Rabbit polyclonal antibodies to IL-6 receptor (IL-6R) and proliferating cell nuclear antigen (PCNA) were acquired from Abcam (Cambridge, UK). Antibodies against signal transducer and activator of transcription 3 (STAT3) and phosphorylated STAT3 (p-STAT3-Tyr705) were obtained from Cell Signaling Technology. Rabbit polyclonal antibodies against VEGF receptor 2 (VEGFR2) and phospho-Tyr1175-VEGFR2 were purchased from Santa Cruz Biotechnology (Santa Cruz, CA, USA). The anti-VEGF antibody was from (R&D Systems, Minneapolis, MN, USA). Antibodies against ERK1/2 and phospho-T202/Y204-ERK1/2 were from BD Biosciences (Franklin Lakes, NJ, USA).

The human hepatoma cell lines HepG2 and HEK293T were purchased from American Type Culture Collection (ATCC; Rockville, MD, USA). The umbilical vein endothelial cells (HUVECs) were obtained from AllCells (Shanghai, China). HepG2 and HEK293T cells were incubated with Dulbecco's modified Eagle's medium (DMEM) (Invitrogen Corp., Grand Island, NY, USA) containing 10% fetal bovine serum. HUVECs were cultured in RPMI-1640 medium (Gibco BRL, Gaithersburg, MD, USA) supplemented with 10% fetal calf serum, 100 U/ml penicillin and 100 g/ml streptomycin. All cells were incubated in a humidified atmosphere at 37°C with 5% CO₂.

Lentivirus-based plasmids for constitutive expression of miR-451 or scrambled miRNA (miR-con, used as NC) and the virus packaging kit were obtained from GeneCopoeia (Rockville, MD, USA). Following co-transfection into HEK293T using EndoFectin Lenti transfection reagent according to the manufacturer's instructions About 48 h later, virus titers were collected and evaluated by the p24 ELISA kit (Cell Biolabs, Inc., San Diego, CA, USA). Then, the obtained lentiviral particles were transduced into HepG2 cells for 24 h. The stable miR-451-overexpressing cells were selected using puromycin for additional 3 days.

HepG2 cells with miR-451 were preconditioning with pcDNA-IL-6R lacking 3′UTR, pIRES-STAT3β [a dominant-negative STAT3 (STAT3D)] or vehicle (GeneChem, Shanghai, China). About 12 h later, cells were incubated in DMEM medium for further 14 h. Then, the TCM was centrifuged sequentially at 500 g to discard the detached cells, followed by 12,000 g centrifugation to remove cell debris at 4°C for 15 min. The supernatant was then gathered and stored at −80°C for the subsequent experiments.

The wild-type (wt) and mutant (mut) 3′UTR of IL-6R predicted to interact with miR-451 were constructed and cloned to the firefly luciferase-expressing vector psiCHECK™ (Promega, Madison, WI, USA). HEK293T cells were seeded into a 96-well plate and co-transfected with wt-IL-6R or mut-IL-6R 3′UTR reporter vector, 5 ng of pRL-TK, and miR-451 or miR-con with the help of Lipofectamine 2000 reagent (Invitrogen Corp., Carlsbad, CA, USA). After 48 h incubation, luciferase activities were detected by Dual-Luciferase Reporter system (Promega).

After treatment under various conditions, the HepG2 cells were lysed with TRI reagent (Sigma) to extract total RNA. Then, the reverse transcription was performed to synthesize the cDNA using a High Capacity cDNA Archive kit (Applied Biosystems, Foster City, CA, USA). The subsequent qRT-PCR was carried out to evaluate the relative expression of mRNA using the miScript SYBR^®-Green PCR kit (Qiagen, China) and SYBR-Green I (Molecular Probes, Invitrogen Corp.) for miR-451 and other molecules. All reaction conditions and processes were implemented according to the manufacturer's instructions. The specific primers for miR-451, IL-6R and VEGF were used as previously published (8,18). The expression levels were normalized using U6 for miR-451 and β-actin for other genes. All data were analyzed using the 2^−ΔΔCt equation.

Cells were solubilized in lysing buffer (Beyotime, Nantong, China) and the extracted protein concentration was measured using the BCA assay (Pierce, Rockford, IL, USA). Then, about 40 µg of proteins was subjected to 12% SDS-PAGE, followed by the transfer to PVDF membrane (Millipore, Bedford, MA, USA). After blocking with 5% non-fat milk, the membrane was probed with primary antibodies against human IL-6R, PCNA, VEGF, STAT3, p-STAT3, VEGFR2, p-VEGFR2, ERK and p-ERK. Then, HRP-conjugated secondary antibodies were added for further incubation of 1 h. To visualize the bound antibodies, the LumiGLO reagent (Pierce) was introduced. The β-actin was used as protein loading control. All band intensities were quantified using a Gel Doc™ XR imaging system and Quantity One (Bio-Rad, USA).

