hsa_circ_102559 Acts as the Sponge of miR-130a-5p to Promote Hepatocellular Carcinoma Progression Through Regulation of ANXA2.

Circular RNAs (circRNAs) are critical regulators in tumor initiation and development and participate in the pathological process of hepatocellular carcinoma (HCC). However, the specific role and mechanism of circRNA, hsa_circ_102559, in HCC remains elusive. First, analysis of HCC-related circRNA expression profile GSE97332 and HCC patients showed a significant upregulation of hsa_circ_102559 in HCC tissues. Upregulation of hsa_circ_102559 in HCC cells was associated with the metastatic properties. Second, hsa_circ_102559 significantly promoted HCC metastasis, while knockdown of hsa_circ_102559 reversed the promotive effects on HCC progression. Functionally, hsa_circ_102559 could target and colocalize with miR-130a-5p in the cytoplasm of HCC cells. Annexin A2 (ANXA2) was identified as a target gene of miR-130a-5p, and overexpression of ANXA2 counteracted with the suppressive effects of hsa_circ_102559 silence on HCC metastasis. Lastly, xenograft experiment was established and results indicated that knockdown of hsa_circ_102559 inhibited HCC growth and metastasis through the downregulation of ANXA2. In conclusion, hsa_circ_102559 inhibited HCC progression via sponging miR-130a-5p to reduce ANXA2 expression, suggesting that hsa_circ_102559 might be a potential biomarker or therapeutic target for HCC.

Introduction

Hepatocellular carcinoma (HCC) is one of the most common pathological types of primary liver cancer and has been listed as the second leading cause of cancer-related deaths[1]. Chronic hepatitis B virus or hepatitis C virus infection, excessive alcohol consumption, immune-related hepatitis, and obesity are generally considered to be major risk factors for HCC[2,3]. Curative resection and chemotherapy are the main therapeutic strategies for HCC which can improve the survival rates for HCC patients[4]. However, most HCC patients who undergo surgical resection experience degeneration and recurrence within 5 years[5]. Therefore, it is necessary to identify new molecular targets and pathways for treatment, the occurrence, and the development of HCC.

Circular RNAs (circRNAs), historically considered as the product of abnormal splicing[6], are widely presented in eukaryotes and associated with human diseases, including neurodegenerative diseases and cancer[7,8]. CircRNAs are characterized by resistance to the RNA exonuclease cleavage and have a longer half-life than linear RNA, making circRNAs attractive candidates for biomarkers of cancer[9]. The role of circRNAs in tumor pathology is one of the research hotspots in recent years. CircRNAs have been reported as either biomarkers or therapeutic targets in HCC. For example, hsa_circ_0001649, downregulated in HCC, might serve as a new potential biomarker for HCC[10]. CircMTO1 inhibited HCC progression and was recognized as a therapeutic target for HCC[11].

Hsa_circ_102559, generated by splice of transmembrane protein 91 gene and located at chr19:41884185-41884424+, was reported to be enhanced in HCC[12]. However, the detailed role of hsa_circ_102559 on HCC progression remains enigmatic. Functional network formed by circRNA-miRNA-mRNA was reported to be implicated in HCC progression[13]. Hsa_circ_103809-miR-377-3p-FGFR1 axis promotes HCC progression[14], while MTO1 inhibits HCC progression through miR-9-p21 axis[11]. However, the miRNA target of hsa_circ_102559 in HCC is still unclear and needs further investigation. MiR-130a, downregulated in HCC[15], could increase drug resistance in cisplatin-treated HCC cells[16]. HCC proliferation, migration, and invasion were inhibited by miR-130a[17]. MiR-130a-3p could regulate insulin signaling and steatosis in the liver[18] and mediate gemcitabine resistant HCC cell migration and invasion[19]. Whether miR-130a is involved in hsa_circ_102559-mediated HCC progression should be investigated to determine the mechanism of hsa_circ_102559 in HCC.

The present study verified the oncogenic role of hsa_circ_102559 in HCC progression and then determined that the potential miR-130a-5p-mediated mRNA target was involved in HCC progression. These results provided the potential therapeutic target for HCC.

Materials and Methods

Tumor Tissues Collection

Seventy-four paired tumor tissues and adjacent normal tissues from HCC patients with written consents were collected at the First Affiliated Hospital of Wenzhou Medical University. The experiment was approved by the Ethics Committee of the First Affiliated Hospital of Wenzhou Medical University. Patients with chemotherapy or radiotherapy were excluded.

