The human OS cell line MG63 was purchased from the American Type Culture Collection (Manassas, VA, USA). MG63 cells were cultured in Dulbecco's modified Eagle's medium (Gibco; Thermo Fisher Scientific, Inc., Waltham, MA, USA) containing 10% fetal bovine serum (FBS, Gibco; Thermo Fisher Scientific, Inc.), penicillin and streptomycin at 37°C under a humidified atmosphere containing 5% CO₂. A total of 32 OS tumor samples, paired with adjacent normal tissues, were obtained from 12 females and 20 males patients aged 19–61 who underwent surgery at Shanghai Putuo District Central Hospital (Shanghai, China) between February 2014 and January 2017. Approval from the Institute Research Ethics Committee of Shanghai Putuo District Central Hospital was obtained for the use of these clinical materials for research purposes and written informed consent was obtained from patients prior to surgery. Bioinformatics analysis (http://starbase.sysu.edu.cn/index.php) was used to predict the potential targets of RNA binding protein HuR.

mRNA expression was confirmed via RT-qPCR analysis. Briefly, total RNA was extracted from the tissues and the MG63 cell line using TRIzol^® reagent (Invitrogen; Thermo Fisher Scientific, Inc.). The first strand cDNA was synthesized using the RevertAid^™ First Strand cDNA Synthesis kit (Thermo Fisher Scientific, Inc.) according to the manufacturer's protocols. The mRNA expression levels were determined on the ABI Prism 7500 Detection system (Applied Biosystems; Thermo Fisher Scientific, Inc.) with Universal SYBR^® Green mix (Vazyme, Piscataway, NJ, USA). The denaturing process was 95°C for 5 min, the annealing process was 58°C for 30 sec and the elongation process was 72°C for 30 sec. There were 35 cycles of this RT-qPCR performed. Relative mRNA expression was calculated with the 2^−∆∆Cq method (13). The primers for RT-qPCR were indicated in Table I. Melting curve analysis and 1% agarose gel electrophoresis were used to check amplification specificity and length of PCR products. GAPDH and U6 snRNA served as an internal control for mRNA and miRNA expression, respectively.

Following different treatments, proteins were extracted using the protein extraction kit (Nanjing KeyGen Biotech Co., Ltd., Nanjing, China). Protein concentration was examined via the Bradford assay. The detailed procedure was referred to in the previous study (9). A total of 30 µg of protein was analyzed on 10% SDS-PAGE and electrotransferred to PVDF membranes (Millipore, Bedford, Massachusetts). The membranes were blocked in 5% non-fat dried milk for 60 min at room temperature. Primary antibodies against HuR (cat. no. ab136542), HMGA1 (cat. no. ab129153), Cleaved-caspase3 (cat. no. ab49822), caspase3 (cat. no. ab13847), epithelial (E)-cadherin (cat. no. ab1416), Vimentin (cat. no. ab8978) and β-actin (cat. no. ab8227) were purchased from Abcam (Cambridge, UK), and the dilution ratio of all antibodies was 1:5,000. Following incubating with primary antibodies, blots were washed with tris-buffered saline with 0.5% Tween 20 and incubated with a peroxidase-conjugated secondary Goat Anti-Rabbit (cat. no., A0208, Beyotime Institute of Biotechnology; dilution rate: 1:5,000) and Goat Anti-Mouse antibody (cat. no., A0216, Beyotime Institute of Biotechnology; dilution rate: 1:5,000), and chemiluminescence was detected using an enhanced chemiluminescence kit (Thermo Fisher Scientific, Inc.) followed by exposure in Bio-Rad ChemiDoc^™ MP system (Bio-Rad Laboratories, Inc., Hercules, CA, USA). The protein expression level was normalized to β-actin.

Cell Counting Kit-8 (CCK-8; Shanghai Yeasen Biotechnology Co., Ltd., Shanghai, China) was used to analyze the cell viability rate of cells with different treatments, in which cells without HuR knockdown were used as control. In brief, cells were suspended and 3,000 cells/well were seeded into 96-well plates, following culturing with DMEM medium with 10% FBS (Gibco; Thermo Fisher Scientific, Inc.) for 24, 48 and 72 h. CCK-8 was added into the DMEM, following the manufacturer's protocol. The absorbance was detected at 450 nm. The cell viability rate was presented with the value relative to the control group. Experiments were repeated at least three times.

