ⅠINTRODUCTION

Dental implants and various types of grafting materials are widely used to rehabilitate the masticatory function and to augment bone deficiencies. However, the use of dental implants and grafting materials are limited by the delay or failure of osteogenesis owing to poor bone quality or deficiency of jaw bone. In addition, there can be unpredictable complications, such as foreign body reaction and bacterial infection, on various grafting materials^1,2).

Many studies have focused on the development of new surfaces, which increased the osteogenesis on titanium implant surfaces³⁾. However, these surface improvements are not critical to the success of dental implantation, and other problems, such as bone deficiencies in the implanted sites and the slow integration of grafted material to the titanium surface, have yet to be resolved. Therefore, considerable efforts have been made to find suitable biomaterials to enhance osteoinduction⁴⁾. Among these, a few bone morphogenetic proteins(BMPs) are known to improve osteoinductive effect by differentiating mesenchymal cells to osteoblasts^5,6) and to play a key role in blood vessel formation in skeletal and tumoral development in mammals^7-9). BMPs were originally identified as inducing ectopic bone formations when implanted into muscle tissues¹⁰⁾. Except for BMP-1, all BMPs are members of the transforming growth factor-β(TGF-β) superfamily^11,12). Among these BMPs, the BMP-2 was demonstrated to have excellent capacity in enhancing bone formation⁶⁾. Although the usefulness of BMP-2 in clinical setting is clear, their isolation methods are quite complicated and expensive for commercial production. In addition, lager amounts of BMP-2 were needed to induce new bone formation in human and monkeys(100-fold larger for rodents)¹³⁾. Today, recombinant human BMP-2(rhBMP-2) is produced by gene transfection to immortalized mammalian cellsorE.colibacteria^12,14). Of the two methods, rhBMP-2 produced by E. coli bacteria is cheaper than rhBMP-2 derived from immortalized human cells. Due to commercialization of the E. coli method, the practical clinical applications of BMP-2 in dental and orthopedic procedure are near market actualization¹⁵⁾. Although some complications such as inflammatory reaction, edema and heterotropic bone formation have been reported¹⁶⁾, many researchers insisted on the safety of BMP-2 application to dental and orthopedic procedures, and several investigators have emphasized the dose and containment of the BMP-2 material as key determinants in decreasing complication rates¹⁷⁾.

The activity of BMPs is suppressed or enhanced by various molecules^18,19). Noggin, chordin, and follistatin are examples of proteins which suppresses BMP activity in the extracellular space by inhibiting the binding of BMPs to cell-surface receptors¹⁹⁾. In contrast, heparin enhances the action of BMPs but does not have any capacity to differentiate from mesenchymal progenitor cells to osteroblasts by itself. The capacity of these sulfated polysaccharides to enhance the actions of BMPs is likely to depend on the size and number of sulfated residues²⁰⁾. Zhao et al.²¹⁾ demonstrated that heparin potentiated ectopic bone formation induced by BMP-2 in vivo, but there still has been no morphologic and histometrical studies about the effects of heparin on the critical sized bony defects, and a large number of studies regarding the effects of heparin to BMP-2 have been conducted in terms of peripheral new bone formation around dental implants or grafted materials²²⁾. Because of this, the pre-existing studies are not specific to osteoinduction as the newly formed peripheral bone may contain both osteoinductive and osteoconductive bones. The actual osteoinductive capacities of certain materials must be expressed by migration of progenitor cells and their differentiation to osteoblasts at the destination sites⁹⁾. Consequently, the osteoinductive effects of a given material measured from peripherally formed new bone cannot be accurate. In this study, we intended to evaluate morphological changes of cultured adipose stem cells obtained from rabbits with the adding of varying concentrations of heparin(0, 0.25, 2.5, and 25 μg/ml) repectively and a constant concentration of rhBMP-2(150 ng/ml) to culture media. In addition, we measured the amount of internally formed bone in grafted material at different concentrations of heparin(0, 0.25, 2.5, and 25 μ g/ml) repectively and constant concentration of rhBMP-2(5 μg/ml), and then the grafts were placed into critical sized calvarial defects in rabbits. Finally, we also investigated the action-pattern of heparin by elapsed time and different concentrations of heparin.

ⅡMATERIALS and METHODS

1Preparation of rhBMP-2, heparin, and anor ganic bovine bone

Purified rhBMP-2 was purchased from Cowell medi Co, Ltd.(Seoul, Korea). The rhBMP-2 produced by E. coli expression system. Purified heparin for injection was obtained from Choong Wae Pharma Co, Ltd.(Seoul, Korea). Anorganic bovine bone was supplied by NIVEC(Seoul, Korea, porous type, particle size 1.0~2.0 mm).

