The human osteosarcoma cell lines MG63 and U-2 OS were obtained from the Shanghai Institute of Cell Biology (Shanghai, China). The cells were grown under standard conditions in RPMI-1640 and supplemented with 10% heat-inactivated FBS and antibiotics/antimycotics (all from Gibco-BRL, Eggenstein, Germany). They were incubated at 37°C in a CO₂ incubator, released from the culture surface using trypsin/EDTA (Gibco-BRL) and counted in a haemocytometer.

IL-6 and sIL-6R were purchased from R&D Systems (Minneapolis, MN, USA). Bestatin was purchased from Sigma (St. Louis, MO, USA). All treatments with IL-6 were at 1 nM and included 15 nM sIL-6R, which was added to achieve a stable effect in osteoblastic cells (36–39).

MG63 or U-2 OS cells (1×10⁶ cells/well) were placed in 6-well plates and incubated with factors (IL-6, sIL-6R or bestatin) for the times indicated. Cells were collected and total RNA was extracted from each treatment using the Qiagen RNeasy Mini kit (Qiagen Inc., Valencia, CA, USA). RNA samples were reverse-transcribed at 42°C using the High Capacity cDNA Reverse Transcription kit for 30 min according to the protocol of the supplier (Applied Biosystems, Foster City, CA, USA).

The internal competitive standard RNA was obtained using the method designed by Kehlen et al and composite primers were synthesized as shown in Table I (40). For cDNA synthesis, 1,000, 500, 50, 10 or 1 pg APN competitor RNA were added to 5 μg total RNA and reverse-transcribed at 42°C using the High Capacity cDNA Reverse Transcription kit for 30 min according to the protocol of the supplier (Applied Biosystems). Two microliters of cDNA was diluted in 50 μl of PCR reaction solution containing primers 3 and 4. The PCR reaction was performed according to the manufacturer’s standard protocol (Qiagen Inc., Hilden, Germany) in a thermal cycler (MaxiCycler PTC-100; MJ Research, Inc., Watertown, MA, USA) for 35 cycles of 60 sec at 94°C, 60 sec at 60°C, and 60 sec at 72°C. Each reaction product (10 μl) was run on a 1.5% agarose gel containing 0.1% μg/ml ethidium bromide in TAE buffer. The relative intensities of the bands corresponding to the target (573 bp) and internal standard (434 bp) PCR products were visualized with UV light. The relative amounts of target and internal standard products were calculated by densitometric analysis using ImageMaster 1D Prime software (Amersham Pharmacia Biotech, Freiburg, Germany). The ratio of standard to target amplification pro ducts was graphed as a function of the initial amount of internal standard, and lines were drawn from a linear regression analysis using InStat (GraphPad Software, Inc., San Diego, CA, USA). The initial amount of target was calculated from the point where the amount of amplified target equals the amount of amplified standard (ratio D 1). The analysis was performed in triplicate.

The following quantitative PCR conditions were used: 2 min at 50°C, 10 min at 95°C, and 40 cycles of 15 sec at 95°C, 1 min at 60°C using 1 μl of the cDNA reverse-transcribed as described above, 2X SYBR-Green PCR Master Mix (Applied Biosystems), and 200 nM forward and reverse primers. The primer sequences were APN forward, GTTCTCCTTCTCCAACCTCATC and reverse, CTGTTTCCTCGTTGTCCTTCT; MMP-2 forward, CCCCAGACAGGTGATCTTGAC and reverse, GCTTGCGAGGGAAGAAGTTG; MMP-9 forward, CGCTGGGCTTAGATCATTCC and reverse, AGGTTGGATACATCACTGCATTAGG; GAPDH forward, ACACCCACTCCTCCACCTTT and reverse, TAGCCAAATTCGTTGTCATACC. Each assay was run on an Applied Biosystems 7300 Real-Time PCR System in triplicate, and expression fold changes were derived using the comparative threshold cycle (CT) method (41,42).

APN activity was assayed using the substrate alanine p-nitroanilide (Ala-pNA) at a final concentration of 1.5 mM. Confluent cell monolayers in 48-well plates were rinsed three times and incubated at 37°C for 20–40 min with pre-warmed substrate. Supernatant p-nitroanilide was measured at an OD of 405 nm by a microplate reader (Anthos Labtec Instruments GmbH, Salzburg, Austria). Assays were run in triplicate, in parallel with cell- and substrate-free blanks. The cells were detached from the plates and counted. Catalytic activity was expressed as pkat/10⁶ cells.

MMP-2 and -9 activity was determined by gelatin zymography after exposure to IL-6 or bestatin. Cells (1×10⁶ cells/well) were plated in 12-well tissue culture plates and incubated in serum-free McCoy’s 5A medium in the presence of 1 nM IL-6 or 100 μM bestatin for 24 and 48 h. The conditioned medium was then collected and separated by electrophoresis on 10% SDS-PAGE containing 0.2% gelatin (Sigma). At the end of electrophoresis, the gels were soaked twice in 2.5% Triton X-100 in dH₂O at 25°C for a total of 60 min, then incubated in substrate buffer (50 mM Tris HCl, 5 mM CaCl₂, 0.02% NaN₃ and 1% Triton X-100, pH 8.0) at 37°C for 18 h. Bands corresponding to MMP-2 and -9 activity were visualized by negative staining using 0.2% Coomassie Blue in 50% methanol and 10% acetic acid as described elsewhere (43,44). Bands were quantified using NIH ImageJ software (NIH, Bethesda, MD, USA) as previously described (45,46).