Cell apoptotic rate was analyzed using the Annexin V-FITC and propidium iodide (PI) kit (Beyotime Institute of Biotechnology, Haimen, China) via flow cytometry. Cells with different treatments were stained with Annexin V-FITC and PI, followed with flow cytometry analysis using a BD FACSCanto II (BD Biosciences, Franklin Lakes, NJ, USA), following the manufacturer's protocol. FlowJo software (version 10.0.7; FlowJo LLC, Ashland, OR, USA) was used to analyze the data which were expressed as cell percentage.

HMGA1 3′UTR sequences were inserted into the PMIR-Reporter plasmid (cat. no., AM5795; Thermo Fisher Scientific, Inc.), denoted as Luc-HMGA1-wt. PCR primers for Luc-HMGA1-wt were indicated in Table I. PMIR-Reporter plasmid holding HMGA1 3′UTR sequences with the mutated binding site of miR-142-3p was obtained with the Site-directed Gene Mutagenesis kit (Beyotime Institute of Biotechnology) and noted as Luc-HMGA1-mut. These aforementioned plasmids were co-transfected with β-gal control plasmid plus Lenti-HuR infection into MG63 cells using Lipofectamine^® 2000 reagent (Invitrogen; Thermo Fisher Scientific, Inc.), according to the manufacturer's protocol. A total of 48 h later, cells were lysed with Reporter lysis buffer (cat. no, E397A; Promega Corporation, Madison, WI, USA) and luciferase activity was measured with VivoGlo Luciferin kit (cat. no., P1041; Promega Corporation) using a luminometer (Thermo Fisher Scientific, Inc.). The luciferase activity was normalized to β-gal activity.

shRNAs against HuR and a scramble non-targeting shRNA were purchased from Santa Cruz Biotechnology, Inc. (Dallas, TX, USA). The shRNA sequences were inserted into pLKO.1 (Sigma-Aldrich, Merck KGaA), termed as Lenti-HuR-shRNA. Additionally, HMGA1 and HuR sequences were inserted into pLVX–IRES-ZsGreen1, termed as Lenti-HMGA1 and Lenti-HuR, respectively. PCR primers for plasmids construction were mentioned in Table I. Lentivirus were packaged in HEK293T cells via co-transfecting Lenti-HuR-shRNA or Lenti-HMGA1 or Lenti-HuR with pCMV-dR8.2 and pMD2.G constructs using Lipofectamine 2000. Following 72 h, the supernatants were collected and ultrafiltrated. The virus supernatants (10 µl in 10⁸ TU/ml) were added to MG63 cells with 2 µg/ml Polybrene (Genomeditech, Shanghai, China).

HuR was knocked down via infecting with Lenti-HuR-shRNA for 48 h at 37°C. Then, 5 µg/ml ActD (ApexBio, Houston, TX, USA) was added to inhibit the de novo RNA synthesis. Total RNA was collected at 2, 4 and 6 h and mRNA expression was determined via RT-qPCR analysis under the same conditions as previously described. The mRNA half-life was determined by comparing to the mRNA level prior to adding ActD.

MG63 cells with or without HuR knockdown were lysed with 25 mM Tris-HCl buffer (pH 7.5) and 100 U/ml RNase inhibitor (Sigma-Aldrich; Merck KGaA, Darmstadt, Germany), and then incubated with protein-A Sepharose beads (Genescript, Nanjing, China) precoated with 2 µg anti-HuR antibody (as previously stated, dilution rate, 1:100), or control rabbit IgG (cat no. ab191867, Abcam, dilution rate, 1:100) was used for negative control for 3 h at 4°C. The RNA-protein complexes were pulled-down by protein A/G agarose beads (Genescript) and RNA was extracted with TRIzol, followed by detecting the HMGA1 expression level with RT-qPCR assay, as previously described. This experiments were repeated at least three times.

