Primary tumor samples from 56 patients with osteosarcoma were referred to the Rizzoli Orthopaedic Institute and diagnosed by expert pathologists. A total of 31 were high-grade tumors and 25 were low-grade. Thirty-one patients were males and 25 females with a mean age of 23 years. At a minimum follow-up of 5 years, 21 patients had lung metastases and 15 died of disease (Table I). For each specimen frozen material was available and in all specimens the percentage of tumor cells estimated after hematoxylin-eosin staining was equal or more than 90%. Thirteen normal bone from non-cancer patients were used as control. The research protocol was approved by ethic committees of the Rizzoli Institute where the tumor samples were collected and all patients provided appropriates informed consent.

Human OS cell lines U2-OS, MG-63 and SAOS, and normal osteoblast cells were obtained from the American Type Culture Collection (ATCC, Rockville, MD, USA) and cultured in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum (FBS), L-glutamine (2 mM), 100 U/ml penicillin and 100 μg/ml streptomycin (Invitrogen, Carlsbad, CA, USA) at 37°C in a 5% CO₂ humidified incubator.

Total-RNA was extracted from cell lines and frozen tissues (150 mg) using TRIzol Reagent (Invitrogen) according to the manufacturer’s protocol and stored at −80°C in RNAsecure reagent (Ambion Inc., Austin, TX, USA). Concentration of total-RNA was measured with spectrophotometer and the 260/280 ratio of RNA was 1.8. Purity and quality were identified by a denatured gel electrophoresis.

Total-RNA of 6 low-grade, 6 high-grade OS and 6 healthy tissues was profiled by Exiqon (www.exiqon.com) using miRCURY LNA array vs 11.0 hsa, mmu and rno (13). Briefly, after RNA quality control performed by Bioanalyser 2100 and RNA measurement on the Nanodrop instrument, using the miRCURY™ Hy3/Hy5 Power Labeling kit (Exiqon, Vedbaek, Denmark), experimental samples and a reference sample were labeled in separate reactions with Hy3 and Hy5, respectively. Labeling efficiency was evaluated by analyzing the signals from control spike-in capture probes. Labeled experimental and reference samples were combined, denatured, and hybridized to microarrays at 56°C for 16 h. Low-stringency and high-stringency washes were carried out and the microarrays dried. Images were acquired using the Axon® GenePix 4000B scanner and GenePix software. The data was pre-processed and normalized using the global locally weighted scatterplot smoothing procedure (14).

Analyses were performed in 56 OS tissues (31 high- and 25 low-grade), 13 normal samples, and in OS cell lines and osteoblasts. Reverse transcription and real-time PCR (RT-PCR) were carried out following TaqMan MicroRNA Assay Protocol (Applied Biosystems, Foster City, CA, USA) and the expression of miR-1 (miRNA assay no. 002222), miR-133b (miRNA assay no. 002247) and miR-378^* (miRNA assay no. 000567) were quantified using 2^−ΔΔCt comparative method (Applied Biosystems, User Bulletin no.2 P/N 4303859) and normalized using RNU44 as endogenous reference (miRNA assay no. 001094). Mesenchymal stem cells (MSC) (ATCC) were used as calibrator for cell lines, while normal osteoblasts resulted the most adequate calibrator for our human OS series. Each miRNA was considered more expressed when the value calculated with the method 2^−ΔΔCt was higher than the value 1.5 ± SD or less expressed when the value was below 0.5 ± SD.

Unsupervised and supervised hierarchical cluster analysis was performed on log2 (Hy3/Hy5) ratios in normal and tumor tissues. P-value for unsupervised analysis was set at p<0.05, for supervised analysis was set at p<3.95E-5. miRNA expression data are shown as median (m) for their strong non-Gaussian distribution. Non-parametric Mann-Whitney U test was performed to compare miRNA expression profile in unpaired samples; p-values were set at <0.05.

OS cells were seeded at a density of 1.5×10⁵ per well in 6-well plates in 2 ml of complete medium containing 10% FCS for 24 h. Transfections were performed using reagent (Lipofectamine 2000, Invitrogen). For each well miR-1 precursor (cod. PM10617, Ambion Inc.), miR-133b precursor (cod. PM10029, Ambion Inc.) or negative control precursor miRNA (scramble) (cod. AM17110, Ambion Inc.) was transfected into U2-OS cells. Pre-miR miRNA precursor molecules are small, chemically modified, double stranded RNA molecules designed to mimic endogenous mature miRNAs once properly transfected and expressed by recipient cells.