Cell mobility was determined in migration and invasion assays through 24-well Transwell inserts (8 mm pore size) coated with 30 μg type 1 collagen (both from Millipore, Billerica, MA, USA) (migration assay) or Matrigel (BD Biosciences, Bedford, MA, USA) (invasion assay) (41,44). U-2 OS and MG63 cells were maintained in serum-free medium for 24 h, after which they were trypsinized, resuspended in serum-free McCoy’s 5A medium and placed in the upper chamber of the Transwell insert (5×10⁴ cells/well). Cells were then treated for 24 or 48 h with 1 nM IL-6 or 100 μM bestatin. McCoy’s 5A medium containing 10% FBS was then placed in the lower chamber. Non-migrating cells in the upper chamber were removed by wiping with a cotton swab. Cells that had penetrated the filter and were located on the lower chamber side of the filter were fixed with 4% formaldehyde in PBS, stained with 2% crystal violet in 2% ethanol and counted under a light microscope at ×200 magnification.

Cells placed on 6-well chamber slides were treated with 1 nM IL-6 or 100 μM bestatin for 24 h, immunofluorescence-labeled using a Cellular NF-κB Translocation kit (Beyotime Institute of Biotechnology, Jiangsu, China) according to the manufacturer’s protocol. Briefly, after fixing, the cells were stained using anti-NF-κB p65 (1:100) overnight and then stained with Cy3 fluorescein-conjugated secondary antibody followed by nuclear counterstain propidium iodide. Photomicrographs were obtained using a Leica TCS SP2 confocal spectral microscope.

U-2 OS cells (1×10⁶ cells/well) were placed in 6-well plates and each well was incubated for 24 or 48 h in 1 nM IL-6 or 100 μM bestatin. Cells were harvested and lysed with ice-cold 50 mM potassium phosphate buffer (pH 7.4) containing 2 mM EDTA and 0.1% Triton X-100, sonicated and then centrifuged at 13,000 g for 10 min at 4°C. The supernatant was collected and total protein was determined using a Bio-Rad Protein Assay kit (Bio-Rad, Hercules, CA, USA) with bovine serum albumin (BSA) as the standard. At the end of electrophoresis, proteins were electrotransferred to nitrocellulose membranes, blotted with primary antibody against GRB2, PI3K, AKT, PKC, p38, NF-κB p65, p38, ERK1/2, JNK, MMP-2, MMP-9, p-p38, p-ERK1/2 or p-JNK (Cell Signaling Technology, Inc., Danvers, MA, USA) then washed and stained with secondary antibody. Bands were visualized with an enhanced chemiluminescence reagent (ECL™; Amersham Biosciences Corp., Piscataway, NJ, USA) and quantified using an NIH Image analyzer (NIH).

Data are presented as the means ± standard deviation (SD) from at least three independent experiments. Differences between groups were determined using Student’s t-test. P<0.05 was considered to indicate a statistically significant result.

The treatment of human osteosarcoma cell lines HOS and MG63 with cytokine IL-6 has been shown to upregulate APN mRNA (24). To confirm and extend these observations we first determined APN mRNA expression in human osteosarcoma cells lines U-2 OS and MG63 after exposure to IL-6 (1 nM) for 24 h. Competitive PCR showed that APN mRNA expression in both cell lines increased significantly (Fig. 1A). We next investigated whether bestatin, a widely used and potent inhibitor of APN, inhibited APN mRNA expression. A 24 h exposure of U-2 OS and MG63 to 100 μM bestatin decreased the mRNA levels slightly, but nevertheless significantly (Fig. 1A).

It has previously been shown that IL-6 treatment of human osteosarcoma cell lines MG63 and HOS significantly increased Ala-pNA hydrolysing activity (24). We confirmed and extended these observations by investigating APN Ala-pNA hydrolysing activity in the cells lines U-2 OS and MG63 after a 24 h treatment with IL-6 (1 nM) or bestatin (100 μM). As expected, IL-6 treatment significantly increased Ala-pNA hydrolysing activity in both cell lines (Fig. 1B). In contrast, the bestatin treatment significantly decreased Ala-pNA hydrolysing activity (Fig. 1B). Together these data confirm that IL-6 treatment activates, whereas bestatin treatment inhibits, APN hydrolysing activity and APN mRNA expression in human osteoblast cell lines.

