Materials Science and Engineering C 44 (2014) 326–335
Contents lists available at ScienceDirect
Materials Science and Engineering C journal homepage: www.elsevier.com/locate/msec
Enhanced healing of rabbit segmental radius defects with surface-coated calcium phosphate cement/bone morphogenetic protein-2 scaffolds Yi Wu a, Juan Hou a, ManLi Yin a, Jing Wang a,⁎, ChangSheng Liu a,b,c,⁎ a b c
Engineering Research Center for Biomedical Materials of Ministry of Education, East China University of Science and Technology, Shanghai 200237, PR China The State Key Laboratory of Bioreactor Engineering, East China University of Science and Technology, Shanghai 200237, PR China Key Laboratory for Ultrafine Materials of Ministry of Education, East China University of Science and Technology, Shanghai 200237, PR China
a r t i c l e
i n f o
Article history: Received 15 April 2014 Received in revised form 23 June 2014 Accepted 2 August 2014 Available online 10 August 2014 Keywords: Calcium phosphate cement Bone regeneration Cellulose Bone morphogenetic protein-2
a b s t r a c t Large osseous defects remain a difficult clinical problem in orthopedic surgery owing to the limited effective therapeutic options, and bone morphogenetic protein-2 (BMP-2) is useful for its potent osteoinductive properties in bone regeneration. Here we build a strategy to achieve prolonged duration time and help inducting new bone formation by using water-soluble polymers as a protective film. In this study, calcium phosphate cement (CPC) scaffolds were prepared as the matrix and combined with sodium carboxymethyl cellulose (CMC-Na), hydroxypropylmethyl cellulose (HPMC), and polyvinyl alcohol (PVA) respectively to protect from the digestion of rhBMP-2. After being implanted in the mouse thigh muscles, the surface-modified composite scaffolds evidently induced ectopic bone formation. In addition, we further evaluated the in vivo effects of surface-modified scaffolds in a rabbit radius critical defect by radiography, three dimensional micro-computed tomographic (μCT) imaging, synchrotron radiation-based micro-computed tomographic (SRμCT) imaging, histological analysis, and biomechanical measurement. The HPMC-modified CPC scaffold was regarded as the best combination for segmental bone regeneration in rabbit radius. © 2014 Elsevier B.V. All rights reserved.
1. Introduction Bone tissue engineering has become an attractive and potential approach for repairing large defects caused by tumors, trauma, surgical resection and congenital malformation [1]. Since the autografting bone substitute has limitations and disadvantages such as insufficient donors and donor site morbidity [2,3] and the potential risk of transmitting infectious for allogeneic bone exists [4], more attentions have been paid on the prosthetic scaffolds for implantation [5,6]. For bone tissue engineering, these scaffolds must require desirable mechanical properties, osteoconductivity, biocompatibility, biodegradability and noncytotoxicity [7,8]. Since its first discovery in the 1980s, calcium phosphate cements (CPCs) have attracted great interest as bone substitutes owing to their excellent biocompatibility and osteoconductivity [9–11]. Moreover, the final composition of hardened CPCs is more similar to the bonelike apatite in vivo than sintered hydroxyapatite ceramics [12], thus exhibiting potential applications in bone regeneration, such as bone augmentation and spinal vertebroplasty [13–16]. However, despite the above advantages, the lack of osteoinductivity is one of the critical drawbacks of CPCs, which might result in bone nonunion, especially in
⁎ Corresponding authors. E-mail addresses: [email protected] (J. Wang), [email protected] (C. Liu).
http://dx.doi.org/10.1016/j.msec.2014.08.020 0928-4931/© 2014 Elsevier B.V. All rights reserved.
the reconstruction of large defect. Incorporation with bioactive growth factors might be an efficient strategy endowing the osteoinductive activity. Bone morphogenetic protein-2 (BMP-2), which is a member of transforming growth factor-β (TGF-β) superfamily, can affect cell differentiation and plays important roles in early embryonal development in adult and organisms [17–19]. As the most representative bone growth factor, recombinant human BMP-2 (rhBMP-2) has been approved by the US Food and Drug Administration (FDA) and the European Medicines Agency (EMEA) for the treatment of spinal fusion and tibial fractures combined with a matrix like collagen or decalcified bone [20–22]. However, rhBMP-2 has a short half-life and always appears bolus release at a very early period which may impair the bone regeneration. To overcome these problems, a number of carriers, including natural and synthetic biomaterials, have been explored in the past decade to maintain BMP retention and provide sustained delivery at the defect site [23–26]. Inorganic calcium phosphate-based scaffolds such as hydroxyapatite (HAP) and tricalcium phosphate (TCP) have already been investigated as carriers of BMP individually. Previous studies demonstrated that the incorporation of BMP into these ceramics greatly accelerates the bone formation [27–30]. CPC can either act as valid carriers of rhBMP-2. When combined with CPCs, the retention of rhBMP-2 can be prolonged to avoid enzymolysis of protease [31,32]. In addition, for the sake of clinic application and prevention of denaturation when
Y. Wu et al. / Materials Science and Engineering C 44 (2014) 326–335
327
Fig. 1. The surgical procedures.
exposed to various organic solvents and undesirable nonphysiologic conditions, BMP-2 is always loaded by physical adsorption instead of conjugation or encapsulation. A series of problems arise due to the weak integration between rhBMP-2 and CPC matrix. The undesirable rapid diffusion and higher initial burst release would cause potential pathological risk. Herein, we present a facile and convenient method to fabricate CPCbased composite scaffolds containing rhBMP-2. Three kinds of hydrophilic excipients, hydroxypropylmethyl cellulose (HPMC), sodium carboxymethyl cellulose (CMC-Na) and polyvinyl alcohol (PVA) were selected for surface-coating. Optimized screening was conducted through the investigation of ectopic bone formation in mouse hindlimb pocket. Finally, a critical-sized defect model was created in a rabbit radius to evaluate the healing capacity of large bone defect. X-ray examination and micro-computed tomographic and synchrotron radiationbased micro-computed tomographic observations were performed. Histological analysis and biomechanics evaluation were also carried out. We hypothesized that via surface-coating, the rhBMP-2-impregnated CPC composite scaffold had enhanced osteogenic efficacy for segmental bone repairing.
2. Materials and methods 2.1. Materials
(DCPA) in an equivalent molar ratio, using preparation methods according to the previous literature [33]. rhBMP-2 was denoted from Shanghai Rebone Biomaterials Co., Ltd. (Shanghai, China). Hydroxypropylmethyl cellulose (HPMC) was purchased from Shandong Liaocheng A Hua Pharmaceutical Co., Ltd. (Shandong, China). Sodium carboxymethyl cellulose (CMC-Na) was purchased from Shanghai Changwei Pharmaceutical Tech. Co., Ltd. (Shanghai, China). Polyvinyl alcohol (PVA) was obtained from Acros organics (USA).
2.2. Fabrication of porous CPC scaffold CPC scaffolds were prepared by a particulate-leaching method [34]. Briefly, the TECP and DCPA powders were mixed with aqueous disodium hydrogen phosphate (4 wt.%) using a spatula at a powder/ liquid mass ratio of 3:1 to form a paste. Sodium chloride particles sieved with diameters of 400–500 μm as porogen were added into the CPC paste. The mixture of CPC/NaCl was placed in stainless steel molds and the mixture was molded under a pressure of 2 MPa. Two different sizes of cylinder scaffolds (Φ3 × 4 mm, Φ4 × 15 mm) were preformed, one is used for ectopic bone formation and the other is applied in segmental radius repairing. After incubating in a constant temperature over at 37 °C and 100% relative humidity, the samples were then immersed in deionized water to leach out the porogen. Finally, they were vacuum-dried to obtain sponge-like scaffolds.
The CPC powder prepared in our laboratory was composed of tetracalcium phosphate (TECP) and dicalcium phosphate anhydrous
Fig. 2. The maximum compressive strength of CPC scaffolds (*p b 0.05, significant difference in the maximum strength as compared to the control group).
Fig. 3. Wet weights and ash content of the ectopic bone induced by rhBMP-2 in vivo for 4 weeks. (*p b 0.05, significant differences were found between HPMC-modified scaffolds and original ones; **p b 0.05, significant differences were found between PVA-modified scaffolds and the other three groups.).
328
Y. Wu et al. / Materials Science and Engineering C 44 (2014) 326–335
Fig. 4. Histological evaluation of the harvested implants at 4 weeks: (a: CPC + rhBMP-2; b: CMC-Na + CPC/rhBMP-2; c: HPMC + CPC/rhBMP-2; d: PVA + CPC/rhBMP-2; B: new bone; F: fibrous tissue).
2.3. Fabrication of rhBMP-2-loaded and surface modified rhBMP-2-loaded CPC scaffolds
300 μL CMC-Na, HPMC and PVA solutions were dropwise added respectively, followed by freeze-drying for 48 h.
Processing was conducted in sterile conditions. The CPC scaffolds were sterilized by ethylene oxide. The CMC-Na solution (1%, w/v), HPMC solution (2%, w/v), and PVA solution (2%, w/v) were sterilized by filtration through a 0.22 μm membrane. Then the rhBMP-2 (0.5 mg/mL) solution 20 μL or 100 μL was dripped onto two kinds of porous scaffolds. After lyophilization, the rhBMP-2 loaded scaffolds were achieved [25]. As for the surface-modified rhBMP-2-loaded scaffold, similar preceding steps were conducted. Before lyophilization,
2.4. Mechanical testing The compressive strengths of both CPC and surface-modified CPC scaffolds (CPC, CMC-Na + CPC/BMP, HPMC + CPC/BMP, PVA + CPC/ BMP) were measured at a loading rate of 1 mm/min using a mechanical testing machine (HY-0230, Shanghai, China) at ambient temperature and humidity. Three replicates were carried out for each group, and the results were expressed as mean ± SD.
