A composited PEG-silk hydrogel combining with

Materials Science and Engineering C 65 (2016) 221–231

Contents lists available at ScienceDirect

Materials Science and Engineering C journal homepage: www.elsevier.com/locate/msec

A composited PEG-silk hydrogel combining with polymeric particles delivering rhBMP-2 for bone regeneration Dakun Ma a,1, Gang An a,1, Min Liang b, Yugang Liu c, Bin Zhang d,⁎⁎, Yansong Wang a,d,e,⁎ a

Department of Spine Surgery, the First Affiliated Hospital of Harbin Medical University, Harbin 150001, China Department of Orthopedic Surgery, Heilongjiang, Provincial Hospital, Harbin, 150001, China Department of Orthopedic Surgery, Affiliated Hospital of Hebei University of Engineering, Hebei, 056002, China d Institute of Hard Tissue Development and Regeneration, Harbin Medical University, Harbin 150001, China e China Orthopedic Regenerative Medicine (CORMed), Hangzhou, Zhejiang 310058, China b c

a r t i c l e

i n f o

Article history: Received 31 January 2016 Received in revised form 20 March 2016 Accepted 11 April 2016 Available online 14 April 2016 Keywords: Hydrogel Bio-composite BMP-2 sustained release Bone regeneration

a b s t r a c t Given the fabulous potential of promoting bone regeneration, BMP-2 has been investigated widely in the bone tissue engineering field. A sophisticated biomaterial loaded with BMP-2, which could avoid the required supraphysiological dose leading to high medical costs and risks of complications, has been considered as a promising strategy to treat non-healing bone defects. In this study, we developed a simple approach to engineer a composited hydrogel consisting polymeric particles (PLA/PLGA) used as a BMP-2 delivery vehicle. Compared with other groups, the introduction of PLA into PEG-silk gels endowed the hydrogel new physicochemical characteristics especially hydrophobicity which inhibited the burst release of BMP-2 and enhanced gel's structural stability. Moreover, such composited gels could stabilize entrapped proteins and maintain their bioactivity fully in vitro. In vivo, the bio-degradability experiment demonstrated this system was biocompatible and the reinforced hydrophobicity significantly decreased degradation rate, and in rat critical-sized cranial defects model, the gel containing PLA promoted the most bone formation. These findings demonstrated the introduction of PLA changed physicochemical features of gels more suitable as a BMP-2 carrier indicated by inducing bone regeneration efficiently in large bone defects at low delivered dose and this system may own translational potential. © 2016 Elsevier B.V. All rights reserved.

1. Introduction Treatments of non-healing osseous defects resulted from challenging fractures and osteotomies present a formidable problem for orthopaedic surgeons and tremendous socioeconomic costs [1]. Currently, autologous and allogeneic bone grafts are the most widely adopted therapies for non-healing bone defects, however, these techniques are limited by lots of drawbacks including the insufficient donor supply, donor site pain and inflammation etc. [2] Recently, bone morphogenetic protein-2 (BMP-2) with osteoinductivity, has sparked great interest in bone tissue engineering and has been a promising alternative to bone grafts. Although BMP-2 is capable of stimulating bone repair, its clinical application is still restricted greatly due to lacking optimal delivery vehicles [3,4]. Commonly, BMP-2 doses are orders of magnitude higher than physiological concentrations causing high costs of treatment and many adverse effects such as ectopic ossification, carcinogenicity and

⁎ Correspondence to: Y. Wang, Youzheng Street, Nangang District, Harbin 150001, China. ⁎⁎ Correspondence to: B. Zhang, Xuefu Road, Nangang District, Harbin 150086, China. E-mail addresses: [email protected] (B. Zhang), [email protected] (Y. Wang). 1 These authors contributed equally to this work.

http://dx.doi.org/10.1016/j.msec.2016.04.043 0928-4931/© 2016 Elsevier B.V. All rights reserved.

inflammation. Therefore developing an optimal BMP-2 delivery vehicle which could promote bone regeneration at low delivered BMP-2 dose to enable safe, efficient and cost-effective BMP-2 treatment is a principally clinical need [5–7]. Various biomaterials have been developed such as films, microspheres, scaffolds and hydrogels engineered by natural or synthetic materials to deliver BMP-2 [7–11]. Hydrogel, water-swollen cross-linked polymer networks, holds some advantages: i) proteins could be entrapped before hydrogels fabrication with 100% loading efficiency; ii) the fabricating process could be simplified, helpful for preparation and disinfection; iii) via regulating the cross-linking degree, it provides various release profiles meeting different situations and iiii) it can be fabricated in situ and used in injection or implantation manner. Herein it's particularly suitable to delivery protein drugs [12–15]. Silk fibroin (SF) has been one of the most attractive choices for tissue engineering. Firstly, biomaterials derived from purified SF own excellent biocompatibility, avoiding immune rejection and pathogen transmission [16,17]. Secondly, the biodegradability and mechanical properties could be modulated via tuning the ratio of α-random coil to β-sheet crystalline structure. Upon easy process, SF-based biomaterials could be processed via an all-aqueous or organic environment [18]. Nowadays, composited biomaterials have attracted great attentions and reveal better traits than those of biomaterials composed of single

222

D. Ma et al. / Materials Science and Engineering C 65 (2016) 221–231

component [19–21]. For instance, Farokhi et al. prepared a SF/calcium phosphate/PLGA nanocomposite scaffold to deliver vascular endothelial growth factor, in which PLGA prolonged the degradation time and inhibited the burst release. Moreover, such system resulted in more bone formations in vivo [22]. Within this context, we designed a simple method to fabricate a composited hydrogel containing PLA/PLGA particles as the BMP-2 carrier. In our study, we utilized PEG-silk hydrogel as the fundamental framework into which the polymeric particles were filled. The method to prepare such a system was simpler compared with in situ-forming hydrogel containing PLA-PEG-PLA copolymer engineered by Zhong et al. [23]. Generally speaking, this study underlines a creative approach to fabricate a composited hydrogel as the BMP-2 carrier for bone regeneration. We set four groups: the pure PEG-silk gel and gel containing PLA, PLGA85/15 and PLGA50/50 particles respectively. Then we investigated what impacts on both physiochemical features and release profile the addition of polymers brought about in vitro and whether this vehicle containing low doses of BMP-2 induced bone formation efficiently in a rat critical-sized cranial defects.

