circRNA hsa_circ_104566 Sponged miR-338-3p to Promote Hepatocellular Carcinoma Progression.

Circular RNAs (circRNAs) could sponge micro-RNAs (miRNAs) to regulate tumor progression of hepatocellular carcinoma (HCC). Hsa_circ_104566 contributes to papillary thyroid carcinoma progression. However, the tumorigenic mechanism of hsa_circ_104566 on HCC remains enigmatic. The role of hsa_circ_104566 on HCC was therefore evaluated in this study. First, the high expression of hsa_circ_104566 was found in HCC tissues, which was significantly associated with poor prognosis in HCC patients. Second, Hsa_circ_104566 promoted HCC progression by decreasing apoptosis and E-cadherin, while increasing cell viability, proliferation, migration, invasion, and N-cadherin. On the other hand, HCC progression was suppressed by knockdown of hsa_circ_104566. Hsa_circ_104566 could target miR-338-3p, and its expression was negatively correlated with miR-338-3p in HCC patients. Moreover, miR-338-3p suppressed protein expression of Forkhead box protein 1 (FOXP1) and had a negative correlation with FOXP1 in HCC patients. Functional assay further indicated that the promotion of HCC progression by hsa_circ_104566 was reversed by miR-338-3p, and miR-338-3p inhibitor could counteract the effect of hsa_circ_104566 knockdown on the suppression of HCC progression. In vivo assay indicated that hsa_circ_104566 knockdown suppressed HCC tumor growth and metastasis. In conclusion, hsa_circ_104566 sponged miR-338-3p to promote HCC progression, providing a potential therapeutic target for cancer intervention.

Introduction

Hepatocellular carcinoma (HCC) is a common pathological type of primary liver cancer and ranks as the second leading cause of cancer-related death[1]. Excessive alcohol consumption or chronic hepatitis B virus or hepatitis C virus infection is considered as the major risk factor for HCC[2,3]. Multiple therapies, including surgical resection and systemic chemical therapies, have improved the survival rates for HCC patients. However, these treatments are unsatisfactory due to the recurrence and metastasis of HCC cells[4]. Moreover, HCC patients are always diagnosed at an advanced stage owing to the lack of effective biomarkers. Therefore, it is necessary to identify the molecular targets and new pathways for the development of HCC.

Circular RNAs (circRNAs) are endogenous RNAs discovered in recent years and historically considered as the product of abnormal splicing[5]. CircRNAs are widely present in eukaryotes[6] and resistant to RNAase, thus, showing longer half-life than linear RNA and making them attractive candidates for biomarkers in cancer[7]. Dysregulated circRNAs have been identified in HCC[8], and circRNAs could function as biomarkers in HCC. For example, hsa_circ_0001649, downregulated in HCC, served as a new potential biomarker for HCC[9]. Furthermore, circRNAs could either promote cell invasion of HCC[10] or inhibit HCC tumor growth[11]. In-depth researches on more circRNAs and the underlying mechanisms in HCC progression may be necessary and could provide clinical value.

Hsa_circ_104566 is a newly identified circRNA produced by a splice of the PSD3 gene[12]. Hsa_circ_104566 was downregulated in gastric cancer[12], while upregulated in papillary thyroid carcinoma[13]. Moreover, hsa_circ_104566 promoted papillary thyroid carcinoma progression[14]. Generally, circRNAs contain micro-RNA (miRNA) response elements and function as miRNA sponges to interact with and mediate miRNA expression[15]. Considering that miRNAs could target the 3′ untranslated region (UTR) of mRNAs[16], circRNA–miRNA–mRNA network is important in the regulation of tumor progression. Hsa_circ_104566 could sponge miR-885-5p to activate RAC1 for papillary thyroid carcinoma progression[14]. The sponging miRNA and target mRNA involved in hsa_circ_104566-regulated HCC progression remain elusive.

This study verified the oncogenic role of hsa_circ_104566 in HCC; the potential miRNAs and mRNAs targets were then identified. In addition, this study revealed that hsa_circ_104566 functioned as a competing endogenous RNA (ceRNA) to enhance HCC progression, providing a potential therapeutic target for HCC.

