Ectopic Expression of MiR-125a Inhibits the Proliferation and Metastasis of Hepatocellular Carcinoma by Targeting MMP11 and VEGF

Background Studies have been shown that miR-125a plays an important role in carcinogenesis, however, the role of miR-125a in hepatocellular carcinoma (HCC) remains elusive. Methodology/Principal Real time-PCR (qRT-PCR) was performed to test the significance of miR-125a in HCC. Ectopic expression of miR-125a was used to test the influences of miR-125a on proliferation and metastasis of HCC cells in vitro and in vivo. Predicted target genes of miR-125a were determined by dual-luciferase reporting, qRT-PCR, and western blot (WB) analyses. Then immunohistochemical staining (IHC) was used to detect the expression of target genes, and the correlations and prognostic values of miR-125a and its target genes were also investigated. Conclusions/Significance Decreased miR-125a was observed in both HCC tissues and cell lines, and associated with patients’ aggressive pathologic features. Up-regulating miR-125a significantly inhibited the malignant phenotypes by repressing the expression of matrix metalloproteinase 11 (MMP11) and vascular endothelial growth factor A (VEGF-A) both in vitro and in vivo. Furthermore, miR-125a expression was inversely correlated with both MMP11 and VEGF-A expression in HCC tissues. Inhibiting miR-125a could increase both MMP11 and VEGF-A expression, and RNA interference targeting MMP11 or VEGF-A mRNA could rescue the loss of miR-125a functions. MiR-125a inhibits the proliferation and metastasis of HCC by targeting MMP11 and VEGF-A. Up-regulation of miR-125a might be a promising approach and a prognostic marker for HCC.


Introduction
Hepatocellular carcinoma (HCC) is the second most frequent cause of cancer death worldwide, and half of HCC diagnoses and deaths were estimated to occur in China [1]. Despite improvements in surgery and other treatments, the postoperative prognosis of HCC patients is still unsatisfactory due to the high rate of recurrence and metastasis. Therefore, identifying the malignant factors and understanding the mechanisms are critical for the improvement of therapeutic strategies for HCC patients.
MicroRNAs (miRNAs) are small non-protein-coding RNAs of 20-24 nucleotides that can down-regulate the expression of their target genes through binding to the 39 untranslated regions (UTRs) of target mRNAs, resulting in mRNA degradation and translation inhibition [2,3]. Several miRNAs have been identified to be abnormally expressed, and to be associated with the malignant progressions of various human cancers [4,5,6]. MiRNAs were recently found to be frequently down-regulated in HCC, and they were also shown to be associated with the carcinogenesis, metastasis, recurrence and prognosis of HCC [7,8,9,10]. However, further investigation is needed to define the molecular mechanisms underlying the effects of miRNAs on the development of HCC. Previous studies have demonstrated that several miRNAs that are located in fragile sites or aberrant genomic regions play oncogenic or suppressive roles in the progression of human cancers [4]. MiR-125a is located at 19q13, which is frequently deleted in several human cancers. MiR-125a over-expression impaired the anchorage-dependent growth, migration and invasion of breast cancer cells by down-regulating ERBB2 and ERBB3 in the ERBB2-dependent SKBR3 cell line [11]. Researchers showed that miR-125a could inhibit the proliferation of human gastric cancer cells in combination with trastuzumab [12]. Furthermore, miR-125a was also downregulated in non-small cell lung cancer and had inverse effects on the invasion and migration of lung cancer cells [13]. Overexpression of miR-125a induces conversion of highly invasive ovarian cancer cells from a mesenchymal to an epithelial morphology, suggesting that miR-125a is a negative regulator of EMT [14]. However, the expression and roles of miR-125a in HCC have not been studied, nor the prognostic value and coexpression of miR-125a with target genes in tumor samples from patients have rarely been investigated.
In the present study, we found that the miR-125a was frequently down-regulated in HCC compared with matched adjacent liver tissues, and was correlated with the malignant progressions of patients. Ectopic expression of miR-125a could repress the proliferation and metastasis of HCC cell lines both in vitro and in vivo. However, over-expression of miR-125a could not significantly down-regulate the expression of ERBB2 or ERBB3 in HCC cells. Therefore, further studied were needed to elucidate the mechanisms underlying the inhibitory effects of miR-125a on HCC cells, and the prognostic values and co-expression of miR-125a with target genes in tumor samples from patients were also investigated.