2Preparation of progenitor cells for morphologic study

Sterile adipose tissue was harvested from the axillae of forelegs of a rabbit (4-month-old; Newzealand White), and was very finely minced and chopped with sterile eye scissors. This prepared adipose tissue was dispersed in 4´5 cm² culture flasks with 0.5ml of α-MEM and 10% fetal bovine serum(Bio Whittaker^Ⓡ, Walkersville, MD, USA) and incubated at 37℃ and 5% CO2 for 4days. After verifying cell migration from prepare dadipose tissue (Fig. 1A), culture media was changed at 2-day intervals for 8days (primary culture).

Cells were detached from flasks by trypsin (Bio Whittaker^Ⓡ, Trypsin-Versene Mixture, Walkersville, MD, USA). The mean cell density were 2x10⁹cells/ml and the cells were dispersed into an other flask at a density of 2x10³cells/ml and cultured (secondary culture).

Verification of stem cell lineage was performed by using monoclonal antibodies for STRO-1 (Cat. No. SC-47733, Fig. 1B) and CD 90 (Santa Cruz, CA, USA, Fig. 1C), which are expressed in mesenchymal stem cells. The verified stem cells were divided into 4 groups of four concentrations of heparin (0, 0.25, 2.5, and 25 μg/ml) with a constant rhBMP-2 concentration (150 ng/ml) and cultured for 2, 4, and 8 days. Naphtol AS phosphate-fast blue BB method for alkaline phosphatase content and Alizarin red stain for calcium content were performed at time schedules and the cultured adipose stem cells were inspected for morphology and osteoblastic activity.

3Graft material preparation

After rhBMP-2 (5 μg/ml) and the aforementioned concentrations of heparin (0, 0.25 2.5, and 25 μg/ml) were mixed, respectively, they were blotted into 0.5 g anorganic bovine bone, freeze dried, and were stored at -60 ℃ until they were grafted into 8-mm diameter rabbit calvarial defects.

4Surgical procedure

The animal experiments was reviewed, and approved by the animal ethics committee at Dankook University (Cheonan, Korea). All procedures were carried out under sterile conditions. Forty-eight rabbits (8-month-old; male; New Zealand White) were housed and acclimated in cages with access to food and water for 1 week ad libitum. Afterwards, rabbits were divided into 4 groups according to heparin concentration (0, 0.25, 2.5, and 25 μg/ml). After general anesthesia with isofluran (Choong Wae Pharma Co, Ltd., Seoul, Korea; imported form ABBOTT, USA), local anesthesia was performed in the calvarial region by lidocaine HCl with epinephrine (1:80,000). An incision was made on the scalp and four 8-mm craniotomy holes prepared on exposed calvaria using a trephine bur. When the dura mater was exposed, the bur was stopped and bony chips removed. Subsequently, the prepared grafting materials were inserted into the holes (Fig. 2). According to the time schedule, four rabbits from each group were sacrificed at 1, 3, and 6 weeks.

5Histologic study

The harvested bony tissues were fixed in 10% buffered neutral formaline and demineralized with 10% formic acid. The prepared sections were treated with Hematoxylin & Eosin (H&E) stain and Masson's trichrome (MT) stain. Photographs of each stained section (2 sections per a rabbit; total 8 specimens on every group) were taken under ´50 magnification, and the images were analyzed using IPTK version 5.0 software (Reindeer company, USA) to measure the total area of internally-formed new bone in the grafted material (1-2 mm) and the total area of grafted material. Subsequently, the total area of internally formed new bone was divided by the total area of grafted material, and the result was expressed as % ratios.

6Statistical analysis

Statistical analysis was performed using SPSS Version 17.0 (IBM Co, USA). The differences in each obtained percent values were evaluated by one-way analysis of variance. Post Hoc tests of ANOVA tests (Scheffe’s method) were performed to verify the significance; A value p < 0.05 was considered significant.

ⅢRESULTS

1Morphologic effects of heparin with rhBMP-2 to adipose stem cells in cultural experiment

In morphologic evaluation, the degrees of morphologic changes and osteoblastic activities showed similar patterns in both the Naphtol AS phosphate-fast blue BB staining and the alizarin r ed s taining. T hese r esults a re summarized i n Fig. 3. All images about experimental results are exhibited at Fig. 4 and 5.

