRNA-104718 expression is correlated with low survival rates.

Figure 1
Increase of circRNA-104718 expression in HCC tissues and its correlation with prognosis of patients

(A) The differential expression of circRNA-104718 in HCC tissues and matched non-tumor tissues of 103 patients as indicated. The median expression level of each circRNA-104718 level was indicated by horizontal line in scatter plot figure (T: tumor tissues, N: adjacent non-tumor tissues). (B) FISH was performed to detect the expression of circRNA-104718 in HCC tissue chip (60 pairs of tissue samples and 68 tumor tissues). (C) The intensity of staining was scored in five separate fields for each sample. (D) Kaplan–Meier survival curve indicated the high circRNA-104718 expression is correlated with low survival rates.

CircRNA-104718 plays a vital role in HCC’s oncogenesis

To identify a potential cell lines to study the effects of circRNA-104718, we checked their expression levels in multiple HCC cell lines such as SK-Hep1, MHCC-LM3, SMMC-7721, and HepG2 (Figure 2A). It was evident that MHCC-LM3 displayed high expression of circRNA-104718, and we used shRNA to silence circRNA-104718 to understand the effects of its loss in HCCs (Figure 2B). Additionally, we used SK-Hep1 cell line, which had the least expression of circRNA-104718, and overexpressed the circRNA to understand its effects parallelly with that of the shRNA silenced cells. When circRNA-104718 was silenced, the proliferation of cells, analyzed with the help of a CCK-8 test, decreased significantly compared with the control cells. In addition, proliferation increased significantly, when cells were overexpressed for circRNA-107418 (Figure 2C). This indicated a potential positive correlation between circRNA-104718 and proliferation. Further, we also performed migration studies with the aid of wound healing assay, which indicated circRNA-104718 overexpression increased SK-Hep1 cell motility significantly. Whereas, silencing of circRNA-104718 decreased migration of cells when compared with its respective controls (Figure 2D). To understand the role of circRNA-104718 in invasion, we performed Transwell chamber assays and observed increased expression of circRNA-104718, increased invasiveness of SK-Hep1. Additionally, silencing of circRNA-104718 decreased the invasion potential of MHCC-LM3 cells (Figure 2E). We were also interested to elucidate the role of circRNA-104718 in apoptosis. We observed using flow-cytometric analysis of the cell cycle, post overexpression of circRNA-104718, the SK-Hep1 cell had decreased apoptosis (2.79%) when compared with the mock control (8.48%). We also further checked the cell cycle for MHCC-LM3 cells after silencing of circRNA-104718. We observed that the percent of cells undergoing apoptosis increased significantly in the silenced cell to 20.91%, indicating the strong role of circRNA-104718 in regulating the cell cycle (Figure 2F). The above-mentioned evidence indicated the potential role of circRNA-104718 in tumorigenesis through the regulation of proliferation, migration, invasion, and apoptosis in HCC cells.

CircRNA-104718 exerts oncogenic effects in HCC cells

Figure 2
CircRNA-104718 exerts oncogenic effects in HCC cells

(A) The expressions of circRNA-104718 were measured in cell lines SK-Hep1, MHCC-LM3, SMMC-7721, HepG2 using RT-qPCR. (B) The expressions of circRNA-104718 were determined with qPCR transfected either with sh-circ or sh-NC in MHCC-LM3 cell line. (C) The growth curves of cells were measured after transfection with indicated vectors by Cell Counting Kit-8 (CCK-8) assays. (D) Cell motility was examined in cells transfected with the indicated vectors by wound healing assay. (E) Cell invasion assays were performed in cells transfected with the indicated vectors using Transwell chamber with matrigel, respectively. (F) Cell apoptosis was analyzed using flow cytometry after transfection with the indicated vectors. The data are presented as means ± SEM. *P<0.05, **P<0.01, ***P<0.01.

Figure 2
CircRNA-104718 exerts oncogenic effects in HCC cells

(A) The expressions of circRNA-104718 were measured in cell lines SK-Hep1, MHCC-LM3, SMMC-7721, HepG2 using RT-qPCR. (B) The expressions of circRNA-104718 were determined with qPCR transfected either with sh-circ or sh-NC in MHCC-LM3 cell line. (C) The growth curves of cells were measured after transfection with indicated vectors by Cell Counting Kit-8 (CCK-8) assays. (D) Cell motility was examined in cells transfected with the indicated vectors by wound healing assay. (E) Cell invasion assays were performed in cells transfected with the indicated vectors using Transwell chamber with matrigel, respectively. (F) Cell apoptosis was analyzed using flow cytometry after transfection with the indicated vectors. The data are presented as means ± SEM. *P<0.05, **P<0.01, ***P<0.01.

CircRNA-104718 promotes the growth and metastasis of liver cancer xenograft

Next, to understand the role of circRNA-104718 in HCC, we developed a BALB/c nude mice model. Initially, circRNA-104718 was overexpressed in the SK-Hep1 cells, which were further transplanted subcutaneously into the nude mice. Increasing tumor volume was traced through 27 days, it was apparent that the tumors in the circRNA-104718 overexpression group had a clear trend in its increase in volume as early as day 10 post implantation. On the contrary, the mice that were transplanted with MHCC-LM3 cells (silenced for circ104718), showed very slow increase in growth, when compared with its control (Figure 3A). Post 27 days, the mice were killed, and the xenograft tumors were analyzed and characterized. It was evident that the tumors increased significantly in size post 27 days in the group of mice, which had over-expression of circRNA-104718 (Figure 3B,C). Post 27 days, the size of the tumors for mice transplanted with MHCC-LM3 (silenced for circ104718) was reduced significantly when compared with shRNA control (sh-NC) (Figure 3B,C). We assessed the tumors isolated post 27 days of implantation using hematoxylin & eosin (H&E), Ki-67 and TUNEL staining. Based on our observations of the Ki-67 staining it was clear that tumors overexpressing circRNA-104718 had increased staining indicating up-regulated proliferation (Figure 3D). Whereas tumors which were silenced (sh-circRNA-104718) had decreased Ki-67 staining, thereby indicating slower proliferation when compared with the shRNA control (sh-NC) (Figure 3D). Moreover, TUNEL staining indicated tumors overexpressing circRNA-104718 had decreased apoptosis, whereas tumors which had been silenced for circRNA-104718 had higher apoptotic index (Figure 3D).

