Human bone tumor samples (n=95) were obtained as anonymous specimens and the normal bone tissues (n=20) were obtained from the resected and discarded bone samples from individuals who underwent total knee arthroplasty with no history of tumors from Hospital. The study was reviewed and approved by the Institutional Review Board of the Hospital. SaO2, U-2OS and HOS human osteosarcoma cell lines used in the experiments were purchased from American Type Culture Collection (ATCC, Manassas, VA, USA). Adenovirus vector with LOX expression (adLOX), negative control vector with GFP expression (adGFP) and virion-packaging elements were purchased from Genechem Company (Shanghai, China). All primary antibodies were purchased from Cell Signaling Technologies (Beverly, MA, USA).

β-APN was purchased from Sigma-Aldrich Chemical Co. (St. Louis, MO, USA); apoptosis detection kit (Hoechst 33258), 3-(4,5-dimethylthiazol-2-yl)-3,5-diphenyl tetrazolium bromide (MTT) and ECL-PLUS kit were purchased from Beyotime (Haimen, China); Dulbecco’s modified Eagle’s medium and fetal bovine serum (FBS) were purchased from Gibco (Gibco, Montevideo, Uruguay); M-MLV Reverse Transcriptase was purchased from Promega (Madison, WI, USA); TRIzol Reagent and Lipofectamine 2000 were purchased from Invitrogen (Carlsbad, CA, USA); SYBR-Green Master Mixture and the primers were purchased from Takara (Otsu, Japan).

SaO2, U-2OS and HOS human osteosarcoma cell lines were cultured in DMEM medium supplemented with 10% FBS, and antibiotics (100 IU/ml of penicillin G and 100 Ag/ml of streptomycin). They were all incubated at 37°C humidified atmosphere of 5% CO₂ and the medium was replaced every 2 days. U-2OS and HOS cell lines were infected with recombinant adenovirus vector expressing LOX or GFP (adLOX and adGFP). In preliminary experiments, cells were co-infected with recombinant adenovirus vector adLOX or negative control adGFP at MOIs of 10–100 pfu/cell according to the manufacturer’s instructions. Cells were subcultured at a 1:5 dilution in 400 μg/ml G418-containing medium and positive stable transfectants were selected. In this study, a recombinant adenovirus vector containing a LOX gene and the negative control adenovirus vector were used to transfect osteosarcoma cells. U-2OS and HOS cells without gene transfection were used as a control group. Subsequently, followed by 24-h recovery after transfection, the proliferation, metastasis and invasion were assessed, respectively.

To quantitatively determine the mRNA expression level of LOX in vivo and in vitro, real-time PCR using SYBR-Premix Ex Taq (Takara, Shiga, Japan) was used according to the manufacturer’s protocol. Total RNA of each specimen was extracted with TRIzol and assessed by an ABI Prism 7500 sequence detection system (Applied Biosystems, San Francisco, CA, USA). The genes were amplified using specific oligonucleotide primer and human β-actin (actin) gene was used as an internal control. The PCR primer sequences were as follows: LOX, 5′-ACTGCACACACACAGGGATTG-3′ and 5′-GCCTT CTAATACGGTGAAATTG-3′; β-actin, 5′-CTGGGACGACA TGGAGAAA-3′ and 5′-AAGGAAGGCTGGAAGAGTGC-3′; Data were analyzed using the comparative Ct method (2^−ΔΔCt).