2. Materials and methods

2.4. Gelation time Upon mixing two kinds solution, the mixture formed into a homogeneous microstructure which caused an increase in light scattering within the visible light range. The gelation kinetics was recorded as the function of optical density (OD) changes to time [18]. The mixture was injected into a 96-well plate with 200 μL and then, the plate was subjected to OD measurements at 550 nm under the kinetics mode on a UV–visible spectrophotometer set at room temperature. The kinetics was monitored over 20 min, at 5-second and 1-minute intervals respectively (n = 3). 2.5. Solubilization of PEG Dried gels (n = 3) were weighed (Wdi) and immersed in glass beakers containing 500 mL deionized water on the shaker with the rate of 80 rpm. At designated time points, gels were taken out, dried at 60 °C, weighed (Wdt) and put back into the beaker with replenished equal volume deionized water. Thereafter on the basis of the known weight of PEG400 (WP), the percentage of PEG release was calculated as follows: PEG400r ð%Þ ¼ ðWdi −Wdt Þ=WP 100:

ð1Þ

2.1. Materials 2.6. The content of residual organic solvent Silk fibers were purchased from Xiehe Silk Corporation (Shengzhou, Zhejiang province, China). Pharmaceutical grade PEG (MW: 400 g/mol) was obtained from Aladdin Industrial Inc. (Shanghai, China); PLA (IV: 0.41) and PLGA (85/15, IV: 0.313 and 50/50, IV: 0.27) were obtained from Daigang Biomaterial Co., Ltd (Shandong province, China); rhBMP-2 (IP: 8.17) and elisa kits were bought from Sinobio Biotech (Shanghai, China); organic solvents, other chemical reagents and pNPP were obtained from Sigma-Aldrich (St Louis, MO, USA), and Dulbecco Modified Eagle's Medium (DMEM) and fetal bovine serum (FBS) were purchased from Gibco (Grand Island, NY, USA).

According to the preparation method, we prepared groups B, C and D viscous solutions of 22 g (W0) PEG400, 4.4 g solvents and 4.4 g polymers and then mixtures were placed into a 50-mL vial to evaporate solvents in a fume hood during which we weighed the mixture until there was almost no decrease in weight (n = 3). The initial weight was called as Wi and the remaining weight as Wr, therefore the percentage of the residual organic solvent (Or) was calculated using the following equation: Or ð%Þ ¼ ð4:4−Wi þ Wr Þ=ð4:4−Wi þ Wr þ W0 Þ 100%:

ð2Þ

2.2. Silk purification

2.7. Scanning electron microscopy (SEM)

Silk fibers purchased were degummed according to published procedures to remove residual sericin contaminants closely related to inflammatory reaction when applied in vivo [24,25]. Briefly, silk fibers were boiled in 0.02 M sodium carbonate solution for 30 min and then rinsed with ultrapure water three times. Then obtained fibers were dissolved in 9.3 M LiBr solution at 60 °C for 48 h following drained and dried in a fuming cupboard 24 h to gain SF solution. Thereafter the solution was dialyzed against deionized water via a dialysis box for 48 h to remove LiBr and centrifuged to drive out insoluble debris, and finally the concentration was 7–8% (w/v) [24,25].

After the removal of PEG400, PEG-silk hydrogels (300 μL) were prefrozen at −80 °C overnight and then lyophilized for 48 h. These freezedried gels were mounted on sample stubs and coated with Au, and images were taken using scanning electron microscopy (JEOL JSM-6300 SEM, Japan) at 12.5 kV (n = 3).

2.3. PEG-silk hydrogel preparation Polymers were dissolved in PEG400 with the help of the ethyl acetate (EA). The weight ratio of polymers, PEG400 and organic solvent was 1:5:5. After 48 h dissolution, solvents were evaporated as much as possible under an aseptic environment. During this procedure, the state of mixtures like turbidity must be supervised. After almost complete evaporation, these four group solutions, namely the pure PEG400 (group A) and PEG400 consisting PLA, PLGA85/15 and PLGA50/50 (groups B, C and D respectively) were stored at 4 °C before being mixed with 30% (w/v) silk fibroin solution following 1:1 ratio. They were mixed via medical syringes which were connected via a connector (Fig. 1a). After several minutes, gels could be squeezed out from the device into 5-mL centrifuge tubes to be stored at 4 °C for other assays.

2.8. Fourier transform infrared spectroscopy (FTIR) To analyze whether the residual solvent and the addition of polymer influence the conformational transformation, after extractions of PEG400 and lyophilized, samples (n = 3) were powdered and subjected to FTIR measurement (Thermo Fisher Scientific, Nicolet 6700 FT-IR) in the spectral region of 400–4000 cm− 1, while background measurements were taken with an empty cell and subtracted from the sample reading. The curves that had absorption bands at the frequency range of 1620–1630 cm−1 and 1525–1535 cm−1 were on behalf of enriched β-sheet structure in silk II form [26]. The β-sheet structure content (contributions of these curves to the amide I band) was assessed by integrating the area under the curve and then normalizing to the total area under the amide I band region (1600–1700 cm−1) as described previously [26]. 2.9. X-ray diffraction (XRD) To evaluate the degree of crystallinity, the rinsed and lyophilized hydrogel (n = 3) was ground into powder and sieved with a 200 mesh sieve before XRD measurement (D/MAX-2500, Rigaku Co., Tokyo,


223

Fig. 1. Gels preparation and determination of gelation properties (a) devices to prepare hydrogel. (b) Different size hydrogels. (c) Time-dependent optical density measurements. Standard deviations range from 0.004 to 0.028. (d) PEG400 release from gels.