Cell viability was monitored by Cell Counting kit (CCK)-8 (Dojindo, Kumamoto, Japan). Briefly, HUVECs were seeded onto 96-well plates at the density of 5×10³ cells/well. Then, cells were incubated with the TCM collected from different background of miR-451 expression for the indicated times (12, 24 and 36 h). Subsequently, 10 µl of CCK8 reagents were added for further 2 h incubation at 37°C. The absorbance at 450 nm was measured to assess the number of viable cells by a Safire 2 microplate reader (Tecan, Switzerland). Relative cell viability was shown as the absorbance percentage of the treatment group to the control group.

Cell migration was evaluated using the scratch wound assay. After seeding in 24-well plates, HUVECs were cultured with various TCM. Then, a single scratch wound was formed by scraping the cell layer using a tip of 200 µl pipette. About 24 h later, the scratch wounds were visualized by an inverted microscope. The scratch wound width was quantified to assess cell migration ability of HUVECs using the ImageJ software.

The 96-well culture plates were precoated with Matrigel (BD Pharmingen, San Jose, CA, USA) overnight. Then, HUVECs were seeded into the plate at the density of 1×10⁴ cells/well and cultured in the absence or presence of various TCM from HepG2 cells. The formation of capillary-like structures were then analyzed at 24-post incubation and photographed under an inverted microscope. Tube formation was evaluated by counting branch points in five random fields per well.

For xenograft implantation experiments, male BALB/c nude mice aged 4–6 weeks were used and obtained from the Hunan Slac Jingda Laboratory Animal Co., Ltd. (Changsha, China). All animals were housed under specific pathogen-free conditions and used according to the guidelines of the National Institutes of Health Guide for the Care and Use of Laboratory Animals. Animal care and procedures were approved by the Institutional Animal Care and Use Committee of the First Affiliated Hospital of Xi'an Jiaotong University. HepG2 cells (5×10⁶) transfected with miR-451 or control were subcutaneously injected into mice to establish xenograft models. Tumor size was detected every 4 days and the tumor volumes (10 animals/group) were calculated with the following formula: Tumor volume = (largest diameter x perpendicular height²)/2. Five weeks later, the animals were euthanized using sodium pentobarbital and tumors were removed.

Tumors from mice were collected and fixed in formalin. Then, the specimens were processed for paraffin embedding and subsequently cut into standard 6-µm sections. To evaluate angiogenesis, immunofluorescence was performed. After blocking endogenous peroxidase activity and non-specific bind, the primary antibodies against CD31 (eBioscience) were added. Then, the samples were incubated with biotin-linked donkey anti-rat and Texas Red Streptavidin (Jackson ImmunoResearch, West Grove, PA, USA), followed by the counterstain with DAPI (Sigma). The specimens were photographed under a Zeiss LSM 510 confocal microscope and blood vessel areas were calculated by the following formula: % Area = total red signal/total DAPI signal.

The serum from the mice and TCM were collected. Then, the equal volume of samples was subjected to ELISA to determine the VEGF concentration using a commercial VEGF ELISA kit (R&D Systems). All procedures were performed according to the manufacturer's instructions.

Data were analyzed by SPSS 11.0 and shown as means ± standard deviation (SD). All experiments were performed at least three times. Comparisons among different groups were analyzed based on Student's t-test and ANOVA. P<0.05 was considered as statistically significant.

To investigate the biological significance of miR-451 in HCC angiogenesis, its effect on cell proliferation, migration and tube formation of HUVECs were explored. As shown in Fig. 1A, a pronounced increase of miR-451 was validated in HepG2 cells after transfection with miR-451. Further analysis demonstrated that incubation with TCM from miR-451-overexpressed HepG2 cells strikingly decreased HUVEC proliferation in a time-dependent manner (Fig. 1B). Consistently, the expression of PCNA, a marker for cell proliferation, was also downregulated in HUVECs when treated with the TCM (Fig. 1C). Moreover, TCM from miR-451-transfected HepG2 cells notably mitigated cell recruitment of HUVECs (Fig. 1D). Importantly, a remarkable inhibition in capillary tube formation of HUVECs was substantiated when HUVECs were grown in TCM obtained from miR-451-elevated HepG2 cells. Accordingly, these data suggested that miR-451 upregulation in HCC cells might inhibit angiogenesis of HUVECs in vitro.

