CircRNA LRIG3 knockdown inhibits the progression of hepatocellular carcinoma by regulating miR-223-3p/MAP2K6 axis and inactivating MAPK/ERK pathway

Background: Emerging evidence suggests that circular RNAs (circRNAs) play critical roles in tumorigenesis. However, the roles and molecular mechanism of circRNA leucine-rich repeat immunoglobulin domain-containing protein 3 (circ_LRIG3) in hepatocellular carcinoma (HCC) have not been investigated. Methods: The expression levels of circ_LRIG3, microRNA-223-3p (miR-223-3p), and mitogen-activated protein kinase kinase 6 (MAP2K6) were determined by quantitative real-time polymerase chain reaction (qRT-PCR). Flow cytometry was applied to determine the cell cycle distribution and cell apoptosis. Cell proliferation, migration and invasion were assessed by methylthiazolyldiphenyl-tetrazolium bromide (MTT) assay and transwell assay, respectively. Western blot assay was employed to measure the protein levels of snail, E-cadherin, MAP2K6, mitogen-activated protein kinase (MAPK), phospho-MAPK (p-MAPK), extracellular signal-regulated kinases (ERKs), and phospho-ERKs (p-ERKs). The relationship between miR-223-3p and circ_LRIG3 or MAP2K6 was predicted by bioinformatics tools and veried by dual-luciferase reporter assay. A xenograft tumor model was established to conrm the functions of circ_LRIG3 in vivo. Results: Circ_LRIG3 and MAP2K6 expression were enhanced while miR-223-3p abundance was reduced in HCC tissues and cells. Knockdown of circ_LRIG3 inhibited the progression of HCC cells via reducing cell proliferation, metastasis and increasing apoptosis. MiR-223-3p was a target of circ_LRIG3 and its downregulation reversed the inhibitory effect of circ_LRIG3 knockdown on progression of HCC cells. Moreover, MAP2K6 could bind to miR-223-3p, and MAP2K6 upregulation also abolished the suppressive with miR-223-3p, anti-miR-NC, or anti-miR-223-3p. (G and H) The mRNA and protein levels of MAP2K6 were examined in HCC cells (Hep3B and Huh7) and THLE-2 cells by qRT-PCR and western blot analyses, respectively. (I and J) QRT-PCR and western blot assays were conducted to measure the mRNA and protein levels of MAP2K6 in HCC tissues and normal tissues, respectively. (K) The association between miR-223-3p abundance and MAP2K6 mRNA level was analyzed in HCC tissues. (L and M) The mRNA and protein levels of MAP2K6 were detected in Hep3B and Huh7 cells transfected with si-NC, si-circ_LRIG3, si-circ_LRIG3 + anti-miR-NC, or si-circ_LRIG3 + anti-miR-223-3p by qRT-PCR and western blot analyses, respectively.

and more stable than linear RNAs [5,6]. Up to now, some studies have shown that circRNAs are extensively expressed in many types of cells and participated in the progression and development of diverse cancers, including HCC [7,8]. For instance, circRNA Cdr1as served as an oncogene in HCC via regulating miR-7 expression [9]. CircRNA cSMARCA5 could restrain HCC cell growth and metastasis [10].
In addition, circRNA leucine-rich repeat and immunoglobulin domain-containing protein 3 (circ_LRIG3; hsa_circ_0027345, chr12:59277301-59308117) has been reported to be overexpressed in HCC tissues [11]. Nevertheless, the functional roles and molecular mechanism of circ_LRIG3 in HCC progression have not been clari ed.
It is widely acknowledged that circRNAs can modulate gene expression via acting as miRNA sponges in eukaryotes, which is one of main mechanisms of physiological and pathological processes [12].
MicroRNAs (miRNAs), a class of non-coding RNAs (~ 22 nucleotides), play regulatory roles in disease through interaction with mRNAs [13]. MiR-223 has been identi ed to play an anti-cancer role in HCC and it might be a possible therapeutic target for treating HCC [14]. However, the connection between circ_LRIG3 and miR-223-3p has not been reported. It has been suggested that mitogen-activated protein kinase kinase 6 (MAP2K6) can serve as a critical regulator in promoting tumorigenesis [15]. However, the role of MAP2K6 in HCC cells has not been reported. In our research, we rst investigated the associations among miR-223-3p, circ_LRIG3 and MAP2K6 in HCC cells.
Here, we measured miR-223-3p, circ_LRIG3 and MAP2K6 expression in HCC tissues and cells, and determined their functions in HCC cells. Besides, we probed the circ_LRIG3/miR-223-3p/MAP2K6 regulatory network in the progression of HCC. The aim of our study was to offer a new insight into the diagnosis and treatment of HCC.