Cell Culture

T84, HCC cell lines (Huh7, HLE, SKHEP-1, Huh6, SNU449) and nontumorigenic normal human hepatocyte cell line (MIHA), from the Chinese Academy of Sciences (Shanghai, China), were preserved in Roswell Park Memorial Institute Medium (RPMI) 1640 medium (Transgene, Beijing, China) containing fetal bovine serum (Gibco, Carlsbad, CA, USA) at 37 °C humidified incubator with 5% carbon dioxide.

Cell Transfection

Full length of hsa_circ_102559 and Annexin A2 (ANXA2) were constructed into pcDNA4.1 (Invitrogen, Carlsbad, CA, USA). Mimics and inhibitor of miR-130a-5p, the corresponding negative controls (NC, NC inh) were synthesized by GenePharma (Shanghai, China). Huh7 or SNU449 cells were transfected with the pcDNA vectors, mimics, or inhibitors via Lipofectamine 3000 (Invitrogen).

For establishment of stable Huh7 cells with hsa_circ_102559 interference, circRNA_102559 sh1# (5′-CACCGCCCCCTGATGTTGAGGAAACAACGGTTTCCTCAACATCAGGGGGCTTTTTG-3′), 2# (5′-CACCGATGTTGAGGAAACCCCTCCTAACGAGGAGGGGTTTCCTCAACATCTTTTTG-3′) or the negative control (shNC) were constructed into pLKO.1 (Biosettia, San Diego, CA, USA), and then co-transfected into HEK-293T cells with psPAX2 and pMD2.G. Specific lentiviruses with 100 multiplicity of infection were harvested 48 h later, and then infected Huh7 cells under 8 mg/ml polybrene. Stable cells were obtained with 5 μg/ml puromycin treatment for 7 days.

Cell Viability

Huh7 or SNU449 cells (1 × 103 cells/well) were seeded and cultured for the indicated time (0, 24, 48, 72, 96 h), followed by addition with 20 µl cell counting kit-8 (CCK8) solution (Dojindo, Tokyo, Japan). Two hours later, the absorbance at 450 nm was determined by Epoch microplate Reader (BioTek, Winooski, VT, USA).

Colony Formation Assay

Huh7 or SNU449 cells (200 cells/well) were seeded and cultured with RPMI 1640 medium for 2 weeks. After removing the supernatant, cells were fixed in 100% methanol and then stained with Giemsa staining liquid. Cell colonies were visualized under a microscope (Olympus, Tokyo, Japan).

Cell Proliferation

Huh7 or SNU449 cells (2 × 104 cells/well) were seeded and cultured in RPMI 1640 medium for 48 h. After incubation with 50 μM EdU for 4 h, cells were fixed by 4% paraformaldehyde and incubated with 0.5% Triton X-100. Cells were observed under a fluorescence microscope (Olympus) before incubation with 500 µl 1× Apollo (EdU Cell Proliferation Assay Kit, Ribobio, Guangzhou, China).

Wound Healing

Huh7 or SNU449 cells (2 × 104 cells/well) were seeded and cultured in RPMI 1640 medium for 24 h. Scratches were made by a plastic pipette tip. The wound width was calculated under a microscope (Olympus) for another 24-h culture.

Transwell Assay

Huh7 or SNU449 cells suspended in serum-free medium were incubated on Matrigel (Shanghai Yeasen Biotechnology Co., Ltd, Shanghai, China)-coated chambers (BD Biosciences, San Jose, CA, USA). Medium with 20% serum was added to the lower chambers. Twenty-four hours later, the invasive cells to the lower chambers were stained with 0.5% methanol-prepared crystal violet and counted under a microscope (Olympus).

Immunofluorescence

Huh7 or SNU449 cells were fixed on glass slides and incubated with specific antibodies, including anti-E-cadherin (Abcam, Cambridge, MA, USA) and anti-N-cadherin (Abcam). After incubation with Alexa Fluor 594-labeled goat antirat IgG antibody, slides were observed under Zeiss axiophot photomicroscope (Carl Zeiss, Oberkochen, Germany) before mounting with 4′,6-diamidino-2-phenylindole (DAPI)-Fluoromount-G (Southern Biotech; Birmingham, AL, USA).