CircRNA-104718 promotes the growth and metastasis of liver cancer xenograft

Figure 3
CircRNA-104718 promotes the growth and metastasis of liver cancer xenograft

(A) After subcutaneous transplantation of SK-Hep1 and MHCC-LM3 with either overexpression or silencing of the circRNA-104718 in BALB/c nude mice. The growth curves of xenograft tumors were measured every 3 days (volume = width2 × length × 1/2). Points represent the mean tumor volumes of three independent experiments. (B) After 27 days, the xenograft tumors were excised from the nude mice. The relative weights of tumors were analyzed. (C) Representative images of the xenograft tumors obtained post 27 days of subcutaneous transplantation of SK-Hep1 and MHCC-LM3 with either overexpression or silencing of the circRNA-104718 in BALB/c nude mice. (D) Cancer cell (Ki67) and apoptosis (TUNEL) in the xenograft tumors were examined by immunohistochemical staining and quantified with IPP software (E) Representative images of intrahepatic metastasis foci. Red arrows indicate metastatic foci. Luciferase signal intensities of mice over time after tail vein injection with 1 × 106 indicated SK-Hep1 or MHCC-LM3 cell clones. n=6. The data are presented as the means ± SEM. *P<0.05, **P<0.01, ***P<0.001.

Figure 3
CircRNA-104718 promotes the growth and metastasis of liver cancer xenograft

(A) After subcutaneous transplantation of SK-Hep1 and MHCC-LM3 with either overexpression or silencing of the circRNA-104718 in BALB/c nude mice. The growth curves of xenograft tumors were measured every 3 days (volume = width2 × length × 1/2). Points represent the mean tumor volumes of three independent experiments. (B) After 27 days, the xenograft tumors were excised from the nude mice. The relative weights of tumors were analyzed. (C) Representative images of the xenograft tumors obtained post 27 days of subcutaneous transplantation of SK-Hep1 and MHCC-LM3 with either overexpression or silencing of the circRNA-104718 in BALB/c nude mice. (D) Cancer cell (Ki67) and apoptosis (TUNEL) in the xenograft tumors were examined by immunohistochemical staining and quantified with IPP software (E) Representative images of intrahepatic metastasis foci. Red arrows indicate metastatic foci. Luciferase signal intensities of mice over time after tail vein injection with 1 × 106 indicated SK-Hep1 or MHCC-LM3 cell clones. n=6. The data are presented as the means ± SEM. *P<0.05, **P<0.01, ***P<0.001.

Further, a lung metastatic model was generated, by injecting either SK-Hep1 (overexpressed for circRNA-104718) or MHCC-LM3 (silenced for circRNA-104718) through the lateral tail vein of the nude mice. The cells were originally labeled for luciferase, thereby enabling direct in vivo imaging of the lung metastasis. It was clear that SK-Hep1 cells overexpressed for circRNA-104718 had more intrahepatic metastases, compared with the control. Whereas, MHCC-LM3 silenced for circRNA-104718 had fewer metastases when compared with both its control and SK-Hep1 (overexpressed for circRNA-104718) (Figure 3E). Quantitatively, we measured the luciferase activity and it was clear that the mice with overexpression for circRNA-104718 at 4 weeks had increased metastasis when compared with the mice silenced for circRNA-104718 (Figure 3E). Taken together these results further strengthen the view that circRNA-104718 plays a vital role in proliferation, apoptosis, and metastasis of HCC.

CircRNA-104718 targets miR-218-5p and inhibits miR-218-5p activity

To further understand circRNA-104718’s various binding targets; we conducted immunoprecipitation studies in circ104718 overexpressing HEK293T cells using a probe specific to circRNA-104718 and a control probe, respectively (Figure 4A). The putative candidate miRNAs binding to circRNA-104718 were predicted using StarBase (http://starbase.sysu.edu.cn/mirMrna.php). The enrichment of circRNA-104718 and microRNAs (miRs) was detected by qRT-PCR and normalized to the control probe. Based on the evidence, it was clear that many miRs were enriched with circRNA-104718. Among these miR-218-5p was most highly expressed, hence we considered this as our potential candidate. Further, we overexpressed miR-218-5p in HEK293T cells and performed RIP assay with anti-AGO2 antibody. We conducted anti-AGO2 immunoprecipitation (RIP) in cells overexpression miR-218-5p to pull down the circRNA-104718 using anti-AGO2 antibodies or control IgG. We observed that circRNA-104718 pulled down with anti-AGO2 antibodies was significantly enriched in HEK293T cells transfected with miR-218-5p mimics compared with controls, and validated the direct binding of circRNA-104718 with miR-218-5p (Figure 4B). Additionally, we also performed luciferase binding assay, wherein labelling vectors were designed for circRNA-104718 wild-type (WT) and mutant (Mut) vectors (Figure 4C) with appropriate luciferase labelling. And the putative miR-218-5p binding sites in the 3′-UTR of TXNDC5 were predicted using TargetScan (http://www.targetscan.org), the relative luciferase activities were analyzed in 293T cells co-transfected with miR-218-5p mimics or mimic-NC. It was evident, in the presence of miR-218-5p, in the cells expressing circRNA-104718 wild-type, there was a significant decrease in the luciferase activity compared with the control (mimics-NC). Further, in the cells expressing mutant circRNA-104718, there was no decrease in the luciferase activity; indicating miR-218-5p has a strong binding and regulating activity on circRNA-104718 (Figure 4D). To check the reverse effect, we overexpressed circRNA-104718 and checked the miR-218-5p expression. It was clear that overexpression of circRNA-104718 significantly decreased the expression of miR-218-5p in both SK-Hep1 and MHCC-LM3 cells (Figure 4E). Additionally, silencing of circRNA-104718 (sh-circRNA-104718) indeed significantly up-regulated the expression of miR-218-5p levels (Figure 4F). To identify the localization of circRNA-107418 and miR-218-5p, we performed FISH on the tumor and its corresponding controls. It was evident the circRNA-104718 was highly expressed, whereas miR-218-5p was expressed in low levels in tumor tissues, when compared with the control (Figure 4G). The combined merging of these two staining did indicate that miR-218-5p was indeed co-localized with circRNA-104718 in both the tumor and non-tumorous tissues.