Specimens collection
In our research, HCC tissues (n=46) and adjacent normal tissues (n=46) were acquired from patients who underwent surgery at Laiyang Central Hospital Of Yantai City. These tissues were harvested and timely frozen in liquid nitrogen, and then preserved at -80℃ until the experiments were performed. These subjects did not receive any treatment and provided informed consents.
Quantitative real-time polymerase chain reaction (qRT-PCR) Trizol reagent (Invitrogen) was utilized to obtain total RNA from tissue samples and cells. For detecting genes expression, Prime Script RT reagent Kit (Takara, Dalian, China) and TaqMan MicroRNA Reverse Transcription Kit (Thermo Fisher Scienti c, Waltham, MA, USA) were used to synthesize the rst strand of complementary DNA (cDNA). All reactions were performed on the ABI 7300 system (Thermo Fisher Scienti c) using SYBR Green PCR kit (Thermo Fisher Scienti c). Primers for circ_LRIG3, LRIG3, miR-223-3p, MAP2K6, U6, and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) were exhibited as followed: MAP2K6, or miR-223-3p expression was assessed using the 2 -ΔΔCt method and standardized by GAPDH or U6, respectively.

RNase R and Actinomycin D treatment
To assess the stability of circ_LRIG3 and its linear isoform (LRIG3), dimethyl sulfoxide solution (DMSO) or actinomycin D (2 mg/mL) was added to cultured medium. RNase R (3 U/μg, Epicentre Technologies, Madison, WI, USA) was utilized to incubate the total RNA (2 μg) at 37℃ for 30 min. After treatment with RNase R or Actinomycin D, these cells were collected and then subjected to qRT-PCR for detecting the expression levels of circ_LRIG3 and LRIG3. Subcellular fractionation location PARIS Kit (Life Technologies Corp., Grand Island, NY, USA) was employed to isolate cytosolic and nuclear fractions. In brief, Hep3B and Huh7 cells were carefully washed by phosphate-buffered saline (PBS) and placed on the ice. Subsequently, these cells were re-suspended in fractionation buffer and centrifuged at 500 × g at 4℃ for 5 min. Subsequently, the cytoplasmic fraction would be separated from the nuclear pellet. After that, the remaining nuclear pellet was again lysed by cell disruption buffer as nuclear fraction. Lastly, the abundance of U6, GAPDH and circ_LRIG3 was examined by qRT-PCR in the nuclear and cytoplasmic fractions. GAPDH and U6 were served as controls for the cytoplasmic and nuclear, respectively.

Cell cycle assay
Hep3B and Huh7 cells were collected following transfection for 48 h, and xed by ice-cold ethanol (70%) at −20℃ for 24 h. Afterward, these cells were centrifuged and washed with PBS, followed by staining with 25 μg/mL propidium iodide (PI ) solution in PBS supplemented with Triton X-100 (0.2%) and RNase A(50 μg/mL) for 20 min in the dark. Lastly, ow cytometry (Guava Technologies, Hayward, CA, USA) was employed to examine the distribution of cell cycle.

Cell proliferation assay
Cell proliferation ability was evaluated using methylthiazolyldiphenyl-tetrazolium bromide (MTT) assay.

Cell apoptosis assay
Annexin V-uorescein isothiocyanate (FITC)/PI apoptosis detection kit (Sangon Biotech) was applied to detect cell apoptosis according to the recommendations. In short, Hep3B and Huh7 cells were harvested and double stained by Annexin V-FITC and PI for 20 min in the darkness. Afterward, apoptotic cells were detected using a ow cytometer.

