Osteopontin receptors are expressed on human cancer cells. Cultured human C8161 melanoma and KRIB osteosarcoma cells were immunostained for two of the major osteopontin receptors: α_v integrins and CD44. Abundant punctuate α_v integrin immunoreactivity was detected in all cell types. The same cells displayed also CD44 immunoreactivity. Negative staining controls with irrelevant isotype control plus secondary antibodies displayed no immunoreactivity. Scale bar, 20 μm.

The RGD-dependent pro-metastatic effect is specific for osteopontin. A, C8161 melanoma cells were pre-incubated for 10 min with fibrinogen and fibronectin, both of which contain the integrin-binding RGD motif, or GST-osteopontin. Treatment with GST-osteopontin significantly increased metastasis formation relative to fibrinogen (p<0.05; asterisk) or fibronectin (p<0.05). Average normal lung weight (0.175 g) is indicated by the line. B, In a similar experiment, mice received C8161 melanoma cells treated with DMEM, 1 mM linear GRGDSP peptide, 1 mM double-cyclic RGD-4C peptide, or GST-osteopontin. The cohort receiving cells treated with GST-osteopontin exhibited a significant increase in lung metastasis relative to the other groups (p<0.001). Bars indicate mean ± SD.

Osteopontin treatment is associated with Src dephosphorylation of detached tumor cells. A, KRIB osteosarcoma cells were detached from tissue culture substrates. Some of the cells were allowed to adhere to the culture dish, while others were treated for 1 min with GST-osteopontin, GST-osteopontin-RGE, GST, RGD-4C peptide, or DMEM only. The cells were subsequently lysed, and 30 μg of the lysate was analyzed for the phosphorylation of Tyr-418 c-Src. Adhesion of the cells to plastic induced transient activation of c-Src. Treatment of the cells with GST-osteopontin induced dephosphorylation of c-Src. Analysis of total Src confirmed equal loading of the proteins. B, Analysis of the tyrosine phosphorylation of the same lysates (30 μg) revealed general dephosphorylation of tyrosine in proteins (note range 50–98 kDa proteins). Arrow indicates c-Src; cellular actin (40kDa) was detected at the same time to ensure equal loading of the proteins.

Tumor cell exposure to exogenous osteopontin increases experimental metastasis that is RGD-dependent. A, Typical gross appearance of lungs after intravenous administration of tumor cells. Arrows indicate tumor foci in the lungs. Scale bar, 0.5 cm. B, C8161 melanoma cells were pre-incubated for 10 min with DMEM, GST or GST-osteopontin and were subsequently intravenously injected into the tail vein of nude mice (10⁶ cells/mouse). After 12–16 weeks, animals were killed and the lung weights were measured (18). The weight was significantly increased (p<0.01; asterisk) in nude mice receiving human melanoma cells previously incubated with GST-osteopontin relative to DMEM- or GST-treated cells. C, Lung weight of mice receiving human KRIB osteosarcoma cells, which were pretreated for 10 min with GST or GST-osteopontin. Average normal lung weight (0.175 g) is indicated by the line. A significant increase in lung metastasis was observed when tumor cells were treated with GST-osteopontin relative to GST (p<0.01). D and E, Tumor cells were pre-incubated for 10 min with GST-osteopontin-RGE or GST-osteopontin. The Asp→Glu (termed RGE) mutation, which disrupts the integrin-binding RGD-domain in osteopontin, abolished its effects on tumor cells as seen in a statistically significant difference in lung weights of mice receiving C8161 melanoma cells (p<0.001) or KRIB osteosarcoma cells (p<0.001). Average normal lung weight (0.175 g) is indicated by the line. F, In a related experiment, mice received KRIB osteosarcoma cells treated with either GST-osteopontin or GST-osteopontin-RGE prior to administration. Survival of tumor-bearing mice was monitored, and the experiment was terminated at day 80 as indicated.

Osteopontin reduces actin fibers in adherent tumor cells. Adherent KRIB osteosarcoma cells were treated with GST-osteopontin or GST-osteopontin-RGE for 10 min, after which the cells were fixed and stained for cytoskeletal actin. Tumor cells treated with GST-osteopontin displayed a reduced actin network, with a decrease in transcytoplasmic actin cables, in comparison with cells treated with GST-osteopontin-RGE (arrows). Scale bar, 20 μm.

Promotion of tumor cell binding to vascular endothelium by osteopontin. A and B, KRIB osteosarcoma cells were treated with DMEM, GST-osteopontin, GST-osteopontin-RGE, or phosphatase inhibitors prior to GST-osteopontin and were subsequently incubated with the endothelium of the dorsal aorta. Relative to the GST-osteopontin-RGE and phosphatase inhibition groups, GST-osteopontin significantly increased cell binding (arrows) to the aortas (p<0.05; asterisk). Bars indicate mean ± SD. Scale bar, 20 μm. C, The number of tumor foci was increased in nude mice receiving human sarcoma cells previously incubated with GST-osteopontin relative to DMEM-treated cells.