MicroRNA-182 downregulates metastasis suppressor 1 and contributes to metastasis of hepatocellular carcinoma

Informed consent was obtained from all the patients for the collection of liver specimens, and the study protocol was approved by the Ethics Committee of Tianjin Medical University. The investigations were conducted according to the Declaration of Helsinki Principles. The clinical pathological data were collected as described in our earlier study [2]. Eighty-six primary HCC patients treated in Cancer Hospital of Tianjin Medical University between 2004 and 2007 were selected according to the following criteria: (1) The diagnosis of HCC was confirmed by pathology; (2) No preoperational chemotherapy or TAE were performed; (3) All of the samples were from the hepatectomy for the first time; (4) Incisal margins were negative; and (5) Clinicopathologic data of the cases could be collected. Among the 86 patients, there were 67 men (77.9%) and 19 women (22.1%). The mean age at diagnosis was 50.7 ± 9.7 years, ranging from 29 to 78 years. HBV was positive in 72 patients (83.7%). The percentage of AFP (>100 ng/ml) was 80.2%. Moreover, one tumor was detected in 79.1% (68/86) patients and multiple tumors (≥2) were found in 20.9% (18/86) patients with totally 29 metastatic leisions. The average tumor size was 5.8 ± 2.7 cm (0.4-16 cm). Histologically, 32.6% (28/86), 46.5% (40/86) and 20.9% (18/86) tumors were grade 1, 2 and 3, respectively. No chemotherapy was performed after radical resection. Patients were followed-up at the outpatient clinic with measurement of the serum alpha-fetoprotein level and hepatic ultrasonography every 2–4 months from the date of initial treatment. The mean time of follow-up was 28.3 months (range 3– 56 months). When recurrence was suspected, further evaluations were performed by abdominal computed tomography (CT) scan, if necessary, by ultrasound-guided biopsy to confirm the diagnosis. Recurrence was observed in 46.5% (40/86) patients. HCC and non-neoplastic tissues were collected and stored at −80°C until analysis. For every frozen tumor tissue, we cut frozen slide and did HE staining and evaluated the percentage of tumor cells. The percentage of tumor cells was about 90%. In addition, paraffin-embedded HCC tissues were also collected.

Total RNA, including miRNA, was extracted using TRIzol reagent (Invitrogen, Carlsbad, CA, USA) according to the manufacturer’s instructions. Total RNA was reversely transcribed using the corresponding RT Primer and the TaqMan MicroRNA Reverse Transcription Kit (Applied Biosystems). The expression of miR-182 and its control RNU44 were detected using TaqMan miRNA assay system (Applied Biosystems, Foster City, CA, USA). The median miRNA intensity value of 86 patient samples was used as the threshold, and patients were divided into two groups (below median, group low miR-182 and above median, group high miR-182) according to the expression of miR-182.

Immunohistochemistry (IHC) was used to detect MTSS1 expression in paraffin-embedded HCC tissues. Five-μm sections of paraffin-embedded HCC tissue were baked at 65°C for 2 h, followed by deparaffinization using standard procedures. After antigen retrieval, MTSS1 antibody (Cell Signaling Technology, Inc. Danvers, MA, USA) was applied to slides, followed by the secondary antibody conjugated with horseradish peroxidase. Signals were revealed by using the Histostain Plus kit (Invitrogen, Grand Island, NY, USA) according to the manufacturer's instruction. 3, 3-Diaminobenzidine (DAB) was used as a chromogen. The sections were counter-stained with hematoxylin. We prepared a negative control by substituting PBS for the antibody.

MTSS1 protein expression was evaluated by two pathologists. MTSS1-positive samples were defined as those with brown staining in the cytoplasm. The results of MTSS1 immunohistochemical analysis were estimated with semi-quantity method. The staining intensity was graded on a scale from 0 to 3 (0 for no staining, 1 for weak immunoreactivity, 2 for moderate immunoreactivity, and 3 for strong immunoreactivity) The percentage of immunoreactivity was scored on a scale from 0 to 4 (0, no positive cells; 1, <25% of cells positive; 2, 25%–50% of cells positive; 3, 50– 75% of cells positive; and 4, >75% cells positive). Finally, a total score (negative: 0; weak: 1–2; medium: 3–5; strong: 6–7) was obtained by adding the scores of staining intensity and percentage positivity.

Cell lysates were harvested with 2% sodium dodecyl sulfate (SDS)-125 mM Tris/HCl (pH 7.4). Cell lysates (25–30 ug of protein) were resolved in Tris/glycine SDS/PAGE gels and transferred to PVDF membranes. Membranes were probed with primary antibodies overnight at 4°C and incubated with horseradish-peroxidase-coupled secondary antibodies (Santa Cruz Biotechnology, Santa Cruz, CA, USA). The background was subtracted, and the signals of the detected bands were normalized to the amount of loading control β-actin (Cell Signaling Technology, Inc. Danvers, MA, USA) band. The protein levels were quantified using ImageJ software (National Institute of Mental Health, Bethesda, MD, USA. http://​rsb.​info.​nih.​gov/​ij).

