To investigate the effect of the sequential delivery of bone morphogenetic proteins BMP-2 and BMP-7 on bone regeneration in rat calvarial defects (40 Sprague-Dawley rats, 8 mm defect size), all animals were treated with a hydroxyapatite (HA)/tricalcium phosphate (TCP) bone graft covered with a collagen membrane. The experimental groups were as follows: (1) control group: unmodified collagen (no treatment); (2) BMP-2 group: 5 μg of BMP-2; (3) hep-BMP-7 group: 5 μg BMP-7 chemically bound to heparinized collagen; and (4) BMP-2/hep-BMP-7 group: 2.5 μg BMP-7 bound to heparinized collagen and subsequently treated with 2.5 μg BMP-2. Defect healing was examined at 2 and 8 weeks after surgery. The BMP-2 group showed the largest new bone area at week 2 (29.3 ± 7.3%; P = 0.009); new bone areas in the hep-BMP-7 and BMP-2/hep-BMP-7 groups were similar (11.8 ± 3.4% and 12.9 ± 5.71%, respectively; P = 0.917). After 8 weeks, the BMP-2/hep-BMP-7 group showed the largest new bone area (43.3 ± 6.2%), followed by the BMP-2 and hep-BMP-7 groups ( P = 0.013). Accordingly, in comparison with single deliveries of BMP-2 and BMP-7, sequential delivery of BMP-2 and BMP-7 using a heparinized collagen membrane significantly induced new bone formation with a smaller quantity of BMP-2 in rat calvarial defects.
Bone morphogenetic proteins (BMPs) trigger angiogenesis, the migration and proliferation of mesenchymal cells, and their differentiation into osteoblasts and chondroblasts. Recently, recombinant human (rh) BMPs have been produced using BMP gene transfected mammalian cell cultures (Chinese hamster ovary (CHO) cells), and rhBMP-2 and rhBMP-7 are now commercially available for the treatment of bony defects. One of the problems associated with the clinical application of CHO cell-derived rhBMP-2 (CrhBMP-2) is its high cost due to the need for high doses. One possible way of solving this problem is to produce monomer rhBMPs from BMP gene transfected Escherichia coli , as this would offer greater production efficiency at lower cost. Recently, Huh et al. conducted a randomized clinical trial to examine the bone-inducing ability of E. coli -derived rhBMP-2 (ErhBMP-2) on socket preservation, and determined the outcomes and effects of ErhBMP-2 on osseointegration.
Several studies have reported that BMP-2 improves local bone regeneration, but its application in vivo has not been shown to improve bone regeneration significantly. These poor in vivo results were considered to be due to the rapid degradation of BMP-2 by proteinases, and thus it was suggested that BMP-2 be administered in more than milligram amounts. Such high concentrations of BMP-2 could cause systemic as well as local side effects, such as ectopic bone formation, osteoclast activation, cyst-like bone void formation, and soft tissue swelling.
BMP-2 is an initial factor in the expression of the transforming growth factor beta (TGF-β) superfamily during fractured bone healing, whereas BMP-7 is usually expressed 2 weeks after bone fracture. Yilgor et al. conducted a study to determine the degree of proliferation and differentiation of mesenchymal cells that occurs when rhBMP-2 and rhBMP-7 are released simultaneously or sequentially from nanocapsules. It was found that the sequential delivery of BMP-2 then BMP-7 facilitated mesenchymal cell proliferation and differentiation into osteoblasts. Based on the results of their study, we expected that the sequential application of BMP-2 and BMP-7 would effectively enhance bone regeneration as compared to BMP-2 alone. However, their study was conducted at the cellular level using an acidic delivery medium and nanocapsules, and thus the technique used is likely to be difficult to apply. On the other hand, collagen membranes are commonly used to treat defects of the craniofacial region, including alveolar bone, in a process called guided bone regeneration (GBR). Furthermore, in many studies and clinical situations, collagen has been used successfully as a carrier for protein delivery.
To achieve the sequential delivery of BMP-2 and BMP-7, heparin can be used to bind BMP-7 and slow down its release. Heparin is a sulphated mucopolysaccharide and binds various growth factors, such as TGF-β and vascular endothelial growth factor (VEGF), using its negatively charged sulphate group. Many authors have described the controlled release of growth factors from heparin-conjugated material, but no previous study has utilized collagen membranes for dental GBR. Therefore, we investigated the effect of the sequential delivery of BMP-2 and BMP-7 using a heparinized collagen membrane delivery system on GBR in rat calvarial defects.
