All procedures were performed in accordance with institutional guidelines from the Fourth People’s Hospital of Jinan (Shandong, China) and the Department of Pathology in Qilu Hospital (Shandong, China). A total of 55 patients diagnosed with osteosarcoma, including 4, 13, 12 and 26 cases of parosteal, fibroblastic, chondroblastic and osteoblastic osteosarcomas, respectively, from the Department of Pathology (Qilu Hospital) were selected for the study. Osteosarcoma types were termed groups 1–4, respectively. Tumor tissues were removed and sent for paraffin embedding. Informed consent was obtained from all patients.

Paraffin-embedded tissue samples were routinely prepared, producing 3-μm tissue sections mounted on slides. Sections were fixed with paraformaldehyde, slides were washed 3 times for 3 min in PBS and then endogenous peroxidase activity was blocked using 1% H₂O₂. Citrate Antigen Retrieval Buffer (Beijing Zhongshan Goldenbridge Biotechnology, Beijing, China) was used to retrieve antigens and then CD133 mouse anti-human antibody (Beijing Biosynthesis Biotechnology, China) was added to sections and incubated at 37°C for 60 and 15 min, respectively, according to the manufacturer’s instructions for the SP9000 immunohistochemical kit (Beijing Zhongshan Goldenbridge Biotechnology). DAB (Fuzhou Maixin Biotechnology Development Co., Ltd., Fuzhou, China) was added to develop staining and slides were observed under a microscope. Hematoxylin staining (Beijing Zhongshan Goldenbridge Biotechnology) was performed and slides were observed under a light microscope.

Each section was analyzed by two observers and judged by a double-blind method. Twelve high-power fields were randomly selected under microscope for immunohistochemical staining. Sections were scored according to the following criteria: i) no color in cytoplasm, 0; ii) cytoplasm presented light yellow cloudiness, (+); iii) cytoplasm presented yellow granular state, (++); iv) cytoplasm presented uniform deep yellow, (+++). (++) and (+++) were considered positive cells and the percentage of positive cells in each section was calculated.

The Saos-2 osteosarcoma cell line was purchased from the American Type Culture Collection (Manassas, VA, USA). Cells were cultured in low-glucose DMEM (Hyclone, Logan, UT, USA) containing 10% fetal bovine serum (FBS; Hangzhou Sijiqing Biological Engineering Materials Co., Ltd., Hangzhou, China) and incubated in a 37°C, 5% CO₂ incubator. Cells were digested with trypsin and passaged every 3 days.

Saos-2 cells in logarithmic growth phase were inoculated to prepare cell-attached coverslips and slides were prepared as described for osteosarcoma tissue samples.

Saos-2 cells in logarithmic growth phase were digested with trypsin (Sino-American Biotechnology, Guangzhou, China), centrifuged, collected and washed twice with PBS. Samples were then centrifuged and the supernatant was removed. Pellets were resuspended in 5 ml PBE buffer from a magnetic bead kit (Miltenyi Biotec Ltd., Surrey, UK), filtered using a 100 mesh screen to obtain a single cell suspension and then centrifuged at 1,000 rpm for 5 min. The supernatant was removed and 100 μl antibody-coated magnetic beads/l×10⁷ cells was added and incubated for 15 min at room temperature under dark conditions, followed by 2 washes in combining buffer. Cells were resuspended in 500 μl combining buffer and separated on a magnetic separation column (Miltenyi Biotec Ltd.). Fractions were collected from the column and dead cells in the original cell suspension were filtered by the column and removed.

Cells were collected, washed twice with PBS at 4°C and the cell concentration was adjusted to 1×10⁶/ml. Cell suspension (100 μl) was put into 5 ml flow tubes and then 20 μl CD133-PE monoclonal fluorescent antibody was added and incubated for 30 min at 4°C under dark conditions. Tubes were washed twice with 5 ml PBS, centrifuged, the supernatant was removed and the precipitate was resuspended in 0.5 ml PBS. Samples were analyzed using a FACSCalibur flow cytometer (BD Biosciences, San Jose, CA, USA). CD133-PE flow fluorescence detection antibody was purchased from Miltenyi Biotec Ltd.

Cell suspension was collected as described and centrifuged for 5 min at 1,000 rpm. The supernatant was removed, 100 μl CD133 monoclonal antibody was added to directly label the magnetic beads and 400 μl PBE was added, mixed and incubated for 30 min at 4°C under dark conditions. Uncombined magnetic beads were removed by two PBS washes and the pellet was resuspended in 500 μl PBE and separated on a magnetic separation column (Miltenyi Biotec Ltd.). Column flow-through contained CD133 cells. Cells retained by the column were washed with PBE and collected, these cells were CD133⁺ cells. CD133⁺ subpopulation cells were labeled using CD133-PE fluorescent antibody and purity was determined by flow cytometry.