Japan) which was operated at 40 kV tube voltage and 40 mA tube current with diffraction angle 2θ = 5°–45° and scanning rate of 4°/min. The main diffraction peak at 19.7° and 24.7° represent the silk I, while those at 9.1° and 20.5° represent silk β-sheet structure [23,27]. 2.10. Mechanical analysis Hydrogels were subjected to an unconfined compression strain-tofailure experiment with 8 mm in diameter and height (n = 3). Samples were placed onto an Instron 5848 Microtester between stainless steel parallel plates and the upper plate was lowered continuously at a rate of 1 mm/min until the compressive force didn't increase owing to the gel's fracture. By normalizing against the sample geometries, the compressive stress and strain were established and the elastic modulus was calculated as reported previously [28].

58 and 61) to verify the cumulative amounts of the released rhBMP-2 and then tubes were replenished with fresh buffer continuously for up to 61 days. Another experiment in which the supernatant was withdrawn at 1st, 3rd, 5th, 7th, 9th, 11th, 13th, 15th, 17th, 19th and 21th day was performed to determine the released amount per day. These collected samples were stored at − 20 °C until for ELISA assay using commercial kits following the manufacturer's instructions (n = 3). Additionally, release profiles could be explained using mathematical kinetic models: zero-order (3), first-order (4), Higuchi (5) [29]. F ¼ kt þ b;

ð3Þ

logQ t ¼ logQ 0 þ Kt =2:303;

ð4Þ

pffiffi F ¼ KH t :

ð5Þ

2.11. The stability of gels in fluid 2.13. The bioactivity of released rhBMP-2 in vitro Gels were cut into blocks with a 1.2-mm diameter and a 1-mm thickness and each sample was immerged in a 100-mL vial with 50 mL PBS (pH 7.4). Before vibration, samples were pre-soaked in PBS for 0, 4, 7 and 14 days respectively. The vial was put into the shaker at a rate of 400 rpm at 37 °C, and taken out until gels fractured thoroughly (n = 3). The time consumed was recorded to evaluate the gel stability in the fluid.

The bioactivity of rhBMP-2 released was assessed via the ALP activity of cultured rat bone marrow mesenchymal stem cells (BMSC). Gel loaded with 5 μg rhBMP-2 was immersed into 2-mL microcentrifuge tubes containing 1-mL complete cell culture medium (DMEM Culture Medium) [30]. The supernatant was withdrawn at 1st, 5th, 10th, 15th, 20th, 25th, 30th, 35th and 40th day and the same volume medium was added to tubes. The collected samples containing rh-BMP2 released

2.12. The release profile in vitro Primarily rhBMP-2 was added into PEG400 after solvents evaporation and according to preparation methods, hydrogels loaded with rhBMP-2 were prepared using 1-mL medical syringes. The hydrogel of 60 μL, containing 5 μg rhBMP-2 was immersed in a 2-mL microcentrifuge tube containing 1 mL PBS (pH 7.4) and 0.02% (w/v) sodium azide, and these samples were incubated at 37 °C with continuous gentle agitation. At various time points, the supernatant was withdrawn (at days 1, 4, 7, 10, 13, 16, 19, 22, 25, 28, 31, 34, 37, 40, 43, 46, 49, 52, 55,

Table 1 Silk gelation time. Group

Gelation time (S)

A

B

C

D

Pure gel

PLA

PLGA85/15

PLGA50/50

360 ± 21.51

484 ± 19.81⁎

435 ± 16.58⁎

421 ± 12.94⁎

Note: gelation time is defined as the time until optical density plateaued. ⁎ p b 0.05 compared to group A.

224


for 15 min and absorbance was read at 570 nm [22]. We also carried out PCG, NCG and gels of no BMP-2 (n = 3).

Table 2 The proportion of residual organic solvents. Group

Residual content (%)

B

C

D

EA

EA

EA

2.55 ± 0.31

2.33 ± 0.33

2.32 ± 0.35

Note: With the premise of ensuring the stability of mixtures.

per day were used to culture rat BMSC while the first day samples were diluted ten times whose concentration was much too higher than that of subsequent samples (n = 3). For this purpose, 4 × 104 cells/well were seeded in 24-well tissue culture plates and after 24 h incubation the medium was replaced with extraction solutions [31]. Then following three culture days, cells were processed to evaluate the ALP activity as manufacturer's protocols of ALP's kit at 37 °C. Additionally we used the complete culture medium with 12 ng free rhBMP-2 as a positive control group (PCG) but without rhBMP-2 as a negative control group (NCG) [22].

