MiR‐205 suppresses tumor growth, invasion, and epithelial‐mesenchymal transition by targeting SEMA4C in hepatocellular carcinoma

Growing evidence indicates that microRNAs are involved in tumorigenesis and progression of hepa tocellular carcinoma (HCC). However, the functional mechanisms of miR‐205 in HCC remain largely unknown. Here, we demonstrate that miR‐205 expression was significantly down‐regulated in HCC tissues and cell lines and was correlated with metastatic pathologic features and shorter disease‐free and overall survival. Overexpression of miR‐205 dramatically inhibited HCC cell proliferation, apoptosis, migration, invasion, epithelial‐mesenchymal transition (EMT) in vitro, and tumor growth in vivo. We subsequently identified semaphorin 4C (SEMA4C) as a novel target of miR‐205. Furthermore, high expression levels of SEMA4C were frequently found in HCC tissues and were associated with poor prognosis. Ectopic expression of SEMA4C restored the suppressive effect of overex pressed miR‐205 on migration, invasion, and EMT. Taken together, our findings provide new insight into the critical role of miR‐205 in regulating tumor growth, invasion, and EMT of HCC, suggesting miR‐205 may serve as a promising therapeutic target and novel prognostic indicator for patients with HCC.—Lu, J., Lin, Y., Li, F., Ye, H., Zhou, R., Jin, Y., Li, B., Xiong, X., Cheng, N. MiR‐205 suppresses tumor growth, invasion, and epithelialmesenchymal transition by targeting SEMA4C in hepatocellular carcinoma. FASEB J. 32, 6123–6134 (2018). www.fasebj.org

Hepatocellular carcinoma (HCC) is one of the most common malignancies and the third leading cause of cancerrelated mortality worldwide (1). Despite numerous therapeutic strategies used to improve the prognosis of HCC, the long-term survival rate is still unsatisfactory because of high rates of recurrence and metastasis (2,3). Several factors are involved in the pathogenesis of HCC; however, the roles and underlying molecular mechanisms of those factors remain poorly understood. Consequently, there is an urgent need to explore the molecular mechanisms responsible for HCC metastasis, which may contribute to identifying novel therapeutic targets and improving the prognosis of patients with HCC.
MicroRNAs (miRNAs) are a class of highly conserved, small, noncoding RNAs that negatively control gene expression at the posttranscriptional level by resulting in mRNA degradation or repressing protein translation (4). Increasing evidence indicates that miRNAs are aberrantly expressed in various types of cancer, and the dysfunction of miRNAs has a crucial role in tumor initiation, invasion, and metastasis (5,6). Recent studies have demonstrated that miRNAs are correlated with the proliferation, angiogenesis, drug resistance, and prognosis of HCC (7)(8)(9)(10).
The roles of miR-205 in human malignancies remain controversial (11). Several reports have indicated that miR-205 is frequently down-regulated in several types of cancer and serves as a tumor suppressor by inhibiting tumor proliferation, invasion, and metastasis (12)(13)(14)(15). However, miR-205 also exhibits oncogenic properties in endometrial carcinoma and cervical cancer (16,17). Therefore, miR-205 appears to carry out various functions depending on tumor types or its target genes. Nevertheless, the function of miR-205 and its molecular mechanisms in HCC are largely unknown.
Epithelial-mesenchymal transition (EMT)-a series of events during which epithelial cells acquire phenotypic characteristics of mesenchymal cells-has been recognized as the key process facilitating invasion and metastasis of tumors (18,19). Numerous studies reveal that EMT is involved in HCC invasion and metastasis (20,21). The molecular mechanisms of the EMT process are complex, and multiple molecules have been shown to be capable of orchestrating EMT programs, including miRNAs (22). However, the functional roles of miR-205 in the regulation of EMT in HCC remain to be elucidated.
In this study, we found that miR-205 was significantly down-regulated in HCC, which was associated with poor prognosis. Overexpression of miR-205 inhibited malignant behaviors and EMT of HCC cells in vitro as well as tumor growth in vivo. Notably, SEMA4C was identified as a direct target of miR-205 and mediated the biologic functions of miR-205 in HCC cells. Furthermore, we revealed that miR-205 may inhibit the EMT process to influence migration and invasion of HCC cells by regulating SEMA4C expression.

