Tie-2 regulates the stemness and metastatic properties of prostate cancer cells

Abstract

Ample evidence supports that prostate tumor metastasis originates from a rare population of cancer cells, known as cancer stem cells (CSCs). Unfortunately, little is known about the identity of these cells, making it difficult to target the metastatic prostate tumor. Here, for the first time, we report the identification of a rare population of prostate cancer cells that express the Tie-2 protein. We found that this Tie-2High population exists mainly in prostate cancer cell lines that are capable of metastasizing to the bone. These cells not only express a higher level of CSC markers but also demonstrate enhanced resistance to the chemotherapeutic drug Cabazitaxel. In addition, knockdown of the expression of the Tie-2 ligand angiopoietin (Ang-1) led to suppression of CSC markers, suggesting that the Ang-1/Tie-2 signaling pathway functions as an autocrine loop for the maintenance of prostate CSCs. More importantly, we found that Tie-2High prostate cancer cells are more adhesive than the Tie-2Low population to both osteoblasts and endothelial cells. Moreover, only the Tie-2High, but not the Tie-2Low cells developed tumor metastasis in vivo when injected at a low number. Taken together, our data suggest that Tie-2 may play an important role during the development of prostate tumor metastasis.

Keywords: Tie-2; cancer stem cells; metastasis; prostate cancer.

Figure 1. Expression of Tie-2 in prostate cancer cells

A.&B. Flow cytometry analysis of Tie-2 expression in a panel of prostate cancer cell lines (LAPC4, 22Rv1, DU145, LNCaP, C42B, MDA-PCa-2b and PC-3). The results are presented as the mean ± SD from triplicate experiments. C. Tie-2 mRNA was analyzed in these different prostate cancer cell lines using qRT-PCR. Note that Tie-2 protein and mRNA expression were both increased in bone metastatic prostate cancer cell lines (highlighted). Results quantified represented as fold change normalized to DU145. D. Immunohistochemical staining was performed on humanized bone scaffold containing the PC-3 metastastic tumor (5 cases) (left panel) and human bone section containing the metastatic prostate tumor (1 case) (right panel). All the sections were stained with antibody against Tie-2. Note that, all the sections were positive for the Tie-2 proteins and the arrows showed the positive staining within the tumor cells (40X magnifications).

Figure 2. The Tie-2^High population possessed stem cell characteristics

A. Analysis of the population sorted by flow cytometry confirmed the successful enrichment of Tie-2^High cells. B. Validation of the selected candidate genes (i.e., KITLG, CXCR4, and FGF1) with qRT-PCR. Results were normalized with internal control and are presented as fold change relative to Tie-2^Low population. C. Flow cytometry analysis revealed that quiescent cells were increased by more than 3-fold in the Tie-2^High population when compared to the Tie-2^Lowpopulation. D. Flow cytometry analysis of apoptotic cells by Annexin V staining in Tie-2^Low and Tie-2^High cells that were treated with 100nM Cabazitaxel for 72hrs (p value = 0.0018 for apoptosis). Note that a high percentage of apoptotic cells were detected in the Tie-2^Low population when compared to the Tie-2^High prostate cancer cells. (p values: * < 0.05, ** < 0.005, *** < 0.0005).

Figure 3. Ang-1 upregulated prostate CSC and quiescent markers in prostate cancer cell lines

Western blotting A. of prostate CSC markers (CD49f and Bmi-1) and a quiescence marker (p27) after Ang-1 treatment in PC-3 cells (left panel). Ang-1 was found to upregulate both stem cell and quiescence markers in a dose-dependent manner. The Tie-2 inhibitor, on the other hand, suppressed both types of markers in a dose-dependent manner in PC-3 cells (right panel). B. Transfection of Tie-2 siRNAs led to downregulation Tie-2 mRNA levels in PC-3 cells. Results were normalized with internal control and are presented as fold change relative to scramble. C. Effect of Tie-2 knockdown on CSC and quiescence marker expression in PC-3 cells. D. The experiment was repeated in the presence of Tie-2 inhibitor (left panel) or a Tie-2 neutralizing antibody (right panel), which showed that both the Tie-2 inhibitor and Tie-2 neutralizing antibody abolished the effect of Ang-1 on PC-3 cells. E. Quantitation of the quiescent population in PC-3 cells with or without Ang-1 treatment. Note that Ang-1 treatment led to a 4-fold induction of the quiescent population in PC-3 cells when compared to the vehicle control. Each experiment was repeated at least three times, and the results are presented as the mean ± SD. (p values: * < 0.05, ** < 0.005, *** < 0.0005).

Figure 4. Ang-1 functioned as an autocrine factor in prostate cancer cells

A. Ang-1 mRNA expression was determined in different prostate cancer cell lines (DU145, C42B and PC-3) using qRT-PCR analysis. Results were normalized with internal control and are presented as fold change relative to DU145. B. Knockdown of Ang-1 in PC-3 cells by siRNA transfection was confirmed with qRT-PCR and ELISA. Note that Ang-1 expression was suppressed by >80% in PC-3 cells. Results were normalized with internal control and are presented as fold change relative to scramble. C. Downregulation of Ang-1 in PC-3 cells by siRNA was associated with suppression of CSC (CD49f and Bmi-1) and quiescence (p27) markers. D. Ang-1 secretion (pg/ml) by different prostate cancer cell lines (DU145, C42B and PC-3) was determined with an ELISA. E. Knockdown of Ang-1 for >50% was confirmed with RT-PCR in C42B cells. F. Ang-1 knockdown suppressed both CSC (CD49f and Bmi-1) and quiescence (p27) markers in C42B cells. G. The response of C42B cells to exogenous Ang-1 treatment (600ng/ml) was restored when endogenous Ang-1 was knocked down by siRNA. (p values: * < 0.05, ** < 0.005, *** < 0.0005).

Figure 5. Tie-2 facilitated the adhesion of prostate cancer cells to osteoblasts and endothelial cells

A cell adhesion assay was performed with the sorted Tie-2^High and Tie-2^Low PC-3 cells prestained with Hoechst 33342. Cells that adhered to osteoblasts (MG-63) A. or endothelial cells (HUVECs) D. were quantified by measuring the fluorescence intensity, and the results are presented as the mean ± SD. Note that Tie-2^High PC-3 cells were more adhesive to MG-63 and HUVECs when compared to Tie-2^LowPC-3 cells. B&E. Effect of Tie-2 inactivation on the adhesive ability of Tie-2^High PC-3 cells. Treatment with a Tie-2 inhibitor (5 μM) prior to the adhesion assay significantly suppressed the adhesion ability of Tie-2^High cells, while the same treatment failed to affect the Tie-2^Low population. C&F. Ectopic Tie-2 expression promoted the adhesion of DU145 cells to osteoblast MG-63 cells and HUVECs. Each experiment was repeated at least three times, and the results are presented as the mean ± SD. (p values: * < 0.05, ** < 0.005, *** < 0.0005).

Figure 6. The Tie-2^High population was highly metastatic in vivo

A. Tie-2^High and Tie-2^Low populations sorted out by FACS were injected into mice via intracardiac injection (top). An experimental regimen showing the intracardiac implantation and monitoring of tumor metastasis is shown below. B. Representative bioluminescence images of a mouse from each group 4 and 8 weeks after the implantation. C. Ex vivo imaging of the Tie-2^High metastatic tumors. D. Summary of the metastatic tumors detected in each mouse. Three out of the eight mice that were injected with Tie-2^High cells exhibited metastasis (two in the bone and one in the kidney) (p < 0.05). (See also Suppl Figure 3).