When only 150 ng/ml of rhBMP-2 was added to culture (control group), cells did not show morphological characteristics of osteoblasts at 2 days. Only few cells reacted to Naphtol AS phosphate-fast blue BB staining and alizarin red staining. The reaction was very mild such that the osteoblastic differentiation seemed to be uncertain (Fig. 4A, 5A). With 150 ng/ml rhBMP-2 and 0.25 μg/ml heparin (Group 1) at 2 days, some cells displayed an early form of osteoblasts and mildly reacted with Naphtol AS phosphate-fast blue BB staining and alizarin red staining (Fig. 4D, 5D). With 150 ng/ml rhBMP-2 and 2.5 μg/ml heparin (Group 2) at 2 days, many cells exhibited morphological characteristics of osteoblasts, and showed some positive reaction in Naphtol AS phosphate-fast blue BB staining and alizarin red staining (Fig. 4G, 5G). With 150 ng/ml rhBMP-2 and 25 μg/ml heparin (Group 3) at 2 days, most cells exhibited morphological characteristics of osteoblasts, and also showed strong positive reaction in Naphtol AS phosphate-fast blue BB staining and alizarin red staining (Fig. 4J, 5J). At 4 days, the control group developed morphologic similarity to the osteoblasts, and reacted positively with Naphtol AS phosphate-fast blue BB staining and alizarin red staining (Fig. 4B, 5B). At 4 days, group 1 cells produced features more similar to osteoblasts and reacted to Naphtol AS phosphate-fast blue BB staining and alizarin red staining more strongly than at 2 days (Fig. 4E, 5E). In the group 2 at 4 days, the morphologies of cells had a more advanced form of osteoblasts and features similar to morphology of group 3 at 2 days (Fig. 4H, 5H). At 8 days, the cell shapes of group 1, 2, and 3 exhibited morphologically degenerating forms, except for the control group, and the degree of morphological degeneration increased with increasing concentrations of heparin. The cellular reactivity to the Naphtol AS phosphate-fast blue BB and alizarin red also decreased even more at 0.25 μg/ml. Especially at 25 μg/ml heparin, they were most inactive in the morphology and reactivity (Fig. 4, 5; C, F, I, L).

2Histologic effects of heparin with rhBMP-2 in rabbit ’ s calvarial defects

The control group (administered only 5 μg/ml of rhBMP-2, no heparin). In one week, newly formed collagen fibers (Fig. 6A; NC: New Collagen-blue color) were observed around the multiporous grafted material (GM: Grafted Material). We observed a few lymphocytes (Fig. 6D; thick arrow), some macrophages (Fig. 6D; Mø) and many capillaries (Fig. 6D; BV: Blood Vessel) around the graft material. Cuboidal cells with a prominent nucleus (Fig. 6D; thin arrow) that resembled osteoblasts were observed in the contact area between the grafted material and connective tissue. However, no changes were observed in the grafted material (Fig. 6D). At 3 weeks, the amount of newly formed collagen fibers increased in the connective tissue around the grafted material, but no changes were observed in the grafted material (Fig. 6B). At 6 weeks, collagen fibers and peripheral bone had formed around the grafted material (Fig. 6C; PB: Peripheral Bone), and some changes were found within the graft material (thin arrow), but these internal changes could not be accepted as histologic standards of new bone formation because these changes did not have clear boundary, specific color for bony material (newly formed bone shows blue color and formed bone exhibits brown to red color in MT stain), and lacunae (Fig. 6C, thin arrow). On the contrary, the peripheral bone around grafted material exhibited standards of newly formed bone (Fig. 6E).

The first group (treated with mixture of 5 μg/ml of rhBMP-2 and 0.25 μg/ml of heparin). In a week, the amount of newly formed bone around the grafted material was very small, but clear changes were observed in the internal areas of grafted materials (Fig. 6F; IB: Internal Bone). These newly formed internal structures exhibited a ring or linear shape, and higher magnified examination revealed newly formed bone containing osteocytes and lacunae (Fig. 6I; OB: Osteoblast). The center of newly formed bone showed a space containing some cells, which were believed to be red blood cell (RBC), white blood cell (WBC), and mesenchymal cells. At 3 weeks, the amount of newly formed peripheral bone and newly formed internal bone had increased from what was observed at 1 week (Fig. 6G). At 6 weeks, the amount of peripheral new bone and internal new bone had increased from 3 weeks (Fig. 6H).