Patients and specimens
Fifty paired human HCC and adjacent nontumor liver tissues were collected from the West China Hospital of Sichuan University between January 2010 and December 2011.
Histopathologic diagnosis was performed according to the TNM Staging Guide [based on the extent of the tumor (T), the extent of spread of the lymph nodes (N), and the presence of metastasis (M); American Joint Committee on Cancer, Chicago, IL, USA]. The tissues were stored at 280°C or embedded in paraffin. None of the patients received any preoperative chemotherapy or radiotherapy. Informed consent was obtained from each patient, and all protocols of this study were approved by the Ethics Committee of West China Hospital. The demographic and clinicopathologic features of the patients are described in the Table 1.

Cell culture and miRNA transfection
Human HCC cell lines (MHCC97-L, HepG2, Huh-7, Bel-7402, MHCC97-H, and HCCLM3) and the human normal hepatocyte cell line (L02) were purchased from the Institute of Biochemistry and Cell Biology (Chinese Academy of Sciences, Shanghai, China). All cells were cultured in complete DMEM (Thermo Fisher Scientific, Waltham, MA, USA) containing 10% fetal bovine serum (Thermo Fisher Scientific) and incubated at 37°C in a humidified incubator with 5% CO 2 .
MiR-205 overexpression and its negative control (NC) lentiviral vectors were designed and synthesized by Shanghai R&S Biotechnology Co. (Shanghai, China). The SEMA4C overexpression vector and its NC vector were obtained from GeneCopoeia (Rockville, MD, USA). Cells were transiently transfected with above vectors using Polo Deliverer 3000 Transfection Reagent (Shanghai R&S Biotechnology Co.) in accordance with the manufacturer's protocol. RNA extraction and real-time, quantitative RT-PCR analysis Total RNA was extracted from fresh-frozen HCC tissues or cell lines with Trizol reagent (Thermo Fisher Scientific). cDNA synthesis was performed with 2 mg total RNA using a reverse transcription kit (Takara Bio, Kusatsu, Japan). Real-time, quantitative RT-PCR (qRT-PCR) was performed with All-in-One miRNA qRT-PCR Detection Kit (GeneCopoeia) on an ABI 7700 Real-Time PCR System (Thermo Fisher Scientific). Real-time qRT-PCR was performed with SYBR green 23 master mixture. Hsa-miR-205 primer, U6 small nuclear RNA qPCR primer, and SEMA4C were purchased from GeneCopoeia. U6 was used as an internal control for microRNA, and b-actin was used for mRNA. The DC t and 2 2DDCt methods were used to analyze gene expression relative to the endogenous control. The specific primer sequences used are listed in Supplemental Table S1.

Transfection and vector construction
The miR-205-mimic and NC miRNA sequences (Gene-Copoeia) were amplified and cloned into a pcDNA6. Immunofluorescence and confocal laser scanning microscopy After handling with 0.5% Triton X-100 and blocking in 4% bovine serum albumin, HCC cells were incubated with antibodies against E-cadherin (1:50; Santa Cruz Biotechnology) or vimentin (1:50; Santa Cruz Biotechnology), followed by incubation with matched secondary antibodies (1:1000; Santa Cruz Biotechnology). The nucleus of HCC cells were stained with DAPI. Confocal image series were collected with an LSM 5 Pascal Laser Scanning Microscope (Carl Zeiss, Oberkochen, Germany).

Immunohistochemical analysis
All the paraffin-embedded tissues were sliced into 5-mm-thick consecutive sections. Tissue sections were subsequently deparaffinized, rehydrated, and quenched in 3% hydrogen peroxide. After blocking with goat serum, the slides were incubated with anti-SEMA4C and, then, incubated with secondary horseradish peroxidase-labeled rabbit anti-goat IgG antibodies. The sections were visualized with a 3,39diaminobenzidine kit (Beyotime) and counterstained with hematoxylin. Immunoreactive scoring was assessed according to the percentage of positively staining cells and the staining intensity. To determine the average scores, each section was identified for 5 independent high-magnifications (3400) fields.
Cell proliferation, colony formation, and cell apoptosis assays Cell proliferation was measured with the cell counting kit 8 (CCK-8) and colony-formation assays. For the CCK-8 assay, cells were seeded in 96-well plates at 4000 cells/well. After 24 h, 10 ml of CCK-8 was added into each well. Subsequently, the optical density was measured at a wavelength of 450 nm with a microplate reader (Bio-Rad). For the colony-formation assay, 500 cells were seeded into each well of a 6-well plate. Colonies were fixed with 20% methanol and stained with 0.1% crystal violet, and the colonies were counted. For the apoptosis assay, cells were stained with 0.5% propidium iodide (Shanghai R&S Biotechnology) and annexin V (Shanghai R&S Biotechnology). The percentage of apoptotic cells was detected by flow cytometry.