Fluorescence In Situ Hybridization

Cultured Huh7 or SNU449 cells were fixed by 4% paraformaldehyde and treated with 2 μg/ml protease K (Sigma-Aldrich; St. Louis, MO, USA). Cells were hybridized with fluorescent in situ hybridization kit (RiboBio, Guangzhou, China) solution with 300 ng/ml hsa_circ_102559 or miR-130a-5p probes (Invitrogen) overnight at 55 °C. Cells were counterstained with DAPI and photographed by the fluorescent microscope.

Dual-Luciferase Reporter Assay

Sequences of hsa_circ_102559 or 3′-UTR of ANXA2 or the mutant sequences without binding abilities to miR-130a-5p were subcloned into psiCheck2 vector (GenePharma, Suzhou, China). Huh7 or SNU449 cells were co-transfected miR-130a-5p mimics or NC, miR-130a-5p inhibitor or NC inh, with the luciferase reporter vectors. Luciferase activities were performed 48 h after transfection via dual-luciferase reporter assay system (Promega, Madison, WI, USA).

RNA Isolation, RNase R Digestion, Actinomycin D Treatment, and Quantitative Reverse Transcription Polymerase Chain Reaction (qRT-PCR)

Total RNAs from HCC tissues or cells were extracted by Trizol (Invitrogen). Nuclear or cytoplasmic RNAs of Huh7 or SNU449 cells were separated via PARIS Kit (Thermo Fisher, Waltham, MA, USA). miRNAs were extracted via miRcute miRNA isolation kit (Tiangen, Beijing, China).

For the RNase R digestion, 5 µg RNAs were incubated with or without 3 U/ µg RNase R (Epicentre Technologies, Madison, WI, USA) at 37 °C for 15 min. The products were then carried out with qRT-PCR. For actinomycin D treatment, Huh7 or SNU449 cells were incubated with 5 μg/ml actinomycin D at indicated times (0, 4, 8, 12, 24 h). The expressions of hsa_circ_102559 and TMEM91 were determined via qRT-PCR. For the qRT-PCR analysis, RNAs were first reverse-transcribed into cDNAs and then conducted qRT-PCR with SYBR Green Master (Roche, Mannheim, Germany). Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) or U6 were used as endogenous controls. The primer sequences were shown in Table 1.

Table 1.

Primer Sequence.

ID	Sequence(5′-3′)
GAPDH F	CCGGGAAACTGTGGCGTGATGG
GAPDH R	AGGTGGAGGAGTGGGTGTCGCTGTT
hsa_circ_102559 F	TGCAGTATCTGGAGGAAGCA
hsa_circ_102559 R	CCATCTGCTTCTGGAACACC
TMEM91 F	GAGTCACGAAGACGTCCTGCC
TMEM91 R	TGCCCAAAGCCATTACAGTCC
miR-130a-5p F	CCAGGGCTTTTCAAAAATGA
miR-130a-5p R	CCGATCCAATCTGTTCTGGT
ANXA2 F	TGAGCGGGATGCTTTGAAC
ANXA2 R	ATCCTGTCTCTGTGCATTGCTG
U6 F	CTCGCTTCGGCAGCACATA
U6 R	AACGATTCACGAATTTGCGT

Western Blot

Protein lysates were extracted from HCC tissues or cells (30 µg) by radioimmunoprecipitation lysis buffer (Beyotime, Ningbo, China). Lysates were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and electrotransferred onto a nitrocellulose membrane. Membrane was blocked in 5% nonfat milk and incubated overnight with anti-ANXA2, anti-proliferating cell nuclear antigen (PCNA), anti-p21, anti-E-cadherin, anti-N-cadherin, or anti-GAPDH antibodies (Abcam) at 4 °C. After incubation with a horseradish peroxidase-labeled secondary antibody (Abcam), the signals were determined by enhanced chemiluminescence (KeyGen, Nanjing, China).

Tumorigenicity Assay

Animal study was reviewed and approved by the Ethics Committee of First Affiliated Hospital of Wenzhou Medical University and in accordance with the National Institutes of Health Laboratory Animal Care and Use Guidelines. Twelve female BALB/c nude mice (4-weeks, 20 to 25 g), obtained from the laboratory animal center of the Chinese academy of sciences (Shanghai, China), were randomly divided into two groups. Huh7 with circ_102559 sh2# or shNC (5 × 106 cells) suspended in 0.1 ml phosphate-buffered saline were subcutaneously inoculated in the right flank of nude mice. Ten days later, tumor volume was calculated every 2 days. Twenty days later, the mice were sacrificed with 40 mg/kg sodium pentobarbital, and the tumor tissues were dissected, weighted, and photographed.