CircRNA-104718 targets miR-218-5p and inhibits miR-218-5p activity

Figure 4
CircRNA-104718 targets miR-218-5p and inhibits miR-218-5p activity

(A) circRIP assays was performed in circ104718-overexpressing 293T cells using a circ104718-specific probe and control probe, respectively. The enrichment of circ104718 and microRNAs was detected by qRT-PCR and normalized to the control probe. (B) 293T cells were transfected with miR-218-5p mimic or mimic-NC for 48 h. The association between circ104718 and miR-218-5p was assessed by RIP assay. (C,D) Schematic of circRNA-104718 wild-type (WT) and mutant (Mut) luciferase reporter vectors. The relative luciferase activities were analyzed in 293T cells co-transfected with miR-218-5p mimics or mimic-NC. (E) The expressions of miR-218-5p were analyzed using RT-qPCR in cells transfected with circRNA-104718 or mock vector respectively. (F) The expressions of miR-218-5p were detected using RT-qPCR in cells transfected with sh-circRNA-104718 or sh-NC vector. (G) miR-218-5p co-localized with circRNA-104718 in HCC adjacent non-tumor and tumor tissues was detected by FISH. The data are presented as the means ± SEM. *P<0.05, **P<0.01.

Figure 4
CircRNA-104718 targets miR-218-5p and inhibits miR-218-5p activity

(A) circRIP assays was performed in circ104718-overexpressing 293T cells using a circ104718-specific probe and control probe, respectively. The enrichment of circ104718 and microRNAs was detected by qRT-PCR and normalized to the control probe. (B) 293T cells were transfected with miR-218-5p mimic or mimic-NC for 48 h. The association between circ104718 and miR-218-5p was assessed by RIP assay. (C,D) Schematic of circRNA-104718 wild-type (WT) and mutant (Mut) luciferase reporter vectors. The relative luciferase activities were analyzed in 293T cells co-transfected with miR-218-5p mimics or mimic-NC. (E) The expressions of miR-218-5p were analyzed using RT-qPCR in cells transfected with circRNA-104718 or mock vector respectively. (F) The expressions of miR-218-5p were detected using RT-qPCR in cells transfected with sh-circRNA-104718 or sh-NC vector. (G) miR-218-5p co-localized with circRNA-104718 in HCC adjacent non-tumor and tumor tissues was detected by FISH. The data are presented as the means ± SEM. *P<0.05, **P<0.01.

miR-218-5p interacts and regulates TXNDC5 expression

Further, we identified TXNDC5 (thioredoxin domain-containing protein 5), to be an important target of miR-218-5p. To assess the binding capacity of miR-218-5p, we designed vectors for wild type 3′-UTR of TXNDC5 (TXNDC5 3′-UTR-WT) and a mutant 3′-UTR of TXNDC5 with mutations at the predicted miR-218-5p binding site (TXNDC5 3′-UTR-MUT) and with appropriate luciferase reporter at the 3′-UTR (Figure 5A). In addition, the putative miR-218-5p binding sites in the 3′-UTR of TXNDC5 were predicted using TargetScan (http://www.targetscan.org). Based on the observations it was evident that WT TXNDC5 in the presence of miR-218-5p had a decreased luciferase activity. Whereas, the TXNDC5 with mutant miR-218-5p site showed no decrease in luciferase activity, indicating a strong regulation of TXNDC5 by miR-218-5p (Figure 5B). Additionally, we performed overexpression or inhibition of miR-218-5p and did qRT-PCR analysis of TXNDC5. It was evident that cells overexpressing miR-218-5p showed a down-regulated expression for TXNDC5 (Figure 5C). On the contrary, when we inhibited the miR-218-5p, it was clear that TXNDC5 expression was highly up-regulated in the both SK-Hep1 and MHCC-LM3 cell lines (Figure 5C). We also confirmed this evidence using western blot analysis (Figure 5D). This indicated miR-218-5p could target and regulate TXNDC5 expression.

MiR-218-5p interacts and regulates TXNDC5 expression

Figure 5
MiR-218-5p interacts and regulates TXNDC5 expression

(A) Schematic representation of the wild type 3′-UTR of TXNDC5 (TXNDC5 3′-UTR-WT) and a mutant 3′-UTR of TXNDC5 with mutations at the predicted miR-218-5p binding site (TXNDC5 3′-UTR-MUT). (B) The luciferase reporter vector carrying TXNDC5 3′-UTR-WT or TXNDC5 3′-UTR-MUT or the empty vector was co-transfected with miR-218-5p mimic or mimic-NC. (C) The mRNA expression levels of TXNDC5 by qRT-PCR analysis. (D) Western blot analysis of TXNDC5 protein. The data are presented as the means ± SEM. *P<0.05, **P<0.01, ***P<0.001.

Figure 5
MiR-218-5p interacts and regulates TXNDC5 expression

(A) Schematic representation of the wild type 3′-UTR of TXNDC5 (TXNDC5 3′-UTR-WT) and a mutant 3′-UTR of TXNDC5 with mutations at the predicted miR-218-5p binding site (TXNDC5 3′-UTR-MUT). (B) The luciferase reporter vector carrying TXNDC5 3′-UTR-WT or TXNDC5 3′-UTR-MUT or the empty vector was co-transfected with miR-218-5p mimic or mimic-NC. (C) The mRNA expression levels of TXNDC5 by qRT-PCR analysis. (D) Western blot analysis of TXNDC5 protein. The data are presented as the means ± SEM. *P<0.05, **P<0.01, ***P<0.001.

miR-218-5p inhibits oncogenic effects in HCC cells by targeting TXNDC5 in vitro.