All data were obtained from at least three independent experiments (n≥3), and presented as the mean ± standard deviation. Datasets with only two groups were analyzed using Student's t-test. Differences among multiple groups were analyzed using one-way analysis of variance with the Tukey's post-hoc test, and P<0.05 was considered to indicate a statistically significant difference.

Firstly, the expression level of HuR in OS tissues and normal adjacent tissues was detected. RT-qPCR and western blot analysis demonstrated that HuR expression was increased in OS tissues, compared with normal adjacent tissues (Fig. 1A and B). An identical result was obtained in a further immunohistochemical assay (Fig. 1C). These results indicated that HuR promoted OS progression.

Since the expression level of HuR was upregulated in OS tissues, HuR expression was knocked down in OS cells. RT-qPCR and western blot analysis confirmed the knockdown efficiency of Lenti-HuR-shRNA (Fig. 2A and B). CCK-8 analysis indicated that HuR knockdown significantly inhibited MG63 cell viability (Fig. 2C) and the expression of proliferation marker Ki67 (Fig. 2D). A further cell apoptosis assay indicated that the knockdown of HuR facilitated cell apoptosis (Q2 and Q3) in MG63 cells (Fig. 2E) and promoted the expression of the apoptosis executor, cleaved caspase 3 (Fig. 2F). Additionally, HuR knockdown notably suppressed the EMT process in MG63 cells, characterized by the downregulation of the expression of the mesenchymal marker Vimentin and upregulation of the expression of the epithelial marker E-cadherin (Fig. 2G and H). These results demonstrated that HuR knockdown inhibited OS progression.

As HuR belongs to the group of RBPs, which bind to mRNA, and enhance mRNA stability and expression (7), targets were identified via bioinformatics analysis (http://starbase.sysu.edu.cn/index.php). HMGA1 has been identified to promote the progression of OS (11). RT-qPCR and western blot analysis demonstrated that the knockdown of HuR significantly decreased HMGA1 expression in MG63 cells (Fig. 3A and B). Furthermore, HMGA1 mRNA stability was reduced with HuR knockdown (Fig. 3C). Thus, it was speculated that HuR may directly bind to HMGA1 in MG63 cells. HuR expression was induced in MG63 cells, with the HuR-binding complex knocked down via the HuR antibody, followed by examination of the bound mRNAs via RT-qPCR. RIP results indicated that HuR had a significantly greater association with HMGA1 compared with the control (Fig. 3D). Since the miR-142-3p/HMGA1 axis has been confirmed in OS cells (9), whether HuR was involved in the miR-142-3p/HMGA1 regulatory pathway in MG63 cells was further investigated. As depicted in Fig. 3E and F, overexpression of miR-142-3p significantly attenuated the promoter effect of HuR on HMGA1 expression. By contrast, knockdown of miR-142-3p reversed the inhibitory effects of HuR knockdown on HMGA1 expression (Fig. 3G and H). Additionally, mutation of the miR-142-3p binding site on HMGA1 3′UTR partially reversed the enhancement of HuR on the luciferase activity of Luc-HMGA1-wt (Fig. 3I). Overall, these results indicated that HuR promoted HMGA1 expression via competitively binding to HMGA1 3′UTR with miR-142-3p.

Whether miR-142-3p/HMGA1 axis was involved in the promotion of HuR in OS progression was further investigated. As depicted in Fig. 4A, infection with Lenti-HMGA1 significantly upregulated the HuR expression level, reversing the suppressive effects of HuR knockdown on HMGA1 expression. Cell viability and apoptosis assays indicated that overexpression of HMGA1 or transfection with miR-142-3p inhibitor attenuated the inhibition of HuR knockdown on cell proliferation and promotion on cell apoptosis (Fig. 4B-D). Additionally, HMGA1 overexpression or miR-142-3p knockdown notably attenuated or even reversed EMT process inhibited by HuR knockdown (Fig. 4E). Notably, HuR expression was positively associated with HMGA1 expression, and negatively associated with miR-142-3p expression in OS tissues (Fig. 4F and G). Therefore, the results of the present study results indicated that HuR promoted OS progression at least partly through the miR-142-3p/HMGA1 axis.