The second group (treated with a mixture of 5 μg/ ml of rhBMP-2 and 2.5 μg/ ml of heparin). In a week, significantly increased formation of new internal bone was observed in greafted material (Fig. 6J; IB). At 3 weeks, the amount of peripheral and internal new bone was more than those observed at 1 week. In addition, internally formed bone were partly interconnected with peripheral bone (Fig. 6K; arrow). At 6 weeks, the amount of newly formed peripheral bone and newly formed internal bone were more than those observed in 3 weeks (Fig. 6L). An image at higher magnification indicated that the internal new bone had complete bony structures, i.e. osteocytes (OC), osteoblasts (OB), and lacunae (Fig. 6M). The central area contained blood vessels which presumably contained RBC, WBC, and progenitor cells (Fig. 6M; BV).

The third group (treated with a mixture of 5 μg/ml of rhBMP-2 and 25 μg/ml of heparin). At one week, significant amounts of new internal bone were observed in the grafted material (Fig. 6N). At 3 weeks, the amounts of new peripheral and internal bones were more increased than those observed after the first (Fig. 6O). At 6 weeks, the amounts of new peripheral bone and internal bone were higher than those observed at 3 weeks (Fig. 6P). The formed spaces in the internal bone exhibited both types of hematopoietic marrow (HM) and fatty marrow (FM) (Fig. 6P).

3Histometric Analysis

The control group did not show critical changes of bone formation in the grafted material throughout the experimental period. In general, the % ratio values of internal new bone to grafted material increased with time elapsed in the first, second, and third groups (Table. 1, Fig. 7). The overall % ratio of internal new bone to graft material also increased with increasing heparin concentration up to a maximum of 5 f old (Table. 1 , Fig. 7 ). I n t he f irst g roup, t here w as a significant difference between 1 week and 6 weeks (p=0.002), and no significant differences were observed for other comparisons (Fig. 7). Between 1 and 3 weeks, there were significant differences across the first, second, and third group. However, at 6 weeks, there were no significant differences across the first, second, and third group (Table. 1, Fig. 7).

ⅣDISCUSSION

This study evaluated the action-pattern of heparin by observing the culture of adipose stem cells at different concentrations of heparin with a fixed concentration of rhBMP-2 and by observing critical sized rabbit calvarial defects, which were grafted with multiporous anorganic bovine bone blotted with different concentration of heparin and fixed concentration of rhBMP-2. In culture experiments, the concentration of rhBMP-2 was fixed as 150 ng/ml and heparin concentrations were 0 (control), 0.25 (group 1), 2.5 (group 2), and 25 (group 3) μg/ml. In the histological study, the concentration of rhBMP-2 was fixed with 5μg/ml, and the adopted concentrations of heparin were 0 (control), 0.25 (group 1), 2.5 (group 2), and 25 μg/ml (group 3), because Hall²³⁾ had achieved ectopic bone formation in a rat model by coating the implant surfaces with 5 μg/ml rhBMP-2, and Zhao²⁴⁾ induced a different amounts of ectopic bone by mixing rhBMP-2 of 5 μg/ml with 0, 0.25, 2.5, 25 μg of heparin i n t he s ame a nimal model. T he g roup without a heparin dose (0 μg/ml) was used as the control. Because anorganic bovine bone substitutes (i.e. Bioss) have prominent osteoconducting capacity, it has been used as bone substitutes in this study²⁵⁾.