It has previously been demonstrated that a 24 h treatment of human osteosarcoma cell lines HOS and MG63 with IL-6 (1 nM) or bestatin (100 μM) enhances or represses their invasive potential, respectively (24). To confirm and extend these observations we assessed the invasive potential of cell lines U-2 OS and MG63 exposed to IL-6 (1 nM) or bestatin (100 μM) for 24 or 48 h. The invasiveness of both MG63 and U-2 OS cells significantly increased upon exposure to IL-6 and significantly decreased upon exposure to bestatin after both 24 and 48 h (Fig. 2). We also assessed the migration potential of MG63 and U-2 OS when exposed to IL-6 (1 nM) or bestatin (100 μM) for 24 or 48 h. The migration of both U-2 OS and MG63 cells significantly increased upon exposure to IL-6 and significantly decreased upon exposure to bestatin after both 24 and 48 h (Fig. 3). The data shown here confirm that APN stimulator IL-6 and APN inhibitor bestatin enhance and reduce cell invasiveness and migration, respectively. Since the invasive potential of cell lines HOS and MG63 correlates with the relative enzymatic activity of APN (24), it is possible that the increased invasive potential and migration of U-2 OS and MG63 cells observed here is also due to increased APN activity.

To investigate whether APN regulates MMP-2 and -9 activity in osteosarcoma cell lines, we measured MMP-2 and -9 enzyme activity by gelatin zymography after inhibiting APN in MG63 and U-2 OS cells with bestatin (100 μM) for 24 and 48 h. We found that both 24- and 48-h bestatin treatments decreased MMP-2 and -9 activity (Fig. 4A and B). These results suggest that the decreased invasiveness of bestatin-treated cells could be due APN’s reduced activation of MMP-2 and -9. By comparison, a 24 and 48 h incubation of MG63 and U-2 OS cells with IL-6 (1 nM) increases MMP-2 and -9 activity (Fig. 4A and B). Together these results raise the possibility that the increased invasiveness of MG63 and U-2 OS by IL-6 could be due to IL-6’s activation of MMP-2 and -9 via APN activation.

Cell invasion and migration are orchestrated by multiple signalling cascades, including the mitogen-activated protein kinase (MAPK) and phosphatidylinositol 3-kinase (PI3K) pathways (47,48). We determined the effect of APN inhibition on protein levels and phosphorylation states of MAPK and PI3K signalling pathway members by western blotting. The treatment of MG63 (Fig. 5) and U-2 OS (Fig. 6) cell lines with bestatin (100 μM) for 24 h resulted in decreased levels of GRB2, PI3K, AKT, PKC, NF-κB p65 (Figs. 5A and 6A), p38, ERK1/2, JNK, p-p38, p-ERK1/2 and p-JNK (Figs. 5B and 6B). By comparison, the treatment of MG63 and U-2 OS cell lines with APN stimulator IL-6 for 24 h resulted in increased levels of GRB2, PI3K, AKT, PKC, NF-κB p65 (Figs. 5A and 6A), p38, ERK1/2, JNK, p-p38, p-ERK1/2 and p-JNK (Figs. 5B and 6B). Together these results show that a reduction of APN activity in osteosarcoma cells dampens MAPK and PI3K signalling, as shown by the reduced levels of phosphorylated p38, ERK1/2 and JNK. This raises the possibility that APN might have a signalling function in osteosarcoma invasiveness by activating the transcription factor NF-κB through the MAPK and PI3K pathways. NF-κB would then activate downstream target genes such as MMP-2 and -9.

We next investigated whether the enzymatic activity changes in MMP-2 and -9 upon bestatin or IL-6 treatment could be due to reduced MMP-2 and -9 protein levels. Indeed, we showed by western blotting that the treatment of MG63 (Fig. 5) and U-2 OS (Fig. 6) cell lines with bestatin (100 μM) for 24 h resulted in decreased levels of MMP-2 and -9 (Figs. 5C and 6C). By comparison, the treatment of with APN stimulator IL-6 resulted in increased levels of MMP-2 and -9 (Figs. 5C and 6C). We next determined MMP-2 and -9 mRNA levels by quantitative PCR and found that these were reduced upon bestatin treatment and increased upon IL-6 treatment. The changes in MMP-2 and -9 protein and mRNA levels (Figs. 5D and 6D) are therefore in accordance with the changes in MMP-2 and -9 activity (Fig. 4).

Western blotting revealed that IL-6 and bestatin treatments increased and decreased, respectively, the levels of the transcription factor NF-κB, which has been shown to activate MMP-1, -3 and -9 transcription (49). We therefore investigated the expression and localization of NF-κB in IL-6 or bestatin-treated MG63 and U-2 OS cells by immunocytochemisty with an antibody against the p65 subunit. In untreated cells, NF-κB was visible in both the cytosol and nucleus (control panel in Fig. 7A and B). The cytoplasmic form is inactive due to its association with inhibitory protein IκB. IL-6 treatment concentrated the NF-κB p65 in the nucleus (Fig. 7A and B), while bestatin reduced overall NF-κB p65 levels, consistent with our western blotting data (Fig. 5). These results support our model of MMP-2 and -9 activation by NF-κB, which has been activated by APN signalling via the MAPK or PI3K pathway.