Hydroxyapatite nanoparticle-containing scaffolds for the study of breast cancer bone metastasis

Abstract

Breast cancer frequently metastasizes to bone, where it leads to secondary tumor growth, osteolytic bone degradation, and poor clinical prognosis. Hydroxyapatite Ca(10)(PO(4))(6)(OH)(2) (HA), a mineral closely related to the inorganic component of bone, may be implicated in these processes. However, it is currently unclear how the nanoscale materials properties of bone mineral, such as particle size and crystallinity, which change as a result of osteolytic bone remodeling, affect metastatic breast cancer. We have developed a two-step hydrothermal synthesis method to obtain HA nanoparticles with narrow size distributions and varying crystallinity. These nanoparticles were incorporated into gas-foamed/particulate leached poly(lactide-co-glycolide) scaffolds, which were seeded with metastatic breast cancer cells to create mineral-containing scaffolds for the study of breast cancer bone metastasis. Our results suggest that smaller, poorly-crystalline HA nanoparticles promote greater adsorption of adhesive serum proteins and enhance breast tumor cell adhesion and growth relative to larger, more crystalline nanoparticles. Conversely, the larger, more crystalline HA nanoparticles stimulate enhanced expression of the osteolytic factor interleukin-8 (IL-8). Our data suggest an important role for nanoscale HA properties in the vicious cycle of bone metastasis and indicate that mineral-containing tumor models may be excellent tools to study cancer biology and to define design parameters for non-tumorigenic mineral-containing or mineralized matrices for bone regeneration.

Fig. 1

HA nanoparticle synthesis schematic showing the two-step precipitation reaction to obtain poorly crystalline HA followed by hydrothermal aging of the precipitate to obtain crystalline HA nanoparticles.

Fig. 2

TEM images of the different HA particles used to prepare mineral-containing PLG-scaffolds: Commercial Sigma Aldrich particles, (shown at low [SIGa], and high [SIGb] magnification), HA particles synthesized from Ca(NO₃)₂ and (NH₄)₂HPO₄ precursors with hydrothermal aging times of 0 hrs (A1), 0.5 hrs (A2), and 72 hrs (A3), and particles synthesized from CaCl₂ and Na₂HPO₄ precursors aged for 1.5 hrs (B1).

Fig. 3

pXRD patterns of particles SIG, A1-A3, and B1. All samples are confirmed to be HA. Major peaks have been labeled with HA crystal indices. Peaks not characteristic of HA pXRD pattern in SIG are indicated by (*), (ICDD PDF no. 09-0432).

Fig. 4

FTIR spectra of particles SIG, A1-A3, and B1. Positions of characteristic HA peaks are indicated.

Fig. 5

SEM micrographs of A3-scaffold (A, C) and SIG-scaffold (B,D) show the distribution of A3 and SIG particles on a fractured surface of the scaffold in which pore surfaces are visible. Regions of high particle density (*), low particle density (O), and pores in the scaffold (#) were observed for both scaffolds (A, B). Residual salt crystals are indicated (x). Arrows point to individual A3 particles (C) and SIG particles (D).

Fig. 6

TEM micrograph of an A3-scaffold shows the distribution of the particles (indicated by arrows) in the bulk (indicated by ‘O’) and the pores (indicated by ‘#’) of the polymeric matrix. Inset: Higher magnification image of the area indicated by a box.

Fig. 7

(A) Serum protein adsorption on non-mineral-containing (NM) and mineral-containing (A1-A3, B1, SIG) scaffolds as quantified via colorimetric BCA analysis of scaffold lysates prepared after incubation with serum-containing cell culture media. (B) Immunofluorescence analysis of fibronectin (FN) adsorption onto non-mineral-containing (NM) and mineral-containing scaffolds containing small, poorly crystalline (A1) and highly crystalline HA particles (B1). Significance between groups and NM, A2, and B1 scaffolds are denoted by (*), (#), and (o) respectively. In all cases, P<0.01 is indicated by a single symbol and P<0.005 is denoted by double symbols.

Fig. 8

Calcium (white) and phosphorus (black) content of media harvested from nonmineral-containing (NM) and mineral-containing (A1) scaffolds after 2 days of culture in the presence and absence of MDA-MB231 breast cancer cells as determined by ICP. Culture media incubated for the same period of time, but not exposed to scaffolds or cells was used as a control (Media).

Fig. 9

(A) MDA-MB231 cell adhesion onto non-mineral-containing (NM) and mineral-containing scaffolds (A1-A3, B1, SIG) as quantified via PicoGreen DNA assay of cell lysates 3 hours after seeding. (B) MDA-MB231 growth 3 days after seeding of non-mineral-containing (NM) and mineral-containing scaffolds (A1-A3, B1, SIG) as analyzed by PicoGreen DNA assay. Values are depicted relative to day 0 to account for variations in initial cell adhesion between the individual groups. Significance between groups and NM, A2, and B1 scaffolds are denoted by (*), (#), and (o) respectively. In all cases, P<0.01 is indicated by a single symbol and P<0.005 is denoted by double symbols.

Fig. 10

(A) IL-8 secretion by MDA-MB231 breast cancer cells as analyzed through ELISA of conditioned media collected from non-mineral-containing (NM) and mineral-containing scaffold cultures (A1-A3, B1, SIG). Values were normalized to cell numbers as determined via PicoGreen DNA assay to account for changes in cell proliferation. (B) IL-8 mRNA expression by MDA-MB231 breast cancer cells as analyzed through quantitative real-time RT-PCR of lysates collected from non-mineral-containing (NM) and representative mineral-containing scaffold cultures (A1, B1, SIG). Values are shown as normalized to expression levels in NM scaffold cultures. Significance between groups and NM, A2, and B1 scaffolds are denoted by (*), (#), and (o) respectively. In all cases, P<0.01 is indicated by a single symbol and P<0.005 is denoted by double symbols.