2.14. MTT assay MTT assay was accomplished to evaluate biocompatibility of gels. Briefly, 1 × 104 cells/well were seeded on 96-well plate and after 24 h incubation, the culture media was replaced with extraction solutions and remained 72 h. The cells were then stained with a 2% MTT solution in PBS for 4 h. The formed formazan was solubilized with isopropanol

2.15. Biodegradations in vivo For the animal experiments, protocols were approved by the Institutional Animal Care and Use Committee of the Harbin Medical University. Wistar rats of 170–220 g and male were used for experiments and then anaesthetized with intraperitoneal sodium pentobarbital (50 mg/kg). Gel blocks of 0.2 mL prepared using 2.5-mL syringes were implanted at the upper back both sides. All experiments were performed under aseptic conditions and all rats survived without any complications (n = 3). Biodegradations of gel blocks at the implantation site were analyzed at both 2 and 8 weeks via histological analysis. 2.16. In vivo bone regeneration in rat calvarial defect model Wistar rats of 170–220 g and male were anaesthetized with intraperitoneal sodium pentobarbital (50 mg/kg) (n = 3). After the scalp hair shaved, a longitudinal incision was made in the midline of cranium and the periosteum was elevated to expose the surface of parietal bones. Using a trephine bur, a circular and transosseous defect of 8 mm diameter was produced in the parietal bone [32]. The drilling site was irrigated with cold saline and the defect was filled with a 60 μL gels loaded with 2.5 μg BMP-2 but the control group with no gel. The periosteum and skin were then closed in layers with resorbable sutures. The rats were housed singly after surgery and received humane care, and the

Fig. 2. Silk hydrogel morphologies determined by SEM.


225

Fig. 3. Structural analysis. (a) the FTIR spectra of hydrogels. (b) conformational fraction calculated after the deconvolution of amide I. (c) XRD measurements.

transplant was retrieved at 2 and 8 weeks after implantation for X-ray and histological analysis.

2.19. Statistical analysis One way analysis of variance (ANOVA) with Tukey honestly significant difference post hoc test and Independent-Sample t-test using SPSS software (SPSS Inc., Chicago, IL, USA) was performed to statistical analysis of all quantitative data which were expressed as means ± standard deviations simultaneously. Statistical significance was represented by p b 0.05.

2.17. Soft X-ray analysis Following euthanasia and harvesting the calvaria, bone regeneration at the calvarial defect site was assessed using soft X-ray (CMB-2, Softex Co., Japan), which was performed at the conditions of 30 kV, 3 mA for 90 s exposure time [33]. The extent of the defects' closure and newly formed bone was observed from the radiographic images and percentages of newly formed bone within the defect were calculated using computerized image analysis system, SPOT version 4 (Diagnostic instrument, Michigan).

3. Results 3.1. Result of gelation time Upon mixing, the mixture turned into opaque and shortly a visually homogeneous and wax-like gel formed. Using different-sized syringes different-sized hydrogel could be prepared (Fig. 1b). The gelation dynamics was determined according to the change of OD and the time consumed from the beginning of mixing to the point where the optical density kept unchanged anymore was defined as the gelation time (Fig. 1c) [27]. We found that hybrid gels consumed more time than group A (p b 0.05) and for group B, it consumed the most time (Table 1). Dependent on the mechanism of how PEG forces the

2.18. Histological analysis For bone regeneration, specimens at week 8 were fixed in 10% buffered formalin, and then decalcified, dehydrated, embedded in paraffin, and stained with hematoxylin and eosin (H&E). For biodegradation, gels were excised at weeks 2 and 8, embedded in paraffin and then subjected to H&E staining. All samples were observed by optical microscopy (Olympus, IX 70, Japan).

Table 3 Mechanical properties obtained via unconfined compression strain-to-failure test. Group

Compressive strength (kPa) Elastic modulus (kPa) Strain-to-failure (%) ⁎ p b 0.05 compared to group B. # p b 0.05 compared to group A.

A

B

C

D

Pure gel

PLA

PLGA85/15

PLGA50/50

202.14 ± 2.54 13.76 ± 0.59 13.89 ± 0.27

151.21 ± 1.54⁎,# 10.28 ± 0.46⁎

158.63 ± 1.73⁎,# 10.43 ± 0.59⁎ 13.56 ± 0.44

134.21 ± 2.51⁎ 9.42 ± 0.74⁎ 13.31 ± 0.48

13.69 ± 0.34

226


Fig. 4. The stability of gels in PBS, release profiles and ALP activity of BMSC in vitro. (a) The consumed time from an integral gel to fractures. ⁎p b 0.05 compared to group D. (b) rhBMP-2 per day release. (c) rhBMP-2 cumulative release amount. (d) Alkaline Phosphatase activity. PCG: positive control group; NCG: negative control group. ⁎p b 0.05 compared to group A, #p b 0.05 compared to group B.

hydrophobic domains to assemble and further form gel network, the addition of polymers reduced the rate of gelling.

completely. However, EA has been approved by the U.S. FDA as a common ingredient in pharmaceutical industries and in case of any polymeric precipitation we chose the above proportion as terminal point.

3.2. The extraction of PEG400 3.4. The morphology of lyophilized gels In the extraction study, the amount of PEG released from pure silkPEG gels was less than that of composited gels within 12 h, followed by gentle PEG leaching out until 48 h when a plateau was achieved (Fig. 1d). Ultimately, the cumulative released amount was 93.5, 98.8, 98.1 and 98.4% from group A to D, demonstrating a discrepancy between the pure and composited groups. 3.3. The content of residual organic solvent After evaporation until the weight was constant, following the equation, we calculated the residual proportion shown in Table 2. From group B to D, the mean residual content was about 2.55, 2.33 and 2.32% respectively and actually the solvent could be evaporated out Table 4 Analysis on release profiles using zero-order, first-order and Higuchi models. Zero-order

First-order

Higuchi

Group

R2

R2

R2

A B C D

0.987 0.996 0.986 0.99

0.975 0.983 0.971 0.981

0.979 0.989 0.981 0.983

Morphological features of hydrogels after lyophilization were imagined using SEM (Fig. 2). Each gel demonstrated different morphologies in terms of micro-architecture and polymeric particles. The pure gel revealed high porosity and interconnectivity, and the diameter of pores whose walls was extraordinarily thin, varied from several to dozens of micrometers. However, for PLA gels, although they revealed compact porous structure, what's the most prominent was that sphere-like particles with several micrometers in diameter distributed in the whole gel system homogeneously while for group C and D, the particle was larger and spread nonuniformly. Additionally, the structure of both group C and D were also irregularly porous and poorly interconnected, which was hazardous to tissue regeneration. Hence from the geographic prospective, the network of A and B was more optimal as BMP-2 vehicle than that of C and D. 3.5. The structural analysis of FTIR As shown in Fig. 3a, β-sheet structure formed accompanying with gelation indicated by a major band at 1623 cm−1 in the amide I region and at 1515 cm−1 in the amide II region. The residual PEG has no characteristic absorption band in the amide I and II region, which results a negligible influence on structural analysis [34]. Moreover, in contrast