To understand the role of miR-218-5p in oncogenesis, we performed multiple overexpression and inhibition experiments. Western blot analysis was used to determine the TXNDC5 expression (Figure 6A). This data suggested that, when TXNDC5 was overexpressed it could supersede the inhibitory effect of miR-218-5p. In addition, when miR-218-5p was inhibited if TXNDC5 gets silenced, then the TXNDC5 stays highly down-regulated. Further, to check the effect of such inhibition and overexpression on proliferation, we performed CCK-8 assay on both SK-Hep1 and MHCC-LM3 cells (Figure 6B). These observations indicated that overexpression of miR-218-5p decreased proliferation, whereas when cells were co-transfected with both miR-218-5p and TXNDC5, there was a significant increase in proliferation. In addition, when cells were inhibited for miR-218-5p there was significant increase in proliferation. However, when both miR-218-5p and TXNDC5 were silenced, there was a significant decrease in proliferation. Wound healing and invasion assays were also performed to understand the migration abilities of SK-Hep1 and MHCC-LM3 (Figure 6C–E). It was clear that overexpression of miR-218-5p decreased the migration and invasion of cells, whereas overexpression of miR-218-5p and TXNDC5 increased cell migration and invasion. On the contrary, inhibition of mIR-218-5p increased migration and invasion but simultaneous inhibition and silencing of miR-218-5p and TXNDC5 decreased cell migration and invasion. Such trends were also observed with apoptosis (Figure 6F).

MiR-218-5p inhibits oncogenic effects in HCC cells through targeting TXNDC5 in vitro

Figure 6
MiR-218-5p inhibits oncogenic effects in HCC cells through targeting TXNDC5 in vitro

(A) TXNDC5 expressions were evaluated by Western blot analysis after transfection with indicated vectors. SK-Hep1 cells were transfected with miR-218-5p mimics alone or co-transfected with the indicated vectors, and MHCC-LM3 cells were transfected with miR-218-5p inhibitors or co-transfected with the indicated vectors. (B) The indicated HCC cell proliferation was measured by CCK-8 assays on the fourth day after co-transfection. (C–E) Cell motility was examined in cells transfected with the indicated vectors by wound healing assay. Cell invasion assays were performed in cells transfected with the indicated vectors using Transwell chamber with matrigel, respectively. (F) Apoptosis was analyzed using flow cytometry after transfection with the indicated vectors. The data are presented as the means ± SEM. *P<0.05, **P<0.01, ***P<0.01.

Figure 6
MiR-218-5p inhibits oncogenic effects in HCC cells through targeting TXNDC5 in vitro

(A) TXNDC5 expressions were evaluated by Western blot analysis after transfection with indicated vectors. SK-Hep1 cells were transfected with miR-218-5p mimics alone or co-transfected with the indicated vectors, and MHCC-LM3 cells were transfected with miR-218-5p inhibitors or co-transfected with the indicated vectors. (B) The indicated HCC cell proliferation was measured by CCK-8 assays on the fourth day after co-transfection. (C–E) Cell motility was examined in cells transfected with the indicated vectors by wound healing assay. Cell invasion assays were performed in cells transfected with the indicated vectors using Transwell chamber with matrigel, respectively. (F) Apoptosis was analyzed using flow cytometry after transfection with the indicated vectors. The data are presented as the means ± SEM. *P<0.05, **P<0.01, ***P<0.01.

CircRNA-104718 promotes the growth and metastasis of HCC via targeting microRNA-218-5P/TXNDC5 pathway

To understand the interplay between circRNA-104718, miR-218-5p and TXNDC5, luciferase activity was analyzed in 293T cells co-transfected with circRNA-104718 or mock, miR-218-5p mimics or mimic-NC and luciferase reporter vector in TXNDC5 3′-UTR-WT sequence. Interestingly, the presence of circRNA-104718 increases the luciferase activity, thus representing increased TXNDC5 expression. But, additional expression of miR-218-5p decreased the luciferase signal. Strikingly, increased expression of miR-218-5p alone decreased the luciferase activity, confirming that the binding of miR-218-5p in the absence of circRNA-104718 is sufficient to decrease expression of TXNDC5. Overexpression of circRNA-104718 increases the expression of TXNDC5, whereas silencing of circRNA-104718 decreases the TXNDC5 expression (Figure 7A). Additionally, such evidence was also clearly also observed using Western blotting analysis (Figure 7B,C). Other studies were performed wherein cells overexpressing both circRNA-104718 and miR-218-5p showed a decreased TXNDC5 expression. This indicated that miR-218-5p is sufficient to down-regulate TXNDC5 expression potentially by regulating circRNA-104718 expression (Figure 7D). Additionally, we observed overexpression of circRNA-104718 increased proliferation, migration, invasion, and decreased apoptosis of cells (Figure 7E–I). Nevertheless, overexpression of circRNA-104718 and miR-218-5p decreased proliferation, migration, invasion, and increased apoptosis of cells (Figure 7E–I). This further strengthens our evidence of the strong role of circRNA-104718 and miR-218-5p in tumor progression. Further, from the “The Cancer Genome Atlas” (TCGA) sample analysis, we observed TXNDC5 levels were significantly up-regulated in 103 paired HCC patient samples. In addition to the above-mentioned evidence, when we performed correlation analysis between circRNA-104718 and TXNDC5 expression levels in 103 HCC tissues, it was evident that samples that expressed low circRNA-104718 expression also had a majorly low TXNDC5 RNA levels (Figure 7J–L). All the above-mentioned data indicated that circRNA-104718 plays a major role in oncogenesis through its regulation of TXNDC5 and miR-218-5p.