For metastases analysis, a tail vein injection model with the T84 cells was established. After intraperitoneal injection of 4 mg luciferin (Promega) suspended in 50 µl phosphate-buffered saline for 10 min, the metastases were photographed by IVIS@ Lumina II system (Caliper LifeSciences, Hopkinton, MA, USA). Lung tissues were dissected and subjected to hematoxylin and eosin staining, and the representative images were observed via a microscope (Olympus).

Immunohistochemistry

Formalin-fixed and paraffin-embedded tumor tissue sections were blocked and then incubated overnight with anti-ANXA2, anti-Ki67, anti-E-cadherin, or anti-N-cadherin antibodies (Abcam). After incubation with biotinylated secondary antibody and then Vectastain Elite ABC reagent, the slides were examined under a light microscope (Olympus) with counterstaining with hematoxylin.

Statistical Analysis

Data were expressed as mean ± SD and performed with SPSS 19.0 software. Survival curve was analyzed by the Kaplan-Meier method and the log-rank test. Statistical analysis was determined by a one-way analysis of variance or Student’s t-test. Value of P < 0.05 was considered to be statistically significant.

Results

Upregulation of hsa_circ_102559 in HCC

To determine the differently expressed genes in HCC, HCC-related circRNA expression profile GSE97332 was downloaded from the Gene Expression Omnibus database. Results showed that a total of 133 differently expressed circRNAs were found in GSE97332 with 84 upregulated and 49 downregulated (Fig. 1A). A total of 20 circRNAs, including 10 enhanced and 10 reduced, were found to be in the top rankings (Fig. 1B). Among the enhanced circRNAs, a new circRNA, hsa_circ_102559 was significantly upregulated in HCC (Fig. 1B). Analysis of HCC patients revealed that hsa_circ_102559 was increased in HCC tissues compared with adjacent normal tissues (Fig. 1C). Moreover, patients with high hsa_circ_102559 expression demonstrated lower overall survival than low hsa_circ_102559 expression (Fig. 1D), and high hsa_circ_102559 expression was closely associated with tumor size (P = 0.032), lymph node metastasis (P = 0.017), and TNM stage (P = 0.030) (Table 2), suggesting that hsa_circ_102559 might be related to metastatic property of HCC. Analysis of HCC cells also revealed that hsa_circ_102559 was increased in HCC cells compared with MIHA (Fig. 1E). RNase R digestion confirmed the circular nature of hsa_circ_102559 in both Huh7 or SNU449 (Fig. 1F). Moreover, following actinomycin D treatment in both SNU449 and Huh7 cells, the half-life of hsa_circ_102559 was longer than that of TMEM91 (supplemental Figure S1), further suggesting the circular characteristic of hsa_circ_102559. Nuclear or cytoplasmic RNAs from Huh7 or SNU449 cells were separated and conducted with qRT-PCR analysis. Results showed that hsa_circ_102559 was mainly distributed in the cytoplasm of HCC cells (Fig. 1G), suggesting that hsa_circ_102559 might regulate nuclear genes through sponging of miRNAs.

Fig. 1.

Upregulation of hsa_circ_102559 in HCC. (A) Volcano plot showed the differentially expressed circRNAs between HCC and normal tissues from the GSE97332 dataset. (B) Heatmap for the 10 upregulated and 10 downregulated differentially expressed circRNAs in HCC determined using the RobustRankAggreg method with an adjust P < 0.05. (C) The expression of hsa_circ_102559 in HCC tissues and adjacent noncancer tissues detected by qRT-PCR (N = 74). (D) Overall survival analysis of HCC patients with high hsa_circ_102559 expression and low levels of hsa_circ_102559. (E) The expression of hsa_circ_102559 in HCC cell lines and MIHA detected by qRT-PCR. (F) Hsa_circ_102559 was resistant to RNase R digestion compared to linear TMEM91 in both SNU449 and Huh7 cells. (G) Expression of hsa_circ_102559, GAPDH, and U6 in cytoplasm or nucleus of SNU449 and Huh7 cells detected by qRT-PCR, suggesting the mainly cytoplasm subcellular localization of hsa_circ_102559. circRNA: circular RNA; GAPDH: glyceraldehyde 3-phosphate dehydrogenase; HCC: hepatocellular carcinoma; qRT-PCR: quantitative reverse transcription-polymerase chain reaction.

