Initial Adhesion Behavior of Fibroblasts onto Hydroxyapatite Nanocrystals

The initial adhesion, spreading and cytoskeleton changes of fibroblast NIH3T3 cells onto hydroxyapatite (HAp) and oxidized poly(styrene) (PSox) sensors preadsorbed fetal bovine serum were analyzed by using a quartz crystal microbalance with dissipation technique and an atomic force microscopy (AFM). The frequency shift (Δf ) and the dissipation shift (ΔD) curves on HAp nanocrystals showed the decrease in Δf with increasing ΔD for 80 min and the subsequent increase in Δf with decreasing ΔD, while those on PSox showed the decrease in Δf for 120 min with increasing ΔD for 50 min and then with subsequent decreasing ΔD. The different adhesion behavior dependent on the surfaces was attributed to the cell-surface interactions; the cells on HAp had rough fibrous pseudopods and those on PSox had dense particulate pseudopods. The different structures indicated that the cytoskeleton changes and the rearrangement of the extracellular matrix at the interface caused the different adhesion behavior.


Introduction
Hydroxyapatite (Ca 10 (PO 4 ) 6 (OH) 2 ; HAp) shows a good biocompatibility for fibroblasts; down growth of percutaneous device [1] and catheter [2] is improved by using HAp sintered body and coatings. The biocompatible features are attributed to the surface property of HAp. Therefore, understanding protein adsorption and cell adhesion onto HAp surface is of great importance for controlling cell functions.
To detect the interfacial phenomena, a quartz crystal microbalance with dissipation (QCM-D) technique is one of excellent in situ analytical methods. The HAp sensor applicable for the QCM-D technique was recently fabricated with an electrophoretic deposition method in our group to analyze protein adsorption [3,6]. Although a few studies about the cell adhesion on the surfaces have been reported, no cell adhesion behavior on the HAp surface with the QCM-D technique has been investigated [4,5].
In this study, the QCM-D technique and atomic force microscopy (AFM) were employed for the initial cell adhesion, spreading and cytoskeleton change of fibroblast NIH3T3 cells onto HAp and oxidized poly(styrene) (PSox) sensors.
The HAp sensor was fabricated by the electrophoretic deposition method based on our previous reports [3,6]. The poly(styrene) sensor oxidized by a UV/OZONE treatment (PSox) was used as a reference. The fibroblast cells were cultured in a plastic culture flask (BD Bioscience, USA) at 37°C in a humidified atmosphere of 5% CO 2 . The cells washed with 15 mL of PBS and treated with 1 mL of trypsin-EDTA for 5 min were dispersed into in 10% FBS/DMEM.  a saturated ΔD/(Δf n=3 /3) value from the ΔD−Δf n=3 /3 plot. The adsorption of FBS dispersed into DMEM at 10 vol% was measured for 1 hr after stabilizing a baseline of serum-free DMEM. Subsequently, the cell suspensions were seeded at 0.5 mL on the FBS-adlayer on HAp or PSox sensors, and cultured for 2 hr in air, and rinsed with 0.5 mL of 10 vol% FBS/DMEM. The cultured cells on the sensors were fixed with 3.7 vol% formaldehyde in PBS. The cells fixed were soaked into 1 mL of ethanol/ultrapure water series at 50, 60, 70, 80, 90, 100 vol% for each 5 min, and were into 1 mL of t-butyl alcohol three times at 37°C. The samples were kept at 4°C for 0.5 h and then freeze-dried at 4°C for 4-5 h.
The wettability of the sensors was evaluated in air by a sessile drop method of distilled water with a contact angle meter (CA-W200, Kyowa Interface Science Inc.). The morphology of the cells cultured on the sensors was observed with a confocal laser scanning microscope (CLSM: OLS-3000, OLYMPUS Inc.). The number, area, and volume of the adherent cells were calculated from the 2-D and the 3-D images (n = 10) obtained with scanning Z-range at a z-step of 10 nm. The sensor surfaces before and after the cell adhesion were observed with an atomic force microscope (AFM: SPM-9500, Shimazu Inc.). Silicon nitride probe mounted on cantilever (OMCL-AC160TS, OLYMPUS Inc.) was employed for the dynamic mode. The surface roughness was calculated by root mean squares (RMS) in the Z-range images.    (Figures 2(a) and (b)). On the contrary, the cell adhesion on PSox showed the monotonic decrease in the Δf n=3 /3 for 120 min with increasing the ΔD for 50 min and subsequently decreasing the ΔD (Figures 2(c) and (d)). These results clearly indicate the different adhesion process depending on the surfaces. The decrease in mass and viscoelastic property for fibroblast adhesion onto the PSox and Ta has been reported [4,5]; the cell spreading and cytoskeleton changes depending surface properties would lead to an increase in rigidity of the adherent cells and result in the decrease in the ΔD. Therefore, the different adhesion process depending on the surface was successfully in situ monitored by the QCM-D technique, and the cellsurface interactions through the protein adsorption were distinguished.

Conclusions
In this study, the QCM-D technique was employed for evaluating the initial adhesion, spreading and cytoskeleton changes of fibroblast on HAp and PSox sensors. The Δf and ΔD curves on HAp and PSox showed the different behavior on the surfaces, indicating the adhesion process affected by cell-surface interactions through the protein adsorption. The Δf and ΔD showed the linear relationship against the number of adherent cell. The AFM images showed the different pseudopod structures depending on the cell adhesion places. Therefore, these results indicate that the cytoskeleton changes and the rearrangement of extracellular matrix at the interfaces caused the different binding behavior.