Cell Lines and Human Samples
Human HCC cells (SMMC7721, MHCC97L, MHCC97H and HCCLM3) were kindly provided by Dr. Tang ZY (Liver Cancer Institute, Zhongshan Hospital, Fudan University, Shanghai, China) [15]. HepG2 cell line was obtained from American Type Culture Collection (ATCC number: HB-8065, Manassas, VA). QZG cell line was obtained from the Cell Research Institute of the Chinese Academy of Sciences (Shanghai, China), and both of them are maintained in our lab [16]. All of the cell lines were cultured in Dulbecco's modified Eagle medium (DMEM) supplemented with 10% FBS, 100 mg/ml penicillin, and 100 mg/ml streptomycin at 37uC in a 5% CO 2 incubator. Fresh cancerous and matched adjacent liver tissues were obtained from 80 patients who received primary HCC resection between January 2006 and December 2006 at the Department of General Surgery in Xijing Hospital (Xi'an, China). The tissues were immediately frozen in liquid nitrogen after surgical removal and stored at 280uC until use. Paraffin-embedded HCC tissues from the 80 patients were also obtained from Department of Pathology at Xijing Hospital. This study was approved by the Hospital's Protection of Human Subjects Committee, and informed consent was also obtained from all patients.

RNA Extraction and Real Time-PCR Analysis (qRT-PCR)
Total RNA was extracted from cell lines and tissue samples using Trizol (Invitrogen), and the concentration of total RNA was quantitated by measuring the absorbance at 260 nm. The miScript SYBR H Green PCR Kit from QIAGEN was used for qRT-PCR analysis. Primers of miR-125a and U6 snRNA (internal control) were also purchased from QIAGEN (MS00003423 and MS00033740). The fold-change for mRNA in HCC tissues relative to the adjacent non-tumor (NT) tissues was calculated using the 2 2DDCT method [17], where DDCt = DCt HCC/NT and DCt = Ct miR-125a-Ct U6. The relative mRNA levels of genes in HCC cells were also determined by qRT-PCR, and the expression of GAPDH was used as the internal controls. Each PCR was performed in triplicate. The primers for the genes examined were presented in Table 1.

MTT Assay
MTT assays were performed to evaluate the speed of cell proliferation as described previously [19]. Briefly, log phase cells were trypsinized into a single cell suspension and seeded into 96well plates at a density of 1610 3 cells/well. After 1, 2, 3, 4, 5, 6 or 7 days of cell cultivation, 20 ml of 3-(4,5)-dimethylthiahiazol (-z-y1)-3,5-di-phenytetrazo-libromide (MTT, 5 mg/ml; Sigma, St. Louis, MO) was added into each well and incubated for 4 hours at 37uC. The supernatant was then aspirated, and 150 ml dimethylsulfoxide (DMSO) was added to dissolve the crystals with agitation for 10 minutes at room temperature. The absorbance values were measured using an ELISA reader (Bio-Rad Laboratories, Richmond, CA) at a wave length of 490 nm. Each experiment was repeated three times.

Soft Agar Assay
Soft agar assays were performed to evaluate the in vitro tumorigenicity of cells. Approximately 2610 3 cells from each cell line were seeded in 0.3% agarose with a 0.5% agarose underlay in a six-well plate. The cells in agarose were cultured at 37uC in an atmosphere of 5% CO 2 at mosphere for 17 days. Cell colonies were counted by two independent observers. Each cell line was tested in triplicate, and the results shown are the mean of three repeated experiments.