Transwell assay
Transwell chambers (pore size 8 μm) (Corning Incorporation, Corning, NY, USA) coated without and with Matrigel (BD Biosciences, San Jose, CA, USA) were utilized to assess Hep3B and Huh7 cell migration and invasion abilities, respectively. In brief, cells were suspended in serum-free medium (DMEM, 100 µL) and then placed in the top surface of the chamber, and DMEM mixed with FBS (10%) was placed in the bottom surface of the chamber. Non-migrated or non-invaded cells from the top surface were gently wiped off using a cotton wool after incubation for 24 h. After that, the migrated or invaded cells were xed using paraformaldehyde (4%) and stained using crystal violet (0.1%). Lastly, a microscope (Olympus, Tokyo, Japan) was utilized to photograph and count the migrated and invaded cells.

Western blot assay
To extract the total protein, tissues or transfected cells were lysed by RIPA lysis buffer (Sigma-Aldrich, St.

Statistical analysis
In this study, all data from at least three independently experiment were displayed as mean ± standard deviation (SD). The signi cance of differences between groups was analyzed with Student's t-test (for 2 groups) or a one-way analysis of variance (ANOVA; for more than 2 groups). Correlation between miR-223-3p and circ_LRIG3 or MAP2K6 was detected by Spearman rank correlation. Statistical analyses were performed by Graphpad Prism version 6.0 software (GraphPad Software, San Diego California, USA). P<0.05 was considered to be a statistically signi cant difference.

