Self-setting particle-stabilized emulsion for hard-tissue engineering

Colloids and Surfaces B: Biointerfaces 126 (2015) 394–400

Contents lists available at ScienceDirect

Colloids and Surfaces B: Biointerfaces journal homepage: www.elsevier.com/locate/colsurfb

Self-setting particle-stabilized emulsion for hard-tissue engineering Yasuhiko Iwasaki a,∗ , Yusuke Takahata a , Syuji Fujii b,1 a Department of Chemistry and Materials Engineering, Faculty of Chemistry, Materials and Bioengineering, Kansai University, 3-3-35 Yamate-cho, Suita-shi, Osaka 564-8680, Japan b Department of Applied Chemistry, Faculty of Engineering, Osaka Institute of Technology, 5-16-1 Omiya, Asahi-ku, Osaka 535-8585, Japan

a r t i c l e

i n f o

Article history: Received 3 October 2014 Received in revised form 26 November 2014 Accepted 2 December 2014 Available online 9 December 2014 Keywords: Calcium phosphate cement Particle-stabilized emulsion Polyphosphoesters Self-setting Bone regeneration Tissue engineering

a b s t r a c t Injectable self-setting materials have recently attracted interest for use in minimally invasive medical treatments and tissue engineering. In particular, calcium phosphate cements (CPCs) offer certain specific advantages for the treatment of bone defects. Although the inner structures of set CPCs are important for the apposition and remodeling of new bone, there are still limitations to the design of cements with a well-controlled inner structure. In the present study, we explored self-setting CPCs that generate interconnected macroporous matrices using solid-particle-stabilized emulsion templates. ␣-Tricalcium phosphate (␣-TCP) and poly(ethylene phosphate) sodium salt-coated poly(d,l-lactide-co-glycolide) (PLGA) microparticles were mixed with castor oil and water to form an oil-in-water (o/w) emulsion. The ␣-TCP and PLGA microparticles functioned as an effective particulate emulsifier by adsorption at the oil–water interface. The resulting emulsion spontaneously set in a humidified atmosphere at ambient temperature. The setting behaviors of different emulsions were characterized through X-ray diffraction analysis and compressive-strength measurements. The PLGA microparticles did not hinder the rate of hardening of the emulsions, and they improved the compressive strengths of the set cements. The PLGA particles incorporated within the set cements were hydrolytically degraded, and the degradation of the PLGA particles resulted in the formation of an interconnected pore structure in the set cement. Finally, mouse osteoblastic (MC3T3-E1) cells were cultivated on the set CPCs. The adherent MC3T3-E1 cells adopted a spindle shape, and significant cellular invasion into the set CPCs was observed after degradation of the PLGA microparticles. In conclusion, self-setting emulsions stabilized with ␣-TCP and PLGA microparticles constitute a novel candidate material for bone regeneration. © 2014 Elsevier B.V. All rights reserved.

1. Introduction The aging of the population is an issue of considerable concern worldwide. Bone and joint diseases are the primary causes of chronic pain and disability among elderly people. To improve the quality of life of the elderly, the design of novel biomaterials that can be used for hard-tissue repair is important. Self-setting calcium phosphate cements (CPCs) are currently used to repair bone defects and vertebral fractures because these cements are minimally invasive and have a mineral composition similar to that of hard tissues. Brown and Chow discovered fundamental systems for generating CPCs [1]. These authors used an equimolar mixture of tetracalcium phosphate (TTCP; Ca(PO4 )2 O) and dicalcium

∗ Corresponding author. Tel.: +81 6 6368 0090. E-mail address: [email protected] (Y. Iwasaki). 1 Tel.: +81 6 6954 4274. http://dx.doi.org/10.1016/j.colsurfb.2014.12.003 0927-7765/© 2014 Elsevier B.V. All rights reserved.

phosphate anhydrous (DCPA; CaHPO4 ) or dicalcium phosphate dihydrate (DCPD; CaHPO4 ·H2 O). Several compositions have been proposed for such cements to improve the cements’ properties [2,3]. ␣-Tricalcium phosphate (␣-TCP; Ca3 (PO4 )2 ) is one of the most reliable CPC materials for the production of self-setting singlephase calcium orthophosphate formulations [4,5]. The hydrolysis of ␣-TCP produces calcium-deficient hydroxyapatite (CDHA) through the following reaction (1): 3Ca3 (PO4 )2 + H2 O → Ca9 (HPO4 )(PO4 )5 OH