The localization of LOX protein in bone tumor tissues and normal tissues were assessed by immunohistochemistry. Bone tumor tissues and normal tissues excised from patients were embedded in paraffin and cut into 6 μM sections. After deparaffinization, the sections were incubated in 3% H₂O₂ and 0.1% Triton X-100 and microwaved in Tris-buffer for 10 min to retrieve the antigen. The sections were then blocked in 5% goat serum (Invitrogen) for 30 min and incubated with LOX antibody (diluted 1:200) for 2 h at room temperature. While a negative control was performed using a section without primary antibody. Following incubation with secondary antibodies, bound antibodies were visualized by a Power Vision Two-step Histostaining Reagent (Dako, Carpinteria, CA, USA) and antibody binding was visualized by incubation with DAB (Boster, Wuhan, China) for 3 min at room temperature followed by counterstaining with hematoxylin (Basa, Zhuhai, China). The LOX immunoreactive and non-immunoreactive cells with a nucleus were counted in 3 sections (with interval 3 sections) per patient in each group. The positive staining of LOX was shown as brown particles. The sections were observed and photographed with the optical microscope (Olympus, Tokyo, Japan). A percentage was calculated by the immunostained/total neuronal ratio × 100%.

The cells were harvested and extracted at indicated time using lysis buffer (Tris-HCl, mercaptoethanol, SDS and glycerol). Cell extracts were boiled for 8 min in loading buffer and then equal amount of cell extracts were separated on 8–15% SDS-PAGE gels according to the molecular weight. Separated protein bands were transferred into polyvinylidene fluoride (PVDF) membranes and the membranes were blocked in 5% skim milk powder. The primary antibodies against LOX, Ki-67, PCNA, COX-2, MMP-2, MMP-9, PI3Kp85α and p-AKT were diluted according to the instructions of antibodies and incubated overnight at 4°C. Then, horseradish peroxidase-linked secondary antibodies were added at a dilution ratio of 1:800, and incubated at room temperature for 3 h. The membranes were washed with PBS for three times and the immunoreactive bands were visualized using ECL-PLUS kit according to the kit instructions. The relative protein level in different cell lines was normalized to β-actin (actin) concentration.

Cell proliferation was analyzed with the MTT assay. Briefly, the cells were incubated in 96-well plates at a density of 1×10⁵ cells per well with DMEM medium supplemented with 10% FBS. Cells were treated with 20 μl MTT dye at 0, 12, 24, 48, 72 and 96 h and then incubated with 150 μl of DMSO for 5 min). The absorbance of optical densities at each time point was measured at 570 nm with enzyme immunoassay analyzer (Bio-Rad, Hercules, CA, USA).

For cell cycle analysis, 10⁶ cells were harvested, re-suspended, washed with phosphate-buffered saline and fixed in 70% ethanol. Cells were then treated with RNase (10 μg/ml) and propidium iodide (50 μg/ml) for 30 min. The cell cycle phase distribution was determined by analytical DNA flow cytometry. The percentage of cells in each phase of the cell cycle was analyzed using Modfit software (Verity Software House, Topsham, ME, USA).

U-2OS and HOS cells were plated in each well of a 6-well culture plate and allowed to grow to 80% confluence. Treatment with adLOX was then performed. On the next day, a wound was created using a 10 μl micropipette tip. The migration of cells towards the wound was monitored at 24 and 48 h.

Briefly, the cells were removed by trypsinizing and suspended with medium supplemented with 10% FBS, then 1×10⁵ were added into the Transwell innserts with 8-μm pore size (BD Biosciences) for 24-well plates coated with 50 μl Matrigel (BD Biosciences), and 200 μl medium containing 15% FBS were added in the bottom chamber. Undergoing migration for 24 and 48 h, a cotton swab was used to remove the non-migrated cells in the upper chamber then the filters were individually fixed with 4% paraformaldehyde and were determined with hematoxylin and eosin staining. The cell number was counted in five random fields of each chamber under the microscope.

The cells were prepared at a density of 5,000 cells per well in a 24-well plate. Apoptotic cells were detected by using the Hoechst 33258 staining. The cells were fixed with 4% paraformaldehyde for 10 min, washed with PBS for three times and then stained with 2 μg/ml Hoechst 33258 for 5 min. Morphologic changes in apoptotic nuclei were evaluated under a fluorescence microscope (excitation wavelength 350 nm, emission filter 460 nm) (FluoView, Olympus).