The mixture Lipofectamine-miRNA (miR-1, 50 nM; miR-133b, 100 nM) was then administered to cells at 37°C in presence of serum-free medium up to 72 h. Optimal transfection conditions were obtained by using Lipofectamine 2000 (1 mg/ml, Invitrogen) at a weight ratio miRNA-transfectant of 1:1. Transfection efficiency was monitored by flow cytometry (FACSCalibur, BD Biosciences, San Jose, CA, USA) using Cy™3 and FAM™ dye-labeled Pre-miR negative controls (cod. AM17120, Ambion Inc.) and by RT-PCR using ΔCt values. All in vitro assays were performed at 24, 48 and 72 h of transfection using scramble and non-treated cell lines (U2-OS wt) as negative controls. All experiments were repeated at least three times.

The number of adherent, viable cells was assessed microscopically using an improved Neubauer hemocytometer and viability was assessed as the percentage of cells that excluded 0.2% trypan blue. After 24, 48 and 72 h from transfection, the cells were washed once with IX Dulbecco’s phosphate buffered saline (PBS), harvested by trypsinization and cell number was determined using trypan blue.

Cell migration assay was carried out using Trans well Permeable Support (Corning Incorporated, Corning, NY, USA). After transfection U2OS cells were carefully transferred on the top chamber of each transwell apparatus at a density of 1×10⁶ per ml (100 μl per chamber). Cells were allowed to migrate for 24, 48 and 72 h at 37°C. Cells that had penetrated to the bottom side of the membrane were then fixed in methanol, stained using hematoxylin and counted at microscope.

Cell invasion was analyzed by using Cultrex 24-well BME Cell Invasion Assay (Trevigen Inc., Gaithersburg, MD, USA) according to standard procedures. Briefly, after transfection, 10³ OS cells were seeded in 100 μl serum-free media into the upper wells previously coated with Matrigel basement extract, and 500 μl of media were added in the bottom wells. After 24,48 and 72 h of CO₂ incubation at 37°C, the non-invasive cells on the upper surface were removed and the cells migrated to the lower surface were fixed in 500 μl of Cell Dissociation Solution/Calcein-AM, incubated al 37 °C in CO₂ incubator for 1 h and quantified by fluorimetric analysis (485 excitation, 520 nm emission).

Apoptotic and necrotic cell death were analyzed by double staining with fluorescein isothiocyanate (FITC)-conjugated Annexin V and PI, in which Annexin V bound to the apoptotic cells with exposed phosphatidyl-serine, while PI labeled necrotic cells with membrane damage using an Annexin V-FITC apoptosis detection kit (MEBCYTO Apoptosis kit, MBL International, Woburn, MA, USA). The green (FL1) and red (FL2) fluorescence of Annexin/PI-stained live cells and PI-stained fixed cells was analyzed with a FACSCalibur flow cytometer (BD Biosciences) using a peak fluorescence gate to exclude cell aggregates during cell cycle analysis. The numbers of viable (Annexin⁻/PI⁻), apoptotic (Annexin⁺/PI⁻) and necrotic (Annexin⁺/PI⁺) cells were determined with CellQuest Software (BD Biosciences). According to the protocol, after 24, 48 and 72 h from transfection, adherent cells were briefly trypsinized, detached, and combined with floating cells from the original growth medium, centrifuged and washed twice with phosphate-buffered saline (PBS). Cells were resuspended in 500 μl of staining solution containing FITC-conjugated Annexin V antibody and propidium iodide (PI). After incubation on ice for 30 min, cells were analyzed by flow cytometry. Basal apoptosis and necrosis were identically determined on untreated cells.

Flow cytometric analyses were performed to define cell cycle distribution for transfected and not transfected cells. After 24, 48 and 72 h from transfection, cells were harvested by trypsinization and fixed with 70% ethanol. Cells were stained for total DNA content with a solution containing 20 μg/ml propidium iodide. Cell cycle distribution was then analyzed with a FACScalibur flow cytometer (BD Biosciences).

Reverse transcription of mRNA from U2-OS wt, U2-OS transfected with miR-1, miR-133b and scramble was carried out in 100 μl final volume from 400 ng total-RNA using High Capacity cDNA Archive kit (Applied Biosystems) according to manufacturer’s instructions. Quantitative RT-PCR was performed using ABI 7900 sequence detection system (Applied Biosystems) according to manufacturer’s protocol. Met gene was quantified by ΔCt method using TaqMan Expression Assays (Hs01565584_ml) (Applied Biosystems) and normalized to a housekeeping ACTB (Hs99999903_ml gene; TaqMan Expression Assays, Applied Biosystems).

According to standard procedures, 50 μg of protein extracts from cell lysate of U2-OS wt, U2-OS transfected with miR-1, U2-OS transfected with miR-133b and scramble were prepared and analyzed by 6% SDS-PAGE. Western blot analysis (WB) was performed by using anti-MET(C-28) (Santa Cruz Biotechnology, Santa Cruz, CA, USA) diluted 1:200. The signal was visualized by Immobilon Western Chemiluminiscent HRP substrate (Millipore, Billerica, MA, USA) and quantified by densitometric analysis using GS-800 imaging densitometer and Quantity One software (Bio-Rad, Hercules, CA, USA). A rabbit anti-actin antibody (Sigma Chemical Co., St. Louis, MO, USA) was used as control.

