As previously described, rhBMP-2 was expressed in E. coli at the Research Institute of Cowellmedi (Busan, Korea) [25]. Briefly, total RNA from human osteosarcoma cells was reverse-transcribed with reverse transcriptase (Gibco BRL, Grand Island, NY, USA). The cDNA encoding of the mature form of the BMP-2 protein was amplified via polymerase chain reaction. The cDNA of human BMP-2 (hBMP-2) was then subcloned into a pRSET(A) vector (Invitrogen, Paisley, UK) to produce the pRSET(A)/hBMP-2 expression vector, which was used to transform the E. coli BL21(DE3) strain. A high-cell density cultivation of E. coli was crushed twice in a French press and then centrifuged. The pellet was then resuspended at 25 mg wet weight/mL in a suspension buffer (20 mM Tris-HCl [pH 8.5], 0.5 mM ethylenediaminetetraacetic acid [EDTA], and 2% v/v Triton X-100), and then centrifuged again. The inclusion bodies (pellets) were resuspended and incubated overnight.

For the in vitro dimerization, the solubilized rhBMP-2 was incubated in a renaturation buffer (0.5 M guanidine-HCl, 50 mM Tris-HCl [pH 8.5], 0.75 M N-cyclohexyl-2-aminoethanesulfonic acid, 1 M NaCl, 5 mM EDTA, and 3 mM total glutathione). Purification of the active rhBMP-2 (dimer) was performed with a heparin column (Heparin Sepharose 6 Fast Flow, GE Healthcare, Milwaukee, WI, USA). The active rhBMP-2 protein was eluted and separated using a stepped NaCl gradient (0.15, 0.3, and 0.5 M). Finally, this rhBMP-2 was reconstituted and diluted in a buffer to a concentration of 0.1 mg/mL.

An experiment to determine the release kinetics of rhBMP-2 in vitro was conducted using a slight modification of a previously described method [26]. Briefly, bMBCPs (n=5) loaded with 0.1 mL of rhBMP-2 solution were placed into 2 mL microcentrifuge tubes containing 1.0 mL of phosphate-buffered saline solution (pH 7.4) and 0.02% (w/v) sodium azide. The tubes were incubated at 37℃ with continuous agitation. The supernatant medium was collected and completely replaced with a fresh buffer solution at the scheduled time points. The amount of rhBMP-2 in the supernatant was determined by a competitive indirect enzyme-linked immunosorbent assay (ELISA) for rhBMP-2 (Human BMP-2 R&D Systems, Minneapolis, MN, USA) according to the manufacturer's protocol. The absorbance of the samples was read at 450 nm using an ELISA plate reader (Model 680, Bio-Rad Laboratories Inc., Hercules, CA, USA). The amount of rhBMP-2 was calculated with the aid of a calibration curve (Fig. 1).

Titanium-surfaced mini-implants (1.8 mm in diameter and 12 mm long; Cheil Pharma and Instrument, Seoul, Korea) were placed into the bMBCPs (4×5×15 mm; Biomatlante, Vigneux de Bretagne, France). The blocks had one longitudinal hole to allow for implant placement and three horizontal holes to facilitate bone healing (Fig. 2). The diameter of each horizontal hole was 2 mm, while that of the vertical hole was 1.5 mm. Implants were installed manually along the vertical hole, with subsequent confirmation that they did not rotate after fixation. The bMBCP was treated overnight with either a phosphate buffer solution (control group) or rhBMP-2 solution at 0.1 mg/mL (test group) before transplantation.

Ten male Sprague-Dawley rats (body weight 250 to 300 g) were assigned to either the test or control group (n=5 animals per group). The animals were maintained in plastic cages in a room with a 12-hour day/night cycle, an ambient temperature of 21℃, and ad libitum access to water as well as a standard laboratory pellet diet. Animal selection, management, surgical protocol, and preparation followed routines approved by the Institutional Animal Care and Use Committee of Yonsei Medical Center in Seoul, Korea.