Materials and Methods

Tumor Tissue Collection

Protocols were approved by the Ethics Committee of The First Hospital of Jilin University. Eighty-seven pairs of HCC and adjacent nontumorous tissues were collected from patients who underwent hepatectomy at The First Hospital of Jilin University. Patients with written informed consent were assigned with stage I, II, III, and IV based on tumor node metastasis (TNM) classification.

Cell Culture

HCC cell lines (SK-HEP-1, HLE, SNU449, Hep-3B, Huh7) and MIHA normal liver cell line, obtained from the Chinese Academy of Sciences (Shanghai, China), were cultured in Roswell Park Memorial Institute Medium (RPMI) 1640 medium (Lonza, Basel, Switzerland) supplemented with 10% fetal bovine serum (FBS; Gibco, Waltham, MA, USA) at 37 °C constant temperature and incubated with 5% carbon dioxide.

RNase R Digestion

RNAs from SK-HEP-1 and Huh7 cells were isolated via Trizol (Invitrogen, Carlsbad, CA, USA). RNAs (5 μg) were then incubated with RNase R (3 U/μg; Epicentre Technologies, Madison, WI, USA) at 37 °C for 20 min. The products were purified and identified via quantitative real-time polymerase chain reaction (qRT-PCR).

Cell Transfection

Hsa_circ_104566 and FOXP1 full-length cDNA were constructed to pcDNA4.1 (Invitrogen). MiR-338-3p mimics, inhibitor, and negative controls (miR-NC, NC inh) were synthesized by GenePharma (Suzhou, China). SK-HEP-1 or Huh7 were transfected with the vectors, mimics, or inhibitors via Lipofectamine 2000 (Invitrogen).

Small hairpin RNAs targeting hsa_circ_104566 (shcircRNA 1# or 2#) or the negative control (Scramble) were inserted to pLKO.1 (Biosettia, San Diego, CA, USA) and then co-transfected with psPAX2 and pMD2.G into HEK-293 T cells. The lentiviruses were harvested 2 days after transfection. SK-HEP-1 was infected with lentiviruses under 8 mg/mL polybrene with ViraPower Packaging Mix (Thermo Fisher, Waltham, MA, USA). Stable cell lines were then obtained after 7 days of treatment with puromycin (5 μg/ml).

Cell Viability Assay

Two days after transfection, SK-HEP-1 or Huh7 cells were seeded and incubated with 20 µl cell counting kit-8 (CCK8) solution (Dojindo, Tokyo, Japan) for 2 h. Absorbance at 450 nm was then determined at 24-h intervals (0, 24, 48, 72, and 96 hours).

Cell Proliferation Assay

Two days after transfection, SK-HEP-1 or Huh7 cells were seeded and incubated with 50 µM EdU (Ribobio, Guangzhou, China) for 8 h. Cells were fixed and then permeabilized with Triton X-100. After incubation with Apollo Staining reaction liquid and staining with 4′,6-diamidino-2-phenylindole, cells were visualized under a fluorescent microscope (Olympus inverted microscope IX71, Tokyo, Japan).

Flow Cytometer

Two days after transfection, SK-HEP-1 or Huh7 cells were harvested and resuspended with 100 µl KeyGEN binding buffer (Jiangning, Nanjing, China) containing 1 U/ml ribonuclease and 0.5 µg propidium iodide (PI). Thirty minutes later, cells were incubated with 5 µl fluorescein isothiocyanate conjugated annexin V. Cells were analyzed by fluorescence-activated cell sorting flow cytometer (Attune, Life Technologies, Darmstadt, Germany). The apoptosis rate was calculated as apoptotic cells (Q2 + Q3)/total cells × 100%.

Wound Healing

Two days after transfection, SK-HEP-1 or Huh7 cells were seeded, and then scratched with a plastic pipette tip. After removing the detached cells, cells were cultured in RPMI 1640 medium for 24 h. Wound width was calculated under an inverted microscope.