227

Fig. 5. OD values of MTT assay. (a) extraction solutions of gels without rhBMP-2. (b) extraction solutions of gels loading with rhBMP-2. ⁎p b 0.05 compared to group B (p b 0.05).

to group A, an intense band emerged for composited gels at 1752 cm−1, the characteristic absorption band of polymers, which implied components were only blended but phase separated. After deconvolution of the amide I band, the proportion of β-sheet structure was similar in total secondary structure elements (Fig. 3b). 3.6. XRD analysis X-ray diffraction was carried out to study the crystallinity of gels (Fig. 3c). A strong peak at 20.3° implied silk existed in silk II form mainly, further confirming results of FTIR.

3.7. The mechanical property Samples were subjected to an unconfined compression strain-tofailure experiment to evaluate the mechanical property. In the Table 3, compressive strength and elastic modulus of composited gels increased when compared with those of group A, and PLA group revealed the best mechanics among them. Although after polymeric intervention mechanical properties were altered significantly (p b 0.05), the strain-tofailure was similar.

Fig. 6. Histological study of degradation at weeks 2 and 8. Asterisk represents remaining gels. (HE × 20).

228


Fig. 7. Radiographic results of the cranial defects at weeks 2 and 8.

3.8. The stability in the PBS

3.10. The bioactivity of rhBMP-2 in vitro

Under the vibration of 400 rpm in the shaker, gel blocks would be broken with time and the PBS changed into nepheloid. The pure gel consumed the least time while group B took the most time (Fig. 4a). After different time of pre-soaking, results demonstrated that the longer pretreatment, the less consumed time and that group B still took the most time. With hydrophobic polymer particles, the consumed time of composited gels was far more than that of group A (p b 0.05), and hence the stability and capability of resisting water invasion of group B was most outstanding.

The results concerning the ALP activity were shown in Fig. 4d. As well known, with the DMEM containing 10% FBS the BMSCs also demonstrate some degree of ALP activity as implied by NCG to which results of extraction solutions of gels unloading rhBMP-2 were similar (data was not shown), while for PCG and experimental groups, the ALP activity was promoted significantly as a consequence of rhBMP-2. Although with time, the osteoinductivity of released rhBMP-2 to BMSC became depressed because of potty loss of bioactivity, results suggested that rhBMP-2 was still integral and active until the end of the assay. In the meanwhile, the ALP activity of group B was distinctly higher than that of group A at 15th, 25th, 30th, 35th and 40th day, mainly relevant to the enhanced hydrophobicity, despite not all significant.

3.9. The release profile in vitro Fig. 4c shows the cumulative amounts of delivered rhBMP-2 in an incubator-shaker following 61 days release, about 24.72% of the initial rhBMP2 was released from the pure hydrogel, of which amount was more than that of composited gels. Additionally, the per day release revealed that except the first day, the discrepancy became more and more impalpable among four groups with time, which was consistent with our speculation that the high concentration hydrogel made sustained release more methodical and orderly (Fig. 4b). Additionally the inhibited burst release of group B within 24 h demonstrated that the gel with enhanced hydrophobicity would control delivery better. As shown in the Table 4, the rhBMP-2 release profile could be best interpreted using zero-order model (R2 varying between 0.987 and 0.996).

3.11. Results of MTT assay As shown in the Fig. 5a, in contrast to the NCG, results of extraction solutions of no rhBMP-2 revealed this system was almost no cytotoxic and compatible with cells. The viability of PCG was higher than other groups significantly (p b 0.05) owing to the presence of free rhBMP-2 and moreover the MTT assay not only revealed this carrier was noncytotoxic but proved proteins released bioactive, which were consistent with ALP test (Fig. 5b). 3.12. In vivo biodegradation The histological results from two representative time points are shown in Fig. 6. At week 2, a capsule membrane mainly consisting

Table 5 The percentage of new bone to the total defect. Time

Control

A Pure gel

B PLA

C PLGA85/15

D PLGA50/50

2 W (%) 8 W (%)

1.32 ± 0.38 4.43 ± 0.59

6.48 ± 0.55# 13.77 ± 1.19#

9.71 ± 1.35⁎ 32.46 ± 2.35⁎

7.18 ± 1.26 20.62 ± 0.84⁎,#

6.73 ± 0.73# 15.38 ± 1.91#

⁎ p b 0.05 compared to group A. # p b 0.05 compared to group B.


229

Fig. 8. Histological study of transverse bone defect sections at week 8. *remaining gels. NB: new regenerated bone. CG: the control group. (HEx10).