Results
Circ_LRIG3 expression was increased in HCC tissues and cells To investigate the potential roles of circ_LRIG3 in HCC, its expression was examined in HCC tissues and cells by qRT-PCR. Results displayed that circ_LRIG3 level was strikingly enhanced in HCC tissues in comparison with normal tissues ( Figure 1A). Similarly, its expression was also increased in HCC cells (Hep3B and Huh7) compared with THLE-2 cells ( Figure 1B). Next, we evaluated the stability of circ_LRIG3 in HCC cells. According to the qRT-PCR analysis, circ_LRIG3 was resistant to RNase R relative to linear LRIG3 in Hep3B and Huh7 cells ( Figure 1C and 1D), implying that circ_LRIG3 formed a loop structure.
Subsequently, Actinomycin D assay demonstrated that the half-life of circ_LRIG3 transcript exceeded 24 h, indicating that circ_LRIG3 transcript was more stable than the linear LRIG3 transcript in Hep3B and Huh7 cells ( Figure 1E and 1F). Moreover, the localization of circ_LRIG3 was analyzed in HCC cells. As presented in Figure 1G and 1H, most of the circ_LRIG3 was located in the cytoplasm. These results suggested that circ_LRIG3 might play critical roles in the progression of HCC.
Knockdown of circ_LRIG3 inhibited cell proliferation, metastasis and induced apoptosis in HCC cells To explore the effects of circ_LRIG3 on proliferation, metastasis and apoptosis of HCC cells, si-NC or si-circ_LRIG3 was transfected into Hep3B and Huh7 cells. The results from qRT-PCR analysis showed that the expression of circ_LRIG3 was evidently reduced in Hep3B and Huh7 cells after transfection with si-circ_LRIG3, suggesting that transfection of si-circ_LRIG3 was successful (Figure 2A). Cell cycle progression was analyzed by ow cytometry and cell proliferation was determined by MTT assay. Results displayed that the percentage of G0/G1 phase cells was increased by downregulating circ_LRIG3, while the percentage of cells in S and G2/M phases were reduced after interference of circ_LRIG3 ( Figure  2B and 2D), suggesting that the cell cycle was arrested at the G0/G1 phase. And MTT analysis proved that cell proliferation was obviously inhibited in Hep3B and Huh7 cells transfected with si-circ_LRIG3 compared with those cells transfected with si-NC ( Figure 2C and 2E). Moreover, we found that cell apoptosis was enhanced in Hep3B and Huh7 cells transfected with si-circ_LRIG3 in contrast to sh-NC group ( Figure 2F). Transwell assay showed that interference of circ_LRIG3 inhibited Hep3B and Huh7 cell migration and invasion ( Figure 2G and 2H). Western blot assay was applied to measure the metastasisrelated proteins (snail and E-cadherin). As depicted in Figure 2I and 2J, circ_LRIG3 silence decreased the protein level of snail (a mesenchymal marker) while increased the protein expression of E-cadherin (an epithelial marker) in Hep3B and Huh7 cells. These data collectively indicated that downregulation of circ_LRIG3 could inhibit the progression of HCC cells.
MiR-223-3p was a direct target of circ_LRIG3 Previous study indicated that circRNAs could act as molecular sponges of miRNAs in HCC cells [16], so the possible target miRNAs of circ_LRIG3 were predicted by circinteractome tool. As shown in Figure 3A, miR-223-3p was predicted as a target of circ_LRIG3. To investigate whether miR-223-3p was a direct target of circ_LRIG3, we performed dual-luciferase reporter assay in HCC cells. Results showed that transfection of miR-223-3p mimic resulted in a signi cant reduction in luciferase activity of circ_LRIG3-wt compared to miR-NC group, while the luciferase activity of circ_LRIG3-mut was unaffected by transfection of miR-223-3p ( Figure 3B and 3C). Next, we explored the impact of circ_LRIG3 on miR-223-3p expression. The results of qRT-PCR demonstrated that transfection of si-circ_LRIG3 led to an obvious promotion of miR-223-3p expression, while co-transfection of anti-miR-223-3p abated this effect ( Figure  3D). Subsequently, the expression of miR-223-3p was detected by qRT-PCR in HCC tissues and cells. As illustrated in Figure 3E and 3F, the expression of miR-223-3p was downregulated in HCC cells and tissues compared with their corresponding controls. Moreover, correlation between miR-223-3p and circ_LRIG3 expression was analyzed in HCC tissues. As displayed in Figure 3G, miR-223-3p expression was negatively correlated with circ_LRIG3 level in HCC tissues (r=-0.5054, P=0.0003). Thus, these results demonstrated that miR-223-3p was a target of circ_LRIG3 in HCC cells.
Knockdown miR-223-3p reversed the inhibitory effect of circ_LRIG3 downregulation on progression of HCC cells To explore whether the functions of circ_LRIG3 was mediated by miR-223-3p, Hep3B and Huh7 cells were transfected with si-NC, si-circ_LRIG3, si-circ_LRIG3 + anti-miR-NC, or si-circ_LRIG3 + anti-miR-223-3p. As shown in Figure 4A-4D, the effects of si-circ_LRIG3 on promotion of G0/G1 phase cells and reduction of S and G2/M phases cells as well as cell proliferation were abolished by downregulating miR-223-3p. Moreover, the promotive effect of circ_LRIG3 knockdown on apoptosis was abated by downregulation of miR-223-3p ( Figure 4E). Transwell assay indicated that interference of miR-223-3p reversed the inhibitory effects of circ_LRIG3 silence on migration and invasion ( Figure 4F and 4G). Likewise, downregulating miR-223-3p also could abrogate the effects of si-circ_LRIG3 on decrease of snail expression and increase of E-cadherin expression in Hep3B and Huh7 cells ( Figure 4H and 4I). Taken together, these ndings disclosed that circ_LRIG3 knockdown inhibited the progression of HCC cell by upregulating miR-223-3p.