Transwell Assay

Two days after transfection, SK-HEP-1 or Huh7 cells were cultured with the serum-free medium in the top chamber (BD Biosciences, Bedford, MA, USA) loaded with Matrigel. RPMI 1640 with 10% FBS was added to the bottom chamber. Medium in the top chamber and the filters were removed 8 h later. Invasive cells were fixed in 100% methanol 24 h later and stained with 0.1% crystal violet. Cells were counted under a microscope (Olympus).

Immunofluorescence

Two days after transfection, SK-HEP-1 or Huh7 cells were fixed in 100% methanol and permeabilized with 0.3% Triton X-100. Cells were incubated overnight with the following antibodies: anti-E-cadherin (1:200; Abcam, Cambridge, MA, USA) and anti-N-cadherin (1:200; Abcam) at 4 °C after blocking in 5% bovine serum albumin. After incubation with anti-mouse immunoglobulin G (IgG)-Alexa Fluor488 secondary antibody for 2 h, cells were imaged under a microscope (Olympus).

Luciferase Reporter Assay

Wildtype or mutant sequences of hsa_circ_104566 or FOXP1 3′-UTR were inserted to pmirGLO luciferase reporter vector (Promega, Madison, WI, USA). SK-HEP-1 or Huh7 cells were seeded and co-transfected miR-338-3p mimics or miR-NC with the luciferase reporter vectors. Luciferase activities were performed 48 h after transfection.

Fluorescence In Situ Hybridization (FISH)

Fixed and permeabilized SK-HEP-1 or Huh7 cells were hybridized with fluorescence-conjugated probes targeting hsa_circ_104566 or miR-338-3p (40 nM) at 37 °C for 5 h and then imaged under laser scanning confocal microscopy (Carl Zeiss, Jena, Germany).

qRT-PCR

RNAs from HCC tissues or cell lines were isolated with Trizol; RNAs in cytoplasm and nucleus were separated via PARIS Kit (Thermo Fisher). miRNAs were extracted via miRcute miRNA isolation kit (Tiangen, Beijing, China). RNAs were reverse-transcribed into cDNAs, and qRT-PCR was conducted with SYBR Green Master (Roche, Mannheim, Germany). Glyceraldehyde 3-phosphate dehydrogenase or U6 was used as endogenous controls. The primer sequences were shown in Table S1.

Western Blot

Proteins extracted from SK-HEP-1 or Huh7 cells (30 µg) were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and electrotransferred onto polyvinylidene difluoride membrane. Membrane was incubated overnight with anti-FOXP1, anti-proliferating cell nuclear antigen (PCNA), anti-c-Myc, anti-Bcl-2, anti-cleaved caspase-3, anti-E-cadherin, anti-N-cadherin, or anti-β-actin antibodies (Abcam) at 4 °C and then incubated with horseradish peroxidase (HRP)-labeled secondary antibody (1:5000; Abcam). Immunoreactivities were determined by enhanced chemiluminescence (KeyGen, Nanjing, China).

Mouse Xenograft Assay

The animal study was approved by the Animal Ethics Committee of The First Hospital of Jilin University in accordance with the National Institutes of Health Laboratory Animal Care and Use Guidelines. Twelve female, 4-week-old BALB/c nude mice (20 to 25 g) were randomly separated into two groups. SK-HEP-1 cells (5 × 106) in 100 µl PBS with stable hsa_circ_104566 knockdown or Scramble were subcutaneously injected into the right flank of nude mice. Mice were sacrificed with 40 mg/kg sodium pentobarbital 20 days later. Tumor tissues and lung tissues were isolated and weighted. The extracted RNAs and proteins were used for other analyses.

Hematoxylin and Eosin (H&E) Staining

Fixed and paraffined lung tissues were sectioned into 7-µm sections. After deparaffinizing in xylene, the sections were stained with H&E after hydration in serially diluted ethanol. Representative images were observed via a microscope (Olympus).