The animals were placed under general anesthesia using an intramuscular injection (5 mg/kg body weight) of ketamine hydrochloride (Ketalar, Yuhan, Seoul, Korea). The surgical site was shaved and scrubbed with iodine, and a vertical incision was then made in the skin of the back. After flap reflection, a subcutaneous pocket was prepared by blunt dissection, and a bMBCP with a titanium implant, treated or untreated with rhBMP-2, was implanted into the pocket. The skin was sutured for primary closure with 4-0 Vicryl sutures (Polyglactin 910 braided absorbable suture, Ethicon, Johnson & Johnson, Edinburgh, UK). The animals were sacrificed after a healing period of 8 weeks, and samples were later obtained for histological analysis.

After the bMBCP transplants were retrieved from the animals after necropsy, clinical photographs were taken (Fig. 3). Block samples were fixed in a 10% formalin solution, dehydrated in an ascending graded ethanol series (70%, 80%, 90%, 95%, and 100%) and pure acetone, and then impregnated and embedded in glycol methylmethacrylate. Polymerization was achieved by increasing the temperature from 20℃ to 80℃. Each resin block was cut into two sections (15 µm thickness) along the longitudinal axis of the biopsy sample using an internal circular diamond saw (Leitz, Wetzlar, Germany) [25]. After staining with hematoxylin and eosin (H&E), all of the samples were examined with the aid of light and polarized-light microscopy (BX50, Olympus Co., Tokyo, Japan), as well as scanning electron microscopy (SEM; S-4300, Hitachi, Tokyo, Japan). The SEM observations were performed after the sections were polished and sputtered with gold palladium, using secondary and back-scattered electrons at 15 kV (Leo 1450VP, Zeiss, Germany). The most central section from each block was selected to compare the findings between the groups. New bone formation and osseointegration in H&E-stained histological slides were analyzed quantitatively using an image analysis system (Image-Pro Plus; Media Cybernetics, Silver Spring, MD, USA). The following parameters were measured: 1) Bone-to-implant contact (BIC): The length of osseointegration that was along the thread profile and in direct contact with the newly formed bone was calculated for both sides of the implant on each of the two central sections obtained from each implant; 2) Bone density within the implant threads: The area of the newly formed bone within the implant threads was measured. The value of each was expressed as a percentage.

The general pattern of bone formation was favorable in the test group, which was in line with the results from previous studies [24,27,28]. Newly formed bone stained with H&E exhibited a high-density, predominantly lamellated bone pattern along the implant surface and within the micropores of the bMBCP. This pattern was different from the control group, in which no mineralized tissue had formed. The distribution of the newly formed bone was equally observed along the implant and bMBCP surfaces, while the bridging pattern between the implant and bMBCP surfaces was clearly evident. The newly formed bone did not appear to be homogeneous and exhibited cement lines separating areas of bone that had been deposited at different times. The new bone appeared to be well mineralized and had numerous lacunae. Numerous osteocytes were observed, as well as a few osteoblasts that were aligned along the surface of the newly formed bone.

The results of the histomorphometric analysis are shown in Table 1. The percentage of direct BIC was 41.23±4.13% (mean±standard deviation) in the test group (n=5), while the control group exhibited no bone formation along the titanium mini-implants (n=5). The bone density in the test group was 33.47±5.73%.

Polarized-light microscopy was used to evaluate the patterns of lamellation in the newly formed bone, which is considered to be an important hallmark of bone maturation. In the treated group, groups of mineralized collagen fibers with numerous osteocytes were clearly observed under polarized light (Fig. 4C), while there appeared to be new bone formation in direct contact with the bMBCP and implant surfaces. Conversely, no mineralized tissue was found in the control group (Fig. 4F).

The back-scattered SEM images confirmed that the area of newly formed bone was significantly larger in the rhBMP-2-treated group than in the untreated control group 8 weeks after transplantation (Fig. 5). Newly formed bone was generally observed within the bMBCP pores and along the implant surface, appearing as light-gray areas. However, no mineralized tissue was observed in the sections from the control group. Collectively, mineralized bony tissue on the implant and bMBCP surfaces (including osteocytes, a Haversian system, and cement lines) was observed exclusively in the test group both during polarized light microscopy and in back-scattered SEM images.