MAP2K6 was a target gene of miR-223-3p in HCC cells
To further elucidate the mechanism of miR-223-3p in HCC cells, target prediction was performed by TargetScan and MAP2K6 was identify as a candidate target for miR-223-3p ( Figure 5A). To further whether MAP2K6 was a direct target of miR-223-3p, dual-luciferase reporter assay was carried out. We observed that the luciferase activity of MAP2K6-wt was markedly suppressed in cells transfected with miR-223-3p, but luciferase activity of MAP2K6-mut was not changed ( Figure 5B and 5C). Transfection e ciency of miR-223-3p and anti-miR-223-3p was measured by qRT-PCR. Results showed that miR-223-3p expression was increased in cells transfected with miR-223-3p while its expression was decreased in cells transfected with anti-miR-223-3p ( Figure 5D), implying that miR-223-3p and anti-miR-223-3p were successfully transfected in Hep3B and Huh7 cells. Subsequently, the effect of miR-223-3p on expression of MAP2K6 was explored. The results from qRT-PCR and western blot analysis showed that overexpression of miR-223-3p reduced the MAP2K6 mRNA and protein expression while knockdown of miR-223-3p presented an opposite effect ( Figure 5E and 5F). Next, the MAP2K6 mRNA and protein expression were examined by qRT-PCR and western blot assays in HCC cells and tissues. The results indicated that the mRNA and protein levels of MAP2K were overexpressed in HCC cells and tissues compared with their matched controls ( Figure 5G-5J). In addition, we found that MAP2K6 mRNA expression was negatively correlated with miR-223-3p abundance ( Figure 5K) (r=-0.5090, P=0.0003).
Furthermore, we investigated whether circ_LRIG3 functioned as a molecular sponge of miR-223-3p to regulate the expression of MAP2K6. We observed that circ_LRIG3 de ciency decreased the mRNA and protein expression of MAP2K6 while interference of miR-223-3p reversed this effect ( Figure 5L and 5M). Collectively, these data elaborated that circ_LRIG3 regulated MAP2K6 expression by sponging miR-223-3p in HCC cells.
Overexpression of MAP2K6 abated the suppressive effect of si-circ_LRIG3 on progression of HCC cells To investigate whether MAP2K6 was involved in si-circ_LRIG3-mediated functions in HCC cells, Hep3B and Huh7 cells were transfected with si-NC, si-circ_LRIG3, si-circ_LRIG3 + pcDNA, or si-circ_LRIG3 + MAP2K6. As presented in Figure 6A and 6B, mRNA and protein expression of MAP2K6 were reduced in cells transfected with si-circ_LRIG3 compared to si-NC group, which was abated by addition of MAP2K6.
Flow cytometry and MTT analysis showed that upregulation of MAP2K6 reversed the effects of si-circ_LRIG3 on promotion of G0/G1 phase cells and reduction of S and G2/M phases cells as well as cell proliferation ( Figure 6C-6F). Additionally, overexpression of MAP2K6 abolished the pro-apoptosis, antimigration and anti-invasion effects caused by silencing circ_LRIG3 (Figure 6G-6I). Western blot assay proved that co-transfection of MAP2K6 attenuated the suppression of snail expression and promotion of E-cadherin expression in Hep3B and Huh7 cells transfected with si-circ_LRIG3 ( Figure 6J and 6K). Therefore, we concluded that circ_LRIG3 knockdown suppressed the progression of HCC cell by downregulating MAP2K6.
Silencing circ_LRIG3 inhibited the activation of MAPK/ERK pathway through upregulating miR-223-3p and downregulating MAP2K6 MAPK/ERK signaling pathway is known to be activated in many cancers [17]. MAPK/ERK-related proteins were analyzed by western blot assay. Results demonstrated that knockdown of circ_LRIG3 reduced the protein levels of p-MAPK and p-ERKs, which was reversed by interference of miR-223-3p or overexpression of MAP2K6, but we observed no change of total MAPK and ERKs protein in Hep3B and Huh7 cells ( Figure 7A and 7B). These ndings indicated that circ_LRIG3 modulated the MAPK/ERK pathway by affecting miR-223-3p and MAP2K6 expression.