Immunohistochemistry

Sectioned HCC tissues from patients or mice were incubated with 3% hydrogen peroxide and immersed with 0.05% Tween 20 in Tris-EDTA buffer. After incubation in 4% dry milk and 0.3% goat serum, the sections were incubated overnight with anti-FOXP1, anti-Ki67, anti-E-cadherin, or anti-N-cadherin primary antibodies (Abcam). Slides were imaged under a light microscope (Olympus) after incubation with HRP goat antirabbit IgG secondary antibody.

Statistical Analysis

Data were presented as mean ± SEM and processed by GraphPad Prism software. The statistical analyses were determined by one-way analysis of variance, and the survival curves were evaluated via the Kaplan–Meier method and log-rank test. Values of P < 0.05, P < 0.01, or P < 0.001 were considered to be statistically significant.

Results

Hsa_circ_104566 was Upregulated in HCC

Hsa_circ_104566 was upregulated in HCC tissues (Fig. 1A). A significant upregulation of hsa_circ_104566 was verified in high pathological grade (TNM III and IV) compared with low grade (TNM I and II) (Fig. 1A), suggesting that hsa_circ_104566 might be related to the metastatic property of HCC. Moreover, patients with high expression of hsa_circ_104566 demonstrated shorter survival than patients with low expression (Fig. 1B). Further analysis between hsa_circ_104566 and clinicopathological features of HCC patients showed that high expression of hsa_circ_104566 was dramatically related to tumor size, TNM stage, and lymph node involvement (Table 1), confirming that hsa_circ_104566 contributed to metastasis of HCC. The upregulation of hsa_circ_104566 was also verified in HCC cell lines (Fig. 1C). The circular nature of hsa_circ_104566 was validated by RNase R treatment in Huh7 and SK-HEP-1 cells (Fig. 1D). qRT-PCR showed that hsa_circ_104566 was expressed in both cytoplasm and nucleus of HCC cells and mainly expressed in the cytoplasm (Fig. 1E). These results showed the possible relationship between hsa_circ_104566 and HCC progression.

Fig. 1.

Expression of hsa_circ_104566 was upregulated in HCC. (A) The expression levels of circRNA hsa_circ_104566 in HCC tissues, including high pathological grade (TNM III and IV) and low grade (TNM I and II), and adjacent noncancer tissues measured by qRT-PCR (N = 87). (B) Overall survival analysis of HCC patients with high and low levels of hsa_circ_104566 expression. (C) The expression levels of hsa_circ_104566 in HCC cell lines (SK-HEP-1, HLE, SNU449, Hep-3B, and Huh7) and MIHA measured by qRT-PCR. (D) Hsa_circ_104566 was resistant to RNase R digestion compared with linear PSD3 in both Huh7 and SK-HEP-1 cells. (E) Expression levels of hsa_circ_104566, GAPDH, and U6 in cytoplasm or nucleus of Huh7 and SK-HEP-1 cells measured by qRT-PCR, suggesting the mainly cytoplasm subcellular localization of hsa_circ_104566. circRNA: circular RNA; GAPDH: glyceraldehyde 3-phosphate dehydrogenase; HCC: hepatocellular carcinoma; qRT-PCR: quantitative real-time polymerase chain reaction; TNM: tumor node metastasis.

Table 1.

Relationship Between hsa_circ_104566 and Clinicopathological Parameters.

Parameters	Number of patients	hsa_circ_104566 expression		P value
		Low (<median)	High (≥median)
Number	87	43	44
Gender
Male	45	23	22	0.745
Female	42	20	22
Age (years)
≥Mean (55)	38	18	20	0.735
<Mean (55)	49	25	24
Tumor size
<5 (cm)	49	30	19	0.012
≥5 (cm)	38	13	25
Tumor node metastasis
I and II	46	28	18	0.024
III and IV	41	15	26
Lymph node involvement
Absent	37	26	11	0.001
Present	50	17	33
Hepatitis B virus infection
Yes	46	26	20	0.161
No	41	17	24