Currently, the most serious and obvious clinical complication of using rhBMP-2 in human is the development of swelling or edema, which is proportionate to the amount of rhBMP-2 applied²⁶⁾. We believe the edema occurs due to two mechanisms. The first is that BMP-2 promotes angiogenesis. Several studies have claimed that BMP-2 promotes angiogenesis by activating endothelial cells through the stimulation of Smad1/5, Erk1/2, and Id expression^27,28). Also, newly formed vessels are known to cause edema in the surrounding tissue because of higher permeability to blood plasma than mature vessels²⁹⁾. The second possible mechanism is that the differentiated osteoblasts produces macrophage-colony stimulating factor (M-CSF) and receptor activated nuclear factor κB ligand (RANKL), which stimulate the action and proliferation of macrophages³⁰⁾. Denosumab, which is also called AMG162 or Prolia, is an antibody to RANKL and inhibits the proliferation of macrophages. This biologic agent has been used to treat osteoporosis, even though it often causes severe infections. This complication might be caused by the destruction of RANKL affecting the immune system³¹⁾. In other words, RANKL may promote an inflammatory and immune reaction, and cause swelling or edema. Other reported complications of rhBMP-2 are transient renal insufficiency and ectopic bone formation, which can occur with increasing dose of rhBMP-2^32,33). In cases of renal insufficiency, Latzman³²⁾ reported that patients suffering from these conditions had been administered higher doses of rhBMP-2 (24 mg) than unaffected patients (12 mg). Boakye³³⁾ reported that there was no heterotopic bone formation in patients after reduction of rhBMP-2 dose from half to one-quarter. Dickerman¹⁷⁾ explained that dose and containment was the key to avoiding complications of rhBMP-2. Therefore, the side effects of rhBMP-2 can be reduced by using smaller doses and improving delivery mechanisms¹⁷⁾. If the clinically applicable materials or methods which can reduce the dosage of rhBMP-2 by enhancing of rhBMP-2 action can be invented, the use of rhBMP-2 for bony defects will be more economical and will reduce morbidity when compared to the use of autogeneous bone grafts³⁴⁾. The activity of BMPs is enhanced or inhibited by various molecules^35,36). Noggin, chordin, follistatin, and the DNA family proteins bind to BMPs in the extracellular space and inhibit the binding of igands (BMPs) to the ell-surfacer eceptors³⁶⁾. In contrast, heparin enhances BMP-induced osteoblast differentiation in C2C12 myoblasts in vitro³⁷⁾. Heparin itself does not induce osteoblast differentiation from mesenchymal progenitor cells and only enhances the action of BMPs^24,37). Therefore, more definite confirmation of the enhancing pattern of heparin on rhBMP-2 is needed.

Rabbit adipose stem cells differentiated into osteoblasts in proportion to heparin concentration in morphological and histochemical evaluations. However, with the lapse of time, the differentiated osteoblasts showed fast degenerating changes in p roportion to heparin concentration. T o put i t another way, the heparin enhanced the effect of rhBMP-2 for osteoblastic differentiation and simultaneously stimulated degenerating changes of the completely differentiated osteoblasts (Fig. 3, 4, 5).

In the past, the osteoinductive effect of rhBMP-2 was evaluated by measuring the amount of peripheral total bone around the grafted material or the implanted sites. The term ‘osteoinduction’ theoretically suggests the recruitment of progenitor cells and the stimulation of these cells to develop into preosteoblasts³⁸⁾. Thus, the osteoinductive process must be evaluated by undifferentiated cell migration phenomenon. The osteoinductive effect of a certain material, not osteoconductive effect, must be evaluated by the internally formed bone in the grafted material because the peripherally formed bone around the grafted material can be produced by both osteoconductive and osteoinductive effects. Accordingly, we measured the area of internally formed bone in grafted material as a more accurate measurement for evaluating osteoinduction. In this study, the total area of newly formed internal bone was divided by the total area of grafted material, which was expressed as a % ratio (Fig. 7).

Although several studies reported that rhBMP-2 caused inflammation and edema, only a few lymphocytes and some macrophages had infiltrated the connective tissue around the grafted material and little edema was observed in the connective tissue only during the first week in our study (Fig. 6D). After 1 week, no inflammatory responses had developed in any of the experiment rabbits. In this study, the reduced inflammatory responses can be explained by three ways. The first is that half-lives of rhBMP-2 and heparin are very short. Most cytokines, including BMP-2 and heparin, have a short half-life. Hence, the action of cytokines is transient and limited to a local area^39-41). The second reason is the delayed release of rhBMP-2 and heparin. According to Hall²³⁾, rhBMP-2 coated on the surface of titanium with a thin porous oxide layer or an open porous oxide layer was released more slowly than that on a machined titanium surface layer. This meant that porosity can assist in prolonged release of rhBMP-2. Dickerman¹⁷⁾ reported that the most important factors in safe use of BMP were dose, delivery mechanism, and containment within the graft/bone margins¹⁷⁾. In this study, rhBMP-2 and heparin were blotted into the multiporous anorganic bovine bone, and the multiporous structure of the anorganic bovine bone could allow slow release into the adjacent environment. It could also prevent rapid increases of rhBMP-2 and heparin concentrations at the periphery of grafted material. Therefore, it is believed that inflammatory reactions were not severely provoked in peripheral connective tissue around the grafted material containing rhBMP-2 and heparin, by inhibiting sudden rising of concentration of rhBMP-2. The third reason is that rhBMP-2 potentiated the immune system and prevented infections in surgical wounds. Osteoblasts can produce M-CSF and RANKL⁴²⁾. Denosumab has been used clinically to treat osteoporosis but has been reported to provoke severe infection³¹⁾. This means that the destruction of RANKL by this antibody retards the immune system. However, an increase of RANKL by osteoblastic differentitation through the application of BMP-2 can potentiate the immune system. In the study by Hiremath⁴³⁾, wound infections had occurred in 12% of patients who were not administered rhBMP-2 during cervical fusion therapy. In contrast, no wound infection occurred in patients administered with rhBMP-2. Miller⁴⁴⁾ reported that 4 out of 14 rabbits surgically infected with Staphylococcus aureus died early but no early deaths occurred for 12 rabbits grafted with rhBMP-2. In addition, Chen⁴⁵⁾ reported that BMPs stimulated the healing of a critical defect in rat femurs in the presence of both chronic and acute infections.