Table 2.

Association Between Clinical Features and hsa_circ_102559 Expression of Hepatocellular Carcinoma Patients (n = 74).

Clinical features	Number of patients	hsa_circ_102559 expression		P value
		Low (<median)	High (≥median)
Number	74	37	37
Age (years)
<Mean (55)	51	24	27	0.451
≥Mean (55)	23	13	10
Sex
Male	39	20	19	0.816
Female	35	17	18
Tumor size
<5 (cm)	29	19	10	0.032*
≥5 (cm)	45	18	27
Lymph node metastasis
Absent	19	14	5	0.017*
Present	55	23	32
Tumor, node, metastasis
I/II	27	18	9	0.030*
III/IV	47	19	28
Hepatitis B virus infection
Yes	40	22	18	0.351
No	34	15	19

Hsa_circ_102559 Promoted Malignant Behavior of HCC

To investigate function role of hsa_circ_102559 in HCC progression, SNU449 was transfected with pcDNA-hsa_circ_102559, and Huh7 was transfected with specific shRNAs (circ102559 sh1# or 2#) (Fig. 2A). Data from CCK8 (Fig. 2B) and colony formation assay (Fig. 2C) indicated that hsa_circ_102559 promoted cell viability and proliferation of SNU449 cell, while knockdown of hsa_circ_102559 suppressed the cell viability and proliferation. In addition, EdU-positive cells in SNU449 cells transfected with pcDNA-hsa_circ_102559 were more than the control (Fig. 2D), while EdU-positive cells in Huh7 cells transfected with circ102559 sh2# were less than shNC (Fig. 2D), confirming the promotive role of hsa_circ_102559 on HCC proliferation. Furthermore, the metastatic capacities of HCC were suppressed by circ102559 sh2# (Fig. 3A, B), while promoted by pcDNA-hsa_circ_102559 (Fig. 3A, B). Protein expression of E-cadherin was decreased in SNU449 cells transfected with pcDNA-hsa_circ_102559 while enhanced in Huh7 cells transfected with circ102559 sh2# (Fig. 3C). However, hsa_circ_102559 showed an opposite effect on protein expression of N-cadherin compared with E-cadherin (Fig. 3C). Therefore, hsa_circ_102559 demonstrated promotive effects on the malignant behavior of HCC.

Fig. 2.

Hsa_circ_102559 promoted cell proliferation of HCC. (A) Transfection efficiency of pcDNA-hsa_circ_102559 in SNU449 cells, and circ_102559 sh1# or 2# in Huh7 cells was detected by qRT-PCR. (B) The influence of hsa_circ_102559 on cell viability of SNU449 and Huh7 cells detected by CCK8. (C) The influence of hsa_circ_002144 on cell proliferation of SNU449 and Huh7 cells detected by colony formation assay. (D) The influence of hsa_circ_002144 on cell proliferation of SNU449 and Huh7 cells detected by EdU staining. **P < 0.01, ***P < 0.001. CCK8: cell counting kit 8; HCC: hepatocellular carcinoma; qRT-PCR: quantitative reverse transcription-polymerase chain reaction.

Fig. 3.

Hsa_circ_102559 promoted cell migration and invasion of HCC. (A) The influence of hsa_circ_102559 on cell migration of SNU449 and Huh7 cells detected by wound healing assay. (B) The influence of hsa_circ_102559 on cell invasion of SNU449 and Huh7 cells detected by transwell assay. (C) The influence of hsa_circ_102559 on protein expression of E-cadherin and N-cadherin in SNU449 and Huh7 cells.

hsa_circ_102559 Bind to miR-130a-5p

The sponging miRNA of hsa_circ_102559 was predicated as miR-130a-5p (Fig. 4A), and fluorescence in situ hybridization confirmed the cytoplasmic distribution of hsa_circ_102559 and miR-130a-5p in HCC (Fig. 4B). To determine the binding ability between hsa_circ_102559 and miR-130a-5p, luciferase reporter assay was conducted. Results showed that luciferase activity of psiCheck2-hsa_circ_102559 wild-type luciferase reporter vector was substantially enhanced in SNU449 and Huh7 cells transfected with miR-130a-5p inhibitor, while reduced by miR-130a-5p mimics (Fig. 4C). However, the luciferase activity of psiCheck2-hsa_circ_102559 mutant-type luciferase reporter vector did not show any obvious changes in either miR-130a-5p mimics or inhibitor (Fig. 4C). Moreover, miR-130a-5p expression was decreased by pcDNA-hsa_circ_102559, while increased by circ102559 sh2# (Fig. 4C). Therefore, hsa_circ_102559 could bind to miR-130a-5p and inhibit its expression in HCC.