Hsa_circ_104566 Promoted Malignant Behaviors of HCC

Huh7 with hsa_circ_104566 overexpression and SK-HEP-1 with stable hsa_circ_104566 knockdown were generated to evaluate the potential role of hsa_circ_104566 in the regulation of HCC progression (Fig. 2A). Cell viability of Huh7 was increased by hsa_circ_104566 overexpression, while shcircRNA 1# or 2# decreased cell viability of SK-HEP-1 (Fig. 2B). Similarly, proliferation was also promoted by hsa_circ_104566 overexpression and decreased by shcircRNA 1# or 2# (Fig. 2C). Moreover, apoptosis was suppressed by hsa_circ_104566 overexpression and promoted by shcircRNA 1# or 2# (Fig. 2D), suggesting that hsa_circ_104566 contributed to the proliferation of HCC cells.

Fig. 2.

Hsa_circ_104566 promoted cell proliferation of hepatocellular carcinoma. (A) Efficiency of pcDNA-hsa_circ_104566 in Huh7 cells and sh-hsa_circ_104566 1# or 2# in SK-HEP-1 cells was measured by quantitative real-time polymerase chain reaction. (B) The influence of hsa_circ_104566 on cell viability of Huh7 and SK-HEP-1 cells measured by the cell counting kit-8. (C) The influence of hsa_circ_104566 on proliferation of Huh7 and SK-HEP-1 cells detected by the EdU staining assay. (D) The influence of hsa_circ_104566 on apoptosis of Huh7 and SK-HEP-1 cells detected by flow cytometry.

Hsa_circ_104566 promoted cell migration (Fig. 3A) and invasion (Fig. 3B) of Huh7. On the other hand, hsa_circ_104566 knockdown inhibited cell migration (Fig. 3A) and invasion (Fig. 3B). Moreover, proteins involved in the metastatic property of HCC were evaluated by immunofluorescence. The result showed that overexpression of hsa_circ_104566 decreased E-cadherin while increased N-cadherin in Huh7 (Fig. 3C). Knockdown of hsa_circ_104566 in SK-HEP-1 cells caused opposite effects on E-cadherin and N-cadherin, further suggesting that hsa_circ_104566 contributed to malignant phenotypes of HCC.

Fig. 3.

Hsa_circ_104566 promoted cell migration and invasion of hepatocellular carcinoma (HCC). (A) The influence of hsa_circ_104566 on cell migration of Huh7 and SK-HEP-1 cells measured by wound healing assay. (B) The influence of hsa_circ_104566 on cell invasion of Huh7 and SK-HEP-1 cells measured by Transwell assay. (C) The influence of hsa_circ_104566 on protein expression of E-cadherin and N-cadherin in Huh7 and SK-HEP-1 cells measured by immunofluorescence.

Hsa_circ_104566 Bound miR-338-3p in HCC

miR-338-3p was identified as the sponging miRNA for hsa_circ_104566 via the Circular RNA Interactome (https://circinteractome.nia.nih.gov/) (Fig. 4A). Furthermore, luciferase reporter assay indicated that the luciferase activity was decreased in Huh7 cells co-transfected miR-338-3p mimics and wild-type hsa_circ_104566 (Fig. 4B). However, co-transfection with mutant hsa_circ_104566 showed no significant changes in the luciferase activity (Fig. 4B). Moreover, the FISH analysis confirmed that hsa_circ_104566 was co-localized with miR-338-3p in the cytoplasm of HCC cells (Fig. 4C). In addition, hsa_circ_104566 functioned as a ceRNA to inhibit miR-338-3p expression (Fig. 4D). Downregulation of miR-338-3p was verified in HCC tissues (Fig. 4E), and the negative correlation between hsa_circ_104566 and miR-338-3p in HCC patients (Fig. 4F) confirmed that hsa_circ_104566 bound miR-338-3p.

Fig. 4.