All data are presented as the means ± standard error (SE) for at least three independent experiments. T-test was used for two group comparison and one-way analysis of variance (ANOVA) using Fisher’s exact test was employed for multiple comparisons. The LSD method of multiple comparisons was used when the probability for ANOVA was statistically significant. P<0.05 was considered with statistical significance.

To ascertain whether abnormal LOX expression occurs in osteosarcoma tissues, the mRNA levels of LOX were examined in tumor/normal paired tissue samples by RT-PCR analysis (95/20). As a result, the expression level of LOX in 18% (17/95) was similar to that of the mean level of normal tissues (>0.5-fold and <2-fold). However, the expression level of LOX in 57% (55/95) of the tumor tissue was significantly lower than that of the mean level of normal tissues (<0.5-fold). A higher level of LOX expression (>2-fold) was observed in 30% (23/78) of tumor tissues (Fig. 1A). On the basis of the results of real-time PCR, the cases were divided into group 1 (55 cases, low level of LOX expression) and group 2 (23 cases, high level of LOX).

In addition, the level and localization of LOX protein were further assessed by immunohistochemisty in groups 1 and 2. Consistently, the incidence of LOX-positive cells in the tumor tissues and the mean level of LOX protein expression in group 1 were significantly lower than those in the normal tissues (Fig. 1B). In contrast, the incidence of LOX-positive cells and the mean level of LOX protein expression in group 2 were higher than those in the normal tissues (Fig. 1).

Subsequently, to further evaluate whether the expression levels of LOX mRNA and protein were related to clinical therapeutic outcomes, the clinical status of each specimen was determined according to tumor-node-metastasis classification (TNM) (Table I). Our results indicated that the tumor size in group 1 was significantly larger than that in group 2 and there were more cases in the poor nodal status in group 2. In addition, the patients in group 2 had a significantly higher occurrence of pulmonary metastasis. These results revealed a decrease in LOX mRNA expression in most advanced-stage tumor tissues.

The endogenous expression of LOX in human osteosarcoma cell lines, SaO2, U-2OS and HOS, was first evaluated using RT-PCR and western blot analysis. As shown in Fig. 2A and B, there were different levels of mRNA and protein expression of LOX in SaO2, U-2OS and HOS cell lines, but the expression levels of LOX were significantly higher in SaO2 cell line than those in U-2OS and HOS cell lines. Thus, U-2OS and HOS were used as osteosarcoma cell lines for infection by recombinant adenovirus containing LOX gene (Fig. 2A and B).

In order to enhance the exogenous expression of LOX in osteosarcoma U-2OS and HOS cells, recombinant adenovirus containing LOX gene (adLOX) was constructed and used for upregulation of LOX in the U-2OS and HOS cell lines. The infection efficiency of adLOX (at a multiplicity of infection = 100) in U-2OS and HOS cell lines were both greater than 80% by fluorescence microscopy (Fig. 2C). Real-time PCR and western blot assays were performed to measure the exogenous expression of LOX after 48 h following adenovirus infection and an obvious increase of LOX mRNA and protein expression was observed in adLOX group as shown in Fig. 2D and E.

Deregulation of cell proliferation and cell cycle is one of the hallmarks of cancer (18). In order to detect the effect of LOX on cell proliferation, we investigated the proliferative activities of adLOX-infected U-2OS and HOS cells by MTT assay. As a result, the cell proliferative ability of adLOX-infected U-2OS and HOS was significantly suppressed as compared with the adGFP group and control group. It was indicated that LOX overexpression would reduce the proliferative activities of U-2OS and HOS cell lines. However, no difference was found between adGFP group and control group in U-2OS and HOS cell lines (Fig. 3A). To further confirm whether LOX overexpression inhibited cell cycle progression, flow cytometry was employed. As shown in Fig. 3B, the apoptotic incidence of U-2OS and HOS cells in the adLOX group was remarkably higher than that in the adGFP group and control group. Cell cycle kinetics showed that the G0/G1 phase fraction was increased, while S phase fraction was decreased. Cell cycle was arrested in G0/G1 phase more in the adLOX group than in the adGFP group and control group (Fig. 3B).