Fig. 4.

hsa_circ_102559 bind to miR-130a-5p. (A) The potential binding targets of hsa_circ_102559 predicted as miR-130a-5p via circular RNA Interactome (https://circinteractome.nia.nih.gov/). (B) Subcellular localization of hsa_circ_102559 and miR-130a-5p in SNU449 and Huh7 cells via RNA-FISH. (C) The influence of miR-130a-5p mimics or inhibitor on luciferase activities of psiCheck2-hsa_circ_102559 wild-type or psiCheck2-hsa_circ_102559 mutant-type in SNU449 and Huh7 cells. The influence of hsa_circ_102559 on miR-130a-5p expression in SNU449 and Huh7 cells.

miR-130a-5p Bind to ANXA2

Similarly, ANXA2 was predicted as a potential target of miR-130a-5p (Fig. 5A). MiR-130a-5p inhibitor increased the luciferase activity of the ANXA2 wild-type luciferase reporter vector, while miR-130a-5p mimics decreased the luciferase activity (Fig. 5B). However, both miR-130a-5p mimics and inhibitor had no obviously effect on luciferase activity of ANXA2 mutant-type luciferase reporter (Fig. 5B). Protein expression of ANXA2 was enhanced by miR-130a-5p inhibitor while reduced by miR-130a-5p mimics (Fig. 5C), suggesting that miR-130a-5p could bind to ANXA2 and inhibited its expression in HCC. Moreover, upregulation of ANXA2 in HCC tissues (Fig. 5D, E) revealed a significant positive correlation with hsa_circ_102559 (P < 0.0001) (Fig. 5F) and negative correlation with miR-130a-5p (P = 0.0002) (Fig. 5F) in HCC patients.

Fig. 5.

miR-130a-5p bind to ANXA2. (A) The potential miR-130a-5p binding targets predicted as ANXA2 via Targetscan (http://www.targetscan.org/vert_71/). (B) The influence of miR-130a-5p mimics or inhibitor on luciferase activities of psiCheck2-ANXA2 wild-type or psiCheck2-ANXA2 mutant-type in SNU449 and Huh7 cells. (C) The influence of miR-130a-5p on protein expression of ANXA2 in SNU449 and Huh7 cells detected by western blot. (D) Protein expression of ANXA2 in HCC tissues and adjacent noncancer tissues detected by western blot. (E) The expression of ANXA2 in HCC tissues and adjacent noncancer tissues detected by qRT-PCR (N = 74). (F) Negative correlation between miR-130a-5p and ANXA2 in HCC patients. Positive correlation between hsa_circ_102559 and ANXA2 in HCC patients. ANXA2: Annexin 2; HCC: hepatocellular carcinoma; qRT-PCR: quantitative reverse transcription-polymerase chain reaction.

Hsa_circ_102559 Promoted Malignant Behavior of HCC Through miR-130a-5p-Mediated ANXA2

To determine hsa_circ_102559/miR-130a-5p axis on ANXA2, HCC cells were co-transfected with pcDNA-hsa_circ_102559 and miR-130a-5p mimics or circ102559 sh2# with miR-130a-5p inhibitor. Overexpression of hsa_circ_102559 increased protein expression of ANXA2 (Fig. 6A), while hsa_circ_102559 silence decreased the expression (Fig. 6A). MiR-130a-5p mimics or inhibitor reversed the effect of pcDNA-hsa_circ_102559 or circ102559 sh2# on ANXA2 expression, respectively (Fig. 6A), confirming the regulatory role of hsa_circ_102559/miR-130a-5p axis on ANXA2. Rescue experiment was then conducted in Huh7 cells co-transfected with circ102559 sh2# and pcDNA-ANXA2. Overexpression of ANXA2 counteracted the suppressive effects of circ102559 sh2# on cell viability (Fig. 6B), proliferation (Fig. 6C), migration (Fig. 6D), and invasion (Fig. 6E) of HCC. In addition, knockdown of hsa_circ_102559 decreased protein expression of ANXA2, PCNA, and N-cadherin while increased the expression of p21 and E-cadherin (Fig. 6F). However, miR-130a-5p inhibitor reversed the effects of hsa_circ_102559 on protein expression of ANXA2, PCNA, N-cadherin, p21, and E-cadherin (Fig. 6F). Taken together, hsa_circ_102559 promoted the malignant behavior of HCC through miR-130a-5p-mediated ANXA2.