Hsa_circ_104566 bound miR-338-3p in HCC cells. (A) miR-338-3p was predicted as the potential binding target of hsa_circ_104566 via circular RNA Interactome (https://circinteractome.nia.nih.gov/). (B) The influence of miR-338-3p mimics on luciferase activities of pmirGLO-wt-hsa_circ_104566 or pmirGLO-mut-hsa_circ_104566 in Huh7 and SK-HEP-1 cells. (C) Subcellular localization of hsa_circ_104566 and miR-338-3p in Huh7 and SK-HEP-1 cells examined via RNA-fluorescence in situ hybridization. (D) The influence of hsa_circ_104566 on miR-338-3p expression in Huh7 and SK-HEP-1 cells. (E) The expression of miR-338-3p in HCC tissues and adjacent noncancer tissues measured by qRT-PCR (N = 87). (F) Negative correlation between miR-338-3p and hsa_circ_104566 in HCC patients. HCC: hepatocellular carcinoma; qRT-PCR: quantitative real-time polymerase chain reaction.

miR-338-3p Bound FOXP1 in HCC Cells

FOXP1 was identified as a potential target for miR-338-3p via Targetscan (http://www.targetscan.org/vert_72/) (Fig. 5A). Luciferase activity was decreased in Huh7 co-transfected miR-338-3p mimics and wild-type FOXP1 (Fig. 5B). Co-transfection of mutant FOXP1 showed no significant changes in the luciferase activity (Fig. 5B). In addition, miR-338-3p could decrease protein expression of FOXP1, while miR-338-3p inhibitor enhanced FOXP1 expression (Fig. 5C). Immunohistochemistry (Fig. 5D) and qRT-PCR (Fig. 5E) demonstrated that FOXP1 was upregulated in HCC tissues. The positive correlation between hsa_circ_104566 and FOXP1, as well as a negative correlation between miR-338-3p and FOXP1 (Fig. 5F), confirmed that miR-338-3p bound FOXP1 in HCC cells.

Fig. 5.

miR-338-3p bound FOXP1 in HCC cells. (A) FOXP1 was predicted as a potential miR-338-3p binding targets via Targetscan (http://www.targetscan.org/vert_71/). (B) The influence of miR-338-3p mimics on luciferase activities of pmirGLO-wt-FOXP1 or pmirGLO-mut-FOXP1 in Huh7 and SK-HEP-1 cells. (C) The influence of hsa_circ_104566 on protein expression of FOXP1 in Huh7 and SK-HEP-1 cells measured by western blot. (D) Immunohistochemical analysis of FOXP1 in HCC tissues and adjacent noncancer tissues. (E) The mRNA expression levels of FOXP1 in HCC tissues and adjacent noncancer tissues measured by qRT-PCR (N = 87). (F) Negative correlation between miR-338-3p and FOXP1 in HCC patients. Positive correlation between hsa_circ_104566 and FOXP1 in HCC patients. (G) Protein expression levels of SOX12 in clear cell renal cell carcinoma tissues and adjacent noncancer tissues measured by western blot. FOXP1: forkhead box protein 1; HCC: hepatocellular carcinoma; qRT-PCR: quantitative real-time polymerase chain reaction.

Hsa_circ_104566 Promoted Malignant Behaviors of HCC Through miR-338-3p/FOXP1

The miR-338-3p suppressed the pcDNA-hsa_circ_104566-induced increase of FOXP1 in Huh7, while the miR-338-3p inhibitor suppressed shcircRNA-induced decrease of FOXP1 in SK-HEP-1 (Fig. 6A). SK-HEP-1 cells were then co-transfected with shcircRNA 1# and pcDNA-FOXP1 to evaluate the potential function of the hsa_circ_104566/FOXP1 axis in the regulation of HCC progression. Decrease in cell viability (Fig. 6B) and cell proliferation (Fig. 6C) induced by hsa_circ_104566 knockdown was reversed by FOXP1 overexpression. In addition, FOXP1 overexpression attenuated hsa_circ_104566 knockdown-induced apoptosis (Fig. 6D). Besides, FOXP1 overexpression counteracted the inhibitory effects of hsa_circ_104566 knockdown on cell migration (Fig. 6E) and invasion (Fig. 6F). Overexpression of FOXP1 also weakened the decrease of FOXP1, PCNA, c-Myc, Bcl-2, and N-cadherin and increase of cleaved caspase-3 and E-cadherin induced by hsa_circ_104566 knockdown (Fig. 6G). Taken together, hsa_circ_104566 promoted malignant behaviors of HCC through miR-338-3p/FOXP1.