Cell migration and invasion assays
Cell migration and invasion assays were performed in 24-well Transwell chambers (Corning, Corning, NY, USA) with or without being coated Matrigel (BD Biosciences, Franklin Lakes, NJ, USA). Briefly, a 200-ml transfected-cell suspension, with serum-free medium (1 3 10 6 /ml), was added to the upper chamber, and the lower chamber was filled with a culture medium containing 10% fetal bovine serum as the chemoattractant. After 24 h, the cells that had migrated into or invaded the lower surface were fixed and stained with 0.1% crystal violet. Photographs of 5 randomly selected fields of the fixed cells were taken, and cell numbers were calculated under a microscope.

MiRNA target prediction and dual luciferase reporter assay
The analysis of miR-205-predicted targets was performed with TargetScan, miRDB, and miRanda algorithms. The 39-UTR sequence of SEMA4C, which was predicted to interact with miR-205, was combined with a corresponding, mutated sequence of SEMA4C that contained the miR-205 binding sites, and was synthesized and cloned into the psiCheck2 dual-luciferase reporter vector (GeneCopoeia) called wild-type (wt)-SEMA4C 39-UTR (wt) and mutant (mt)-SEMA4C 39-UTR (mutated). Subsequently, the cells were cultured in 24-well plates and cotransfected with the wt or mt 39-UTR of the SEMA4C vector, together with the miR-205 mimics or NC miRNA with Lipofectamine 2000 Reagent (Thermo Fisher Scientific). Forty-eight hours after cotransfection, firefly and Renilla luciferase activities were measured with the Dual-Luciferase Reporter Assay Kit (Promega, Madison, WI, USA), according to the manufacturer's protocol. The ratio of firefly to Renilla was used to normalize the firefly luciferase activity.

Animal experiments
All animal studies were approved by the Institutional Animal Care and Use Committee of Sichuan University, China. Male immune-deficient nude mice (BALB/C-nu) at 4-6 wk of age were purchased from the Sichuan Laboratory Animal Center (Chengdu, China) and were randomized into 2 groups (n = 5 each). The mice were raised in cages under specific pathogen-free conditions. A total of 1 3 10 7 HCCLM3 cells, which stably transfected with either miR-205 or the NC vector, were subcutaneously injected into the right flank of each mouse. Mice were monitored daily for general condition, and tumor size was measured every 3 d. Tumor volume (V ) was calculated with the following formula: where L is the length, and W is the width of each tumor. Forty days later, all mice were euthanized, and the tumors were removed, weighed, and photographed.

Statistical analysis
All data are expressed as means 6 SD. Statistical analyses were conducted using SPSS software (v.21.0; IBM, Armonk, NY, USA). Statistical differences between the 2 groups were determined by independent Student's t test. The x 2 test or Fisher's exact test was used to analyze the relationship between miR-205 expression and clinicopathologic characteristics. Overall survival (OS) and disease-free survival (DFS) were evaluated WITH Kaplan-Meier survival curves and were compared by the log-rank test. Univariate and multivariate analyses (Cox proportional hazard-regression model) were performed to determine potential prognostic factors for OS and DFS. A value of P , 0.05 was considered statistically significant.