CircRNA-104718 promotes the growth and metastasis of HCC via targeting microRNA-218-5P/TXNDC5 pathway

Figure 7
CircRNA-104718 promotes the growth and metastasis of HCC via targeting microRNA-218-5P/TXNDC5 pathway

(A) The relative luciferase activities were analyzed in 293T cells co-transfected with circRNA-104718 or mock, miR-218-5p mimics or mimic-NC and luciferase reporter vectors TXNDC5 3′-UTR-WT. (B,C) Analyses showed the mRNA and protein levels of TXNDC5 after overexpressing or silencing circRNA-104718. (D) circRNA-104718 could significantly promote the expression of TXNDC5, and the promotion was retarded after co-transfecting with miR-218-5p mimics. (E) Proliferation assessed using a CCK-8 kit in SK-Hep1 cells on the fourth day after co-transfection. (F–H) The wound healing and Transwell invasion assays showed that circRNA-104718 could promote the metastasis and invasion of HCC cells and that the promotion could be blocked by overexpression of miR-218-5p. (I) Apoptosis was measured and the data are presented as the means ± SEM. (J,K) TXNDC5 levels, detected by quantitative RT-PCR, were significantly up-regulated in data from TCGA and 103 paired HCC patient samples (T: tumor tissues, N: adjacent non-tumor tissues). (L) The correlation between the RNA level of circRNA-104718 and the mRNA level of TXNDC5 in 103 HCC tissues. The correlation was measured by Pearson correlation analysis. The data are presented as the means ± SEM. *P<0.05, **P<0.01.

Figure 7
CircRNA-104718 promotes the growth and metastasis of HCC via targeting microRNA-218-5P/TXNDC5 pathway

(A) The relative luciferase activities were analyzed in 293T cells co-transfected with circRNA-104718 or mock, miR-218-5p mimics or mimic-NC and luciferase reporter vectors TXNDC5 3′-UTR-WT. (B,C) Analyses showed the mRNA and protein levels of TXNDC5 after overexpressing or silencing circRNA-104718. (D) circRNA-104718 could significantly promote the expression of TXNDC5, and the promotion was retarded after co-transfecting with miR-218-5p mimics. (E) Proliferation assessed using a CCK-8 kit in SK-Hep1 cells on the fourth day after co-transfection. (F–H) The wound healing and Transwell invasion assays showed that circRNA-104718 could promote the metastasis and invasion of HCC cells and that the promotion could be blocked by overexpression of miR-218-5p. (I) Apoptosis was measured and the data are presented as the means ± SEM. (J,K) TXNDC5 levels, detected by quantitative RT-PCR, were significantly up-regulated in data from TCGA and 103 paired HCC patient samples (T: tumor tissues, N: adjacent non-tumor tissues). (L) The correlation between the RNA level of circRNA-104718 and the mRNA level of TXNDC5 in 103 HCC tissues. The correlation was measured by Pearson correlation analysis. The data are presented as the means ± SEM. *P<0.05, **P<0.01.

Discussion

circRNA has recently gained immense interest in the field of oncology specifically due its deregulated presence in many different cancers, such as gastric cancers [24], colorectal cancer, lung cancer [25], bladder cancer [20], and osteosarcoma [7]. Interestingly in most of these cancers their role is consistently found to be similar, viz proliferation, migration, invasion, epithelial-mesenchymal transition (EMT), and metastasis [26]. Specifically, in HCC many such circRNA have been identified to play an important part in tumorigenesis. CircRNA-001946 (ciRS-7) has been identified to promote cell proliferation and invasion [27,28], whereas circRNA 0005075 has been found to promote cell adhesion [10]. Pro-oncogenic activities such as increased invasion and proliferation by circRNA are achieved by its role as a sponge on miRNAs. One study by Huang et al. [11], had showed that the circRNA-0000130 sponges miR-141-3p and regulates HCC invasion through regulation of epithelial growth factor receptor (EGFR). Many genes and pathways such as NOTCH1 [14], p21 [29], and Wnt/b-catenin pathways [30] which are vital for proliferation and migration are affected by circRNAs. Vascular invasion in HCC is an adverse prognostic clinical factor to predict survival, and has been shown to promote the occurrence of HCC. Tumor size, histological, and vascular invasion and so on are correlate with tumor recurrence and poor survival, understanding the risk for vascular invasion is crucial for assessing the tumor progression [31,32]. In our study, we observed circRNA-104718 to be highly expressed in HCC patients. We discovered that the positive correlation between circRNA-104718 high expression and vascular invasion as well as tumor metastasis, which supported the idea that elevated circRNA-104718 has a poor prognosis and aggressive metastasis in HCC. Then using an in vitro model, we observed circRNA-104718 up-regulation increased proliferation, migration, and invasion of Sk-Hep1 and MHCC-LM3 cells. An in vivo nude mice model was also developed which showed that overexpression of circRNA-104718 increased the tumor size, tumor weight, and metastasis of HCCs. This suggested that circRNAs, specifically circRNA-104718 could be considered as an important prognostic marker for HCC.

Next, we identified through RIP studies, circRNA-104718 binds and regulates miR-218-5p. We also confirmed with the help of luciferase binding assay, circRNA-104718 binds to miR-218-5p and decreases its expression. Further, we also observed they co-localized with each other. MiR-218-5p has been identified previously as an inducer of cartilage destruction through the regulation of PI3K/Akt/mTOR pathway and silencing of miR-218-5p has been identified as a potential treatment strategy for treatment of osteoarthritis [33]. Another study also miR-218 increases osteoblast differentiation through feedforward mechanism of inhibiting Wnt signaling inhibitors [34]. However, other studies have shown that miR-218-5p regulates a gene LHFPL3 and thereby decreases the invasion of glioma cells [35]. A study by Zhu et al., [36] showed that miR-218-5p inhibits cancer cell proliferation and migration by regulating EGFR in non-small-cell lung cancer (NSCLC). Interestingly, in our study as well we observed that overexpression of miR-218-5p decreased proliferation, invasion, and migration of HCC cells. This indicated miR-218-5p could be a potential candidate for development of treatment strategies for HCCs.