Fig. 6.

Hsa_circ_102559 promoted malignant behavior of HCC through miR-130a-5p-mediated ANXA2. (A) The influence of hsa_circ_102559 and miR-130a-5p on protein expression of ANXA2 in SNU449 and Huh7 cells detected by western blot. (B) The influence of hsa_circ_102559 and ANXA2 on cell viability of Huh7 cells detected by CCK8. (C) The influence of hsa_circ_102559 and ANXA2 on cell proliferation of Huh7 cells detected by EdU staining. (D) The influence of hsa_circ_102559 and ANXA2 on cell migration of Huh7 cells detected by wound healing assay. (E) The influence of hsa_circ_102559 and ANXA2 on cell invasion of Huh7 cells detected by transwell assay. (F) The influence of hsa_circ_102559 and miR-130a-5p on protein expression of ANXA2, PCNA, p21, E-cadherin, and N-cadherin in Huh7 cells detected by western blot. ANXA2: Annexin 2; CCK8: cell counting kit 8; PCNA: proliferating cell nuclear antigen.

Knockdown of hsa_circ_102559 Suppressed HCC Tumor Growth and Metastasis

To assess the effect of hsa_circ_102559 on in vivo HCC tumor growth, the establishment of a xenograft mouse model was conducted with the injection of Huh7 cells with stably silence of hsa_circ_102559. Hsa_circ_102559 and ANXA2 were significantly reduced while miR-130a-5p was enhanced in circ102559 sh2# group mice compared with the shNC group (Fig. 7A). Injection of Huh7 cells with stably silence of hsa_circ_102559 suppressed tumor growth, indicated by decreased tumor volume and weight (Fig. 7B). Moreover, metastasis of HCC was also inhibited by circ102559 sh2# (Fig. 7C), and mice injected with Huh7 cells with stably silence of hsa_circ_102559 showed less number of pulmonary metastatic nodules than shNC (Fig. 7D). These results indicated that knockdown of hsa_circ_102559 suppressed HCC tumor growth and metastasis. Immunohistochemical analysis showed decreased ANXA2, Ki-67, and N-cadherin, while increased E-cadherin, in tissues from mice injected with Huh7 cells with stably silence of hsa_circ_102559 compared with shNC (Fig. 7E).

Fig. 7.

Knockdown of hsa_circ_102559 suppressed HCC tumor growth and metastasis. (A) Expression of hsa_circ_102559, miR-130a-5p, and ANXA2 in tissues from mice in circ102559 sh2# and shNC groups. (B) The effect of circ102559 sh2# on HCC tumor growth in xenograft tumor mice. The tumor volume and weight were calculated. (C) Represent bioluminescence images of mice in circ102559 sh2# and shNC groups. (D) Hematoxylin and eosin staining of lung tissues revealed morphologic features of lung tissues from mice in circ102559 sh2# and shNC groups. (E) Immunohistochemistry staining was used to determine the expression of ANXA2, Ki-67, E-cadherin, and N-cadherin in tissues from mice in circ102559 sh2# and shNC groups. ANXA2: Annexin 2; HCC: hepatocellular carcinoma.

Discussion

The high incidence of metastasis and recurrence and the rapid tumor progression lead to unsatisfactory clinical treatment for HCC[20]. Molecular targeted drugs have attracted much attention currently for the treatment of HCC[21]. Considering the close relation between circRNAs and human diseases, including tumorigenesis[22], circRNAs are the hotspot due to the potential role in HCC[23]. Although hsa_circ_102559 was found to be enhanced in HCC[12], the role and mechanism of hsa_circ_102559-mediated HCC growth require further exploration.

Early diagnosis of HCC is particularly important for the patients. For example, downregulation of hsa_circ_0005986[24], hsa_circ_0003570[25], and hsa_circ_0004018[26] in HCC samples could distinguish HCC from liver cirrhosis and chronic hepatitis. Here, analysis of HCC-related circRNA expression profile GSE97332 and HCC patients showed that hsa_circ_102559 was significantly upregulated in HCC and associated with poor overall survival. Besides, hsa_circ_102559 was also correlated with tumor size, lymph node metastasis, and TNM stage, suggesting that hsa_circ_102559 might function as a reliable biomarker for HCC. However, the potential value of hsa_circ_102559 as a diagnostic or prognostic biomarker for HCC needs to be further deciphered.