Fig. 6.

Hsa_circ_104566 promoted malignant behaviors of HCC through miR-338-3p/FOXP1. (A) The influence of hsa_circ_104566 and miR-338-3p on protein expression of FOXP1 in Huh7 and SK-HEP-1 cells measured by western blot. (B) The influence of hsa_circ_104566 and FOXP1 on cell viability of Huh7 and SK-HEP-1 cells measured by cell counting kit-8. (C) The influence of hsa_circ_104566 and FOXP1 on cell proliferation of Huh7 and SK-HEP-1 cells measured by the EdU staining assay. (D) The influence of hsa_circ_104566 and FOXP1 on cell apoptosis of Huh7 and SK-HEP-1 cells measured by flow cytometry. (E) The influence of hsa_circ_104566 and FOXP1 on cell migration of Huh7 and SK-HEP-1 cells measured by wound healing assay. (F) The influence of hsa_circ_104566 and FOXP1 on cell invasion of Huh7 and SK-HEP-1 cells measured by Transwell assay. (G) The influence of hsa_circ_104566 and FOXP1 on protein expression levels of FOXP1, PCNA, c-Myc, Bcl-2, cleaved caspase-3, E-cadherin, and N-cadherin in Huh7 and SK-HEP-1 cells measured by western blot. FOXP1: forkhead box protein 1; HCC: hepatocellular carcinoma; PCNA: proliferating cell nuclear antigen.

Knockdown of hsa_circ_104566 Attenuated Tumor Growth

SK-HEP-1 with stable transfection of shcircRNA 1# or Scramble was injected into nude mice to further investigate the role of hsa_circ_104566 on HCC. After the injection, hsa_circ_104566 was decreased, while miR-338-3p was increased in shcircRNA 1# group mice compared with the Scramble group (Fig. 7A). Tumor growth was inhibited by shcircRNA 1#, as demonstrated by the fact that both tumor volume and weight were smaller in shcircRNA 1# group mice than the Scramble group (Fig. 7B). Moreover, the number of pulmonary metastatic nodules was decreased in shcircRNA 1# group mice compared with the Scramble group (Fig. 7C and 7D). As shown in Fig. 7E, protein expression of FOXP1, Ki-67, and N-cadherin was markedly reduced, and E-cadherin was notably enhanced in tissues from mice in the shcircRNA 1# group compared with that in Scramble group. These results showed that hsa_circ_104566 knockdown attenuated tumor growth and metastasis.

Fig. 7.

Knockdown of hsa_circ_104566 attenuated tumor growth. (A) Expression levels of hsa_circ_104566 or miR-338-3p in tissues from mice in shcircRNA 1# and Scramble groups. (B) The effect of sh-hsa_circ_104566 on hepatocellular carcinoma tumor growth in xenograft tumor mice. The tumor volume and weight were calculated. (C) Representative images of lung tissues from mice in shcircRNA 1# and Scramble groups. (D) Hematoxylin and eosin staining revealed morphologic features of lung tissues from mice in shcircRNA 1# and Scramble groups. The lung metastases have been shown by the arrows. (E) Immunohistochemistry staining was used to determine the expression of FOXP1, Ki-67, E-cadherin, and N-cadherin in tissues from mice in shcircRNA 1# and Scramble groups. Scale bar: 200 μm.

Discussion

Tumor cells are inclined to recurrence and metastasis with resistance to apoptosis and abnormal proliferation, migration, and invasion[17]. CircRNAs could promote or inhibit tumor cell progression, and they have been regarded as potential therapeutic targets for cancer intervention, especially in HCC[18]. Here, a novel upregulated circRNA in HCC tissues: hsa_circ_104566, was verified to be a critical regulator in HCC progression.