Further, we identified circRNA-104718 acts as competitive endogenous RNA on miR-218-5p to allow expression of a gene TXNDC5. This gene has been identified as a member of the protein disulfide isomerase family [37], with functions associated with many diseases such as arthritis [38,39], neurodegenerative disease [40], and cancer [41]. TXNDC5 has been found to be highly expressed in cervix [19], HCC [41], and lung cancer [42]. In our studies, we observed it to be highly expressed in HCC patients and cells. In gastric cancer, TXNDC5 has been identified to promote cell proliferation, migration, and invasion [43]. Another study had shown TXNDC5 to be highly expressed in 62% of NSCLC when compared with control [42]. Potentially, TXNDC5 is considered to be up-regulated in cancer in response to the hypoxic environment created by the cancer micro-environment, as previously a study identified them to be highly expressed in endothelial cells during hypoxic conditions [44]. Thus, it is clear that TXNDC5 could play a vital role in tumor progression in HCC. Further, to elucidate TXNDC5’s role, we performed inhibition of circRNA-104718 and observed down-regulation of TXNDC5, whereas up-regulation of miR-218-5p decreased TXNDC5. We also identified through luciferase activity studies, that miR-218-5p binds and regulates TXNDC5 activity. Thus, this allowed us to identify the circRNA-104718/miR-218-5p/TXNDC5 axis in regulation and progression of tumorigenesis in HCC.

In conclusion, the present study is the first study to identify circRNA-104718 as a competitive endogenous RNA of miR-218-5p and regulates TXNDC5. The present study also established that circRNA-104718 could be a potential diagnostic and prognostic marker for HCC. Regulation of miR-218-5p could be used as a vital treatment strategy to treat HCCs.

Clinical perspectives

  • HCC is among one of the most prevalent causes of liver malignancy and worldwide still stands as the most leading cause for cancer-related deaths. This indicates a strong need for identification and development of various diagnostic and treatment strategies to treat HCC.

  • Current study allowed us to identify the circRNA-104718 /miR-218-5p/TXNDC5 axis in regulation and progression of tumorigenesis in HCC.

  • The present study further established that circRNA-104718 could be a potential diagnostic and prognostic marker for HCC and regulation of miR-218-5p could be used as a vital treatment strategy to treat HCCs.

Competing Interests

The authors declare that there are no competing interests associated with the manuscript.

Authors’ Contribution

J.Y. and M.Y. performed the experiments and drafted the article. B.Z., J.L., and Z.Z. helped perform the research, contributed new reagents/analytic tools and analyzed data. W.Z. and Z.Y. designed research, analyzed data, edited and revised the manuscript, revised and approved the final version of the manuscript.

Funding

The present study was supported by the Shanghai Municipal Planning Commission for advanced technology promotion project [grant no. 16ZR1433000], and National Natural Science Foundation of China [grant no. 81771948].