To further elucidate the function of miR-451 on the development of carcinoma, HepG2 cells stably expressing miR-451 or control were subcutaneously injected into BALB/c nude mice. Interestingly, a noticeable decrease in tumor volume was observed in HepG2-miR-451 tumors in contrast to control groups (Fig. 2A), indicating that miR-451 could suppress tumor growth in vivo. The obvious decrease in blood vessel density was corroborated in HepG2-miR-451 groups by detecting the levels of CD31 (Fig. 2B), a common marker for vascular formation. Concomitantly, ectopic expression of miR-451 in HepG2 tumors also triggered an analogous downregulation of VEGF concentration in serum in comparison with HepG2-control groups (Fig. 2C). These results indicated that miR-451 might act as a critical suppressor of angiogenesis in HCC.

Analysis was performed to predict the potential target of miR-451 using publicly available algorithms (TargetScan, PicTar and microRNA.org). Among these genes, IL-6R was identified as a potential target based on a predicted binding site of miR-451 at its 3′UTR (Fig. 3A). It has been reported that IL-6/IL-6R signaling is involved in tumor angiogenesis (8). To clarify the underlying mechanism related to miR-451-mediated inhibitory effect on angiogenesis in tumors, the expression of IL-6R was assessed. As expected, overexpression of miR-451 markedly reduced the mRNA levels of IL-6R in HepG2 cells (Fig. 3B). Simultaneously, a similar decrease of IL-6R protein was also demonstrated following miR-451 transfection (Fig. 3C). Noticeably, the luciferase activity was markedly diminished following co-transfection of miR-451 with wt-IL-6R-3′UTR vector, but not in mut-IL-6R-3′UTR groups (Fig. 3D), indicating that IL-6R was a direct target of miR-451.

Based on the target relationship between miR-451 and IL-6R, we further explored whether miR-451 elicits its inhibitory role in angiogenesis of HUVECs by directly targeting IL-6R. Following transfection with pCDNA-IL-6R lacking 3′UTR, the inhibitory effect of TCM from miR-451-transfected HepG2 cells on HUVEC proliferation was obviously ameliorated (Fig. 3E). Consistently, the decreased migration of HUVECs triggered by miR-451 elevation was also attenuated in the above culture medium (Fig. 3F). Notably, a similar increase in capillary tube formation of HUVECs was also observed when HUVECs were incubated with TCM from HepG2 cells co-transfected with miR-451 and IL-6R. The above data confirmed that miR-451 could suppress angiogenesis of HUVECs in vitro by targeting IL-6R.

VEGF is widely accepted as a vital regulator for angiogenesis. Accumulation evidence corroborates that IL-6/IL-6R exerts an important role in angiogenesis by activating STAT3-VEGF signaling (19,20). To further illustrate the underlying mechanism involved in miR-451-trigged inhibition on tumor angiogenesis, we investigated the expression of VEGF. Consistent with our hypothesis, elevation of miR-451 noticeably abrogated the mRNA level of VEGF (Fig. 4A). Moreover, the concentration of VEGF in conditioned medium of HepG2 cells was also reduced (Fig. 4B). Additionally, miR-451 upregulation significantly inhibited the STAT3 phosphorylation (Fig. 4C). Interestingly, IL-6R upregulation drastically antagonized the reduction of p-STAT3 and VEGF expression trigged by miR-451 overexpression (Fig. 4D). Furthermore, the increased expression of IL-6R obviously upregulated the concentration of VEGF in TCM collected from miR-451-transfected HepG2 cells (Fig. 4E). Concomitantly, blocking STAT3 signaling with STAT3D noticeably suppressed VEGF expression (Fig. 4F) and concentration (Fig. 4G), implying that miR-451 could dampen VEGF production secreted by HCC cells through IL-6R-STAT3 signaling.

Convincing evidence indicates that tumor cell-produced VEGF can induce endothelial cell proliferation, migration and angiogenesis by activating VEGFR2, which then phosphorylates its down-stream ERK and subsequently induces angiogenesis (21). We further assess whether miR-451-decreased VEGF production by IL-6R-STAT3 signaling can abolish the VEGFR2 pathway in HEVECs. Western blotting confirmed the obvious downregulation of p-VEGFR2 and p-ERK in HUVECs incubated with TCM of HepG2 cells that stably overexpressed miR-451 (Fig. 5A and B). Surprisingly, overexpression of IL-6R could remarkably restore the reduction of VEGF levels in TCM from miR-451-transfected HCC cells, which then ameliorated the inhibitory effect on the phosphorylation of VEGFR2 and p-ERK in HUVECs (Fig. 5C). Together, these results demonstrated that miR-451 could block the VEGFR2 pathway in HUVECs, which will lead to reduction in angiogenesis.