Cell Invasion and Migration Assays
Cell migration and invasion assays were performed as described by others [20]. For migration assays, HepG2-125a, HepG2-NC, LM3-125a and LM3-NC cells were cultured in DMEM with 0.1% FBS. After 24 h, 2610 5 starved cells in DMEM without FBS or HGF were plated in the upper chamber with a non-coated membrane (6.5 mm in diameter, 8-lm pore size; Corning, NY, USA). For invasion assays, the chambers used were coated with 200 mg/ml of Matrigel and dried for 12 h. Then, 2610 5 cells were added to the top chamber. For both invasion and migration  assays, medium with 10% FBS and HGF (20 ng/ml) was placed into the lower chamber. After 24 h at 37uC, the cells on the upper membrane surface were carefully removed. After incubation with 95% ethanol for 20 min, the filters were stained with 0.2% crystal violet solution for 10min, and counted using an inverted microscope. The results presented are the mean of three independent assays.

Luciferase Reporter Assay
For dual luciferase assays, HepG2-125a or HepG2-NC cells in 24-well plates were co-transfected with 0.4 mg of the firefly luciferase report vector and 0.08 mg of pRL-TK control vector containing Renilla luciferase (Promega) using siPORTneoFX (Ambion, Austin, TX) according to the manufacturer's protocol. Lysates were prepared at 48 h post-transfection. Luciferase activity was measured using the Dual-Light luminescent reporter gene assay (Applied Biosystems). Firefly luciferase activity was normal-ized to renilla luciferase activity for each transfected well. Three independent experiments were performed in triplicate. The primers for the 39-UTR of targets containing a miR-125a binding site are shown in Table 2.

Immunohistochemical Staining (IHC)
IHC was used to detect the MMP11 and VEGF expression in HCC tissues as described in our previous work [19]. After being blocked with 3% H 2 O 2 , sections were incubated overnight at 4uC with antibodies (MMP11, 1:50;VEGF, 1:100) diluted in phosphate buffered saline (PBS). After three 5-min washes with PBS-T, sections were incubated with biotinylated goat anti-Rb or mouse IgG/HRP (1:2000, Zhongshan) for 1h, followed by three additional 5-min washes with PBS-T. DAB solution was used for visualization of the samples. As a negative control, the primary antibody was replaced with PBS. As in previous study, scores of 0 and 1 were considered to represent negative expression, and 2 and 3 were considered to represent positive expression [11].

Animal Experiments
Athymic mice (4-5 weeks old) were purchased from the Animal Center of the Chinese Academy of Science (Shanghai, China) and maintained in laminar ow cabinets under specific pathogen-free conditions. A total of 10 mice were used for the subcutaneous study. HepG2-125a or LM3-125a cells were injected into the right anks, and HepG2-NC or LM3-NC cells were injected into the left flanks. Four weeks after inoculation, the tumor-bearing mice were sacrificed, and the tumor volumes were calculated following the formula: L6S2 /2. Moreover, the total RNA and protein of each representative tumor from mice were used for analyses of the expression levels of miR-125a, MMP11 and VEGF-A target genes by qRT-PCR and western blot.
Previous studies have demonstrated that HCC-LM3 is a highly invasive cell line that can be used for in vivo metastasis assays [15], therefore, LM3-125a and LM3-NC cells were chosen for the metastasis experiments. Approximately 2610 6 cells from each cell line were suspended in 0.2 ml sterile PBS and injected into the tail veins of 10 mice. The mice were sacrificed 6 weeks after injection. After the visible tumors numbers in the liver and lung surface were counted, histological examinations were performed.

Statistical Analysis
All statistical analyses were carried out using the SPSS 16.0 software package (SPSS, Chicago, IL), and variables with a value of P,0.05 were considered to be statistically significant. The differential expression of miR-125a in HCC and paired noncancerous tissues was analyzed using the Wilcoxon rank sum test. The Kruskal-Wallis H test or the Mann-Whitney U test was used to investigate the significance of miR-125a, VEGF-A and MMP11 expression as correlated with clinical features in HCC. A one-way ANOVA test was utilized to evaluate the difference between three comparisons in cell proliferation and soft agar clonogenic assays, and the least significant difference-T test was used for the analysis of two groups. Overall survival curves were plotted using the Kaplan-Meier method and were evaluated for the statistical significance using a log-rank test.