Fig. 5. X-ray images of the segmental bone at 2, 4, 8, and 12 weeks post-operation: Group A (control); group B (CPC/rhBMP-2 implants); group C (HPMC + CPC/rhBMP-2 implants); and group D (CMC-Na + CPC/rhBMP-2 implants).
Y. Wu et al. / Materials Science and Engineering C 44 (2014) 326–335
329
Fig. 6. Micro-CT section view images of the segmental bone at 2, 4, 8, and 12 weeks post-operation; (1) 2 w, (2) 4 w, (3) 8 w, and (4) 12 w.
2.5. Ectopic bone formation All surgical procedures conducted in the present study were performed and approved by the Institutional Animal Care and Use Committee. There were four groups of CPC-based scaffolds designed for ectopic bone formation: (1) CPC scaffolds loaded with 10 μg rhBMP-2, denoted as CPC/rhBMP-2 scaffolds; (2) CPC scaffolds loaded with 10 μg rhBMP-2 then coated with 1% (w/v) CMC-Na, denoted as CMC-Na + CPC/rhBMP2; (3) CPC scaffolds loaded with 10 μg rhBMP-2 then coated with 2% (w/ v) HPMC, denoted as HPMC + CPC/rhBMP-2; and (4) CPC scaffolds loaded with 10 μg rhBMP-2 then coated with 2% (w/v) PVA, denoted as PVA + CPC/rhBMP-2. Animal surgeries were carried out under aseptic conditions. Sixteen male KM mice weighing about 25 g were used for the study and divided into 4 groups randomly. Before operation, the mice were weighed and anesthetized by 3% pentobarbital sodium (30 mg kg−1) intramuscularly. Scaffolds were implanted into the right thigh muscle pouches of those mice. The incision sites were then closed around the implant with resorbable continuous suture. All mice were maintained in plastic cages with an ambient temperature of 21 °C, and were given water and standard pellets ad libitum. Animals were sacrificed after 4 weeks to estimate the ectopic bone formation. Three specimens of each group were collected. After stripping off the remnant CPC implantation, the weight of the bony tissue was measured, and then incinerated at 600 °C in a muffle furnace for 6 h to obtain the ash contents of the ectopic bone. The other bony specimens with surrounding tissues were retrieved and assigned to histological analysis. After being fixed in 4% neutral buffered formalin for 48 h, the ectopic bone was decalcified in 12.5% EDTA, dehydrated in a graded series of alcohol, and embedded in paraffin. Serial paraffin sections were stained using hematoxylin/eosin (HE) and Masson's
trichrome for histological assessment. Specimens were observed under a light microscope (Nikon TE2000-U, Japan).
2.6. Segmental defect repair 2.6.1. Scaffold preparation and animal surgery The fabrication of scaffolds was carried out under sterile conditions. In this study, CPC scaffolds were prepared as column shapes with the diameter of 4 mm and length of 15 mm. Ninety-two male New Zealand rabbits with an average weight of 3 kg were randomly divided into four groups: (A) Defect without any materials implanted, as the self-repairing negative control group; (B) CPC scaffolds loaded with 50 μg rhBMP-2, as the CPC/rhBMP-2 scaffold group; (C) CPC scaffolds loaded with 50 μg rhBMP-2 then coated with 2% (w/v) HPMC, as the HPMC + CPC/rhBMP-2 scaffold group; and (D) CPC scaffolds loaded with 50 μg rhBMP-2 then coated with 1% (w/v) CMC-Na, as the CMCNa + CPC/rhBMP-2 scaffold group. Another three intact rabbits served as the normal control group. Animal surgery was carried out under sterile conditions. The rabbits were anesthetized by intramuscular injection of 3% pentobarbital sodium (30 mg kg− 1) and the right forelimb was shaved. A longitudinal skin and musculature incision measuring approximately 30 mm was made at the anteromedial aspect. Using a hard drill, a unilateral segment of the periosteum and radius with critical-sized length of 15 mm was cut off in the middle of the radius. Subsequently, the defect site was rinsed with physiological saline and filled with the scaffold prepared above. Finally the underlying musculature and skin were sutured, and the wound was covered with sterile gauze. The total operation procedure was recorded in Fig. 1. After the surgery, the rabbits were individually caged and normally fed.
330
Y. Wu et al. / Materials Science and Engineering C 44 (2014) 326–335
were obtained by the deduction of the background via Image Pro-Plus software. 2.6.5. Histological analysis The bony specimens with surrounding tissues of each group were harvested and assigned to histological analysis. All of the harvested specimens were rinsed with 0.9% physiological saline. Serial sections were stained using hematoxylin/eosin (HE) and Masson's trichrome after fixation and decalcification and observed under a light microscope (Nikon TE2000-U, Japan). 2.6.6. Biomechanical test After 12 weeks, the soft tissues of specimens were carefully dissected from the limbs. In an attempt to estimate the mechanical stability of the regenerated bones, the defective radius and adjoining ulna were subjected to the three-point bending test by using a mechanical testing machine (HY-0230, Shanghai, China) at ambient temperature and humidity. Briefly, the specimens were positioned on two supports spaced 20 mm apart, and the bending load was applied at the midpoint of the defect at a constant displacement rate of 5 mm min−1 until the breakage. During the bending measurement, the data generated were automatically recorded and stored. (n = 3). 2.7. Statistical analysis All quantitative data were analyzed with Origin 8.0 (OriginLab Corporation, USA) and statistical comparisons were worked out using analysis of variance (ANOVA). Statistical significance of comparison between two groups was attained at p b 0.05. 3. Results and discussion 3.1. Mechanical strength Fig. 7. Quantitative data of orthotopic bone formation with different implants: (a): The bone mineral content of the four groups at the different stages of the bone reconstruction process; (b): the BMC and BMD of the four groups at 12 weeks. (*p b 0.05, significant differences were found).
The rabbits were sacrificed successively at scheduled intervals of 2, 4, 8 and 12 weeks. Each respective series consisted of five animals, and the surplus 3 animals were disposed for the mechanical test at 12 weeks. Another three intact animals were set as controls. 2.6.2. Radiographic examination The obtained bone specimens at 2, 4, 8 and 12 weeks were examined by X-ray machine (Faxitron MX-20, USA) to evaluate the bone defect repairing process. 2.6.3. Micro-computed tomographic evaluation The defective radius and adjoining ulna were harvested and scanned with a micro-computed tomographic (μCT) imaging system (GE Explore Locus SP micro-CT, USA). Three-dimensional μCT images were reconstructed to evaluate the repairing process by using the Microview software. Furthermore, the bone mineral content (BMC) and bone mineral density (BMD) of the bone defects were quantitatively calculated using ABA analysis software. 2.6.4. Synchrotron radiation-based micro-computed tomographic evaluation Synchrotron radiation-based micro-computed tomographic (SRμCT) measurements were performed at beamline BL13W of Shanghai Synchrotron Radiation Facility (SSRF, Shanghai, China) using a monochromatic beam with an energy of 30 keV and a sample-to-detector distance of 1.5 m. In the current study, images of materials remaining
Three specimens from each group were used for the compressive strength test to evaluate the maximum strength compared with unmodified CPC scaffold (control). As the matrix, CPC, the biocompatible and osteoconductive biomaterial, had good mechanical property and was used as both the framework of the implantation and the carrier of bioactive growth factor. The maximum strength of compression in the control group was 1.59 ± 0.16 MPa; this is reasonable that the higher porosity resulted in the lower mechanical strength. By comparison, each surface-modification group displayed an improved compressive strength owing to the surface-doping of polymeric film. As displayed in Fig. 2, the compressive strengths of modified specimens of CMCNa + CPC/BMP, HPMC + CPC/BMP, and PVA + CPC/BMP were 2.16 ± 0.10 N, 2.50 ± 0.14 N, and 1.96 ± 0.08 N, respectively. Particularly, significant mechanical enhancements were obtained in both CMC-Na-modified and HPMC-modified groups. Hence, ameliorate mechanical properties could be achieved via those surface-coating method whereas the PVA-modified group didn't. 3.2. Ectopic bone formation From the data in Fig. 3, all rhBMP-2 loaded groups displayed a growing tendency of wet and ash weights of the bone after 4 weeks. The high weight of the wet bone came from the whole weight of the new bone, blood as well as tissue fluid in the marrow cavity. For the CMC-Na + CPC/rhBMP-2 and the HPMC + CPC/rhBMP-2 groups, new bone formation induced by rhBMP-2 raised at 54.0 ± 6.8 mg and 59.4 ± 6.6 mg respectively, which were higher than the CPC/rhBMP-2 group (51.6 ± 8.2 mg). However, the wet weight of the PVA modified scaffolds was much lower as 36.6 ± 5.9 mg. After being incinerated at 600 °C for 6 h, only inorganic ash contents remained, indicating the mineralization of the ectopic bone. Note that the HPMC + CPC/rhBMP-2 group possessed the highest ash weight of about 5.3 ± 0.6 mg, exceeding the
Y. Wu et al. / Materials Science and Engineering C 44 (2014) 326–335
331
Fig. 8. Virtual slices of the segmental bone of groups A and C at 12 weeks post-operation.