In this study, none of the rabbits grafted with rhBMP-2/heparin-blotted anorganic bovine bone showed any signs of infection. Based on the results of other studies and this study, we suppose that rhBMP-2 enhances the immune responses in local areas, but this postulation must be considered in future studies. In the histologic analysis of the control group, some cells underwent morphological changes only in the vicinity of grafted material, which suggested osteoblastic differentiation. However, this change did not take place inside the grafted material itself (Fig. 6D). This means that 5 μg/ml rhBMP-2 alone (no heparin) cannot induce mesenchymal cell migration or other changes (vessel infiltration) in the grafted material in rabbits. Zhao²⁴⁾ suggested that the heparin protected rhBMP-2 from the Noggin produced from cells through a negative feedback mechanism or that the heparin protected the rhBMP-2 from proteases, and thus the rhBMP-2 level in culture media was sustained when heparin was added to culture media. T he rhBMP-2 unprotected by heparin in the grafted material might have been inactivated or degraded early by Noggin or nonspecific serum-proteases before progenitor cells or endothelial cells reached the inner area of grafted material. Therefore, we thought that sustained level of BMP-2 in the grafted material by heparin could allow progenitor cell migration and/or vessel infiltration into the grafted material. In this study, all experimental groups grafted with rhBMP-2/heparin-blotted anorganic bovine bone showed internal bone formation. We thought that this phenomenon was accomplished by combined effect of rhBMP-2 and heparin. That is to say heparin assists rhBMP-2 mediated osteoinduction in grafted material .

Generally speaking, three essential components of osteoinduction are growth factors for osteoinduction, progenitor cells for osteogenesis^7,8), and blood supply^7,9). In this study, although the morphology of some forming internal bone in the grafted material was a form of simple spicules in part, most internal bone showed ring shapes with a center that might be occupied by newly forming vessels and connective tissue in grafted materials. We suggest that these spicular forms of internal bone in grafted material had resulted from thin tissue sectioning procedure. The center of forming bone was filled with numerous RBC and WBC (Fig. 7M). This means that there were abundant blood vessels in the center of forming bone. rhBMP-2 has not only osteoinductive effects but are also known to have a capacity for inducing blood vessels formation^27,28). Blood vessels formation is the sine qua non of progenitor cell migration which is an important factor for osteoinduction⁴⁶⁾. In this study, histologic evidences showed that these three essential factors were accomplished by combined effect of rhBMP-2 and heparin (Fig. 6).

The results of the histologic study showed that the total amount of internally formed new bone was increased with rising heparin concentration, but the absolute increased amount of internal bone was not proportional to increased heparin concentration (Fig. 7). Also, the effect of heparin decreased as time passed. The important factors that should be considered are that the total amount of newly formed bone was not statistically significant in the third group (the highest heparin concentration: 25 μg/ml) regardless of duration (Fig. 7), and that no statistical differences existed at 6 weeks across all groups (Fig. 7). These results indicate that the final amount of recovering bone in calvarial defects is nearly about the same regardless concentrations of heparin and durations. In summary, heparin enhanced the osteoinductive effects of BMP-2 by protecting rhBMP-2 degradation. However, this enhancing effect was not in proportion to the heparin concentration, and heparin mainly influenced osteoinductive effect of rhBMP-2 in the initial stage of bone formation. The most important factor in grafting procedures is stabilization obtained from initial bone responses. Therefore, the present study suggests that the use of heparin with rhBMP-2 could offer cheaper, safer, and improved clinical results in grafting procedures.