MiR-125a was Down-regulated in HCC Tissues and Cell Lines
The expression of miR-125a in HCC and paired adjacent nontumor liver tissues from 80 patients was detected using qRT-PCR. The results showed that miR-125a expression was decreased in 77.5%(62/80)of HCC tissues compared with matched adjacent liver tissues, with an average of 4.72-fold reduction in expression (median = 0.37 vs 0.95; P = 0.005; Fig. 1A). Further analysis revealed that lower expression of miR-125a in HCC was significantly correlated with the patients' clinical features, including T stage and tumor metastasis (Table 3; P , 0.05).
Consistent with the results observed in tissues, significantly decreased expression of miR-125a was observed in all five HCC cell lines (SMMC-7721, HepG2, MHCC-97L, MHCC-97H and HCC-LM3) compared with the QZG cells (Fig. 1B). MiR-125a expression was low in SMMC-7721 and HepG2 cells, but barely in highly invasive cell lines, such as MHCC-97L, MHCC-97H and HCC-LM3. Therefore, we chose the HepG2 and HCC-LM3 cell lines for further study.

Ectopic Expression of MiR-125a Inhibited HCC Cell Proliferation
To investigate the effects of miR-125a on the proliferation of HCC, HepG2 and HCC-LM3 cell lines were used for the lentivirus-mediated miR-125a transfection. As presented in Figure 2A, qRT-PCR analysis confirmed that miR-125a was markedly up-regulated in both HepG2-125a and LM3-125a cells compared with controls. MTT assays were performed to test the effects of miR-125a on cell growth. Both HepG2-125a and LM3-125a cells exhibited much slower growth than their corresponding controls, and the inhibitory effects showed statistical significance after culture for four days (Fig. 2B). The colony formation assay results showed that ectopic expression of miR-125a markedly decreased the colony numbers of HepG2 and HCC-LM3 cells ( Fig. 2C; P,0.01). These data were further strengthened by the results from the nude mice tumorigenicity experiments. Four weeks after subcutaneous injection, tumors appeared much smaller at the sites injected with HepG2-125a or LM3-125a cells compared with each control ( Fig. 2D; P , 0.01). Thus, the over-expression of miR-125 may inhibit the tumorigenicity of HCC cells in vitro and in vivo. Moreover, the total RNA and protein of each representative tumor from mice were used for analyses of the expression levels of miR-125a, MMP11 and VEGF-A target genes by qRT-PCR and western blot. The result showed that the expression of miR-125a in the right flanks were significantly lower that the left flanks. Inversely, both the expression of MMP11 and VEGF-A were significantly higher in the right flanks were significantly lower that the left flanks (Fig. 2 E and F).

Over-expression of MiR-125a Repressed HCC Cell Invasion and Metastasis
Because the low expression levels of miR-125a in HCC patients were markedly correlated with HCC metastasis, we postulated that increased expression of miR-125a could inhibit the invasion and metastasis of HCC cells. As shown in Figures 3A and 3B, overexpression of miR-125a significantly inhibited the migratory and invasive abilities of HepG2 and HCC-LM3 cells (P , 0.01). Consistent with the in vitro observations, the animal experiment results showed that liver and lung metastases were apparent in mice injected with LM3-NC cells, but few were observed in mice injected with LM3-125a cells ( Figure 3C). Histological analysis revealed that both the number and the size of metastatic nodules in the lungs and livers of mice were significantly lower than those in the controls ( Figure 3D; P , 0.01). Therefore, miR-125a could repress the metastasis of HCC both in vitro and in vivo.

Down-regulation of miR-125a Inceased the Proliferation and Invasion of SMMC-7721 Cells
To study the effects of miR-125a down-regulation on proliferation and invasion of HCC, lentivirus-mediated siRNA targeting the precursor of miR-125a was transfected into SMMC-7721 cells, which showed relative high miR-125a expression among HCC cell lines ( Figure 1B). qRT-PCR analysis showed that miR-125a was significantly deceased in Si-7721 cells ( Figure 4A). The results of MTT revealed that miR-125a inhibition could promote the growth of SMMC-7721 ( Figure 4B). Both migration and invasion assays showed that the migratory and invasive abilities of Si-7721 cells were significantly lower compared with controls ( Figure 4C).