Abbreviations

     
  • AGO2

    Argonaute2

  •  
  • CCK-8

    Cell Counting Kit-8

  •  
  • cDNA

    complementary deoxyribonucleic acid

  •  
  • ceRNA

    competing endogenous RNA

  •  
  • CircRNA

    circular RNA

  •  
  • EGFR

    epithelial growth factor receptor

  •  
  • EMT

    epithelial-mesenchymal transition

  •  
  • FBS

    fetal bovine serum

  •  
  • FISH

    fluorescence in situ hybridization

  •  
  • GAPDH

    Glyceraldehyde-3-phosphate dehydrogenase

  •  
  • H&E

    hematoxylin & eosin

  •  
  • HCC

    hepatocellular carcinoma

  •  
  • IPP

    Image ProPlus

  •  
  • miR

    microRNA

  •  
  • MRE

    miRNA responsive element

  •  
  • Mut

    mutant

  •  
  • ncRNA

    non-coding RNA

  •  
  • NSCLC

    non-small cell lung cancer

  •  
  • PI

    propidium iodide

  •  
  • qRT-PCR

    quantitative real-time polymerase chain reaction

  •  
  • RIP

    RNA immunoprecipitation

  •  
  • RT-PCR

    real-time polymerase chain reaction

  •  
  • sh-NC

    shRNA Control

  •  
  • TBST

    Tris-buffered saline-Tween 20

  •  
  • TCGA

    The Cancer Genome Atlas

  •  
  • TMA

    tissue microarray

  •  
  • TXNDC

    thioredoxin domain-containing protein 5

  •  
  • WT

    wild-type

References

References
1.
Centers for Disease Control and Prevention (CDC).
(
2010
)
Hepatocellular carcinoma - United States, 2001-2006
.
MMWR Morb. Mortal. Wkly. Rep.
59
,
517
520
[PubMed]
2.
Balogh
J.
,
Victor
D.
,
Asham
E.H.
,
Burroughs
S.G.
,
Boktour
M.
,
Saharia
A.
et al. .
(
2016
)
Hepatocellular carcinoma: a review
.
J. Hepatocell. Carcinoma
3
,
41
53
[PubMed]
3.
Siegel
R.L.
,
Miller
K.D.
and
Jemal
A.
(
2018
)
Cancer statistics, 2018
.
CA Cancer J. Clin.
68
,
7
30
[PubMed]
4.
Ebbesen
K.K.
,
Kjems
J.
and
Hansen
T.B.
(
2016
)
Circular RNAs: identification, biogenesis and function
.
Biochim. Biophys. Acta
1859
,
163
168
[PubMed]
5.
Qu
S.
,
Liu
Z.
,
Yang
X.
,
Zhou
J.
,
Yu
H.
,
Zhang
R.
et al. .
(
2018
)
The emerging functions and roles of circular RNAs in cancer
.
Cancer Lett.
414
,
301
309
[PubMed]
6.
Fu
L.
,
Jiang
Z.
,
Li
T.
,
Hu
Y.
and
Guo
J.
(
2018
)
Circular RNAs in hepatocellular carcinoma: functions and implications
.
Cancer Med.
,
7
,
3101
3109
[PubMed]
7.
Jin
H.
,
Jin
X.
,
Zhang
H.
and
Wang
W.
(
2017
)
Circular RNA hsa-circ-0016347 promotes proliferation, invasion and metastasis of osteosarcoma cells
.
Oncotarget
8
,
25571
25581
[PubMed]
8.
Yao
Z.
,
Luo
J.
,
Hu
K.
,
Lin
J.
,
Huang
H.
,
Wang
Q.
et al. .
(
2017
)
ZKSCAN1 gene and its related circular RNA (circZKSCAN1) both inhibit hepatocellular carcinoma cell growth, migration, and invasion but through different signaling pathways
.
Mol. Oncol.
11
,
422
437
[PubMed]
9.
Qin
M.
,
Liu
G.
,
Huo
X.
,
Tao
X.
,
Sun
X.
,
Ge
Z.
et al. .
(
2016
)
Hsa_circ_0001649: a circular RNA and potential novel biomarker for hepatocellular carcinoma
.
Cancer Biomark.
16
,
161
169
[PubMed]
10.
Shang
X.
,
Li
G.
,
Liu
H.
,
Li
T.
,
Liu
J.
,
Zhao
Q.
et al. .
(
2016
)
Comprehensive circular RNA profiling reveals that hsa_circ_0005075, a new circular RNA biomarker, is involved in hepatocellular crcinoma development
.
Medicine
95
,
e3811
[PubMed]
11.
Huang
X.Y.
,
Huang
Z.L.
,
Xu
Y.H.
,
Zheng
Q.
,
Chen
Z.
,
Song
W.
et al. .
(
2017
)
Comprehensive circular RNA profiling reveals the regulatory role of the circRNA-100338/miR-141-3p pathway in hepatitis B-related hepatocellular carcinoma
.
Sci. Rep.
7
,
5428
[PubMed]
12.
Di Leva
G.
,
Garofalo
M.
and
Croce
C.M.
(
2014
)
MicroRNAs in cancer
.
Ann. Rev. Pathol.
9
,
287
314
[PubMed]
13.
Sun
Z.
,
Han
Q.
,
Zhou
N.
,
Wang
S.
,
Lu
S.
,
Bai
C.
et al. .
(
2013
)
MicroRNA-9 enhances migration and invasion through KLF17 in hepatocellular carcinoma
.
Mol. Oncol.
7
,
884
894
[PubMed]
14.
Fu
L.
,
Chen
Q.
,
Yao
T.
,
Li
T.
,
Ying
S.
,
Hu
Y.
et al. .
(
2017
)
Hsa_circ_0005986 inhibits carcinogenesis by acting as a miR-129-5p sponge and is used as a novel biomarker for hepatocellular carcinoma
.
Oncotarget
8
,
43878
43888
[PubMed]
15.
Zheng
Q.
,
Bao
C.
,
Guo
W.
,
Li
S.
,
Chen
J.
,
Chen
B.
et al. .
(
2016
)
Circular RNA profiling reveals an abundant circHIPK3 that regulates cell growth by sponging multiple miRNAs
.
Nat. Commun.
7
,
11215
[PubMed]
16.
Qiu
L.P.
,
Wu
Y.H.
,
Yu
X.F.
,
Tang
Q.
,
Chen
L.
and
Chen
K.P.
(
2018
)
The emerging role of circular RNAs in hepatocellular carcinoma
.
J. Cancer
9
,
1548
1559
[PubMed]
17.
Drakaki
A.
,
Hatziapostolou
M.
,
Polytarchou
C.
,
Vorvis
C.
,
Poultsides
G.A.
,
Souglakos
J.
et al. .
(
2015
)
Functional microRNA high throughput screening reveals miR-9 as a central regulator of liver oncogenesis by affecting the PPARA-CDH1 pathway
.
BMC Cancer
15
,
542
[PubMed]
18.
Cai
L.
and
Cai
X.
(
2014
)
Up-regulation of miR-9 expression predicate advanced clinicopathological features and poor prognosis in patients with hepatocellular carcinoma
.
Diagn. Pathol.
9
,
1000
[PubMed]
19.
Chang
X.
,
Xu
B.
,
Wang
L.
,
Wang
Y.
,
Wang
Y.
and
Yan
S.
(
2013
)
Investigating a pathogenic role for TXNDC5 in tumors
.
Int. J. Oncol.
43
,
1871
1884
[PubMed]
20.
Zhong
Z.
,
Huang
M.
,
Lv
M.
,
He
Y.
,
Duan
C.
,
Zhang
L.
et al. .
(
2017
)
Circular RNA MYLK as a competing endogenous RNA promotes bladder cancer progression through modulating VEGFA/VEGFR2 signaling pathway
.
Cancer Lett.
403
,
305
317
[PubMed]
21.