Noncoding RNAs function as potential biomarkers for the diagnosis or prognosis of HCC[19]. circRHOT1, upregulated in HCC tissues, was shown to predict poor prognosis in HCC[20]. Here, this study identified higher expression of hsa_circ_104566 in high pathological grade than a low grade, suggesting a positive correlation between hsa_circ_104566 and TNM stage of HCC. Moreover, high hsa_circ_104566 expression was associated with shorter overall survival than low expression in HCC patients, suggesting that hsa_circ_104566 might function as a potential biomarker for prognosis of HCC.

Hsa_circ_0001649 functions as a biomarker for HCC and inhibits HCC progression[21], revealing a close relationship between biomarkers and therapeutic targets in HCC. Therefore, circRNAs may not only contribute to the diagnosis of HCC as potential prognosis biomarkers but also be regarded as therapeutic targets of HCC[18]. Consistent with the previous report that upregulated hsa_circ_104566 promoted papillary thyroid carcinoma progression[14], the present study showed that inhibition of hsa_circ_104566 suppressed cell progression, while induced apoptosis of HCC. Oncogenic role of hsa_circ_104566 in HCC was associated with the promotion of cell proliferation via targeting PCNA and c-Myc. Knockdown of hsa_circ_104566 decreased protein expression of PCNA[22] and c-Myc[23] to inhibit cell proliferation of HCC. Moreover, expression levels of proteins important for apoptosis[24] were also determined: Bcl-2 was reduced, while cleaved caspase-3 was enhanced by hsa_circ_104566 knockdown in HCC. In addition, epithelial to mesenchymal transition, important for HCC development[25], was restrained by knockdown of hsa_circ_104566 with the enhancement of E-cadherin and reduction of N-cadherin. In vivo tumor model study indicated that hsa_circ_104566 knockdown suppressed HCC tumor growth and decreased the number of pulmonary metastatic nodules, suggesting potential clinical application in HCC.

The underlying mechanism involved in hsa_circ_104566-mediated HCC progression remains to be clarified. MiR-338-3p was validated as a sponging miRNA for hsa_circ_104566 in HCC. Huang et al. has shown that miR-338 was reduced in HCC[26], and downregulation of miR-338-3p could promote angiogenesis in HCC[27]. Moreover, miR-338-3p inhibited cell invasion[28], epithelial to mesenchymal transition[29], and cell growth[30] of HCC. Results in this study revealed that miR-338-3p was also reduced in HCC tissues and co-localized with hsa_circ_104566 in the cytoplasm of HCC. Besides, miR-338-3p could target genes involved in cell metabolism to regulate HCC progression[31,32]. FOXP1, as a transcription factor to regulate cell proliferation and metabolism, was identified as a target gene of miR-338-3p. FOXP1 was validated as an oncogene in HCC[33], whose downregulation could inhibit cell proliferation of HCC[34]. Results in this study revealed that FOXP1, increased in HCC tissues, could counteract the suppressive effect of hsa_circ_104566 knockdown on HCC progression. CircRNA-miR-338-3p-mRNA network has been reported to participate in HCC progression[35,36]. Considering that miR-338-3p could also target FOXP4 during HCC development[37], whether other proteins of the FOXP family or other functional proteins were involved in hsa_circ_104566/miR-338-3p axis-mediated HCC progression should be investigated in further studies.

In summary, hsa_circ_104566 functioned as an oncogene in HCC, whose knockdown inhibited HCC progression via miR-338-3p/FOXP1 axis. This finding provided the potential application of hsa_circ_104566 in HCC treatment.

Click here for additional data file.

Supplemental Material, Table_S1_Primer_sequence for circRNA hsa_circ_104566 Sponged miR-338-3p to Promote Hepatocellular Carcinoma Progression by Guangming Liu, Wei Guo, Min Rao, Junjie Qin, Feng Hu and Ke Li in Cell Transplantation