CPC/rhBMP-2 group (5.2 ± 0.7 mg), CMC-NA + CPC/rhBMP-2 group (4.8 ± 0.8 mg), and PVA + CPC/rhBMP-2 group (4.1 ± 0.5 mg), indicating the better mineral content of the HPMC-modified group. Since rhBMP-2 located on the surface of the CPC scaffold is apt to escaping, rapidly releasing and depleting usually weaken its osteoinductivity. Formation of polymeric membrane through surface-coating tactics can protect and retard the diffusion of rhBMP-2. Consequently, increased bone generation at ectopic site indicates the promotion of osteoinductive activity of rhBMP-2. Histological images with more detailed information for the rhBMP2-loaded CPC and surface-modified CPC groups were displayed in Fig. 4. Each group displayed the generation of the new bony tissue in the ectopic site, closely surrounding the materials. In Fig. 4(a,b,c) few fibrous tissue formation can be found without obvious adverse inflammation or immunoreaction around implants. Contrastively, plenty of inflammatory cells were observed in the PVA-modified group encircling the fibrous tissue as well as some necrosis and hemorrhagic spots (Fig. 4d). After implantation for 4 weeks, there were plenty of the medulla ossium rubrum and bone trabeculae that appeared. Besides, the residual CPC scaffolds were still visible, and appeared partially degraded. Furthermore, numerous osseous lacuna appeared and the bony tissue was stained with dark blue and red in Masson's trichrome image, which represented the maturity of the bone. However, the CPC/PVA/ rhBMP-2 group exhibited insufficient bone formation, and large amount of the medulla ossium flava existed instead of the bone marrow, as well as the fibrous granulation tissue.
3.3. Bone regeneration in rabbit segmental radius defect Due to the incompatibility of the PVA-modified CPC group according to the ectopic bone formation, both the CMC-Na and HPMC-modified CPC groups, as well as the unmodified CPC/rhBMP-2 group were
selected as carriers of rhBMP-2 for segmental bone regeneration, and the blank sample without any materials implanting was used as control. 3.3.1. Radiographic assessment Qualitative assessment of X-ray radiographs demonstrated the new bone formation during the repairing process, which can be seen in Fig. 5. Group A was set as the negative control group, which represented the self-repairing capability of critical bone defect without any exogenous aids. Little radiopacity can be observed in the defects at 2 and 4 weeks after operation, and a small amount of new bone formations and callus can be found to bridge the distal ends of the defects at 8 weeks. It seemed that both ends of the segmental bone defects were partially joined at 12 weeks, but cannot achieve complete healing. Group B had a much better performance over group A. The implanted material completely filled the defect area and observable callus formation was seen in the distal ends of the defects at 2 weeks. After 4 weeks, part of the scaffold was covered by the new bone. At 8 and 12 weeks, bridging callus formations were observed and consolidated thereafter. The mutual integration between the implant and new bone generated, then the shadow density of the defects increased. The presence of rhBMP-2 was proved to accelerate the resorption of matrix and stimulate accelerated osteoclast differentiation. When combined with CPCs, the retention of rhBMP-2 can be prolonged to avoid enzymolysis of protease. Similar features were observed in groups C and D. The neogenesis bone augmented from the 4th week, and integrated interface between the material and new bone formation can be observed after 12 weeks, especially in group C (red frame). The surface-coating of hydrophilic matrix was aimed to prevent the undesirable rapid diffusion and higher initial burst release of rhBMP-2 physically adsorbed on the CPCs. The surface-coating gave the rhBMP-2-impregnated CPC composite scaffold enhanced osteogenic efficacy for segmental bone repairing and reduced potential pathological risk.
332
Y. Wu et al. / Materials Science and Engineering C 44 (2014) 326–335
Fig. 9. Histological evaluation of orthotopic formed bone sections (HE staining, 100×): Groups A, B, C and D at 2, 4 and 8 weeks. (A: self-repairing without any implants; B: CPC/rhBMP-2 scaffold; C: HPMC + CPC/rhBMP-2 scaffold; D: CMC-Na + CPC/rhBMP-2 scaffold).
3.3.2. Micro-CT analysis The repairing process of the long segmental defected bone in each group with different times was evaluated through 3-dimensional (3D) reconstruction images (Fig. 6). Being consistent with the results of Xray assessment, group A, the control group showed minimal bone binding and low repairing rate without massive bone formation until 8 weeks. The bone defects did not bridge completely even after 12 weeks, indicating the nonunion of the bone. This is not surprising that the 1.5 cm segmental defect has surpassed the self-repairing capability of rabbit, thus led to nonunion or delay of union. In the other three groups, owing to the combinational effect of both the conduction of CPC scaffold and induction of rhBMP-2, more effective new bone formation was present compared to group A throughout the process. Bone defect area was filled with scaffolds and new bone formation connected the ends of the defects at 2 weeks. After 4 weeks, part of the scaffold was covered with the new bone. Remarkable bridging callus formations and excellent repairing capacity of the bone were detected as early as 8 weeks. The scaffolds partially degraded and the mutual integration between the implant and new bone generated. The ingrowth of neogenesis bone increased after 12 weeks, displacing the material gradually. In addition, as it can be seen from the sectional view images,
groups C and D achieved the augmentation of the new bone, suggesting the better repair of bone defects. 3.3.3. BMC and BMD The regenerated bone mineral content (BMC) and bone mineral density (BMD) of defect area were measured by the bone analysis function of the ABA analysis software to quantitatively analyze the repair more precisely (Fig. 7). The BMC indicates the total bone content in the selected region of interest (ROI) and BMD denotes the ratio of BMC to the volume of the selected ROI. Fig. 7(a) compares the BMC of four groups at 2, 4, 8 and 12 weeks. The relatively small amount of BMC in group A (18.46 ± 2.54 mg, 21.04 ± 2.53 mg, 40.97 ± 2.71 mg, 59.43 ± 5.61 mg) showed unsatisfied new bone formation, demonstrating a slow repairing rate without the induction of growth factor. Group B (71.11 ± 4.20 mg, 84.37 ± 2.42 mg, 100.56 ± 3.06 mg, 128.89 ± 4.61 mg) and group D (69.53 ± 2.40 mg, 84.38 ± 4.16 mg, 101.96 ± 3.28 mg, 134.93 ± 5.24 mg) had similar BMC at 2, 4, 8, and 12 weeks after surgery while the new bone volume of group C (77.56 ± 4.05 mg, 87.58 ± 2.59 mg, 104.10 ± 4.78 mg, 142.76 ± 4.56 mg) was slightly higher than the other groups, especially at 12 weeks. Significant differences of new bone mass were
Y. Wu et al. / Materials Science and Engineering C 44 (2014) 326–335
333
Fig. 10. HE and Masson's trichrome staining of orthotopic formed bone sections at 12 w (HE staining, 100×): Groups A, B, C and D. (A: self-repairing without any implants; B: CPC/rhBMP-2 scaffold; C: HPMC + CPC/rhBMP-2 scaffold; D: CMC-Na + CPC/rhBMP-2 scaffold).
found for all groups at 12 weeks, among which groups C and D had a much higher bone mass than groups A and B at each point, confirming the results of 3D reconstruction images in Fig. 6. Moreover, it is clear that group C is superior to both groups B and D during the 12 weeks of the healing process. Furthermore, the resulting BMD was similar to that of the BMC with the lowest in group A (83.95 ± 5.46 mg) and highest in group C (279.42 ± 3.27 mg) due to little individual differences of the radius area of the rabbits. 3.3.4. SRμCT analysis Further subtle observation was obtained via synchrotron radiationbased μCT tomography. SRμCT measures and images have a high spatial resolution and contrast to give an accurate review of bone defect repair. Fig. 8 compared the SRμCT images of the regenerated radial bone defects of groups A and C after 12 weeks. The gray line which was dense and thick represented the cortical bone while the gray wire mesh which was much thinner represented the cancellous bone trabecula. And the implanted materials showed as disorderly white part in the middle. Sparse distribution of the bone tissue and low bone density regardless of a few new bone formation were found in group A. Only a very small
amount of the bones was ingrown in some parts of defects, therefore, callus generated, which represented the incomplete repair of bone defects and the poor effect of self-repairing. As can be seen, the thick compacted cortical bone, as well as the reticular cancellous bone trabecula could be observed in group C. It is noticeable that composite scaffold in group C was partially degradable, along with a lot of the new bone ingrowth becoming mature and displacing the scaffold matrix. The relatively higher bone density and improved bone regeneration demonstrated the expedited and better repairing effect in group C than the self-repairing group A. It is worth mentioning that the promoted degradation of CPC scaffold occurred under the aids of rhBMP-2, and obvious disintegration of CPC was achieved in C-12 w, making a room for new bone ingrowth. In a series of experiments, the presence of rhBMP-2 was proved to accelerate the resorption of CPCs when compared to the blank sample. That was believed to be caused by the ability of rhBMP-2 to stimulate accelerated osteoclast differentiation [35]. However, the results was controversial because this accelerated cement degradation was not found in other works that also use rhBMP-2 in combination with a cement carrier, which deposited the rhBMP-2 on the surface of CPC to investigate healing in critical supra-alveolar
334
Y. Wu et al. / Materials Science and Engineering C 44 (2014) 326–335
3.3.6. Biomechanical property Three specimens from each group were used for the three-point bending test to evaluate the flexural strength compared with the intact normal limb bones at 12 weeks. The maximum load values for fracturing in the normal control group were 248.57 ± 5.88 N, and the regenerational bones of groups A, B, C and D were 135.92 ± 7.05 N, 184.43 ± 13.62 N, 214.91 ± 11.64 N, and 189.06 ± 12.23 N, respectively. In these groups, the maximum load values in group A was extraordinarily low comparatively, and no statistical differences were observed between groups B and D, whereas the HPMC-modified group showed significantly higher flexural strength than the other operated groups at 12 weeks, indicating the better osteogenesis capability. The result was consistent with iconographical and histological observations that HPMC-modified had the optimal healing capability of segmental bone defect. We think this is mainly attributed to more efficient retaining delivery of rhBMP-2 (Fig. 11). Fig. 11. Biomechanical properties evaluated by the three-point bending test at 12 weeks: (A: self-repairing without any implants; B: CPC/rhBMP-2 scaffold; C: HPMC + CPC/ rhBMP-2 scaffold; D: CMC-Na + CPC/rhBMP-2 scaffold). The symbol (*) indicates significant difference in the maximum load as compared to the group C (p b 0.05).