Materials and methods
Toluidine blue O (TBO), (3-dimethylaminopropyl)-3-ethylcarbodiimide (EDC), N-hydroxysuccinimide (NHS), and 2-morpholinoethane sulphonic acid buffer (MES) were purchased from Sigma Aldrich (St. Louis, MO, USA); heparin sulphate salt was obtained from Acros Organics (Morris Plains, NJ, USA). Collagen membranes were purchased from GENOSS (Suwon, South Korea), and ErhBMP-2 (referred to as BMP-2) and ErhBMP-7 (referred to as BMP-7) were obtained from PeproTech (Rocky Hill, NJ, USA). All other chemicals and solvents were of analytical grade and used without further purification.
Heparinization of collagen membranes
Collagen membranes were heparinized using simple EDC/NHS chemistry. In brief, heparin sulphate salt (1 mg/ml) was dissolved in 0.1 M MES buffer (pH 5.3), and 25 mM EDC and 50 mM NHS were dissolved in this mixture to generate an amine-reactive NHS-ester in the heparin. Collagen membranes were soaked in this mixture and stirred for 4 h. Thereafter, membranes were collected, washed repeatedly using double-distilled (DI) water, and freeze-dried ( Fig. 1 ).
Characterization of heparinized collagen membranes
To investigate structural changes in collagen membranes after heparinization, membranes were coated with platinum using a sputter coater for 70 s. Thereafter, scanning electron microscope (SEM) analysis (JSM-6390; JEOL, Kyoto, Japan) was performed using a 10 kV electron beam at a distance of 10–12 mm. TBO staining was used to determine the amounts of heparin on membranes. In brief, a heparinized membrane was stained by placing it in TBO solution (0.1 M HCl, 20 mg NaCl, and 4 mg TBO chloride), which was stirred for 2 h. Unbound TBO was then completely removed by repeat washing in DI water, and the TBO was then extracted from the membrane using 0.1 M NaOH/ethyl alcohol (1:4, v/v). Optical density was measured using a microplate reader at 630 nm (PowerWave XS; Biotek, Winooski, VT, USA). Concentrations were determined using a standard curve prepared using known TBO concentrations.
In vitro release of BMP-2 and BMP-7
Heparinized collagen membranes were pre-wetted with DI water and placed in a BMP-7 solution (100 ng/100 μl) for 24 h, which allowed the BMP-7 to bind chemically with the heparin. Membranes were then removed and freeze-dried at −80 °C for 24 h. To physically adsorb BMP-2, freeze-dried BMP-7-treated membranes were immersed and shaken in BMP-2 solution (100 ng/100 μl) for 1 h prior to in vitro release experiments. To obtain membrane release profiles, BMP-2/BMP-7-treated membranes ( n = 4) were placed in 10 ml phosphate buffered saline (PBS) containing 0.1% bovine serum albumin and in a constant temperature shaking water bath (100 rpm, 37 °C; Lab Companion, Seoul, Korea) for 30 days. Samples (100 μl) of medium were taken at prescribed times and an equivalent volume of fresh medium was added. Subsequent analysis was performed using an enzyme-linked immunosorbent assay (ELISA) kit (PeproTech). Briefly, samples (50 μl) were diluted with diluent (0.5% Tween 20 in PBS), loaded into the wells of a capture antibody-coated microplate, and then treated with biotinylated anti-human BMP antibody, streptavidin, and 3,3′,5,5′-tetramethylbenzidine (TMB) substrate. Optical densities were measured at 450 nm using a microplate reader.
Forty Sprague-Dawley rats (males, weighing 250–300 g) were used. Animals were individually housed in plastic cages under standard laboratory conditions, and had access to water and standard laboratory pellets ad libitum. Animal selection, care, management, the surgical protocol, and animal preparation for surgery were conducted in accordance with the guidelines issued by the Ethics Committee on Animal Experimentation of the Korean Atomic Energy Research Institute (KAERI-IACUC-2013-004).