Saos-2 cells in logarithmic growth phase were selected and separated into CD133⁺ and CD133⁻ subpopulation cells by the method described, fixed with 70% ice-cold ethanol for 24 h and treated with Triton X-100 (Sigma-Aldrich, St. Louis, MO, USA). Following PBS washing and centrifugation, 1 mg/ml RNase A (AppliChem Inc., St. Louis, MO, USA) was added to the cells and incubated for 15 min. Then, 50 μg/ml PI (Shanghai Jingmei Biotech, Shanghai, China) was added and cells were stained for 15 min. Cells were collected and analyzed using the FACSCalibur flow cytometer and Modfit software (Bio-Rad, Hercules, CA, USA) was used to analyze cell cycle stage.

TRIzol (1 ml) was added to the collected cell suspension (5–10×10⁶ cells). Following homogenization, samples were incubated for 5 min at 15–30°C to completely separate nucleic acid-protein complexes. CHCl₃ (0.2 ml; Shanghai Chemical Reagent, Shanghai, China) was added and the tubes were agitated by hand for 15 sec. Tubes were incubated for 2–3 min at 15–30°C and then centrifuged at 12,000 × g for 15 min at 4°C. Following centrifugation, the mixed solution was separated into 3 phases in which RNA was in the clear water phase. The water phase was transferred into a new centrifuge tube and mixed with 0.5 ml isopropanol (Shanghai Chemical Reagent) to completely precipitate RNA. The amount of added isopropanol was determined as follows: if 1 ml TRIzol was added to each sample to develop the homogenate, then the amount of isopropanol was 0.5 ml. The samples were incubated for 10 min at 15–30°C and then centrifuged at 12,000 × g for 10 min at 4°C. Before centrifugation, the invisible RNA precipitate formed gelatinous precipitate at the bottom and side wall of tube. The supernatant was removed and the precipitate was washed in 1 ml 75% ethanol (prepared with DEPC-treated water), agitated and centrifuged at 7,500 × g for 5 min at 4°C. Ethanol was removed and the RNA precipitate was dried for 5–10 min in air. The RNA precipitate was not dried completely, to avoid reducing the solubility. The A_260/280 ratio of partly dissolved RNA samples was <1.6. RNA was dissolved in RNase-free water, incubated for 10 min at 55–60°C and preserved at −70°C. For extraction of low concentrations of RNA, 5–10 μg RNase-free glycogen (<4 mg/ml; Invitrogen Life Technolgies, USA) was added as water phase vector prior to the addition of isopropanol. To decrease the solution viscosity, the samples were passed through a 26-gauge syringe needle twice to slice genomic DNA prior to the addition of CHCl₃. After separating the two phases, glycogen remained in the water phase and coprecipitated with RNA.

Real-time (RT) reaction solution (10 μl) was added to 10 μl annealing mixture, incubated for 60 min in a 37°C water bath, heated to 95°C for 5 min and then placed in an ice bath. cDNA templates confirmed to express target and house-keeping genes were selected. PCR conditions were as follows: 95°C for 3 min, 40 PCR cycles (94°C for 20 sec, 59°C for 20 sec and 72°C for 30 sec) and 72°C for 5 min. Products were run on a 2% agarose gel with a 100-bp DNA ladder and visualized with ethidium bromide to determine whether the correct gene was amplified. The PCR product (set as 1) was diluted to a series of concentrations. DNA templates of these gradient concentrations and all cDNA samples were respectively applied to prepare the real-time PCR system. Primer sequences and reaction conditions are presented in Table I.

The real-time PCR conditions were as follows: GAPDH, 95°C for 5 min, 45 PCR cycles (95°C for 10 sec, 59°C for 15 sec, 72°C for 20 sec and 83.8°C for 5 sec); multidrug resistance protein 1 (MDR1), 95°C for 5 min, 40 PCR cycles (95°C for 10 sec, 59°C for 15 sec, 72°C for 20 sec and 81°C for 5 sec); Oct4, 95°C for 5 min, 45 PCR cycles (95°C for 10 sec, 59°C for 15 sec, 72°C for 20 sec and 84°C for 5 sec); Sox2, 95°C for 5 min, 40 PCR cycles (95°C for 10 sec, 59°C for 15 sec, 72°C for 20 sec, 81°C for 5 sec); and Nanog, 95°C for 5 min, 40 PCR cycles (95°C for 10 sec, 59°C for 15 sec, 72°C for 20 sec, 81°C for 5 sec). In order to establish the melting curve of the PCR products, the products were heated slowly from 72 to 99°C (temperature increased by 1°C every 5 sec) after the amplification reaction. Target and house-keeping genes of each sample were analyzed by real-time PCR to calculate gene concentrations by standard curve. The corrected relative content of the gene in the sample was calculated by dividing the house-keeping gene concentration by the target gene concentration in the sample.