periodontal defects in Hound Labrador mongrels and found slow resorption rate which may prevent its wider use [36]. We speculated that it was mainly due to the rapid depletion of the BMP-2 located on the superficial of the scaffold. Herein, after surface-coating, scattering of BMP-2 is dominated, thus the promotion of bone resorption can be observed. 3.3.5. Histological observations Correspondingly, histological analysis was performed to investigate orthotopic bone formation. The photomicrographs of cross section through the bone defect sites of four groups were presented in Fig. 9. At 2 weeks, massive chondrocytes (CC) were present in segmental bone defect area in group A. At 4 weeks, although partial chondrocytes could be seen, others were transformed into bony matrix. Furthermore, the newly formed bone which consisted of the woven bone was partially formed at 8 weeks. However large vacancies that still existed consisting of fibrous and fatty elements evident from the micrographs indicated the incomplete repair of bone defects. For group B, a better effect on bone formation was observed at 2 weeks; the primary bone trabecula surrounding the non-degraded material (Mat) was observed. At 4 weeks, the newly formed bone tissue partially grew into the material area. While the non-degraded material could still be observed at 8 weeks, obvious transition from the primary bone trabecula to mature bones occurred. Group C and group D had much more dense array of trabecular bone at the same time, and plenty of blood cells were observed in meshes at 2 weeks. After 4 weeks, the materials began to degrade along with the new bone formation. The hairchested bone trabecula was generated and developed into meshwork, and then became woven bone. At 8 weeks, plenty of the mature bone appeared, and obvious bone ingrowth can be observed. Especially in group C, abundant osseous tissue was found to interconnect into the lamellar bone, indicating the acceleration of bone regeneration. Besides, the undegraded residual material decreased. Fig. 10 shows the healing capacity of segmental bone defects after 12 weeks. Along with the reconstruction duration, the trabecular bone was tending to mature and intensive lamellar bone was generated. However, a small amount of cartilage cells was still visible at the outer edge of the lamellar bone in Masson's trichrome staining image of group A, which illustrated the imperfect transformation from the new bone to mature bone. As to group B, the occurrence of callus might indicate the incomplement of bone repairing. Groups C and D showed a sparse distribution of material remainders and the blue strip in trichrome staining revealed the formation of the mature bone. In particular, compact lamellar bones were apparently oriented regularly in group C, demonstrating more conductivity to bone regeneration.
4. Conclusion In this present study, we designed surface-coating strategy to fabricate surface-modified CPC scaffolds with hydrophilic matrix and investigate their capability of repairing large segmental bone defect as a carrier of rhBMP-2. Aqueous CMC-Na, HPMC, and PVA were selected for surface-doping. The rhBMP-2 loaded composite scaffolds were certified to increased ectopic ossification owing to the protection of rhBMP-2 by polymeric membrane. Furthermore, CMC-Na and HPMC-modified rhBMP-2 loaded CPC scaffolds were implanted in the critical-sized segmental bone defect of rabbit. Significantly the promotion and augmentation of bone formation were achieved compared to the self-repairing group and unmodified CPC/rhBMP-2 group. Especially for the HPMCmodified CPC group, the highest bone mineral can be obtained and the reunion of the medullary cavity of rabbit radius can be realized according to iconographical and histological analyses. These results indicate that the HPMC modified CPC/rhBMP-2 scaffold may have potential for clinical applications in bone regeneration. Acknowledgments The authors were indebted to the financial support from the National Basic Research Program of China (973 Program, 2012CB933600), the National Natural Science Foundation of China (Nos. 31271011, 313 30028), and the National Science and Technology Support Program (2012BAI17B02). References [1] H. Petite, V. Viateau, W. Bensaid, A. Meunier, C. de Pollak, M. Bourguignon, K. Oudina, L. Sedel, G. Guillemin, Nat. Biotechnol. 18 (2000) 959–963. [2] A. Merolli, L. Nicolais, L. Ambrosio, M. Santin, Biomed. Mater. 5 (2010) 015008. [3] J.-H. Ye, Y.-J. Xu, J. Gao, S.-G. Yan, J. Zhao, Q. Tu, J. Zhang, X.-J. Duan, C.A. Sommer, G. Mostoslavsky, Biomaterials 32 (2011) 5065–5076. [4] P.D. Sawin, V.C. Traynelis, A.H. Menezes, J. Neurosurg. 88 (1998) 255–265. [5] J.R. Porter, T.T. Ruckh, K.C. Popat, Biotechnol. Prog. 25 (2009) 1539–1560. [6] C. Xu, P. Su, X. Chen, Y. Meng, W. Yu, A.P. Xiang, Y. Wang, Biomaterials 32 (2011) 1051–1058. [7] M. Mastrogiacomo, A. Muraglia, V. Komlev, F. Peyrin, F. Rustichelli, A. Crovace, R. Cancedda, Orthod. Craniofacial Res. 8 (2005) 277–284. [8] R. Langer, Adv. Mater. 21 (2009) 3235–3236. [9] W. Brown, Cements Research Progress, 1987. 351–379. [10] M. Julien, I. Khairoun, R.Z. Legeros, S. Delplace, P. Pilet, P. Weiss, G. Daculsi, J.M. Bouler, J. Guicheux, Biomaterials 28 (2007) 956–965. [11] S.V. Dorozhkin, J. Mater. Sci. 43 (2008) 3028–3057. [12] S. Takagi, L. Chow, K. Ishikawa, Biomaterials 19 (1998) 1593–1599. [13] M. Ginebra, M. Espanol, E. Montufar, R. Perez, G. Mestres, Acta Biomater. 6 (2010) 2863–2873. [14] E. Verron, I. Khairoun, J. Guicheux, J.-M. Bouler, Drug Discov. Today 15 (2010) 547–552. [15] M. Hofmann, A. Mohammed, Y. Perrie, U. Gbureck, J. Barralet, Acta Biomater. 5 (2009) 43–49. [16] M. Espanol, R. Perez, E. Montufar, C. Marichal, A. Sacco, M. Ginebra, Acta Biomater. 5 (2009) 2752–2762.
Y. Wu et al. / Materials Science and Engineering C 44 (2014) 326–335 [17] T. Katagiri, A. Yamaguchi, M. Komaki, E. Abe, N. Takahashi, T. Ikeda, V. Rosen, J.M. Wozney, A. Fujisawa-Sehara, T. Suda, J. Cell Biol. 127 (1994) 1755–1766. [18] A. Yamaguchi, T. Katagiri, T. Ikeda, J.M. Wozney, V. Rosen, E.A. Wang, A.J. Kahn, T. Suda, S. Yoshiki, J. Cell Biol. 113 (1991) 681–687. [19] Y.-T. Xiao, L.-X. Xiang, J.-Z. Shao, Biochem. Biophys. Res. Commun. 362 (2007) 550–553. [20] S. Fujibayashi, M. Takemoto, M. Neo, T. Matsushita, T. Kokubo, K. Doi, T. Ito, A. Shimizu, T. Nakamura, Eur. Spine J. 20 (2011) 1486–1495. [21] J. Even, M. Eskander, J. Kang, J. Am. Acad. Orthop. Surg. 20 (2012) 547–552. [22] S. Kokubo, R. Fujimoto, S. Yokota, S. Fukushima, K. Nozaki, K. Takahashi, K. Miyata, Biomaterials 24 (2003) 1643–1651. [23] H. Hosseinkhani, M. Hosseinkhani, A. Khademhosseini, H. Kobayashi, J. Control. Release 117 (2007) 380–386. [24] Z.S. Patel, M. Yamamoto, H. Ueda, Y. Tabata, A.G. Mikos, Acta Biomater. 4 (2008) 1126–1138. [25] J. Jansen, J. Vehof, P. Ruhe, H. Kroeze-Deutman, Y. Kuboki, H. Takita, E. Hedberg, A. Mikos, J. Control. Release 101 (2005) 127–136. [26] H. Shen, X. Hu, F. Yang, J. Bei, S. Wang, Biomaterials 30 (2009) 3150–3157.
335
[27] H.C. Kroese-Deutman, P.Q. Ruhé, P.H. Spauwen, J.A. Jansen, Biomaterials 26 (2005) 1131–1138. [28] P.Q. Ruhé, E.L. Hedberg-Dirk, N.T. Padron, P.H. Spauwen, J.A. Jansen, A.G. Mikos, Tissue Eng. 12 (2006) 789–800. [29] E.R. Luvizuto, S. Tangl, G. Zanoni, T. Okamoto, C.K. Sonoda, R. Gruber, R. Okamoto, Biomaterials 32 (2011) 3855–3861. [30] T. Crouzier, F. Sailhan, P. Becquart, R. Guillot, D. Logeart-Avramoglou, C. Picart, Biomaterials 32 (2011) 7543–7554. [31] M. Ginebra, T. Traykova, J. Planell, J. Control. Release 113 (2006) 102–110. [32] F.C. van de Watering, J.D. Molkenboer-Kuenen, O.C. Boerman, J.J. van den Beucken, J. A. Jansen, J. Control. Release 164 (2012) 283–290. [33] C. Liu, H. Shao, F. Chen, H. Zheng, Biomaterials 27 (2006) 5003–5013. [34] D. Zou, L. Guo, J. Lu, X. Zhang, J. Wei, C. Liu, Z. Zhang, X. Jiang, Tissue Eng. A 18 (2012) 1464–1478. [35] H. Kaneko, T. Arakawa, H. Mano, T. Kaneda, A. Ogasawara, M. Nakagawa, Y. Toyama, Y. Yabe, M. Kumegawa, Y. Hakeda, Bone 27 (2000) 479–486. [36] R.G. Sorensen, U.M. Wikesjö, A. Kinoshita, J.M. Wozney, J. Clin. Periodontol. 31 (2004) 796–804.