All surgical procedures were performed under general anaesthesia after intramuscular administration of a mixture of tiletamine–zolazepam (Zoletil; Virbac Laboratories, Carros, France) and xylazine (Rompun; Bayer Korea, Seoul, Korea). The cranial surgical site was shaved and disinfected using Betadine; 2% lidocaine HCl (Yu-Han Co., Gunpo, South Korea) was used for local anaesthesia. A U-shaped incision was made and a full-thickness flap, including skin and periosteum, was elevated. A trephine bur (3i Implant Innovation, Palm Beach Garden, FL, USA) was used to create a standardized, circular, transosseous, 8-mm defect in the middle of the cranium. The surgical site was irrigated continuously with saline during drilling. Control and experimental materials were introduced to prepared sites after carefully removing the trephinated bony disc. Ten animals were allocated to each of the four study groups. Each animal was treated with 0.12 mg hydroxyapatite (HA)/tricalcium phosphate (TCP) bone graft material (Bio-C; Cowellmedi Co. Ltd, Seoul, Korea) at the defect site, which was then covered with a 10 mm × 10 mm collagen membrane ( Fig. 2 a) . The four study groups were as follows: (1) the control group: collagen membrane (no treatment); (2) the BMP-2 group: collagen membrane treated with 5 μg BMP-2 just prior to application; (3) the hep-BMP-7 group: heparinized collagen membrane containing bound BMP-7 (5 μg); and (4) BMP-2/hep-BMP-7 group: heparinized collagen membrane containing bound BMP-7 (2.5 μg) and treated with BMP-2 (2.5 μg) just prior to application ( Table 1 ). All surgical sites were sutured with 4–0 absorbable sutures (Vicryl; Ethicon, Somerville, NJ, USA). Five animals in each group were allowed a 2-week healing period, and the remaining five animals an 8-week healing period.
|Group||Number||Collagen membrane treatment||Growth factor (dose)|
|Hep-BMP-7||10||Heparinized collagen membrane||rhBMP-7 (5 μg)|
|BMP-2||10||Collagen membrane||rhBMP-2 (5 μg)|
|BMP-2/Hep-BMP-7||10||Heparinized collagen membrane||rhBMP-7 (2.5 μg)
rhBMP-2 (2.5 μg)
Rat calvaria tissue specimens were decalcified using 14% ethylenediaminetetraacetic acid (EDTA) and rapid acid decalcification reagents. Specimens were then embedded in paraffin and 5-μm sections obtained from the centres of the calvarial defects. The two central most sections in each block were selected and stained with haematoxylin–eosin and Masson’s trichrome. The prepared histology slides were observed under a light microscope (BX50, Olympus, Tokyo), and images were captured digitally. To identify areas of new bone and of residual biomaterials in the images, computer-assisted histometric measurements were done using an image analysis program (Image-Pro Plus; Media Cybernetic, Silver Spring, MD, USA). Percentages of new bone and residual biomaterials in all defect areas were calculated using the equations shown in Fig. 2 b.
The statistical analysis was performed using SPSS v. 18.0 software (SPSS Inc., Chicago, IL, USA). The Kruskal–Wallis test was used to determine differences between histometric measurements taken in the four groups at 2 and 8 weeks after surgery and between measurements taken within groups. The Mann–Whitney U -test was used for post hoc testing. Statistical significance was accepted for P -values <0.05.
Heparinization of collagen membranes
A schematic diagram of the designed sequential BMP delivery carrier is illustrated in Fig. 1 . The morphologies of collagen membranes before and after heparin conjugation are presented in Fig. 3 a . Nanofibrous collagen bundles (fibre diameter 243 ± 56 nm) were observed on the surfaces of membranes. Heparinized membranes had similar nanofibre diameters, and although some rupture of collagen fibres was observed on membrane surfaces, the fibrous structure of the collagen was maintained. The conjugation of heparin on membranes was visualized by TBO staining. As shown in Fig. 3 b, collagen membranes were not stained by TBO, whereas heparinized membranes were stained blue. Amounts of bound TBO were determined by extraction using an alkaline solution. Although it was difficult to determine the degree of staining by eye, only small amounts of TBO were detected on non-modified membranes. Collagen membranes contained 3.1 ± 6.3 nM of TBO, which was significantly lower than that contained by heparinized membranes (194.4 ± 7.8 nM; Fig. 3 c).
In vitro release kinetics of BMP-2 and BMP-7
Prior to treating membranes with BMP-2, BMP-7 was loaded for 24 h to enable it to bind with heparin. The loading efficiency of BMP-7 was approximately 60% (loaded amounts: 58.1 ± 6.2 ng/membrane, n = 4), and the amount of BMP-2 that was subsequently physically adsorbed onto membranes during 1 h of treatment was 41.5 ± 2.8 ng/membrane (loading efficiency: 40%, n = 4). The release profiles of BMPs from heparinized membranes are presented in Fig. 4 . The percentage of BMP-2 released was 75.4 ± 3.4% at 1 h, whereas only 20.8 ± 1.7% of BMP-7 was released at this time. Whereas all of the loaded BMP-2 was released within 30 h, mean BMP-7 release was 36.5 ± 1.4%, 46.9 ± 1.3%, and 53.1 ± 5.5% after 4, 7, and 14 days, respectively. At 28 days, the mean BMP-7 release was 62.5 ± 12.7%.