Knockdown of circ_LRIG3 limited tumor growth by regulating miR-223-3p and MAP2K6 expression
Sh-NC or sh-circ_LRIG3-transfected Huh7 cells were introduced into nude mice to assess the role of circ_LRIG3 in vivo. As displayed in Figure 8A and 8B, interference of circ_LRIG3 reduced tumor volume and weight in xenograft model. We then detected the expression of circ_LRIG3, miR-223-3p and MAP2K6 in tumor tissues. As shown in Figure 8C-8E, silencing circ_LRIG3 decreased the expression of circ_LRIG3 and MAP2K6 while elevated the abundance miR-223-3p in excised tumor masses. Western blot assay also proved that circ_LRIG3 interference led to decrease of MAP2K6 protein expression in tumor tissues ( Figure 8F). To sum up, these results disclosed that circ_LRIG3 de ciency inhibited tumor growth via upregulating miR-223-3p and downregulating MAP2K6 in vivo.

Discussion
HCC is one of the most common deadly cancers in the world. Growing evidence showed that the abnormal expression of circRNAs was tightly related to tumorigenesis and development of tumors, including HCC [18]. Hence, more efforts should be made to deeply explain the functional roles and underlying mechanisms of circ_LRIG3 in HCC. Here, we found that circ_LRIG3 knockdown inhibited the progression of HCC by regulating miR-223-3p/MAP2K6 axis and inactivating MAPK/ERK signaling pathway.
Accumulating evidence has shown that circRNAs are abundant in eukaryotes and aberrantly expressed in human cancers [19]. In addition, because of their covalently closed-structure, circRNAs are more stable and more suitable as e cacious biomarkers than linear-RNAs, such as lncRNAs and miRNAs [20]. For instance, circ_UVRAG [21], circ_BACH2 [22] and circ_ANKS1B [23] have been identi ed to be diagnostic or prognostic biomarker for gastric cancer, papillary thyroid carcinoma and breast cancer, respectively.
Previous report has been demonstrated that hsa_circ_0027345 (a circRNA derived from linear LRIG3) was overexpressed in HCC tissues [11]. However, there is no report on the functions and underlying mechanism of circ_LRIG3 in HCC. Consistent with previous report, we also uncovered that circ_LRIG3 level was enhanced in HCC tissues and cell lines. Additionally, we observed that knockdown of circ_LRIG3 inhibited the progression of HCC cells via reducing cell proliferation, metastasis and promoting apoptosis. These ndings suggested that circ_LRIG3 might act as a tumor promoter in HCC.
Emerging evidence showed that some circRNAs participated in tumorigenesis through functioning as sponges for miRNAs [16,24]. Then, circinteractome was utilized to predict the potential targets of circ_LRIG3, the data showed that circ_LRIG3 might interact with miR-223-3p, which was validated using the dual-luciferase reporter assay in HCC cells. MiR-223 a well-studied miRNA, presented different properties in different cancers, acting as an oncogene in colorectal cancer [25], gastric cancer [26] and prostate cancer [27], or as an anti-oncogene in esophageal carcinoma [28], breast cancer [29] and osteosarcoma [30]. Previous studies have suggested that miR-223 was lowly expressed HCC [31,32].
Moreover, miR-223 has been suggested to repress HCC cell growth and accelerate apoptosis through the Rab1-mediated mTOR activation [33]. In agreement with these ndings, we proved that miR-223-3p abundance was reduced in HCC tissues and cells, and its interference abated the repressive impact of circ_LRIG3 downregulation on progression of HCC cells. These data suggested that circ_LRIG3 exerted its functions by sponging miR-223-3p in HCC cells.
It is well known that miRNAs mediate various cellular activities through regulating their molecular targets [34]. Thus, the possible downstream targets of miR-223-3p were searched through the TargetScan software. Our results revealed that MAP2K6 was a direct target of miR-223-3p. MAP2K6 (important components of MAPK signal pathway) is involved in a variety of physiological and pathological processes as well as drug resistance in human cancer cells, and it has been recognized as an oncogene in many cancers, such as esophageal adenocarcinoma [35], prostate cancer [36] and colon cancers [15]. However, the expression and effect of MAP2K6 in HCC cells have not been clari ed. Here, it was found that the MAP2K6 was overexpressed in HCC tissues and cells, and was positively regulated by circ_LRIG3 and inversely modulated by miR- Availability of data and materials The data sets used and/or analyzed during the current study are available from the corresponding author on reasonable request.
Competing interests