In the first step, we analyzed total-RNA from 6 high-grade OS (patients 1–6; Table IA), 6 low-grade OS (patients 1–6; Table IB) and 6 normal bone samples from non cancer patients used as control. The unsupervised hierarchical clustering, performed on log2 (Hy3/Hy5) ratios, revealed 100 miRNAs with the highest standard deviation across all samples (Fig. 1a). The supervised hierarchical clustering showed that 40 miRNAs were differentially expressed when compared with control with a p-value <3.95E-5, and 80% were underexpressed in the OS samples. This analysis pointed out that the most underexpressed were miR-206, miR-1, miR-133a, miR-133b, all belonging to miR-206 cluster, miR-378^*, and miRplus-c1049, that not being included in miRBase database was excluded from further analysis (Fig. 1b). When comparing low- and high-grade OS, unsupervised analysis identified 12 differentially expressed miRNAs, including miR-378^* (p-value <0.05) (Fig. 1c).

Levels of the most relevant miRNAs, miR-1, miR-133b and miR-378^*, were determined by RT-PCR in a larger cohort including 56 OS (31 high- and 25 low-grade) and 13 normal control samples. Results confirmed a lower expression in OS than in controls for all selected miRNAs when osteoblasts are used as calibrator (value = 1) (Fig. 2). In detail, miR-1 median values (2^−ΔΔCt) were 372.5 for OS samples and 422913 for control group (p=0.0005), miR-133b median values were 7.5 for OS and 2619 for control (p=0.0005), miR-378^* median values were 133 in OS and 1681 in control group (p<0.001).

In addition, within tumor population, slightly lower median values were found in high-grade as compared with low-grade OS for miR-1, miR-133b and miR-378^* (Table II), and in metastatic tumors as compared to non-metastatic ones (294 vs 605 for miR-1; 5 vs 45 for miR-133b; and 103 vs 162 for miR-378^*), with no statistical differences.

The expression of miR-1 and miR-133b, that resulted significantly lower in OS than in normal tissue, was investigated in OS cell lines and normal osteoblasts using MSC as calibrator. RT-PCR data showed that U2-OS, MG-63 and SAOS had, respectively, 300-, 27- and 35-fold lower miR-1 expression and 5-, 17- and 12-fold lower miR-133 expression when compared to normal osteoblasts (Table III).

To test whether increased levels of miR-1 and miR-133b might affect OS cell behavior, we introduced pre-miR miRNA precursor molecules in U2-OS cells. Transfection was validated by RT-PCR that assessed increased ΔCt levels in transfected cells when compared with scramble and U2-OS wt used as controls (Fig. 3a and b) and by flow cytometry.

With a transfection efficiency over 70% we found that the ectopic expression of miR-133b and miR-1 progressively decreased cell proliferation up to 72 h of transfection (67 and 74% of control, respectively) (Fig. 3c). Concomitantly, by Annexin V-FITC assay, induction of cell death, more evident in miR-133b transfected cells (p<0.05), was seen when compared to controls (Fig. 3d). The invasion analysis by Boyden chamber and migration through polycarbonate membrane showed that ectopic expression of miR-133b and miR-1 reduced cell invasiveness up to 48 h of 15 and 29%, respectively (p<0.05) (Figs. 4a and 5a) and cell migration of 22% (p<0.05) and 43% (p<0.01) (Figs. 4b and 5b). However, at 72 h the percentage of invasive and migrating cells shifted towards control levels.

Cell cycle distribution revealed that exposure of U2-OS to miR-133b induced a transient cell accumulation in S and G₂ phase up to 48 h (Fig. 4c), while miR-1 ectopic expression resulted in accumulation of cells in G₁ phase coincident with a cell decrease in S phase up to 48 h of transfection (Fig. 5c), suggesting a block at G₁/S check point.

To investigate the molecular mechanism of the anti-proliferative propriety of miR-1 and miR-133b, we examined the expression of oncogene Met in U2-OS transfected cells that previous data indicated as a candidate target of both miRNAs (15,16). Relative to non-transfected cells and scramble, ectopic expression of both miR-1 and miR-133b led to a reduced expression of MET protein level (Fig. 6). Adjusted volume of the electrophoretic bands quantified by densitometric analysis was 21.83 and 17.52, respectively, for controls, 5.04 for miR-1 and 8.43 for miR-133b transfected cells. However, Met gene mRNA levels were not different in miRNA-1 and miR-133b transfected U2-OS as compared with negative control (ΔCt=6.40 vs 6.30 and 5.20 vs 4.73, respectively), suggesting an incomplete pairing between miRNA and mRNA target.