MiR-125a Directly Down-regulated the MMP11 and VEGF-A
To explore the mechanism underlying miR-125a involvment in the progression and metastasis of HCC, miRanda and Pic Tar algorithms were used to search for the target genes. The analysis revealed several tumorigenicity-related genes as potential targets of miR-125a, including MMP11, ERBB2, ERBB3,EDN1, VEGF-A, MMP14 and BCL-9L. Therefore, we constructed luciferase reporters carrying the 39-UTR with the putative miR-125a binding sites for each of those genes. Luciferase assays showed that the 39-UTRs of both MMP11 and VEGF-A, but not others, caused a significant reduction in the luciferase activity ( Figure 5A); however, these effects disappeared with the deletion of key seed regions in the 39UTR of MMP11 or VEGF-A ( Figure 5B). The miR-125a binding sites in the 39UTRs of MMP11 and VEGF-A are shown in Figure 5C. The qRT-PCR analysis showed that the mRNA expression levels of MMP11 and VEGF-A, were significantly inhibited by the over-expression of miR-125a ( Figure 5D). These findings were further verified by the results of western blot analyses, which revealed that over-expression of miR-125a could markedly suppress MMP11 and VEGF, but not ERBB2 or ERBB3, expression in HepG2 and HCC-LM3 cells, and down-regulation of miR-125a could obviously increase the expression of MMP11 and VEGF in SMMC-7721 cells ( Figure 5E). Then HCC-LM3 cells were infected with the miR-125a lentivirus in different MOI. Western blot analysis showed that miR-125a exhibited a dose-dependency on MMP11 and VEGF-A expression in HCC-LM3 cells ( Figure 5G).
We also asked whether miR-125a was correlated with MMP11 and VEGF-A expression in clinical HCC tissues. IHC was performed to detect MMP11 and VEGF-A expression in HCC samples from the 80 patients. Spearman's rank test showed that a significant negative correlation was found between miR-125a and MMP11 protein expression (r = 20.41, P , 0.01),and downregulation of miR-125a was also strongly correlated with upregulation of VEGF-A expression in these HCC tissues (r = 20.59, P , 0.01) ( Figure 5F; Table 4).

Down-regulation of MMP11 or VEGF-A Rescued the Loss of miR-125a Function
The expression of MMP11 and VEGF-A were markedly upregulated in si-7721 cells comapared with SMCC7721 cells ( Figure 5E ). To investigate whether miR-125a works through MMP11 or VEGF-A, the siRNA of MMP11 or VEGF-A was transfected into Si-7721 cells. Western blot showed that MMP11 or VEGF-A was significantly down-regulated in the Si-MMP11 or Si-VEGF cells compared with Si-7721 cells ( Figure 6A). MTT assays showed that knockdown VEGF could significantly inhibit the growth of Si-7721 cells, but inhibition of MMP11 could not significantly influence the growth of Si-7721 cells ( Figure 6B). What's more, RNA interference targeting MMP11 or VEGF-A mRNA could significantly repress the migration and invasion of Si-7721 cells ( Figure 6C). These data indicated that MMP11 or VEGF gene silencing could rescue the loss of miR-125a function in SMMC-7721 cells.

Survival Analysis
MiR-125a, MMP11 and VEGF-A expression levels in HCC samples were used for a survival analysis. The median survival time of these patients was 32 months after operation. A Kaplan-Meier analysis revealed that low miR-125a expression was significantly associated with poor 5-year overall survival time (25.762.7 months vs 45.462.8 months, P , 0.001) (Fig. 7A), and that higher MMP11 expression was another significant prognostic factor for poor survival (44.763.3 months vs 29.762.7 months, P = 0.005) (Fig. 7B). Interestingly, higher expression of VEGF-A was also correlated with poor survival of these patients (29.862.5 months vs 4863.4 months, P = 0.002) (Fig. 7C). Univariate  analysis showed that the patients' survival time was also significantly correlated with the clinical features such as metastasis (P , 0.001), AFP (P = 0.002) and T stage (P = 0.02), but not with age, gender or HBV infection (Table 5; P . 0.05). Multivariate regression analysis showed that AFP, but not others, was an independent prognostic factor.