(1)

This reaction was first described by Monma and Kanazawa [6]. In ␣-TCP-based CPCs, progressive dissolution of the ␣-TCP particles and the simultaneous formation of an entangled network of precipitated CDHA crystals are responsible for the setting and hardening of the cement paste [7]. Several methods have been proposed to improve cellular invasion into cement generated with ␣-TCP [5,8,9]. Ginebra and co-workers have recently summarized various processing approaches for generating macroporous CPCs

Y. Iwasaki et al. / Colloids and Surfaces B: Biointerfaces 126 (2015) 394–400

[10,11]. Although these processes successfully produce interconnected macroporous structures in matrices, they suffer from several disadvantages, such as poor reproducibility, weak strength, and irritation resulting from additives. In the present work, we propose an alternative self-setting CPC that can form an interconnected macroporous structure using a particle-stabilized emulsion template. The formation of particlestabilized emulsions in the absence of any molecular surfactants is widely used in both research and technology [12–14]. The adsorption of solid particles with predictable and well-defined structures (self-assembly) at oil–water interfaces is of interest for a variety of applications [15]. The surface wettability of the particles at the oil–water interface determines the stability and types of the emulsions that are formed. Moreover, the oil droplet size of the emulsions strongly depends on the concentration of the particles. An effective method of preparing porous solid matrices from emulsions stabilized with solid particles has recently been described. Binks [16] prepared macroporous silica materials using emulsion templates stabilized with a binary mixture of hydrophilized and hydrophobized silica particles. Under certain specific conditions, matrices that contained an interconnected pore structure were obtained after evaporation of the solvents. Bismarck and co-workers have also succeeded in preparing high-internal-phase emulsions (HIPEs) and their polymer forms (poly-HIPE) using functionalized titania nanoparticles as stabilizers [17]. In the present study, we found that ␣-TCP microparticles can function as an effective particulate emulsifier by adsorption at the oil–water interface to form a stable oil-in-water (o/w) emulsion. The spontaneous setting of the emulsions in a humidified atmosphere at ambient temperature resulted in the formation of macroporous matrices. The size of the pores in the cement could be controlled by changing the powder/liquid ratio. To generate an interconnected pore structure, biodegradable polymer particles were added to the emulsions as a co-emulsifier. The biodegradable polymers did not exert any adverse effects on the hydrolytic setting reaction of ␣-TCP. Cellular invasion into the CPCs was also evaluated. 2. Materials and methods 2.1. Materials Poly(d,l-lactide-co-glycolide) (PLGA; supporting Fig. S1) (Mw = 7000–17,000, acid-terminated, L:G = 50:50) and 1,8diazabicyclo[5.4.0]undec-7-ene (DBU) were purchased from Sigma–Aldrich (St. Louis, MO, USA). 2-Methoxy-2-oxo-1,3,2dioxaphospholane (MP) was synthesized following a previously described method [18]. ␤-Tricalcium phosphate (␤-TCP) was purchased from the Taihei Chemical Industrial Co., Ltd. (Osaka, Japan). Other chemicals were purchased from Wako Pure Chemical Industries, Ltd. (Osaka, Japan) and were used without further purification. 2.2. Synthesis of poly(ethylene phosphate) sodium salt MP (20 mmol) was placed into a thoroughly dried 30-mL round-bottom flask equipped with a three-way stopcock. After the mixture was dried under reduced pressure for 2 h, methanol and DBU were added as an initiator and a catalyst, respectively, in an argon gas atmosphere with ice cooling. The reaction was allowed to proceed for 3 h. Poly(2-methoxy-2-oxo-1,3,2-dioxaphospholane) (PMP) was purified through reprecipitation from diethyl ether. The PMP (2.5 g) was dissolved in 50 mL of deionized water (dH2 O) and stirred with two equivalents of trimethyl amine relative to the MP unit. After 24 h, 20 g of cation-exchange resin (Amberlite®

