Thrombospondin-1 (TSP-1) up-regulates tissue inhibitor of metalloproteinase-1 (TIMP-1) production in human tumor cells: exploring the functional significance in tumor cell invasion

Abstract

Thrombospondin-1 (TSP-1), a matrix-bound adhesive glycoprotein, has been shown to modulate tumor progression. We previously demonstrated that TSP-1 up-regulates matrix metalloproteinases MMP-2 and MMP-9. Our studies suggested that the balance between MMPs and tissue inhibitors of metalloproteinases (TIMPs) is a key determinant in tumor cell invasion. We now report that TSP-1 up-regulates TIMP-1 expression in both human breast and prostate cancer cell lines. The effect of TSP-1 on TIMP-1 expression was examined in human breast adenocarcinoma cell lines (MDA-MB-231) and human prostate cancer cell lines (PC3-NI and PC3-ML) treated with exogenous TSP-1. TIMP-1 expression was also examined in TSP-1 stably transfected breast cancer cell line (MDA-MB-435). Northern and western blot analysis revealed TIMP-1 mRNA and TIMP-1 protein expression increased with increasing concentrations of TSP-1. This effect was inhibited by antibodies against the type I repeat domain of TSP-1 further suggesting that TSP-1 mediates TIMP-1 secretion. Inhibition of TSP-1 induced TIMP-1 levels increased tumor cell invasion. We conclude that TSP-1 is involved in influencing the critical balance between MMPs and their inhibitors, maintaining the controlled degradation of the extracellular matrix needed to support metastasis and our results may provide an explanation for the divergent activities reported for TSP-1 in tumor progression.

Fig. 1

TSP-1 up-regulates TIMP-1 production in human tumor cells in a dose-dependant manner. MDA-MB-231 breast cancer cells, PC3-NI, non-invasive human prostate cancer cells and PC3-ML, metastatic human prostate cancer cells were plated in 6 well plates and were allowed to grow to 85 % confluency. Cells were then were treated with TSP-1 at the indicated concentrations in serum-free media containing 0.1% BSA . Cells treated with 100ng/ml of PMA were used as a positive control and those treated with bis-Tris-Phosphate buffer (BTP) were the vehicle buffer control for TSP-1 treatment. Conditioned media were harvested after 72 hours and analyzed by western blot or ELISA for TIMP-1 levels. The amount of conditioned media analyzed was corrected for the total number of cells in each well so that TIMP-1 secretion was compared from the same amount of cells from each treatment group. Experiments were repeated three times and the results of a representative experiment are shown in the figure.

Fig. 2

Quantitative analysis of TSP-1 induced TIMP-1 production of human cancer cells through ELISA analysis. Conditioned media from cells treated with TSP-1 as described in figure legend were analyzed for TIMP-1 levels using an ELISA kit from Oncogene Science of Siemens Healthcare Diagnostics Inc., Cambridge, MA. A: MDA-MB-231 conditioned media. B: PC3-NI and PC3-ML conditioned media. TIMP-1 levels were normalized to 10⁶ cells and are presented as percent of control. Each data point is the average of three replicates and the error bars represent standard deviation. The two curves in panel B are statistically different (p<0.05). Statistical analysis was performed using GraphPad Prizm software, San Diego CA.

Fig. 3

Time dependant effect of exogenous TSP-1 treatment on TIMP-1 mRNA in human breast cancer cells, MDA-MB-231. Cells were treated with either 40 μg/ml of TSP-1 in serum –free media or 100 ng/ml PMA and RNA isolated as described in Material and methods. Lane 1: PMA for 12 hours. Lane 2: BTP buffer for 24 hours. Lanes 3-5: TSP-1 treatment for 6hours, 12hours, and 6-24 hours as noted. A 2.4 kB β-actin probe was used as an equal loading control. Experiments were repeated three times and the results of a representative experiment are shown in the figure.

Fig. 4

Anti-TSP-1 antibody and type 1 repeat peptide (CSVTCG) inhibits TIMP-1 production in human breast cancer cells. In all panels, cells were incubated in serum-free media with TSP-1 alone (40 μg/ml) or with TSP-1 (40 μg/ml) and (10 μg/ml) of the indicated antibodies or peptides for 48 hours. Conditioned media was then collected and analyzed by Western blot analysis using a rabbit polyclonal anti-TIMP-1 antibody. A. Effect of a polyclonal goat anti-TSP-1 antibody on TIMP-1 production. B. Effect of polyclonal rabbit antibody against CSVTCG present in the type I repeat domain of TSP-1. C. Effect of the CSVTCG peptide (from the type 1 repeat of TSP-1) on TIMP-1 production. The last lane in all experiments was conditioned media of HT-1080 cells treated with PMA (positive control). Experiments were repeated three times and the results of a representative experiment are shown in the figure.

Fig. 5

TIMP-1 expression in TSP-1 stably transfected MDA-MB-435 cells as assessed by western blot analysis. Cells were grown in serum-free media for 72 hours and the conditioned media measured for TIMP-1expression by western blot analysis. Lane #1: conditioned media from HT-1080 cells treated with PMA. Lane #2: vector control cell line (TH5). Lane #3: high TSP-1 producers (TH26). Lane #4: the COOH-terminal truncated TSP-1 producers (TH50). Experiments were repeated three times and the results of a representative experiment are shown in the figure.

Fig. 6

TSP-1 inhibits breast cancer cell invasion through TIMP-1 production. MDAMB-231 cells were pretreated with either BTP (buffer control) or TSP-1 (40 μg/ml) for 36 hours. Cells were harvested and placed in the top chamber of the Boyden chamber assay with or without the neutralizing antibodies as indicated. Cells were treated with antibody or control IgG at a concentration of 10 μg/ml. No TSP-1 or BTP buffer was added to the bottom chamber in this experiment. Bottom chambers contained serum-free media with 0.1% BSA. Cells were incubated for 5 hours at 37° C, fixed and then stained. Invasive cells adhering to the bottom of the chamber were counted using a phase contrast microscope (400X). The data were expressed as the summation of the number of invasive tumor cells in five representative fields. Experiments were repeated three times and the results of a representative experiment are shown in the figure.