fibrous connective tissue and vessels had formed on the gel surface and the degree of degradation of pure gels was larger than that of composited gels implied by that neo-tissues started to appear in the pure gel while composited gels remained almost intact especially the PLA group. At week 8, the pure gel degraded into large pieces and for C and D group, cells and micro-vessels have penetrated into the centre of gel. Regarding to the PLA group, the capsule became more intense and a bit little neo-tissue appeared in the gel. Additionally the interface between silk gel and surrounding tissues was clean, with no sign of inflammatory responses. 3.13. Bone regeneration in vivo Soft X-ray allowed mineralized tissues to be distinguished from the remaining soft tissues (Fig. 7). At week 2, dispersed irregular-shaped bone formation adjacent to the defect margins except the control group was found. At week 8, large amounts of new bone formed in defects and in group B, most bone formed. In Table 5, new bone formed areas were significantly larger in the BMP-2-loaded groups than in the control group at different time points and for the PLA group, the area was significantly higher than that of other gels containing BMP-2. At 8 weeks after implantation, histological results demonstrated no severe inflammation, necrosis and hemorrhage appeared around the gel. In the blank control group, large amounts of fibrous connective tissue filled the defect region while new and large bone islands formed in test groups indicated by Fig. 8. Moreover, more mature bone accompanying with pronounced vascularization appeared in the PLA group while sparse neovascularization was observed within the porous structure of gel for groups A, C and D. The boundary between new bone and the gel was unclear in the PLA group because sufficient mature bone tissue that had grown into the pores of gels. Contrarily, for the pure gel,

although new bone filled the defect, an obvious space still appeared between the gel and new bone. 4. Discussion It has been widely recognized that the prolonged and localized retention of BMP-2 in vivo would promote bone regeneration better, herein developing a strategy to deliver BMP-2 in a spatiotemporal way via a biomaterial possessing optimal biodegradability, biocompatibility and biomechanics has been popular in the bone tissue engineering field [3,9]. Considering advantageous biocomposites, we developed an innovative approach to fabricate a composited hydrogel system [3,20]. We used medical syringes to mix PEG400 containing polymers with high concentration silk solution and when PEG inducing silk structural transitions and gel network formation, the polymer precipitated from the organic phase, formed particles with the high shear force and filled within the gel. With the aim of reducing the hydrophilicity of gels to optimize the release profile, to improve the stability of BMP-2 and to prolong the degradation time, we introduce hydrophobic polymers into gels in such a way [27]. Evidently, our method was simpler and more cost-effective and saving-time compared to other bio-composites, like the bio-hybrid SF/calcium phosphate/PLGA nanocomposite scaffold fabricated using freeze-drying and electro-spinning [22,23]. Besides, this composited gel was prepared under mild ambient not harsh conditions like high temperature and poisonous chemical agents etc. [35]. The present study focuses on that what impacts the introduction of hydrophobic polymers had on physicochemical features and whether the composited gel exerted more potential of BMP-2 to induce bone formation in vivo. Primarily the mechanism of gelation is that the hydrophilic hydroxyl of PEG interacts with the water molecular around chains of SF, inducing

230


hydrophobic domains to assemble and resulting in conformation transitions from random coil-dominated structure to a high content of crystalline β-sheet second-structure.[18,25,27] After the dissolution of PL(G)A in PEG400, polymeric chains spread thoroughly leading to increase in hydrophobicity of PEG400 mixtures, which lead composited gels took more time to accomplish the gelation than pure gels (Fig. 1c). In PLA group, polymeric particles distributed throughout the matrix homogeneously, which was resulted from that upon mixing two kinds solutions, polymers precipitated promptly and filled into the silk gel network induced by PEG400 simultaneously (Fig. 2). The reinforced hydrophobicity also resulted in less hydrogen bond between PEG and Serine, which caused the cumulative amount of PEG extracted of PLA group was the most all the way (Fig. 1d). Dependent on the images of SEM, after the introduction of polymers, the network of gels whose pores became smaller was more compact and PLA microparticles distributed homogeneously within the gel, leading to significant improvement in mechanical properties (Table 3). Moreover, the stability of composited gels in fluid was promoted significantly, chiefly caused by enhanced hydrophobicity and improved mechanical properties (Fig. 4a), which was essential to predict the performance in vivo. Despite physiochemical properties has been changed after addition of polymers, based on FTIR and XRD assays, the copolymer had no obvious impact on conformation transition because of no interaction between copolymer and silk or PEG, just phase separation (Fig. 3). Besides, after polymeric interventions, not only was the biocompatibility unaffected by residual solvents revealed by the MTT assay in vitro and H&E staining in vivo (Figs. 5, 6 and 8) but the burst release was inhibited significantly within 24 h as a result of enhanced hydrophobicity. Moreover, the subsequent delivery was methodical and orderly, which was controlled by the high concentration gel indicated by the similar per day release among these four groups (Fig. 4b–c). Commonly, we could divide entrapped BMP-2 into four categories: i) dissolved in free water imbedded in hygrogels; ii) entrapped in solidified PEG400 interacted with silk; iii) interacting with silk and polymers by electrical attractions and iiii) closely encapsulated in compact gel's network (accounting for the most) [9]. According to this point, we could modulate the release profile via modulating loading manner and silk concentration to satisfy different requirements. Besides we could utilize this composited gel to deliver two growth factors simultaneously with synergistic effects (like VEGF and BMP-2) via different incorporation ways to promote better bone repair [36]. Although the extended release period, due to majorities of rhBMP-2 encapsulated in gels, composited gels provided enough protection for entrapped proteins implied by the ALP activity (Fig. 4d). Obviously, hydrogels containing PLA could not only provide better control on BMP-2 delivery but also guarantee its bioactivity better when compared with widely investigated biomaterials such as collagen/chitosan scaffold and PEG/hyaluronic acid hydrogel [37–40]. In vivo, the biodegradation of gels was chiefly initiated by proteolysis and hydrolysis. After implantation subcutaneously, at week 8, the pure gel degraded into small pieces accompanied with large amounts of neo-tissue ingrowth while the PLA gel was just lysised partially due to the reinforced hydrophobicity and the compact microstructure (Fig. 6). In contrast to PLA gels, group C and D degraded faster because of inferior hydrophobicity of PLGA85/15 and PLGA50/50, and nonuniform particles distribution. With this advantage, PLA gels could accomplish the local release and increase the retention time under the prerequisite of stabilizing rhBMP-2 in vivo, which was beneficial for bone repair. The result of radiographic images and H&E staining of defects confirmed our speculation that the PLA gel was the best BMP-2 vehicle (Figs. 7–8). The implanted PLA gels supplied not only the osteoinductivity via loading with BMP-2 but also osteoconductivity via acting as a substrate that provided space for cells ingrowth. Despite with enhanced hydrophobicity for group B, C and D, the PLA gel owned more porous and interconnected microarchitecture that could provide a better substrate for bone tissue ingrowth than other two groups, which lead the most bone formation in group B at week 8

