Solutions of calcium salt and phosphoric acid (Ca/P = 1.66 mol) were used to synthesize HA particles and incorporate them on bovine type I collagen fibrils by a direct precipitation technique in situ. This technique was optimized to produce 3 different ratios of Col-HA composites (20%Col-80%HA; 50%Col-50%HA; and 80%Col-20% HA) as follows: bovine type I collagen (Sigma) was directly dissolved in 0.1 M phosphoric acid (H₃PO₄) and stirred at 40°C for 30 minutes, resulting in a 2.0% (w/w) collagen gel mixture in the final composites. The Ca/P 1.66 mol solution of calcium chloride dehydrate (CaCl₂) (EMD Chemical Inc) and phosphoric acid (H₃PO₄) (Sigma) was added drop-wise into the reagent mixture stirring at 40°C (Stire & Temp) to allow HA precipitate onto bovine type I collagen fibrils in the gel mixture. The pH of the mixture was adjusted to between 8.0 and 9.0 using NaOH solution (J.T. Backer) and hydrochloric acid (HCl) (Sigma), and then lowered to a final pH of 7.0–7.2. After aging and prefreezing at −20°C overnight, the Col-HA mixture was freeze-dried in a freeze dryer (FreeZone, Labconco 117) at −52°C under vacuum pressure of 0.052 mbar for 24 hours, followed by drying at 19°C, 0.97 mbar for 24 hours. Dehydrothermal cross-linking of the residual collagen fibrils was performed in a vacuum oven at 110°C for 24 hours. Final products of collagen-hydroxyapatite composites were cooled down to room temperature. One sample from each product was fractured to observed the internal structures using a scanning electron microscope (SEM) (Supra 40VP FESEM).

To evaluate the biologic properties of prototype Col-HA composites as scaffolds, we performed MSC attachment and proliferation assays in vitro. Mouse MSCs (C₃H₁₀T_1/2) were seeded (2 × 10⁵ cells/well) on 3 different ratios of Col-HA (20%Col-80%HA; 50%Col-50%HA; 80%Col-20%HA) composites in a 12-well plate and cultured in Basal Eagle Media (Invitrogen)which contained 100 U/mL penicillin, 0.1 mg/mL streptomycin, and 10% fetal bovine serum in a CO₂ incubator with 5% carbon dioxide, 95% humidified air at 37°C for 1, 3, 5, and 7 days. A commercial, pure type I collagen wound dressing material (CollaPlug, Zimmer Inc, Carlsbad, Calif) was used as a control. To observe the attachment and proliferation of mouse MSCs under the fluorescence microscope, the cells were cultured by Adi-GFP (3 μL/mL) transfection overnight.

The 50%Col-50%HA composite was selected for further testing as it has the best mechanical properties suitable for bone grafting applications. We performed cytotoxicity assays using human periodontal ligament stem cells (hPDSCs) in vitro. We prepared and extracted solutions from 4 different test materials: 3 pure type I collagen (CollaPlug; Foundation, OTB, Tokyo, Japan; and a prototype collagen without HA) and the 50%Col-50%HA composite. Six samples of each material were prepared with each sample 5 mm in height and 8 mm in diameter. Samples were soaked in 15-mL centrifuge tubes with 10-mL phosphate buffered saline and placed into a shaking incubator (Max Q Mini 4000) at 37°C, 120 rpm for 48 hours.

Human periodontal ligaments were collected from the extracted third molars, and soaked into the Dulbecco's Modified Eagle Medium (DMEM) with 250 μg/mL gentamicin sulfate, glutamine, 5 pμg/mL amphotericin B, 100 U/mL penicillin, 0.1 mg/mL streptomycin, and 10% fetal bovine serum, PDL was digested by a solution of collagenase type I (3 mg/mL) and Dispase (4 mg/mL) for 1 hour. hPDSCs were cultured by DMEM supplemented with 100 U/mL penicillin, 0.1 mg/mL streptomycin, and 10% fetal bovine serum in a CO₂ incubator with 5% carbon dioxide, 95% humidified air at 37°C. Upon nearing confluence, cells are detached using 0.08% trypsin/0.04%EDTA (pH 7.2) and passage 1:3 into fresh culture dishes. The fourth passage hPDSCs (5 × 10⁴ cells/well) were seeded in 60-mm culture dishes and cultured with 50% extracted sample solutions and 50% modified α-MEM supplemented with 100 U/mL penicillin, 0.1 mg/mL streptomycin, and 10% fetal bovine serum in a CO₂ incubator for 1, 3, and 6 days. Living cells were harvested and the cell numbers were counted by a hemacytometer under a light microscope. Each sample had 6 duplicated count cell numbers and the means with SD were calculated. Analysis of variance (ANOVA) statistical analysis was applied to assess the differences of the cells proliferation between the test groups and the control group; a P value of <.05 was considered statistically significantly different between groups.