MiR-205 is down-regulated in HCC and is correlated with a poor prognosis
To investigate the expression and significance of miR-205 in HCC, we first examined the expression level of miR-205 in 50 matched pairs of HCC and adjacent nontumor tissues. We found that miR-205 expression was significantly down-regulated in 76% (36 of 50) HCC tissues when compared with corresponding adjacent nontumor tissues (P , 0.001, Fig. 1A). Moreover, miR-205 expression was obviously lower in aggressive HCC tissues than in nonaggressive HCC tissues (P , 0.01, Fig. 1B). Aggressive HCC tissues were defined as HCC tissues with intrahepatic metastasis and vascular and bile duct invasion. To further confirm the relationship between miR-205 expression and clinicopathologic features, 50 patients with HCC were divided into 2 subgroups based on the cutoff values, which were determined as the median level of miR-205 in HCC tissues. Interestingly, clinicopathologic analysis showed that miR-205 expression was strongly correlated with the number of tumor nodules, microvascular invasion, TNM staging, and Edmondson-Steiner grading (Table 1). In addition, Kaplan-Meier survival analysis showed that patients with HCC who had low miR-205 expression levels had significantly shorter OS (Fig. 1C) and DFS (Supplemental Fig. S1) than those with high miR-205 expression levels. The 1-, 3-, and 5-y OS rates of patients with low miR-205 expression were 56, 16, and 8%, respectively, which were significantly lower than those with high miR-205 expression (84, 52, and 20%, respectively; P = 0.001). A multivariate Cox regression analysis revealed that low miR-205 expression was the independent risk factor for OS (P = 0.026; Table 2) and DFS (P = 0.015; Supplemental Table S2). Consistent with miR-205 expression in HCC tissues, miR-205 expression was also decreased in HCC cell lines with different metastatic potentials as compared with the normal human hepatocyte line LO2 (P , 0.001, Fig. 1D). Importantly, the expression level of miR-205 decreased with an increase in metastatic potentials in the HCC cell lines. These results indicated that down-regulation of miR-205 is correlated with poor prognosis in HCC and might be involved in HCC progression.

MiR-205 not only inhibits cell proliferation, migration, and invasion of HCC cells but also induces HCC cells apoptosis in vitro
To further investigate the biologic role of miR-205 in the development and progression of HCC, gainof-function experiments were performed in HCCLM3 and HepG2 cells that showed different expression levels of endogenous miR-205. Those 2 HCC cell lines were transfected with the miR-205 expression plasmid or the NC miRNA vector. As detected by qRT-PCR, we confirmed that miR-205 was effectively overexpressed in both cell lines (P , 0.001, Fig. 2A). The CCK-8 assay showed that up-regulation of miR-205 in HCCLM3 and HepG2 cells resulted in significant suppression of cell proliferation compared with those in the NC groups (P , 0.01, Fig. 2B). The colony-formation assay also demonstrated that HCCLM3 and HepG2 cells overexpressing miR-205 showed remarkably decreased colony-formation abilities (Fig. 2C). We also investigated whether miR-205 modulated cell apoptosis and found that miR-205 overexpression induced apoptosis of HepG2 cells (15 vs. 5% in the control group; P , 0.001) (Fig. 2D). Similar effects of miR-205 overexpression were found in HCCLM3 cells (P , 0.001, Fig. 2D). To determine the influence of miR-205 on migration and invasion of HCC cells, we performed transwell assays with HCCLM3 and HepG2 cells.

MiR-205 inhibits the EMT process of HCC cells
Considering that EMT has a critical role in the invasion and metastasis of HCC and that miRNAs have been implicated in regulating EMT, we performed experiments to assess whether miR-205 inhibits EMT in HCC cells. qRT-PCR and Western blot results revealed that miR-205 overexpression remarkably increased the mRNA and protein levels of the epithelial marker E-cadherin, whereas the mRNA and protein levels of mesenchymal marker vimentin were significantly reduced (P , 0.05, respectively; Fig. 3A-C).
Immunofluorescence analysis also confirmed this result by showing that overexpression of miR-205 was associated with enhanced expression of E-cadherin and decreased expression of vimentin (Fig. 3D). Taken together, these data suggest that miR-205 suppresses the EMT process in HCC cells.

MiR-205 inhibits tumor growth of HCC in vivo
Based on our results in vitro, we further identified the biologic function of miR-205 in HCC growth in vivo. HCCLM3 cells that stably overexpressed miR-205 or the NC vector were injected subcutaneously into nude mice. Although tumors successfully formed in all mice, the tumor growth rate of HCCLM3-miR-205 group was significantly slower than that of the HCCLM3-NC group (Fig. 3E). In addition, the sizes of tumors from mice injected with miR-205 were significantly smaller than those of mice injected with the NC vector (Fig.  3F). Therefore, these results indicate that overexpression of miR-205 suppresses tumor growth of HCC in vivo.