Human HCC cell lines HLE, HLF, HepG2, Hep3B and HUH-1 were obtained from American Type Culture Collection (Manassas, VA, USA) and cultured in DMEM (Invitrogen) except HepG2 (MEM) supplemented with heat-inactivated 10% fetal bovine serum (Invitrogen) at 37°C in a humidified incubator containing 5% CO₂.

For transfection, 2 × 10⁵ HLF or HUH-1 cells were seeded into each well of a 6-well plate and incubated overnight, then the cells were transfected with Pre-miR miRNA Precursor Molecule pre-182 (pre-miR-182) and anti-miR miRNA inhibitor anti-182 (anti-miR-182) (Applied Biosystems) at a final concentration of 100 nM using the Lipofectamine 2000 transfection reagent (Invitrogen, Carlsbad, CA, USA), according to the manufacturer’s instructions. The specificity of the transfection was verified using the Pre-miR miRNA Precursor Molecule Negative Control #1 (control pre-miR) and Anti-miR miRNA Inhibitors Negative Control #1 (control anti-miR) (Applied Biosystems). The expression levels of miR-182 and MTSS1 were quantified 24 h after transfection, and the cells were used for western blot analysis.

The human MTSS1 3’ UTR luciferase reporter construct (MTSS1-3’UTR WT) was generated by cloning MTSS1 mRNA 3’UTR sequence into downstream of pMIR-Report construct (Ambion, Foster City, CA, USA). The MTSS1 3’ UTR sequence was generated by PCR using primer MTSS1 3’UTR F SpeI: 5’-AAACTAGTTGATTTTTCTGAAGGT GCCAAATTCCATTTAA-3’ and primer MTSS1 3’UTR R SacI: 5’–GGGAGCTCTTTGGCAACATTTTATTTATTCA-3’. The miR-182 target site-mutation MTSS1 3’ UTR luciferase reporter 1 (MTSS1-3’UTR mutation 1) construct was generated by employing direct-site mutagenesis using mutation primers which mutate the miR-182 binding site from TCTGAAGGTGCCAA to GATGAAGGTCGGTA. miR-182 target site-mutation MTSS1 3’ UTR luciferase reporter (MTSS1-3’UTR mutation 2) was mutated from TTGCCAA to TAACGCT in the miR-182 binding site. MTSS1-3’UTR mutation 1, 2 was mutated in these two miR-182 binding sites.

HUH-1 cells were co-transfected with miR-182 plasmid and wild-type or mutant MTSS1 3’ UTR luciferase reporter construct and luciferase activities were measured using the Dual-Glo Luciferase. Data were normalized by dividing Firefly luciferase activity with that of Renilla luciferase.

HLF and HUH-1 cell invasion assays were performed using 24-well Matrigel Invasion Chambers (BD Biosciences, CA, USA). The lower chambers were filled with 0.75 ml of DMEM medium containing 10% fetal bovine serum (FBS). A cell suspension of 2 × 10⁵ in 0.5 ml DMEM medium was added into each well of the upper chamber. After the cells were incubated for 24 h at 37°C in a humidified incubator with 5% CO₂, The invasive cells attached to the lower surface of the membrane insert were fixed in 10% formalin at room temperature for 5 min and stained with 0.05% crystal violet. The non-invading cells that remained on the upper surface of the membrane were removed by scraping. The number of invasive cells on the lower surface of the membrane was then counted under a microscope.

Differences in MTSS1 immunohistochemical staining between groups were compared using chi-square or Fisher exact tests in human samples. The correlation between MTSS1 expression and miR-182 was evaluated by calculating the Spearman rank correlation coefficient. Moreover, mean ± SD of clinicopathological variables were calculated, and differences in the means were analyzed using one-way analysis of variance or Student’s t test. We also used the Kaplan-Meier method and the log-rank test in univariate survival analysis, and we used the Cox proportional hazards regression model in our multivariate analysis. SPSS version 16.0 (IBM) was used to perform our statistical analysis. Two-tailed P values < 0.05 were considered statistically significant.

To investigate the role of miR-182 in HCC development, we tested the expression of miR-182 in 86 HCC and matched non-neoplastic tissues (Figure 1A). The relative expression of miR-182 in HCC samples (2.21 ± 1.29) was significantly higher than that of matched normal tissues (1.12 ± 0.47) (p < 0.01) (Figure 1B). Hence, we considered the up-regulation of miR-182 may contribute to HCC tumorigenesis. Furthermore, the relative level of miR-182 in poorly differentiated HCC (3.28 ± 1.79) was almost one time higher than that in well (1.62 ± 0.68) and medium differentiated cases (2.14 ± 0.83) (Figure 1B), which suggested that miR-182 might also correlate with the progress of HCC.