Discussion
Several studies have revealed that the expression of miRNAs is frequently down-regulated and correlated with the malignant progression of HCC. However, because they can post-transcriptionally regulate hundreds of targets involved in several signal transduction pathways, the roles of miRNAs in the carcinogenesis of HCC are far from being fully be elucidated. Therefore, exploring the clinical potential and molecular mechanism of miRNAs underlying malignant development is significant for beneficial therapies to delay the development and ameliorate the living conditions of HCC patients.
Previous studies have shown that miR-125a expression was down-regulated in several different tumors, including breast cancer, ovarian cancer, lung cancer, medulloblastoma and gastric cancer [11,13,14,21]. The results of human miRNA microarrays have indicated that HCC tissues exhibit lower expression of miR-125a [22,23]. These studies rarely focused on the prognostic values and co-expression of miR-125a with target genes in tumor samples from patients. In the present study, we demonstrated for the first time that miR-125a was frequently down-regulated in HCC compared with matched adjacent liver tissues, and that lower expression of miR-125a was significantly associated with the progression and poor prognosis of patients with HCC. Consistent with the results observed in tissues, significantly lower expression of miR-125a was observed in HCC cell lines compared with QZG cells. Ectopic expression of miR-125a could inhibit the proliferation, migration and invasion of HCC cells both in vitro and in vivo. Inversely, down-regulating miR-125a could increase the proliferation, migration and invasion of SMMC-7721 cells. Interestingly, MTT assays showed that the inhibition of HCC cell proliferation by miR-125a after culture for four days was statistically significant; however the end point of the migration and invasion assays was 24 h, indicating that the inhibitory effects on cell migration and invasion were not caused by a reduction of the cell proliferation. Therefore, further experiments are needed to elucidate the mechanism.
MiRNA-125a can translationally arrest the mRNA of the p53 tumor suppressor gene [24], and interfere with the expression of hepatitis B virus surface antigen [25]. Analyses using the miRanda and Pic Tar algorithms indicated that several cancer-associated genes including MMP11, ERBB2, ERBB3, EDN1, VEGF-A, MMP14 and BCL-9L, were potential target genes of miR-125a. Interestingly, over-expression of miR-125a did not significantly repress the expression of ERBB2 or ERBB3, which were demonstrated to be critical for inhibiting the proliferation and metastasis of breast cancer [11]. Instead, luciferase reporter assays confirmed that MMP11 and VEGF-A, but not other genes tested, were targets genes of miR-125a. These findings were further strengthened by the results of qRT-PCR and western blot analyses. Moreover, the co-expression of miR-125a with its targeted genes MMP11 and VEGF-A were also detected in HCC tissues from surgical ablation. The results showed that miR-125a was inversely correlated with both MMP11 and VEGF-A expression in HCC tissues. Based on these data, we suspected that miR-125a could inhibit the proliferation and metastasis of HCC partly through down-regulating MMP11 and VEGF-A.
Previous studies have shown that increased VEGF expression in HCC is associated with a high proliferative index, poor  encapsulation of tumors, and venous tumor emboli portal vein thrombosis [26,27]. SiRNA targeting VEGF can inhibit HCC growth and tumor angiogenesis in vivo [28], and has anti-tumor efficacy for patients with HCC under hypoxic conditions [29].
Our study showed that down-regulation of MMP11 or VEGF-A could rescue the lost of miR-125a function. VEGF expression was significantly correlated with T stage and metastasis in HCC, and MMP11 expression was significantly correlated with the metastasis of HCC. What's more, our present study showed that lentivirusmediated MMP11 or VEGF siRNA can rescue the lost of miR-125a function on SMMC-7721 cells, further confirming that miR-125a works through MMP11 or VEGF-A.
In conclusion, our present study showed that miR-125a is frequently down-regulated and inversely correlated with aggressiveness and poor prognosis in HCC. Ectopic expression of miR-125a can inhibit the proliferation, invasion and metastasis of HCC by down-regulating MMP11 and VEGF in vitro and in vivo, and thus, miR-125a represents a new prognosis marker and a promising approach for HCC treatment.