(Table 5). Obviously, this system was cyto-compatible not like PLGA scaffolds causing inflammatory action due to their degraded byproducts and could not only provide sufficient space and structural support for tissue ingrowth but release BMP-2 locally to promote the bone formation efficiently at low doses avoiding ectopic ossification associated with collagen scaffold [35,38]. And for pure gels, they couldn't retard BMP-2 for sufficient time locally with overhigh degradation rate, and hence the new bone area was the least. 5. Conclusion In our study, we developed a simple strategy to fabricate a composited gel containing polymeric particles possessing superior physicochemical properties as a BMP-2 vehicle. The enhanced hydrophobicity after adding PLA into PEG-silk gels not only protected the bioactivity of proteins but prolonged their degradation time in vivo. In critical-sized cranial defects, the PLA gel with low dose BMP-2 exerted the potential of BMP-2 to promote bone formation efficaciously. Additionally the composited gel also demonstrated the excellent biocompatibility without obvious inflammation in vivo. Finally we proposed a promising strategy and platform for BMP-2 application in healing large bone defects, even with significant translational potentials. Disclosures The authors have declared that there is no conflict of interest. Acknowledgments This study was supported by the National Nature Science Foundation of China (project no. 81271696). References [1] A. Shekaran, J.R. García, A.Y. Clark, T.E. Kavanaugh, A.S. Lin, R.E. Guldberg, et al., Bone regeneration using an alpha 2 beta 1 integrin-specific hydrogel as a BMP-2 delivery vehicle, Biomaterials 35 (2014) 5453–5461. [2] P.C. Bessa, E.R. Balmayor, J. Hartinger, G. Zanoni, D. Dopler, A. Meinl, et al., Silk fibroin microparticles as carriers for delivery of human recombinant bone morphogenetic protein-2: in vitro and in vivo bioactivity, Tissue Eng. Part C Methods 16 (2010) 937–945. [3] H. Seeherman, J. Wozney, R. Li, Bone morphogenetic protein delivery systems, Spine 27 (2002) S16–S23. [4] H. Seeherman, J.M. Wozney, Delivery of bone morphogenetic proteins for orthopedic tissue regeneration, Cytokine Growth Factor Rev. 16 (2005) 329–345. [5] E.J. Carragee, E.L. Hurwitz, B.K. Weiner, A critical review of recombinant human bone morphogenetic protein-2 trials in spinal surgery: emerging safety concerns and lessons learned, Spine J. 11 (2011) 471–491. [6] S.T. Yoon, S.D. Boden, Osteoinductive molecules in orthopaedics: basic science and preclinical studies, Clin. Orthop. Relat. Res. 395 (2002) 33–43. [7] G.B. Bishop, T.A. Einhorn, Current and future clinical applications of bone morphogenetic proteins in orthopaedic trauma surgery, Int. Orthop. 31 (2007) 721–727. [8] L. Luca, A.-L. Rougemont, B.H. Walpoth, R. Gurny, O. Jordan, The effects of carrier nature and pH on rhBMP-2-induced ectopic bone formation, J. Control. Release 147 (2010) 38–44. [9] O. Jeon, S.J. Song, H.S. Yang, S.-H. Bhang, S.-W. Kang, M.A. Sung, et al., Long-term delivery enhances in vivo osteogenic efficacy of bone morphogenetic protein-2 compared to short-term delivery, Biochem. Biophys. Res. Commun. 369 (2008) 774–780. [10] X.B. Yang, M.J. Whitaker, W. Sebald, N. Clarke, S.M. Howdle, K.M. Shakesheff, et al., Human osteoprogenitor bone formation using encapsulated bone morphogenetic protein 2 in porous polymer scaffolds, Tissue Eng. 10 (2004) 1037–1045. [11] P.D. Mariner, J.M. Wudel, D.E. Miller, E.E. Genova, S.O. Streubel, K.S. Anseth, Synthetic hydrogel scaffold is an effective vehicle for delivery of INFUSE (rhBMP2) to critical-sized calvaria bone defects in rats, J. Orthop. Res. 31 (2013) 401–406. [12] S.R. Van Tomme, G. Storm, W.E. Hennink, In situ gelling hydrogels for pharmaceutical and biomedical applications, Int. J. Pharm. 355 (2008) 1–18. [13] J. Patterson, R. Siew, S.W. Herring, A.S. Lin, R. Guldberg, P.S. Stayton, Hyaluronic acid hydrogels with controlled degradation properties for oriented bone regeneration, Biomaterials 31 (2010) 6772–6781. [14] K.Y. Lee, D.J. Mooney, Hydrogels for tissue engineering, Chem. Rev. 101 (2001) 1869–1880. [15] B.B. Mandal, S. Kapoor, S.C. Kundu, Silk fibroin/polyacrylamide semiinterpenetrating network hydrogels for controlled drug release, Biomaterials 30 (2009) 2826–2836.