For better understanding the potential role of miR-182 in HCC progression, we analyzed its correlation with some clinicopathological variables including age, sex, HBV infection, AFP, tumor number, tumor size, expression of MTSS1, histological grade, portal vein invasion and recurrent time (Table 1). Based on the median value (1.92) of miR-182 expression, all patients were divided into two groups including group with low expression of miR-182 and group with high expression of miR-182. Intra-hepatic metastasis (tumor number ≥2, p = 0.034) and higher recurrence (p = 0.031) tended to occur in the patients with high expression of miR-182. Though the p values did not reach statistical significance, the patients with high expression of miR-182 had a tendency to undergo occur portal vein invasion (p = 0.096) (Table 1).

The recurrent percentage is 46.5% (40/86) for all patients. The median disease-free survival time in group with low miR-182 and group with high miR-182 was 27.0 months and 24.0 months, respectively. The Kaplan–Meier method revealed that higher miR-182 expression level correlated with significantly reduced disease-free survival (42.0 ± 2.93 months in group with low miR-182 versus 31.2 ±2.79 months in group with high miR-182, p = 0.039) (Figure 2). Multivariate survival analysis revealed that multiple tumors (p = 0.023) and high expression of miR-182 (p = 0.022) were significantly correlated with the poor prognosis of HCC patients (Table 2). The result further indicated the importance of miR-182 up-regulation in HCC development.

MTSS1 protein expression was tested with IHC in HCC and paired normal tissues (for some cases, there are tumor and adjacent normal tissue in the same slide). MTSS1 was positive in the cytoplasm of tumor and normal liver cells. MTSS1 was often highly expressed in normal tissue (Figure 3 A, B and E), while drastically reduced MTSS1 expression was shown in the tumor cells (Figure 3 B, D and E). Totally, MTSS1 was positive in 79% (68/86) of normal tissue and in 43% (37/86) of HCCs (43%). The rate of MTSS1 positive case in HCC was significantly lower than that in paired normal tissue (p < 0.001, Figure 3G). Among the 37 cases with MTSS1 positive expression, 13 (35%), 17 (46%) and 7 (19%) cases showed weak, mederate and strong expression of MTSS1, respectively. Moreover, no difference of MTSS1 expression was found among the multiple lesions in the same patient. The MTSS1 positive rate in metastatic HCC (17%, 3/18) was significantly lower than that in non-metastatic HCC (50%, 34/68) (p < 0.001, Figure 3G). In addition, the tumor thrombus in small hepatic vein also showed low expression of MTSS1 (Figure 3 F).

For the MTSS1 positive cases tested by IHC, the examination of Western Blot were further performed and the expressions were quantified (Figure 4 A). Negative correlation between the expression of miR-182 and that of MTSS1 in HCC was indicated in Figure 4 B (r = −0.673; p < 0.01), which suggested MTSS1 maybe one important functional protein contributing to the oncogenic role of miR-182. Meanwhile, the negative correlation between miR-182 and MTSS1 expression were also found in HCC cell lines (r = −0.931, p = 0.021) (Figure 4 C and D).

Next, we sought to investigate the molecular mechanism responsible for the oncogene effect of miR-182 on HCC observed above. As miRNAs function mainly through inhibiting their target mRNAs by binding to the 3’ UTR, we searched the putative target genes of miR-182 in Target Scan and Pictar. In both websites, 841 and 702 conserved targets were found, respectively. Among those targets, human MTSS1, known to have critical roles in the inhibition of cancer metastasis, contained two putative conserved miR-182 binding sites with high context scores (Figure 5A). To verify whether MTSS1 was a direct target of miR-182, a dual-luciferase reporter system was used by co-transfection of miR-182 and a luciferase reporter plasmid containing the 3’ UTR of human MTSS1 into HUH-1. As shown in Figure 5B, the luciferase activity was significantly inhibited by miR-182 co-transfection, mutation either of the two miR-182 binding site, while miR-182 failed to inhibit the expression of luciferase construct with both binding sites mutated, suggesting that miR-182 could directly target on the 3’ UTR of MTSS1.

As one target gene of miR-182 demonstrated above, the expression of MTSS1 was down-regulated in HUH-1 with transfected miR-182 and up-regulated in HLF with transfected anti-miR-182 (Figure 6 A). An in vitro invasion assay indicated that the relative invasiveness of HLF transfected with anti-miR-182 was specifically reduced by approximately 41% (p < 0.05) and the relative invasiveness cells of HUH-1 transfected with miR-182 was increased by approximately 36% (p < 0.05) (Figure 6B). The result in vitro further demonstrated that miR-182 could promote metastasis of HCC and inhibited the expression of MTSS1.