Contents lists available at ScienceDirect
Materials Science and Engineering C journal homepage: www.elsevier.com/locate/msec
Enhanced healing of rabbit segmental radius defects with surface-coated calcium phosphate cement/bone morphogenetic protein-2 scaffolds Yi Wu a, Juan Hou a, ManLi Yin a, Jing Wang a,⁎, ChangSheng Liu a,b,c,⁎ a b c
Engineering Research Center for Biomedical Materials of Ministry of Education, East China University of Science and Technology, Shanghai 200237, PR China The State Key Laboratory of Bioreactor Engineering, East China University of Science and Technology, Shanghai 200237, PR China Key Laboratory for Ultrafine Materials of Ministry of Education, East China University of Science and Technology, Shanghai 200237, PR China
a r t i c l e
i n f o
Article history: Received 15 April 2014 Received in revised form 23 June 2014 Accepted 2 August 2014 Available online 10 August 2014 Keywords: Calcium phosphate cement Bone regeneration Cellulose Bone morphogenetic protein-2
a b s t r a c t Large osseous defects remain a difficult clinical problem in orthopedic surgery owing to the limited effective therapeutic options, and bone morphogenetic protein-2 (BMP-2) is useful for its potent osteoinductive properties in bone regeneration. Here we build a strategy to achieve prolonged duration time and help inducting new bone formation by using water-soluble polymers as a protective film. In this study, calcium phosphate cement (CPC) scaffolds were prepared as the matrix and combined with sodium carboxymethyl cellulose (CMC-Na), hydroxypropylmethyl cellulose (HPMC), and polyvinyl alcohol (PVA) respectively to protect from the digestion of rhBMP-2. After being implanted in the mouse thigh muscles, the surface-modified composite scaffolds evidently induced ectopic bone formation. In addition, we further evaluated the in vivo effects of surface-modified scaffolds in a rabbit radius critical defect by radiography, three dimensional micro-computed tomographic (μCT) imaging, synchrotron radiation-based micro-computed tomographic (SRμCT) imaging, histological analysis, and biomechanical measurement. The HPMC-modified CPC scaffold was regarded as the best combination for segmental bone regeneration in rabbit radius. © 2014 Elsevier B.V. All rights reserved.
1. Introduction Bone tissue engineering has become an attractive and potential approach for repairing large defects caused by tumors, trauma, surgical resection and congenital malformation [1]. Since the autografting bone substitute has limitations and disadvantages such as insufficient donors and donor site morbidity [2,3] and the potential risk of transmitting infectious for allogeneic bone exists [4], more attentions have been paid on the prosthetic scaffolds for implantation [5,6]. For bone tissue engineering, these scaffolds must require desirable mechanical properties, osteoconductivity, biocompatibility, biodegradability and noncytotoxicity [7,8]. Since its first discovery in the 1980s, calcium phosphate cements (CPCs) have attracted great interest as bone substitutes owing to their excellent biocompatibility and osteoconductivity [9–11]. Moreover, the final composition of hardened CPCs is more similar to the bonelike apatite in vivo than sintered hydroxyapatite ceramics [12], thus exhibiting potential applications in bone regeneration, such as bone augmentation and spinal vertebroplasty [13–16]. However, despite the above advantages, the lack of osteoinductivity is one of the critical drawbacks of CPCs, which might result in bone nonunion, especially in
⁎ Corresponding authors. E-mail addresses: [email protected] (J. Wang), [email protected] (C. Liu).
http://dx.doi.org/10.1016/j.msec.2014.08.020 0928-4931/© 2014 Elsevier B.V. All rights reserved.
the reconstruction of large defect. Incorporation with bioactive growth factors might be an efficient strategy endowing the osteoinductive activity. Bone morphogenetic protein-2 (BMP-2), which is a member of transforming growth factor-β (TGF-β) superfamily, can affect cell differentiation and plays important roles in early embryonal development in adult and organisms [17–19]. As the most representative bone growth factor, recombinant human BMP-2 (rhBMP-2) has been approved by the US Food and Drug Administration (FDA) and the European Medicines Agency (EMEA) for the treatment of spinal fusion and tibial fractures combined with a matrix like collagen or decalcified bone [20–22]. However, rhBMP-2 has a short half-life and always appears bolus release at a very early period which may impair the bone regeneration. To overcome these problems, a number of carriers, including natural and synthetic biomaterials, have been explored in the past decade to maintain BMP retention and provide sustained delivery at the defect site [23–26]. Inorganic calcium phosphate-based scaffolds such as hydroxyapatite (HAP) and tricalcium phosphate (TCP) have already been investigated as carriers of BMP individually. Previous studies demonstrated that the incorporation of BMP into these ceramics greatly accelerates the bone formation [27–30]. CPC can either act as valid carriers of rhBMP-2. When combined with CPCs, the retention of rhBMP-2 can be prolonged to avoid enzymolysis of protease [31,32]. In addition, for the sake of clinic application and prevention of denaturation when
Y. Wu et al. / Materials Science and Engineering C 44 (2014) 326–335
327
Fig. 1. The surgical procedures.
exposed to various organic solvents and undesirable nonphysiologic conditions, BMP-2 is always loaded by physical adsorption instead of conjugation or encapsulation. A series of problems arise due to the weak integration between rhBMP-2 and CPC matrix. The undesirable rapid diffusion and higher initial burst release would cause potential pathological risk. Herein, we present a facile and convenient method to fabricate CPCbased composite scaffolds containing rhBMP-2. Three kinds of hydrophilic excipients, hydroxypropylmethyl cellulose (HPMC), sodium carboxymethyl cellulose (CMC-Na) and polyvinyl alcohol (PVA) were selected for surface-coating. Optimized screening was conducted through the investigation of ectopic bone formation in mouse hindlimb pocket. Finally, a critical-sized defect model was created in a rabbit radius to evaluate the healing capacity of large bone defect. X-ray examination and micro-computed tomographic and synchrotron radiationbased micro-computed tomographic observations were performed. Histological analysis and biomechanics evaluation were also carried out. We hypothesized that via surface-coating, the rhBMP-2-impregnated CPC composite scaffold had enhanced osteogenic efficacy for segmental bone repairing.
2. Materials and methods 2.1. Materials
(DCPA) in an equivalent molar ratio, using preparation methods according to the previous literature [33]. rhBMP-2 was denoted from Shanghai Rebone Biomaterials Co., Ltd. (Shanghai, China). Hydroxypropylmethyl cellulose (HPMC) was purchased from Shandong Liaocheng A Hua Pharmaceutical Co., Ltd. (Shandong, China). Sodium carboxymethyl cellulose (CMC-Na) was purchased from Shanghai Changwei Pharmaceutical Tech. Co., Ltd. (Shanghai, China). Polyvinyl alcohol (PVA) was obtained from Acros organics (USA).
2.2. Fabrication of porous CPC scaffold CPC scaffolds were prepared by a particulate-leaching method [34]. Briefly, the TECP and DCPA powders were mixed with aqueous disodium hydrogen phosphate (4 wt.%) using a spatula at a powder/ liquid mass ratio of 3:1 to form a paste. Sodium chloride particles sieved with diameters of 400–500 μm as porogen were added into the CPC paste. The mixture of CPC/NaCl was placed in stainless steel molds and the mixture was molded under a pressure of 2 MPa. Two different sizes of cylinder scaffolds (Φ3 × 4 mm, Φ4 × 15 mm) were preformed, one is used for ectopic bone formation and the other is applied in segmental radius repairing. After incubating in a constant temperature over at 37 °C and 100% relative humidity, the samples were then immersed in deionized water to leach out the porogen. Finally, they were vacuum-dried to obtain sponge-like scaffolds.
The CPC powder prepared in our laboratory was composed of tetracalcium phosphate (TECP) and dicalcium phosphate anhydrous
Fig. 2. The maximum compressive strength of CPC scaffolds (*p b 0.05, significant difference in the maximum strength as compared to the control group).
Fig. 3. Wet weights and ash content of the ectopic bone induced by rhBMP-2 in vivo for 4 weeks. (*p b 0.05, significant differences were found between HPMC-modified scaffolds and original ones; **p b 0.05, significant differences were found between PVA-modified scaffolds and the other three groups.).
328
Y. Wu et al. / Materials Science and Engineering C 44 (2014) 326–335
Fig. 4. Histological evaluation of the harvested implants at 4 weeks: (a: CPC + rhBMP-2; b: CMC-Na + CPC/rhBMP-2; c: HPMC + CPC/rhBMP-2; d: PVA + CPC/rhBMP-2; B: new bone; F: fibrous tissue).
2.3. Fabrication of rhBMP-2-loaded and surface modified rhBMP-2-loaded CPC scaffolds
300 μL CMC-Na, HPMC and PVA solutions were dropwise added respectively, followed by freeze-drying for 48 h.
Processing was conducted in sterile conditions. The CPC scaffolds were sterilized by ethylene oxide. The CMC-Na solution (1%, w/v), HPMC solution (2%, w/v), and PVA solution (2%, w/v) were sterilized by filtration through a 0.22 μm membrane. Then the rhBMP-2 (0.5 mg/mL) solution 20 μL or 100 μL was dripped onto two kinds of porous scaffolds. After lyophilization, the rhBMP-2 loaded scaffolds were achieved [25]. As for the surface-modified rhBMP-2-loaded scaffold, similar preceding steps were conducted. Before lyophilization,
2.4. Mechanical testing The compressive strengths of both CPC and surface-modified CPC scaffolds (CPC, CMC-Na + CPC/BMP, HPMC + CPC/BMP, PVA + CPC/ BMP) were measured at a loading rate of 1 mm/min using a mechanical testing machine (HY-0230, Shanghai, China) at ambient temperature and humidity. Three replicates were carried out for each group, and the results were expressed as mean ± SD.