Separated CD133^+/− subpopulation cells were suspended in DMEM containing 20% FBS. The cell suspension was diluted and the cells were inoculated into culture dishes (diameter, 60 mm) containing 10 ml prewarmed medium (37°C) at densities of 50, 100 and 200 cells/dish and cultured for 3 weeks at 37°C, 5% CO₂ in a saturated humidity incubator. Following incubation, the cells were observed and counted to calculate the colony-forming efficiency using the following formula: Colony-forming efficiency (%) = no. of clones/no. of inoculated cells.

CD133⁺ subpopulation cells were inoculated into two culture flasks in DMEM containing 20% FBS and cultured at 37°C, 5% CO₂ in a saturated humidity incubator. On day 3 and 7, the proportion of CD133⁺ cells in the CD133 cell population was determined by flow cytometry. On day 7, the cells were collected to determine the cell cycle phase, with Saos-2 cells used as a control. All experiments were repeated three times.

All tests were repeated three times. Statistical analysis was carried out using the Student’s t-test with SPSS software. P<0.05 was considered statistically significant.

In the present study, CD133⁺ cells were identified in various types of osteosarcoma (Fig. 1). Cell count analysis indicated that the CD133⁺ ratio in parosteal, fibroblastic, chondroblastic and osteoblastic osteosarcoma was 5.63±1.96, 6.54±1.65, 8.54±1.25 and 13.84±3.81, respectively. Q-analysis identified no significant difference between CD133⁺ ratios in parosteal, fibroblastic and chondroblastic osteocarcinomas (P>0.05). However, the ratio obtained in osteoblastic osteosarcomas was found to be significantly different compared with the other osteosarcoma groups (P<0.01; Tables II and III).

The results of the immunostaining performed in the Saos-2 cell line are presented in Fig. 2A. Analysis by flow cytometry revealed that a mean of 5.73±0.93% of cells were CD133⁺ (Fig. 2B).

Compared with Saos-2 cells, CD133⁺ subgroup cells were identified in the G₀/G₁ stages of the cell cycle and exhibited a lower G₂ peak (Fig. 3). The percentage of Saos-2 cells in G₀/G₁, G₂/M and S stages was 48.86±1.09, 16.99±1.34 and 34.15±2.31, respectively, while CD133⁺ subgroup cells were identified as significantly different at 79.09±1.61, 3.66±0.32 and 17.25±1.29, respectively (P<0.01; Table IV). The number of CD133⁺ subpopulation cells was increased in G₀/G₁ and decreased in G₂/M and S stages compared with Saos-2 cells, indicating that CD133⁺ subgroup cells were quiescent, while Saos-2 cells were proliferating.

After 10 days of culture in complete medium, CD133⁺ subpopulation cell proliferation and cell cycle analyses indicated that cell percentages at G₀/G₁, G₂/M and S stages were 50.24±1.35, 16.09±3.78 and 33.67±4.81, respectively, and were observed to be significantly different compared with the percentages on day 0 (P<0.05). No significant difference in the mean cell percentages at G0/G1, G2/M and S stages compared with Saos-2 cells was noted (P>0.05; Table V). Following complete CD133⁺ subgroup cell culture, the percentages of cells in the G₂/M and S stage increased significantly and the cell cycle shifted from proliferative to the quiescent G₀/G1 stage (P<0.05). These results indicate the differentiation abilities of CD133⁺.

Following isolation using magnetic beads, the percentage of CD133⁺ subgroup cells at day 0, 3 and 10 following cell culture was detected using flow cytometry as 90.02±1.75, 51.31±3.47 and 5.57±1.01, respectively. Statistical analysis indicated that the CD133⁺ cell percentage decreased significantly at days 3 and 10 compared with day 0 (90 to 5%), similar to that observed in the Saos-2 cells (Fig. 4). Results indicate the potential differentiation ability of VD133⁺ cells in complete culture medium.

CD133⁺ cell colonies were larger compared with CD133⁻ cells following 5-week culture. The mean efficiencies of CD133⁺ and CD133⁻ subgroups, calculated from three repetitions, were 61.84±8.39 and 24.77±5.53, respectively, and were significantly different (P<0.05; Fig. 5).

In CD133⁺ Saos-2 cells, Sox2 and MDR1 gene expression was identified to be significantly increased compared with CD133⁻ cells (P<0.05; Fig. 6).