Bone matrix osteonectin limits prostate cancer cell growth and survival

Abstract

There is considerable interest in understanding prostate cancer metastasis to bone and the interaction of these cells with the bone microenvironment. Osteonectin/SPARC/BM-40 is a collagen binding matricellular protein that is enriched in bone. Its expression is increased in prostate cancer metastases, and it stimulates the migration of prostate carcinoma cells. However, the presence of osteonectin in cancer cells and the stroma may limit prostate tumor development and progression. To determine how bone matrix osteonectin affects the behavior of prostate cancer cells, we modeled prostate cancer cell-bone interactions using the human prostate cancer cell line PC-3, and mineralized matrices synthesized by wild type and osteonectin-null osteoblasts in vitro. We developed this in vitro system because the structural complexity of collagen matrices in vivo is not mimicked by reconstituted collagen scaffolds or by more complex substrates, like basement membrane extracts. Second harmonic generation imaging demonstrated that the wild type matrices had thick collagen fibers organized into longitudinal bundles, whereas osteonectin-null matrices had thinner fibers in random networks. Importantly, a mouse model of prostate cancer metastases to bone showed a collagen fiber phenotype similar to the wild type matrix synthesized in vitro. When PC-3 cells were grown on the wild type matrices, they displayed decreased cell proliferation, increased cell spreading, and decreased resistance to radiation-induced cell death, compared to cells grown on osteonectin-null matrix. Our data support the idea that osteonectin can suppress prostate cancer pathogenesis, expanding this concept to the microenvironment of skeletal metastases.

Figure 1. Osteonectin-null bone matrices synthesized in vitro are disorganized and hypomineralized

Collagen fibrer morphology in wild type (A) and osteonectin-null (B) matrices, imaged by SHG (representative fields, individual optical sections). Arrows indicate sites in the matrix formerly occupied by osteoblasts. MicroCT imaging of mineral in wild type (C) and osteonectin-null (D) matrices (representative image of entire well). Collagen fiber length (E) and relative SHG intensity (F) in wild type vs. osteonectin-null matrix. * = significantly different from WT, p<0.01.

Figure 2. Prostate cancer cell-induced osteoblastic lesions have a woven bone phenotype

Osteoblastic lesion in a mouse tibia injected with LAPC-9 human prostate carcinoma cells: Hematoxylin and eosin staining, where “b” indicates bone and T indicates tumor (A and B). SHG imaging of collagen fibril morphology in osteoblastic lesions (C and D). For comparison, SHG imaging of trabecular bone (E and F) and cortical bone (G and H) in wild type (E and G) and osteonectin-null (F and H) mice. (representative fields, individual optical sections).

Figure 3. The abundance of selected extracellular matrix proteins is not altered in osteonectin-null matrices

(A) Data for acid soluble collagen was obtained using a Sircol dye binding assay, and data for tenascin C, vitronectin and fibronectin were obtained by Western blot analysis, and representative lanes are shown in (B).

Figure 4. Osteonectin-null matrices impair prostate cancer cell spreading

Representative confocal micrographs of PC-3 cells grown for 7 days on wild type (A) or osteonectin-null matrix (B). Phalloidin and DAPI staining (original magnification 100×). Cell perimeters were quantified, as a measure of cell spreading, after 3 and 7 days of culture (C). * = significantly different from day 3, p<0.05.

Figure 5. Osteonectin-null bone matrices promote prostate cancer cell proliferation

(A) PC-3 cells were grown on wild type or osteonectin-null matrices, and pulsed with ³H-thymidine on the indicated days. (B) PC-3 cells were grown on osteonectin-null matrices, in the presence or absence of 3 µg/mL recombinant human osteonectin, then pulsed with ³H-thymidine on the indicated days. (C) PC-3 cells (PC) or MG-63 (MG) cells were grown to confluence. Osteonectin in serum-free conditioned medium was analyzed by Western blot analysis. Molecular weight standards (kDa) are indicated on the left of the panel. (D) Serum-deprived cultures of PC-3 cells were treated for 24 hours with conditioned medium from PC-3 cells that had been cultured for 7 days on wild type or osteonectin-null matrices. Cells were then pulsed with ³H-thymidine. Dilutions of medium tested: 100% conditioned medium, 1:2 and 1:3 dilution of conditioned medium in 10% FBS. * = significantly different from WT, p<0.05.

Figure 6. Osteonectin-null bone matrices promote resistance to radiation-induced cell death

PC-3 cells were grown on wild type or osteonectin-null matrices for 6 days, then exposed to 0, 2, or 4 Gy gamma radiation. Cells were re-plated and colony formation was quantified by crystal violet staining. * = significantly different from WT, p<0.05.