Fig. 5. X-ray images of the segmental bone at 2, 4, 8, and 12 weeks post-operation: Group A (control); group B (CPC/rhBMP-2 implants); group C (HPMC + CPC/rhBMP-2 implants); and group D (CMC-Na + CPC/rhBMP-2 implants).
Y. Wu et al. / Materials Science and Engineering C 44 (2014) 326–335
329
Fig. 6. Micro-CT section view images of the segmental bone at 2, 4, 8, and 12 weeks post-operation; (1) 2 w, (2) 4 w, (3) 8 w, and (4) 12 w.
2.5. Ectopic bone formation All surgical procedures conducted in the present study were performed and approved by the Institutional Animal Care and Use Committee. There were four groups of CPC-based scaffolds designed for ectopic bone formation: (1) CPC scaffolds loaded with 10 μg rhBMP-2, denoted as CPC/rhBMP-2 scaffolds; (2) CPC scaffolds loaded with 10 μg rhBMP-2 then coated with 1% (w/v) CMC-Na, denoted as CMC-Na + CPC/rhBMP2; (3) CPC scaffolds loaded with 10 μg rhBMP-2 then coated with 2% (w/ v) HPMC, denoted as HPMC + CPC/rhBMP-2; and (4) CPC scaffolds loaded with 10 μg rhBMP-2 then coated with 2% (w/v) PVA, denoted as PVA + CPC/rhBMP-2. Animal surgeries were carried out under aseptic conditions. Sixteen male KM mice weighing about 25 g were used for the study and divided into 4 groups randomly. Before operation, the mice were weighed and anesthetized by 3% pentobarbital sodium (30 mg kg−1) intramuscularly. Scaffolds were implanted into the right thigh muscle pouches of those mice. The incision sites were then closed around the implant with resorbable continuous suture. All mice were maintained in plastic cages with an ambient temperature of 21 °C, and were given water and standard pellets ad libitum. Animals were sacrificed after 4 weeks to estimate the ectopic bone formation. Three specimens of each group were collected. After stripping off the remnant CPC implantation, the weight of the bony tissue was measured, and then incinerated at 600 °C in a muffle furnace for 6 h to obtain the ash contents of the ectopic bone. The other bony specimens with surrounding tissues were retrieved and assigned to histological analysis. After being fixed in 4% neutral buffered formalin for 48 h, the ectopic bone was decalcified in 12.5% EDTA, dehydrated in a graded series of alcohol, and embedded in paraffin. Serial paraffin sections were stained using hematoxylin/eosin (HE) and Masson's
trichrome for histological assessment. Specimens were observed under a light microscope (Nikon TE2000-U, Japan).
2.6. Segmental defect repair 2.6.1. Scaffold preparation and animal surgery The fabrication of scaffolds was carried out under sterile conditions. In this study, CPC scaffolds were prepared as column shapes with the diameter of 4 mm and length of 15 mm. Ninety-two male New Zealand rabbits with an average weight of 3 kg were randomly divided into four groups: (A) Defect without any materials implanted, as the self-repairing negative control group; (B) CPC scaffolds loaded with 50 μg rhBMP-2, as the CPC/rhBMP-2 scaffold group; (C) CPC scaffolds loaded with 50 μg rhBMP-2 then coated with 2% (w/v) HPMC, as the HPMC + CPC/rhBMP-2 scaffold group; and (D) CPC scaffolds loaded with 50 μg rhBMP-2 then coated with 1% (w/v) CMC-Na, as the CMCNa + CPC/rhBMP-2 scaffold group. Another three intact rabbits served as the normal control group. Animal surgery was carried out under sterile conditions. The rabbits were anesthetized by intramuscular injection of 3% pentobarbital sodium (30 mg kg− 1) and the right forelimb was shaved. A longitudinal skin and musculature incision measuring approximately 30 mm was made at the anteromedial aspect. Using a hard drill, a unilateral segment of the periosteum and radius with critical-sized length of 15 mm was cut off in the middle of the radius. Subsequently, the defect site was rinsed with physiological saline and filled with the scaffold prepared above. Finally the underlying musculature and skin were sutured, and the wound was covered with sterile gauze. The total operation procedure was recorded in Fig. 1. After the surgery, the rabbits were individually caged and normally fed.
330
Y. Wu et al. / Materials Science and Engineering C 44 (2014) 326–335
were obtained by the deduction of the background via Image Pro-Plus software. 2.6.5. Histological analysis The bony specimens with surrounding tissues of each group were harvested and assigned to histological analysis. All of the harvested specimens were rinsed with 0.9% physiological saline. Serial sections were stained using hematoxylin/eosin (HE) and Masson's trichrome after fixation and decalcification and observed under a light microscope (Nikon TE2000-U, Japan). 2.6.6. Biomechanical test After 12 weeks, the soft tissues of specimens were carefully dissected from the limbs. In an attempt to estimate the mechanical stability of the regenerated bones, the defective radius and adjoining ulna were subjected to the three-point bending test by using a mechanical testing machine (HY-0230, Shanghai, China) at ambient temperature and humidity. Briefly, the specimens were positioned on two supports spaced 20 mm apart, and the bending load was applied at the midpoint of the defect at a constant displacement rate of 5 mm min−1 until the breakage. During the bending measurement, the data generated were automatically recorded and stored. (n = 3). 2.7. Statistical analysis All quantitative data were analyzed with Origin 8.0 (OriginLab Corporation, USA) and statistical comparisons were worked out using analysis of variance (ANOVA). Statistical significance of comparison between two groups was attained at p b 0.05. 3. Results and discussion 3.1. Mechanical strength Fig. 7. Quantitative data of orthotopic bone formation with different implants: (a): The bone mineral content of the four groups at the different stages of the bone reconstruction process; (b): the BMC and BMD of the four groups at 12 weeks. (*p b 0.05, significant differences were found).
The rabbits were sacrificed successively at scheduled intervals of 2, 4, 8 and 12 weeks. Each respective series consisted of five animals, and the surplus 3 animals were disposed for the mechanical test at 12 weeks. Another three intact animals were set as controls. 2.6.2. Radiographic examination The obtained bone specimens at 2, 4, 8 and 12 weeks were examined by X-ray machine (Faxitron MX-20, USA) to evaluate the bone defect repairing process. 2.6.3. Micro-computed tomographic evaluation The defective radius and adjoining ulna were harvested and scanned with a micro-computed tomographic (μCT) imaging system (GE Explore Locus SP micro-CT, USA). Three-dimensional μCT images were reconstructed to evaluate the repairing process by using the Microview software. Furthermore, the bone mineral content (BMC) and bone mineral density (BMD) of the bone defects were quantitatively calculated using ABA analysis software. 2.6.4. Synchrotron radiation-based micro-computed tomographic evaluation Synchrotron radiation-based micro-computed tomographic (SRμCT) measurements were performed at beamline BL13W of Shanghai Synchrotron Radiation Facility (SSRF, Shanghai, China) using a monochromatic beam with an energy of 30 keV and a sample-to-detector distance of 1.5 m. In the current study, images of materials remaining
Three specimens from each group were used for the compressive strength test to evaluate the maximum strength compared with unmodified CPC scaffold (control). As the matrix, CPC, the biocompatible and osteoconductive biomaterial, had good mechanical property and was used as both the framework of the implantation and the carrier of bioactive growth factor. The maximum strength of compression in the control group was 1.59 ± 0.16 MPa; this is reasonable that the higher porosity resulted in the lower mechanical strength. By comparison, each surface-modification group displayed an improved compressive strength owing to the surface-doping of polymeric film. As displayed in Fig. 2, the compressive strengths of modified specimens of CMCNa + CPC/BMP, HPMC + CPC/BMP, and PVA + CPC/BMP were 2.16 ± 0.10 N, 2.50 ± 0.14 N, and 1.96 ± 0.08 N, respectively. Particularly, significant mechanical enhancements were obtained in both CMC-Na-modified and HPMC-modified groups. Hence, ameliorate mechanical properties could be achieved via those surface-coating method whereas the PVA-modified group didn't. 3.2. Ectopic bone formation From the data in Fig. 3, all rhBMP-2 loaded groups displayed a growing tendency of wet and ash weights of the bone after 4 weeks. The high weight of the wet bone came from the whole weight of the new bone, blood as well as tissue fluid in the marrow cavity. For the CMC-Na + CPC/rhBMP-2 and the HPMC + CPC/rhBMP-2 groups, new bone formation induced by rhBMP-2 raised at 54.0 ± 6.8 mg and 59.4 ± 6.6 mg respectively, which were higher than the CPC/rhBMP-2 group (51.6 ± 8.2 mg). However, the wet weight of the PVA modified scaffolds was much lower as 36.6 ± 5.9 mg. After being incinerated at 600 °C for 6 h, only inorganic ash contents remained, indicating the mineralization of the ectopic bone. Note that the HPMC + CPC/rhBMP-2 group possessed the highest ash weight of about 5.3 ± 0.6 mg, exceeding the
Y. Wu et al. / Materials Science and Engineering C 44 (2014) 326–335
331
Fig. 8. Virtual slices of the segmental bone of groups A and C at 12 weeks post-operation.