CircRNAs could participate in the regulatory network of competitive endogenous RNA during HCC development[27]. The oncogenic role of hsa_circ_102559 on HCC progression was validated in this study. Knockdown of hsa_circ_102559 could either suppress in vitro HCC progression or repress in vivo tumor growth. Moreover, both in vitro and in vivo experiments revealed that proteins involved in epithelial-mesenchymal transition were implicated in hsa_circ_102559-mediated HCC progression. Knockdown of hsa_circ_102559 increased protein expression of E-cadherin while decreased N-cadherin to inhibit epithelial-mesenchymal transition of HCC. Epithelial-mesenchymal transition could be associated with physiological and pathophysiological events of HCC, including tumor invasion and cancer cell stemness[28]. Inhibition of epithelial-mesenchymal transition by reducing N-cadherin and enhancing E-cadherin could prevent HCC progression[29,30]. Therefore, these results present the implication of hsa_circ_102559 for the carcinogenesis of HCC.

CircRNAs could sequestrate miRNAs to fine-tune expression of target genes during HCC development[31]. The validated hsa_circ_102559-miR-130a-5p-ANXA2 axis was clarified in this study. Targeted inhibition of miR-130a-5p could promote the progression of breast cancer[32], esophageal squamous cell carcinoma[33], and glioma[34]. MiR-130a could repress HCC progression through the downregulation of Rho-kinase 2[17]. Our results also revealed that hsa_circ_102559 promoted HCC progression through inhibition of miR-130a-5p, and miR-130a-5p could target ANXA2. ANXA2, a calcium-dependent phospholipid-binding protein, belonging to the Annexin family, could regulate cellular growth[35]. Abnormal expression of ANXA2 in HCC tissues provided diagnostic value of ANXA2 in HCC[36], and upregulated ANXA2 was regarded as a discriminative serological candidate for the early diagnosis of HCC[37] or predicted poor prognosis in HCC[38]. This study indicated that ANXA2 was upregulated in HCC tissues and positively correlated with hsa_circ_102559 in HCC patients. Moreover, ANXA2 could bind to STAT3 to promote epithelial to mesenchymal transition during HCC metastasis[39], thereby enhancing the malignant behavior of HCC through remodeling the cell motility[40] or upregulation of regulatory T cells for immune escape of HCC[41]. ANXA2 could induce inflammation in the development of HCC[42], and inhibition of ANXA2 could inhibit HCC growth through regulation of cell cycle[43]. Functional assays in this study revealed that overexpression of ANXA2 counteracted the suppressive effect of hsa_circ_102559 interferences on HCC progression. Knockdown of hsa_circ_102559 suppressed in vivo HCC growth via downregulation of ANXA2. Therefore, hsa_circ_102559 acts as the sponge of miR-130a-5p to promote HCC progression through the regulation of ANXA2. Phosphorylation of ANXA2 promotes breast cancer cell proliferation and invasion through activation of STAT3[44]. The downstream signaling pathways involved in hsa_circ_102559/miR-130a-5p/ANXA2-mediated HCC progression needs to be further investigated.

In summary, as shown in Fig. 8, upregulated hsa_circ_102559 in HCC inhibited the expression of miR-130a-5p, thus promoting ANXA2 expression to induce epithelial-mesenchymal transition, cell proliferation, and migration of HCC. This finding provided the potential application of hsa_circ_102559 in HCC.

Fig. 8.

Potential mode of action of the hsa_circ_102559-miR-130a-5p-ANXA2 network in HCC. (A) Upregulated hsa_circ_102559 in HCC inhibited expression of miR-130a-5p, thus promoting ANXA2 expression to induce epithelial-mesenchymal transition, cell proliferation, and migration of HCC.

Supplemental Fig. S1. Expression of hsa_circ_102559 and TMEM91 in SNU449 and Huh7 cells following actinomycin D treatment at indicated times (0, 4, 8, 12, 24 h).

Click here for additional data file.

Supplemental Material, figs1 for hsa_circ_102559 Acts as the Sponge of miR-130a-5p to Promote Hepatocellular Carcinoma Progression Through Regulation of ANXA2 by Junjian Li, Zhengpin Yu, Qiandong Zhu, Chonglin Tao and Qigang Xu in Cell Transplantation