SEMA4C is directly targeted by miR-205 and up-regulated in HCC
To reveal the underlying molecular mechanisms by which miR-205 exerts its functional effects on HCC cells, we predicted and identified the potential target genes of miR-205 by searching 3 databases, including TargetScan, miRDB, and miRNAda. SEMA4C was screened and selected as the candidate target gene of miR-205 because the complementary sequence of miR-205 was found in the 39-UTR of SEMA4C (Fig. 4A). Furthermore, previous studies have demonstrated that SEMA4C is highly expressed in breast cancer and lung cancer cells and can prompt the invasion and metastasis of tumors by inducing EMT (23,24). To further verify that SEMA4C was directly targeted by miR-205 in HCC, we investigated whether miR-205 directly interacted with the predicted 39-UTR of SEMA4C mRNA by a dual-luciferase reporter system. Results showed that miR-205 overexpression significantly inhibited the luciferase activity of wt-SEMA4C-39-UTR compared with that of the NC miRNAs (P , 0.05, Fig. 4B). In contrast, the luciferase activity of mt-SEMA4C-39-UTR was not influenced by miR-205 or the NC miRNA, indicating that SEMA4C was directly targeted by miR-205. Consistent with the above findings, qRT-PCR and Western blot also demonstrated that overexpression of miR-205 markedly suppressed the mRNA and protein levels of SEMA4C (Fig. 4C, D).
The mRNA levels of SEMA4C were also validated by qRT-PCR in the previous cohort of 25 paired HCC clinical specimens. Compared with adjacent, healthy liver tissues, SEMA4C mRNA was significantly up-regulated in 16 (64%) HCC tissues (P , 0.05, Fig. 5A). Additionally, SEMA4C protein expression in HCC tissues detected by immunohistochemical staining analysis was dramatically increased when compared with adjacent, noncancerous tissues (P , 0.05, Fig. 5B). We also performed a Kaplan-Meier survival analysis to evaluate the prognostic value of SEMA4C in patients with HCC. As shown in Fig. 5C, patients with high SEMA4C expression levels had shorter OS and DFS (Supplemental Fig. S3) than those with low expression levels of SEMA4C (P , 0.05). These data suggest that SEMA4C is upregulated in HCC and correlated with poor prognosis in patients with HCC.

SEMA4C mediates the functional effects of miR-205 and EMT on HCC cells
To ascertain whether miR-205 elicits inhibitory effects on HCC cells through SEMA4C, SEMA4C was  restored by overexpression plasmids in HCCLM3 cells stably overexpressing miR-205. Subsequent Western blotting results confirmed the overexpression of SEMA4C in miR-205-overexpressing HCCLM3 cells (Fig. 5D). As expected, SEMA4C restoration abrogated the inhibitory effects on migration and invasion of HCCLM3 cells induced by miR-205 (P , 0.05, Fig. 5E). Likewise, the repressive effects of miR-205 on EMT were rescued by overexpression of SEMA4C, leading to molecular changes associated with upregulation of E-cadherin and reduced expression of vimentin (Fig. 5D). Thus, these data provided evidence that reexpression of SEMA4C could rescue miR-205-mediated migration, invasion, and EMT of HCC cells.