CPC/rhBMP-2 group (5.2 ± 0.7 mg), CMC-NA + CPC/rhBMP-2 group (4.8 ± 0.8 mg), and PVA + CPC/rhBMP-2 group (4.1 ± 0.5 mg), indicating the better mineral content of the HPMC-modified group. Since rhBMP-2 located on the surface of the CPC scaffold is apt to escaping, rapidly releasing and depleting usually weaken its osteoinductivity. Formation of polymeric membrane through surface-coating tactics can protect and retard the diffusion of rhBMP-2. Consequently, increased bone generation at ectopic site indicates the promotion of osteoinductive activity of rhBMP-2. Histological images with more detailed information for the rhBMP2-loaded CPC and surface-modified CPC groups were displayed in Fig. 4. Each group displayed the generation of the new bony tissue in the ectopic site, closely surrounding the materials. In Fig. 4(a,b,c) few fibrous tissue formation can be found without obvious adverse inflammation or immunoreaction around implants. Contrastively, plenty of inflammatory cells were observed in the PVA-modified group encircling the fibrous tissue as well as some necrosis and hemorrhagic spots (Fig. 4d). After implantation for 4 weeks, there were plenty of the medulla ossium rubrum and bone trabeculae that appeared. Besides, the residual CPC scaffolds were still visible, and appeared partially degraded. Furthermore, numerous osseous lacuna appeared and the bony tissue was stained with dark blue and red in Masson's trichrome image, which represented the maturity of the bone. However, the CPC/PVA/ rhBMP-2 group exhibited insufficient bone formation, and large amount of the medulla ossium flava existed instead of the bone marrow, as well as the fibrous granulation tissue.
3.3. Bone regeneration in rabbit segmental radius defect Due to the incompatibility of the PVA-modified CPC group according to the ectopic bone formation, both the CMC-Na and HPMC-modified CPC groups, as well as the unmodified CPC/rhBMP-2 group were
selected as carriers of rhBMP-2 for segmental bone regeneration, and the blank sample without any materials implanting was used as control. 3.3.1. Radiographic assessment Qualitative assessment of X-ray radiographs demonstrated the new bone formation during the repairing process, which can be seen in Fig. 5. Group A was set as the negative control group, which represented the self-repairing capability of critical bone defect without any exogenous aids. Little radiopacity can be observed in the defects at 2 and 4 weeks after operation, and a small amount of new bone formations and callus can be found to bridge the distal ends of the defects at 8 weeks. It seemed that both ends of the segmental bone defects were partially joined at 12 weeks, but cannot achieve complete healing. Group B had a much better performance over group A. The implanted material completely filled the defect area and observable callus formation was seen in the distal ends of the defects at 2 weeks. After 4 weeks, part of the scaffold was covered by the new bone. At 8 and 12 weeks, bridging callus formations were observed and consolidated thereafter. The mutual integration between the implant and new bone generated, then the shadow density of the defects increased. The presence of rhBMP-2 was proved to accelerate the resorption of matrix and stimulate accelerated osteoclast differentiation. When combined with CPCs, the retention of rhBMP-2 can be prolonged to avoid enzymolysis of protease. Similar features were observed in groups C and D. The neogenesis bone augmented from the 4th week, and integrated interface between the material and new bone formation can be observed after 12 weeks, especially in group C (red frame). The surface-coating of hydrophilic matrix was aimed to prevent the undesirable rapid diffusion and higher initial burst release of rhBMP-2 physically adsorbed on the CPCs. The surface-coating gave the rhBMP-2-impregnated CPC composite scaffold enhanced osteogenic efficacy for segmental bone repairing and reduced potential pathological risk.
332
Y. Wu et al. / Materials Science and Engineering C 44 (2014) 326–335
Fig. 9. Histological evaluation of orthotopic formed bone sections (HE staining, 100×): Groups A, B, C and D at 2, 4 and 8 weeks. (A: self-repairing without any implants; B: CPC/rhBMP-2 scaffold; C: HPMC + CPC/rhBMP-2 scaffold; D: CMC-Na + CPC/rhBMP-2 scaffold).
3.3.2. Micro-CT analysis The repairing process of the long segmental defected bone in each group with different times was evaluated through 3-dimensional (3D) reconstruction images (Fig. 6). Being consistent with the results of Xray assessment, group A, the control group showed minimal bone binding and low repairing rate without massive bone formation until 8 weeks. The bone defects did not bridge completely even after 12 weeks, indicating the nonunion of the bone. This is not surprising that the 1.5 cm segmental defect has surpassed the self-repairing capability of rabbit, thus led to nonunion or delay of union. In the other three groups, owing to the combinational effect of both the conduction of CPC scaffold and induction of rhBMP-2, more effective new bone formation was present compared to group A throughout the process. Bone defect area was filled with scaffolds and new bone formation connected the ends of the defects at 2 weeks. After 4 weeks, part of the scaffold was covered with the new bone. Remarkable bridging callus formations and excellent repairing capacity of the bone were detected as early as 8 weeks. The scaffolds partially degraded and the mutual integration between the implant and new bone generated. The ingrowth of neogenesis bone increased after 12 weeks, displacing the material gradually. In addition, as it can be seen from the sectional view images,
groups C and D achieved the augmentation of the new bone, suggesting the better repair of bone defects. 3.3.3. BMC and BMD The regenerated bone mineral content (BMC) and bone mineral density (BMD) of defect area were measured by the bone analysis function of the ABA analysis software to quantitatively analyze the repair more precisely (Fig. 7). The BMC indicates the total bone content in the selected region of interest (ROI) and BMD denotes the ratio of BMC to the volume of the selected ROI. Fig. 7(a) compares the BMC of four groups at 2, 4, 8 and 12 weeks. The relatively small amount of BMC in group A (18.46 ± 2.54 mg, 21.04 ± 2.53 mg, 40.97 ± 2.71 mg, 59.43 ± 5.61 mg) showed unsatisfied new bone formation, demonstrating a slow repairing rate without the induction of growth factor. Group B (71.11 ± 4.20 mg, 84.37 ± 2.42 mg, 100.56 ± 3.06 mg, 128.89 ± 4.61 mg) and group D (69.53 ± 2.40 mg, 84.38 ± 4.16 mg, 101.96 ± 3.28 mg, 134.93 ± 5.24 mg) had similar BMC at 2, 4, 8, and 12 weeks after surgery while the new bone volume of group C (77.56 ± 4.05 mg, 87.58 ± 2.59 mg, 104.10 ± 4.78 mg, 142.76 ± 4.56 mg) was slightly higher than the other groups, especially at 12 weeks. Significant differences of new bone mass were
Y. Wu et al. / Materials Science and Engineering C 44 (2014) 326–335
333
Fig. 10. HE and Masson's trichrome staining of orthotopic formed bone sections at 12 w (HE staining, 100×): Groups A, B, C and D. (A: self-repairing without any implants; B: CPC/rhBMP-2 scaffold; C: HPMC + CPC/rhBMP-2 scaffold; D: CMC-Na + CPC/rhBMP-2 scaffold).
found for all groups at 12 weeks, among which groups C and D had a much higher bone mass than groups A and B at each point, confirming the results of 3D reconstruction images in Fig. 6. Moreover, it is clear that group C is superior to both groups B and D during the 12 weeks of the healing process. Furthermore, the resulting BMD was similar to that of the BMC with the lowest in group A (83.95 ± 5.46 mg) and highest in group C (279.42 ± 3.27 mg) due to little individual differences of the radius area of the rabbits. 3.3.4. SRμCT analysis Further subtle observation was obtained via synchrotron radiationbased μCT tomography. SRμCT measures and images have a high spatial resolution and contrast to give an accurate review of bone defect repair. Fig. 8 compared the SRμCT images of the regenerated radial bone defects of groups A and C after 12 weeks. The gray line which was dense and thick represented the cortical bone while the gray wire mesh which was much thinner represented the cancellous bone trabecula. And the implanted materials showed as disorderly white part in the middle. Sparse distribution of the bone tissue and low bone density regardless of a few new bone formation were found in group A. Only a very small
amount of the bones was ingrown in some parts of defects, therefore, callus generated, which represented the incomplete repair of bone defects and the poor effect of self-repairing. As can be seen, the thick compacted cortical bone, as well as the reticular cancellous bone trabecula could be observed in group C. It is noticeable that composite scaffold in group C was partially degradable, along with a lot of the new bone ingrowth becoming mature and displacing the scaffold matrix. The relatively higher bone density and improved bone regeneration demonstrated the expedited and better repairing effect in group C than the self-repairing group A. It is worth mentioning that the promoted degradation of CPC scaffold occurred under the aids of rhBMP-2, and obvious disintegration of CPC was achieved in C-12 w, making a room for new bone ingrowth. In a series of experiments, the presence of rhBMP-2 was proved to accelerate the resorption of CPCs when compared to the blank sample. That was believed to be caused by the ability of rhBMP-2 to stimulate accelerated osteoclast differentiation [35]. However, the results was controversial because this accelerated cement degradation was not found in other works that also use rhBMP-2 in combination with a cement carrier, which deposited the rhBMP-2 on the surface of CPC to investigate healing in critical supra-alveolar
334
Y. Wu et al. / Materials Science and Engineering C 44 (2014) 326–335
3.3.6. Biomechanical property Three specimens from each group were used for the three-point bending test to evaluate the flexural strength compared with the intact normal limb bones at 12 weeks. The maximum load values for fracturing in the normal control group were 248.57 ± 5.88 N, and the regenerational bones of groups A, B, C and D were 135.92 ± 7.05 N, 184.43 ± 13.62 N, 214.91 ± 11.64 N, and 189.06 ± 12.23 N, respectively. In these groups, the maximum load values in group A was extraordinarily low comparatively, and no statistical differences were observed between groups B and D, whereas the HPMC-modified group showed significantly higher flexural strength than the other operated groups at 12 weeks, indicating the better osteogenesis capability. The result was consistent with iconographical and histological observations that HPMC-modified had the optimal healing capability of segmental bone defect. We think this is mainly attributed to more efficient retaining delivery of rhBMP-2 (Fig. 11). Fig. 11. Biomechanical properties evaluated by the three-point bending test at 12 weeks: (A: self-repairing without any implants; B: CPC/rhBMP-2 scaffold; C: HPMC + CPC/ rhBMP-2 scaffold; D: CMC-Na + CPC/rhBMP-2 scaffold). The symbol (*) indicates significant difference in the maximum load as compared to the group C (p b 0.05).