In order to investigate the mechanisms of proliferative inhibition of LOX, we evaluated effects of LOX on two independent proliferative markers: Ki-67 and proliferating cell nuclear antigen (PCNA). Ki-67 is identified as a critical mediator in tumor progression in Ewing’s sarcoma (19) and is widely used to evaluate the outcome of anticancer treatment (1,20,21). Similarly, PCNA has been employed to evaluate cell proliferation (22). The expression of Ki-67 and PCNA was examined by western blot assay and the results indicated that the expression level of Ki-67 and PCNA was significantly inhibited by LOX overexpression in U-2OS and HOS cell lines and no difference was found between adGFP group and control group (Fig. 3C). These data suggested that overexpression of LOX might inhibit osteosarcoma cell proliferation via downregulation of Ki-67 and PCNA.

To determine the effect of LOX on U-2OS and HOS cell migration, wound-healing assay was performed and the results showed that the migrative capacities of osteosarcoma cells with adLOX infection were markedly suppressed. However, no significant changes were detected between adGFP group and control group (Fig. 4A). Moreover, Transwell invasion assay was performed to examine the ability of adLOX-infected 2U-2OS and HOS cells to traverse the Matrigel-coated Transwell chamber. As shown in Fig. 4B, compared with the adGFP group and control group, invasion through Matrigel in adLOX group was reduced by 48%.

It has been reported that cyclooxygenase-2 (COX-2) (23) and MMPs were correlated with tumor invasive progression or metastasis in several types of cancers (22,24,25). To investigate whether the activities of cyclooxygenase-2 and MMPs were affected by LOX, western blot analysis was used to examine the expressions of COX-2, MMP-9 and MMP-2. As shown in Fig. 4C, the expression of MMP-9 and MMP-2 proteins was significantly suppressed in adLOX group as compared with the adGFP group and control group. However, no significant difference in expression of COX-2 was found among the three groups. These data indicated that overexpression of LOX might inhibit the migration of osteosarcoma cells via downregulation of MMP-9 and MMP-2 expression.

It has been shown that the mechanism of action of tumorigenesis factors is associated with their ability to suppress apoptosis. Thus, in this study, apoptosis was detected by Hoechst 33342 staining. After culture for 24 and 48 h, the cells were stained with Hoechst 33342 and their cell nucleus was observed under the microscope for apoptosis. The representative micrographs (Fig. 5A) and quantitative scoring of all data (Fig. 5B) showed that the number of U-2OS and HOS apoptotic cells and necrotic cells (strong blue staining) in adLOX group significantly increased compared with that in adGFP group and control group in a time-dependent manner. No difference was found between adGFP group and control group (P>0.05). These data suggested that overexpression of LOX could decrease the ability of osteosarcoma cells against apoptosis.

The PI3K/Akt signaling pathway plays a critical role in regulating cellular proliferation, growth, survival and motility for normal cells and its dysfunction is deeply connected with tumorigenesis (20,26). To directly test the effects of PI3K/AKT signaling, the expression of PI3Kp85α and p-AKT was assessed by western blot assay. As shown in Fig. 6, an obvious decrease of PI3Kp85α and p-AKT was observed in adLOX group compared with that in the adGFP group and control group (^*P<0.05). But, no difference was found between adGFP group and control group (P>0.05). To further confirm this result, 300 μM β-aminopropionitrile (β-APN), a LOX inhibitor, was used to treat the adLOX-infected U-2OS and HOS cells. Western blot results showed that the expressions of PI3Kp85α and p-AKT were significantly suppressed by β-APN. In addition, the inhibitory effect of LOX on proliferation and migration of human osteosarcoma cells also could be reversed by β-APN (data not shown). These data suggested that overexpression of LOX might inhibit the proliferation and migration of human osteosarcoma cells through PI3K/AKT signaling pathway.