DISCUSSION
Emerging evidence has demonstrated that dysregulation of miRNAs has a crucial role in the carcinogenesis and progression of HCC (25)(26)(27). A previous study (28) showed that miR-205 expression levels decrease in HCC tissues as compared with those in matched healthy tissues, and down-regulation of miR-205 promotes stemness of HCC by targeting PLCb1 and increasing CD24 expression. However, the relationship between miR-205 expression and prognosis in patients with HCC, as well as its functional roles in HCC are still largely unknown. In this study, we found that miR-205 was significantly down-regulated in HCC, especially in aggressive HCC phenotypes. In addition, reduced miR-205 expression was correlated with poor clinicopathologic features and shorter DFS and OS. These data strongly suggest that miR-205 may serve as a tumor suppressor in HCC and is a promising predictor for prognosis in patients with HCC. MiRNAs have been recognized as key regulators in the invasion and metastasis of human cancers (29,30). Thus, we explored whether miR-205 was involved in the progression of HCC through modulating the metastatic behaviors of HCC cells. Our results showed that overexpression of miR-205 inhibited the metastatic ability of HCC cells in vitro. Additionally, to confirm the similar biologic function of miR-205 in a complex microenvironment, we evaluated the suppressive effect of miR-205 in vivo. In agreement with our in vitro findings, miR-205 also inhibited tumor growth of HCC grafted in mice.
EMT, a key process characterized by loss of epithelial marker E-cadherin, acquisition of mesenchymal marker vimentin, and enhanced migratory and invasive behaviors, is attracting increasing attention as a crucial mechanism that facilitates progression and metastasis of cancer (18,31). Previous studies have revealed that up-regulation of miR-205 promoted EMT process in cervical cancer and lung cancer cells (32), whereas other reports showed the opposite results, that down-regulation of miR-205 resulted in acceleration of EMT in prostate cancer and breast cancer cells (13,33). In the present study, we found that ectopic expression of miR-205 significantly increased E-cadherin expression and concomitantly decreased vimentin expression. These findings suggest miR-205 may inhibit invasion and metastasis of HCC by hindering the EMT process.
SEMA4C, a member of the semaphorin family, has been reported to be highly expressed in several human cancers and encourages proliferation, migration, and EMT of cancer cells (23,34). Moreover, miR-25-3p can reverse the EMT process in cisplatin-resistant cervical cancer cells via targeting SEMA4C (35). However, the functional role of SEMA4C in HCC remains unclear.
Here, SEMA4C was confirmed as a direct downstream target of miR-205 and mediated the biologic function of miR-205 in HCC. Interestingly, this interaction between miR-205 and SEMA4C has not, to our knowledge, been previously reported. Our conclusion was based on the following evidence. First, miR-205 overexpression remarkably decreased the expression of SEMA4C at both the mRNA and protein levels in HCC cells. Second, the complementary sequence of miR-205 was identified in the 39UTR of SEMA4C mRNA, and, importantly, the luciferase reporter activity of wt-39UTR instead of mt-39UTR was specifically attenuated by miR-205 overexpression. Third, SEMA4C expression was up-regulated in HCC tissues, and that up-regulation was inversely correlated with downregulation of miR-205, and elevated SEMA4C expression was closely related to poor prognosis in patients with HCC. Fourth, we discovered that recovery of SEMA4C expression reversed the inhibitory effects of miR-205 on migration, invasion, and EMT process of HCC cells. These results indicate that SEMA4C is directly targeted by miR-205 and functions as an oncogene in HCC.
However, the molecular mechanisms by which miR-205/SEMA4C contributes to the metastasis and EMT of HCC are still entirely unknown. Previous studies have demonstrated that SEMA4C promoted the proliferation and migration of cancer cells by activating PlexinB2-MET signaling. Interestingly, aberrant activation of Met signaling was confirmed to be closely associated with tumor growth, angiogenesis, metastasis, and poor survival in HCC (36,37). Furthermore, Met can also activate an EMT program in human HCC cells (38). Consequently, miR-205/ SEMA4C probably promotes the metastasis and EMT of HCC by activating Met signaling. Further studies and their matched tumor-adjacent tissues. C ) Kaplan-Meier analysis of overall survival between high (n = 13) and low (n = 12) SEMA4C expression levels of patients with HCC. D) Western blot analysis of SEMA4C and EMT markers E-Cadherin and vimentin in HCCLM3 cells that stably express miR-205 or NC vector after restoration of SEMA4C expression. E ) Transwell assays in HCCLM3 cells that stably expressing miR-205 or NC vector after restoration of SEMA4C expression. *P , 0.05, **P , 0.01.
should identify the pathways that miR-205/SEMA4C mediates for its inhibitory effects on metastasis and EMT in HCC.
In summary, we discovered that miR-205 is frequently down-regulated in HCC and is significantly correlated with poor prognosis in patients. Our results indicate that miR-205 inhibits the metastatic behaviors and EMT of HCC cells by directly targeting SEMA4C. Therefore, this study provides new insight into the mechanism involved in HCC progression and suggests that miR-205 may act as a novel biomarker and promising therapeutic target in HCC.