periodontal defects in Hound Labrador mongrels and found slow resorption rate which may prevent its wider use [36]. We speculated that it was mainly due to the rapid depletion of the BMP-2 located on the superficial of the scaffold. Herein, after surface-coating, scattering of BMP-2 is dominated, thus the promotion of bone resorption can be observed. 3.3.5. Histological observations Correspondingly, histological analysis was performed to investigate orthotopic bone formation. The photomicrographs of cross section through the bone defect sites of four groups were presented in Fig. 9. At 2 weeks, massive chondrocytes (CC) were present in segmental bone defect area in group A. At 4 weeks, although partial chondrocytes could be seen, others were transformed into bony matrix. Furthermore, the newly formed bone which consisted of the woven bone was partially formed at 8 weeks. However large vacancies that still existed consisting of fibrous and fatty elements evident from the micrographs indicated the incomplete repair of bone defects. For group B, a better effect on bone formation was observed at 2 weeks; the primary bone trabecula surrounding the non-degraded material (Mat) was observed. At 4 weeks, the newly formed bone tissue partially grew into the material area. While the non-degraded material could still be observed at 8 weeks, obvious transition from the primary bone trabecula to mature bones occurred. Group C and group D had much more dense array of trabecular bone at the same time, and plenty of blood cells were observed in meshes at 2 weeks. After 4 weeks, the materials began to degrade along with the new bone formation. The hairchested bone trabecula was generated and developed into meshwork, and then became woven bone. At 8 weeks, plenty of the mature bone appeared, and obvious bone ingrowth can be observed. Especially in group C, abundant osseous tissue was found to interconnect into the lamellar bone, indicating the acceleration of bone regeneration. Besides, the undegraded residual material decreased. Fig. 10 shows the healing capacity of segmental bone defects after 12 weeks. Along with the reconstruction duration, the trabecular bone was tending to mature and intensive lamellar bone was generated. However, a small amount of cartilage cells was still visible at the outer edge of the lamellar bone in Masson's trichrome staining image of group A, which illustrated the imperfect transformation from the new bone to mature bone. As to group B, the occurrence of callus might indicate the incomplement of bone repairing. Groups C and D showed a sparse distribution of material remainders and the blue strip in trichrome staining revealed the formation of the mature bone. In particular, compact lamellar bones were apparently oriented regularly in group C, demonstrating more conductivity to bone regeneration.
4. Conclusion In this present study, we designed surface-coating strategy to fabricate surface-modified CPC scaffolds with hydrophilic matrix and investigate their capability of repairing large segmental bone defect as a carrier of rhBMP-2. Aqueous CMC-Na, HPMC, and PVA were selected for surface-doping. The rhBMP-2 loaded composite scaffolds were certified to increased ectopic ossification owing to the protection of rhBMP-2 by polymeric membrane. Furthermore, CMC-Na and HPMC-modified rhBMP-2 loaded CPC scaffolds were implanted in the critical-sized segmental bone defect of rabbit. Significantly the promotion and augmentation of bone formation were achieved compared to the self-repairing group and unmodified CPC/rhBMP-2 group. Especially for the HPMCmodified CPC group, the highest bone mineral can be obtained and the reunion of the medullary cavity of rabbit radius can be realized according to iconographical and histological analyses. These results indicate that the HPMC modified CPC/rhBMP-2 scaffold may have potential for clinical applications in bone regeneration. Acknowledgments The authors were indebted to the financial support from the National Basic Research Program of China (973 Program, 2012CB933600), the National Natural Science Foundation of China (Nos. 31271011, 313 30028), and the National Science and Technology Support Program (2012BAI17B02). References [1] H. Petite, V. Viateau, W. Bensaid, A. Meunier, C. de Pollak, M. Bourguignon, K. Oudina, L. Sedel, G. Guillemin, Nat. Biotechnol. 18 (2000) 959–963. [2] A. Merolli, L. Nicolais, L. Ambrosio, M. Santin, Biomed. Mater. 5 (2010) 015008. [3] J.-H. Ye, Y.-J. Xu, J. Gao, S.-G. Yan, J. Zhao, Q. Tu, J. Zhang, X.-J. Duan, C.A. Sommer, G. Mostoslavsky, Biomaterials 32 (2011) 5065–5076. [4] P.D. Sawin, V.C. Traynelis, A.H. Menezes, J. Neurosurg. 88 (1998) 255–265. [5] J.R. Porter, T.T. Ruckh, K.C. Popat, Biotechnol. Prog. 25 (2009) 1539–1560. [6] C. Xu, P. Su, X. Chen, Y. Meng, W. Yu, A.P. Xiang, Y. Wang, Biomaterials 32 (2011) 1051–1058. [7] M. Mastrogiacomo, A. Muraglia, V. Komlev, F. Peyrin, F. Rustichelli, A. Crovace, R. Cancedda, Orthod. Craniofacial Res. 8 (2005) 277–284. [8] R. Langer, Adv. Mater. 21 (2009) 3235–3236. [9] W. Brown, Cements Research Progress, 1987. 351–379. [10] M. Julien, I. Khairoun, R.Z. Legeros, S. Delplace, P. Pilet, P. Weiss, G. Daculsi, J.M. Bouler, J. Guicheux, Biomaterials 28 (2007) 956–965. [11] S.V. Dorozhkin, J. Mater. Sci. 43 (2008) 3028–3057. [12] S. Takagi, L. Chow, K. Ishikawa, Biomaterials 19 (1998) 1593–1599. [13] M. Ginebra, M. Espanol, E. Montufar, R. Perez, G. Mestres, Acta Biomater. 6 (2010) 2863–2873. [14] E. Verron, I. Khairoun, J. Guicheux, J.-M. Bouler, Drug Discov. Today 15 (2010) 547–552. [15] M. Hofmann, A. Mohammed, Y. Perrie, U. Gbureck, J. Barralet, Acta Biomater. 5 (2009) 43–49. [16] M. Espanol, R. Perez, E. Montufar, C. Marichal, A. Sacco, M. Ginebra, Acta Biomater. 5 (2009) 2752–2762.
Y. Wu et al. / Materials Science and Engineering C 44 (2014) 326–335 [17] T. Katagiri, A. Yamaguchi, M. Komaki, E. Abe, N. Takahashi, T. Ikeda, V. Rosen, J.M. Wozney, A. Fujisawa-Sehara, T. Suda, J. Cell Biol. 127 (1994) 1755–1766. [18] A. Yamaguchi, T. Katagiri, T. Ikeda, J.M. Wozney, V. Rosen, E.A. Wang, A.J. Kahn, T. Suda, S. Yoshiki, J. Cell Biol. 113 (1991) 681–687. [19] Y.-T. Xiao, L.-X. Xiang, J.-Z. Shao, Biochem. Biophys. Res. Commun. 362 (2007) 550–553. [20] S. Fujibayashi, M. Takemoto, M. Neo, T. Matsushita, T. Kokubo, K. Doi, T. Ito, A. Shimizu, T. Nakamura, Eur. Spine J. 20 (2011) 1486–1495. [21] J. Even, M. Eskander, J. Kang, J. Am. Acad. Orthop. Surg. 20 (2012) 547–552. [22] S. Kokubo, R. Fujimoto, S. Yokota, S. Fukushima, K. Nozaki, K. Takahashi, K. Miyata, Biomaterials 24 (2003) 1643–1651. [23] H. Hosseinkhani, M. Hosseinkhani, A. Khademhosseini, H. Kobayashi, J. Control. Release 117 (2007) 380–386. [24] Z.S. Patel, M. Yamamoto, H. Ueda, Y. Tabata, A.G. Mikos, Acta Biomater. 4 (2008) 1126–1138. [25] J. Jansen, J. Vehof, P. Ruhe, H. Kroeze-Deutman, Y. Kuboki, H. Takita, E. Hedberg, A. Mikos, J. Control. Release 101 (2005) 127–136. [26] H. Shen, X. Hu, F. Yang, J. Bei, S. Wang, Biomaterials 30 (2009) 3150–3157.
335
[27] H.C. Kroese-Deutman, P.Q. Ruhé, P.H. Spauwen, J.A. Jansen, Biomaterials 26 (2005) 1131–1138. [28] P.Q. Ruhé, E.L. Hedberg-Dirk, N.T. Padron, P.H. Spauwen, J.A. Jansen, A.G. Mikos, Tissue Eng. 12 (2006) 789–800. [29] E.R. Luvizuto, S. Tangl, G. Zanoni, T. Okamoto, C.K. Sonoda, R. Gruber, R. Okamoto, Biomaterials 32 (2011) 3855–3861. [30] T. Crouzier, F. Sailhan, P. Becquart, R. Guillot, D. Logeart-Avramoglou, C. Picart, Biomaterials 32 (2011) 7543–7554. [31] M. Ginebra, T. Traykova, J. Planell, J. Control. Release 113 (2006) 102–110. [32] F.C. van de Watering, J.D. Molkenboer-Kuenen, O.C. Boerman, J.J. van den Beucken, J. A. Jansen, J. Control. Release 164 (2012) 283–290. [33] C. Liu, H. Shao, F. Chen, H. Zheng, Biomaterials 27 (2006) 5003–5013. [34] D. Zou, L. Guo, J. Lu, X. Zhang, J. Wei, C. Liu, Z. Zhang, X. Jiang, Tissue Eng. A 18 (2012) 1464–1478. [35] H. Kaneko, T. Arakawa, H. Mano, T. Kaneda, A. Ogasawara, M. Nakagawa, Y. Toyama, Y. Yabe, M. Kumegawa, Y. Hakeda, Bone 27 (2000) 479–486. [36] R.G. Sorensen, U.M. Wikesjö, A. Kinoshita, J.M. Wozney, J. Clin. Periodontol. 31 (2004) 796–804.