Meng
J.
,
Chen
S.
,
Han
J.X.
,
Qian
B.
,
Wang
X.R.
,
Zhong
W.L.
et al. .
(
2018
)
Twist1 regulates vimentin through Cul2 circular RNA to promote EMT in hepatocellular carcinoma
.
Cancer Res.
78
,
4150
4162
[PubMed]
22.
Yu
J.
,
Xu
Q.G.
,
Wang
Z.G.
,
Yang
Y.
,
Zhang
L.
,
Ma
J.Z.
et al. .
(
2018
)
Circular RNA cSMARCA5 inhibits growth and metastasis in hepatocellular carcinoma
.
J. Hepatol.
68
,
1214
1227
[PubMed]
23.
Yuan
J.H.
,
Yang
F.
,
Wang
F.
,
Ma
J.Z.
,
Guo
Y.J.
,
Tao
Q.F.
et al. .
(
2014
)
A long noncoding RNA activated by TGF-beta promotes the invasion-metastasis cascade in hepatocellular carcinoma
.
Cancer Cell
25
,
666
681
[PubMed]
24.
Chen
J.
,
Li
Y.
,
Zheng
Q.
,
Bao
C.
,
He
J.
,
Chen
B.
et al. .
(
2017
)
Circular RNA profile identifies circPVT1 as a proliferative factor and prognostic marker in gastric cancer
.
Cancer Lett.
388
,
208
219
[PubMed]
25.
Zhu
X.
,
Wang
X.
,
Wei
S.
,
Chen
Y.
,
Chen
Y.
,
Fan
X.
et al. .
(
2017
)
hsa_circ_0013958: a circular RNA and potential novel biomarker for lung adenocarcinoma
.
FEBS J.
284
,
2170
2182
[PubMed]
26.
Kristensen
L.S.
,
Hansen
T.B.
,
Veno
M.T.
and
Kjems
J.
(
2018
)
Circular RNAs in cancer: opportunities and challenges in the field
.
Oncogene
37
,
555
565
[PubMed]
27.
Yu
L.
,
Gong
X.
,
Sun
L.
,
Zhou
Q.
,
Lu
B.
and
Zhu
L.
(
2016
)
The circular RNA Cdr1as act as an oncogene in hepatocellular carcinoma through targeting miR-7 expression
.
PLoS ONE
11
,
e0158347
[PubMed]
28.
Yang
X.
,
Xiong
Q.
,
Wu
Y.
,
Li
S.
and
Ge
F.
(
2017
)
Quantitative proteomics reveals the regulatory networks of circular RNA CDR1as in hepatocellular carcinoma cells
.
J. Proteome Res.
16
,
3891
3902
[PubMed]
29.
Han
D.
,
Li
J.
,
Wang
H.
,
Su
X.
,
Hou
J.
,
Gu
Y.
et al. .
(
2017
)
Circular RNA circMTO1 acts as the sponge of microRNA-9 to suppress hepatocellular carcinoma progression
.
Hepatology
66
,
1151
1164
[PubMed]
30.
Li
F.
,
Zhang
L.
,
Li
W.
,
Deng
J.
,
Zheng
J.
,
An
M.
et al. .
(
2015
)
Circular RNA ITCH has inhibitory effect on ESCC by suppressing the Wnt/beta-catenin pathway
.
Oncotarget
6
,
6001
6013
[PubMed]
31.
Han
B.
,
Zheng
Y.
,
Wang
L.
,
Wang
H.
,
Du
J.
,
Ye
F.
et al. .
(
2019
)
A novel microRNA signature predicts vascular invasion in hepatocellular carcinoma
.
J. Cell. Physiol.
1
10
,
[PubMed]
32.
Fransvea
E.
,
Mazzocca
A.
,
Antonaci
S.
and
Giannelli
G.
(
2009
)
Targeting transforming growth factor (TGF)-betaRI inhibits activation of beta1 integrin and blocks vascular invasion in hepatocellular carcinoma
.
Hepatology
49
,
839
850
[PubMed]
33.
Lu
J.
,
Ji
M.L.
,
Zhang
X.J.
,
Shi
P.L.
,
Wu
H.
,
Wang
C.
et al. .
(
2017
)
MicroRNA-218-5p as a potential target for the treatment of human osteoarthritis
.
Mol. Ther.
25
,
2676
2688
[PubMed]
34.
Hassan
M.Q.
,
Maeda
Y.
,
Taipaleenmaki
H.
,
Zhang
W.
,
Jafferji
M.
,
Gordon
J.A.
et al. .
(
2012
)
miR-218 directs a Wnt signaling circuit to promote differentiation of osteoblasts and osteomimicry of metastatic cancer cells
.
J. Biol. Chem.
287
,
42084
42092
[PubMed]
35.
Taipaleenmaki
H.
,
Farina
N.H.
,
van Wijnen
A.J.
,
Stein
J.L.
,
Hesse
E.
,
Stein
G.S.
et al. .
(
2016
)
Antagonizing miR-218-5p attenuates Wnt signaling and reduces metastatic bone disease of triple negative breast cancer cells
.
Oncotarget
7
,
79032
79046
[PubMed]
36.
Liu
Y.
,
Huang
D.
,
Wang
Z.
,
Wu
C.
,
Zhang
Z.
,
Wang
D.
et al. .
(
2016
)
LMO2 attenuates tumor growth by targeting the Wnt signaling pathway in breast and colorectal cancer
.
Sci. Rep.
6
,
36050
[PubMed]
37.
Horna-Terron
E.
,
Pradilla-Dieste
A.
,
Sanchez-de-Diego
C.
and
Osada
J.
(
2014
)
TXNDC5, a newly discovered disulfide isomerase with a key role in cell physiology and pathology
.
Int. J. Mol. Sci.
15
,
23501
23518
[PubMed]
38.
Chang
X.
,
Zhao
Y.
,
Yan
X.
,
Pan
J.
,
Fang
K.
and
Wang
L.
(
2011
)
Investigating a pathogenic role for TXNDC5 in rheumatoid arthritis
.
Arthritis Res. Ther.
13
,
R124
[PubMed]
39.
Chang
X.
,
Cui
Y.
,
Zong
M.
,
Zhao
Y.
,
Yan
X.
,
Chen
Y.
et al. .
(
2009
)
Identification of proteins with increased expression in rheumatoid arthritis synovial tissues
.
J. Rheumatol.
36
,
872
880
[PubMed]
40.
Lin
S.H.
,
Liu
C.M.
,
Liu
Y.L.
,
Shen-Jang Fann
C.
,
Hsiao
P.C.
,
Wu
J.Y.
et al. .
(
2009
)
Clustering by neurocognition for fine mapping of the schizophrenia susceptibility loci on chromosome 6p
.
Genes Brain Behav.
8
,
785
794
[PubMed]
41.
Park
M.S.
,
Kim
S.K.
,
Shin
H.P.
,
Lee
S.M.
and
Chung
J.H.
(
2013
)
TXNDC5 gene polymorphism contributes to increased risk of hepatocellular carcinoma in the Korean male population
.
Anticancer Res.
33
,
3983
3987
[PubMed]
42.
Vincent
E.E.
,
Elder
D.J.
,
Phillips
L.
,
Heesom
K.J.
,
Pawade
J.
,
Luckett
M.
et al. .
(
2011
)
Overexpression of the TXNDC5 protein in non-small cell lung carcinoma
.
Anticancer Res.
31
,
1577
1582
[PubMed]
43.
Wu
Z.
,
Zhang
L.
,
Li
N.
,
Sha
L.
and
Zhang
K.
(
2015
)
An immunohistochemical study of thioredoxin domain-containing 5 expression in gastric adenocarcinoma
.
Oncol. Lett.
9
,
1154
1158
[PubMed]
44.
Sullivan
D.C.
,
Huminiecki
L.
,
Moore
J.W.
,
Boyle
J.J.
,
Poulsom
R.
,
Creamer
D.
et al. .
(
2003
)
EndoPDI, a novel protein-disulfide isomerase-like protein that is preferentially expressed in endothelial cells acts as a stress survival factor
.
J. Biol. Chem.
278
,
47079
47088
[PubMed]

Author notes

*

These authors contributed equally to this work.