D. Ma et al. / Materials Science and Engineering C 65 (2016) 221–231 [16] G.H. Altman, F. Diaz, C. Jakuba, T. Calabro, R.L. Horan, J. Chen, et al., Silk-based biomaterials, Biomaterials 24 (2003) 401–416. [17] M. Fini, A. Motta, P. Torricelli, G. Giavaresi, N.N. Aldini, M. Tschon, et al., The healing of confined critical size cancellous defects in the presence of silk fibroin hydrogel, Biomaterials 26 (2005) 3527–3536. [18] A. Matsumoto, J. Chen, A.L. Collette, U.-J. Kim, G.H. Altman, P. Cebe, et al., Mechanisms of silk fibroin sol-gel transitions, J. Phys. Chem. B 110 (2006) 21630–21638. [19] M. Swetha, K. Sahithi, A. Moorthi, N. Srinivasan, K. Ramasamy, N. Selvamurugan, Biocomposites containing natural polymers and hydroxyapatite for bone tissue engineering, Int. J. Biol. Macromol. 47 (2010) 1–4. [20] C. Soundrapandian, B. Sa, S. Datta, Organic–inorganic composites for bone drug delivery, AAPS PharmSciTech 10 (2009) 1158–1171. [21] V. Mouriño, J.P. Cattalini, J.A. Roether, P. Dubey, I. Roy, A.R. Boccaccini, Composite polymer-bioceramic scaffolds with drug delivery capability for bone tissue engineering, Expert Opin. Drug Deliv. 10 (2013) 1353–1365. [22] M. Farokhi, F. Mottaghitalab, M.A. Shokrgozar, J. Ai, J. Hadjati, M. Azami, Bio-hybrid silk fibroin/calcium phosphate/PLGA nanocomposite scaffold to control the delivery of vascular endothelial growth factor, Mater. Sci. Eng. C 35 (2014) 401–410. [23] T. Zhong, C. Deng, Y. Gao, M. Chen, B. Zuo, Studies of in situ-forming hydrogels by blending PLA-PEG-PLA copolymer with silk fibroin solution, J. Biomed. Mater. Res. Part A 100 (2012) 1983–1989. [24] H. Yamada, H. Nakao, Y. Takasu, K. Tsubouchi, Preparation of undegraded native molecular fibroin solution from silkworm cocoons, Mater. Sci. Eng. C 14 (2001) 41–46. [25] D.N. Rockwood, R.C. Preda, T. Yücel, X. Wang, M.L. Lovett, D.L. Kaplan, Materials fabrication from Bombyx mori silk fibroin, Nat. Protoc. 6 (2011) 1612–1631. [26] X. Hu, D. Kaplan, P. Cebe, Determining beta-sheet crystallinity in fibrous proteins by thermal analysis and infrared spectroscopy, Macromolecules 39 (2006) 6161–6170. [27] X. Wang, B. Partlow, J. Liu, Z. Zheng, B. Su, Y. Wang, et al., Injectable silkpolyethylene glycol hydrogels, Acta Biomater. 12 (2015) 51–61. [28] X. Wang, J.A. Kluge, G.G. Leisk, D.L. Kaplan, Sonication-induced gelation of silk fibroin for cell encapsulation, Biomaterials 29 (2008) 1054–1064. [29] P. Shi, S.A. Abbah, K. Saran, Y. Zhang, J. Li, H.-K. Wong, et al., Silk fibroin-based complex particles with bioactive encrustation for bone morphogenetic protein 2 delivery, Biomacromolecules 14 (2013) 4465–4474.

231

[30] A. Abarrategi, A. Civantos, V. Ramos, J.V. Sanz Casado, J.L. López-Lacomba, Chitosan film as rhBMP2 carrier: delivery properties for bone tissue application, Biomacromolecules 9 (2007) 711–718. [31] A. Lochmann, H. Nitzsche, S. von Einem, E. Schwarz, K. Mäder, The influence of covalently linked and free polyethylene glycol on the structural and release properties of rhBMP-2 loaded microspheres, J. Control. Release 147 (2010) 92–100. [32] S.S. Kim, K.M. Ahn, M.S. Park, J.H. Lee, C.Y. Choi, B.S. Kim, A poly (lactide-coglycolide)/hydroxyapatite composite scaffold with enhanced osteoconductivity, J. Biomed. Mater. Res. Part A 80 (2007) 206–215. [33] S. Jo, S. Kim, T.H. Cho, E. Shin, S.J. Hwang, I. Noh, Effects of recombinant human bone morphogenic protein-2 and human bone marrow-derived stromal cells on in vivo bone regeneration of chitosan–poly (ethylene oxide) hydrogel, J. Biomed. Mater. Res. Part A 101 (2013) 892–901. [34] T. Nguyen, M. Raupach, L.J. Janik, Fourier-transform infrared study of ethylene glycol monoethyl ether adsorbed on montmorillonite: implications for surface area measurements of clays, Clay Clay Miner. 35 (1987) 60–67. [35] X. Liu, P.X. Ma, Polymeric scaffolds for bone tissue engineering, Ann. Biomed. Eng. 32 (2004) 477–486. [36] D.W. Hutmacher, Scaffolds in tissue engineering bone and cartilage, Biomaterials 21 (2000) 2529–2543. [37] Y.-J. Seol, J.-Y. Lee, Y.-J. Park, Y.-M. Lee, I.-C. Rhyu, S.-J. Lee, et al., Chitosan sponges as tissue engineering scaffolds for bone formation, Biotechnol. Lett. 26 (2004) 1037–1041. [38] V. Karageorgiou, D. Kaplan, Porosity of 3D biomaterial scaffolds and osteogenesis, Biomaterials 26 (2005) 5474–5491. [39] J. Kim, I.S. Kim, T.H. Cho, K.B. Lee, S.J. Hwang, G. Tae, et al., Bone regeneration using hyaluronic acid-based hydrogel with bone morphogenic protein-2 and human mesenchymal stem cells, Biomaterials 28 (2007) 1830–1837. [40] C.-C. Lin, K.S. Anseth, PEG hydrogels for the controlled release of biomolecules in regenerative medicine, Pharm. Res. 26 (2009) 631–643.