395

IR-120, Merck KGaA, Darmstadt, Germany) was added to the polymer solution, followed by stirring for 1 h. This solution was then filtered to remove the cation-exchange resin. The filtrate was dialyzed overnight using a dialysis membrane (MWCO = 1000) in distilled water to remove formic acid. The dialyzed solution was then freeze-dried. The product, poly(ethylene phosphate) (PHP), was obtained as a viscous colorless liquid. To form a sodium salt of the polymer, the PHP was dissolved in an aqueous solution (20 mg/mL), and the pH of the solution was adjusted to 7.0 via the addition of 0.1 N and 0.01 N sodium hydroxide (NaOH). The solution was dialyzed with distilled water for 1 day. Freeze-drying of the aqueous solution afforded the poly(ethylene phosphate) sodium salt (PHP-Na), as shown in supporting Fig. S1. PHP-Na is a polyelectrolyte and is water soluble. 2.3. Preparation of ˛-TCP microparticles Bulk ␣-TCP was synthesized by annealing ␤-TCP for 5 h at 1200 ◦ C, followed by quenching to room temperature [19,20]. The ␣-TCP thus obtained was milled in a planetary ball mill (PM301, Retsch, Hahn, Germany) with stainless balls 2.5 mm in diameter for 1 h at 450 rpm and sieved through a 25-␮m sieve. 2.4. Preparation of PLGA microparticles covered with PHP-Na PLGA particles were prepared using the water-in-oil-in-water (w/o/w) emulsion solvent evaporation method [21]. PLGA (0.2 g) was dissolved in 0.6 g of dichloromethane in a 10-mL glass tube. Then, 60 ␮L of dH2 O was added, and the mixture was vortex mixed for 90 s. Subsequently, 6 mL of a 0.1% aqueous PHP-Na solution was added, and the mixture was emulsified at 4000 rpm for 5 min (IKA® T25 digital Ultra Turrax® , Staufen, Germany) to form a water-in-oil emulsion. The mixture was then added to 94 mL of a 0.1% aqueous PHP-Na solution and 100 mL of a 2% aqueous isopropanol solution and subjected to rapid stirring for 5 h. The particles were allowed to settle for 15 min, and the resulting clear solution was then decanted. The remaining suspension was centrifuged at 1000 rpm for 5 min to collect the particles, and the collected particles were washed three times with dH2 O. Finally, the particles were lyophilized and stored at −30 ◦ C. Rhodamine6G-labeled PLGA particles were prepared by simply replacing 60 ␮L of dH2 O with 60 ␮L of an aqueous rhodamine-6G solution. The size of the PLGA particles was determined via laser diffraction, and Xray photoelectron spectroscopy (XPS; ESCA-3400, Shimadzu Co., Japan) measurements were conducted to confirm the presence of PHP-Na on the surfaces of the particles. 2.5. Preparation of particle-stabilized emulsion ␣-TCP particles (powder), PLGA particles (powder), castor oil and water were vortex mixed and homogenized using a probetype ultrasonic homogenizer (Sonifier® S-250A, Branson, USA) for 5 min to form emulsions. The following conditions were employed: ␣-TCP particles/PLGA particles/castor oil/water (wt%) = 30/0/35/35 (CPC-P0), 30/10/30/30 (CPC-P10), or 30/20/25/25 (CPC-P20). Each emulsion was placed in a 2.5-mL plastic syringe and injected into a silicone rubber ring of 5 mm height and 5 mm diameter. The samples were then allowed to set at 37 ◦ C and 95% humidity. 2.6. Characterization of self-setting particle-stabilized emulsion In the following investigations, the cements were immersed in 2-propanol to remove the castor oil and were subsequently dried under reduced pressure prior to each measurement.

396

Y. Iwasaki et al. / Colloids and Surfaces B: Biointerfaces 126 (2015) 394–400

2.6.1. Phase analysis The hydrolysis of ␣-TCP to CDHA was identified based on X-ray powder diffraction (XRD) patterns (RAPID II, Rigaku Co., Japan; Cu K␣, 125 mA, 40 kV, step time of 60 s). 2.6.2. Mechanical properties The compressive strength was measured using a universal testing machine (Autograph AGS-J, Shimadzu Co., Japan) at a crosshead speed of 0.5 mm/min. 2.6.3. Porosity The percentage porosity in each set substrate was determined using the following equation: Porosity (%) =



1−

M V × 2.79



× 100

(2)

where M and V are the weight and volume of the porous substrate, respectively. The theoretical density of CDHA is 2.79 g/cm3 [22,23]. 2.6.4. Morphology The pore structures of the set substrates were observed via scanning electron microscopy (SEM; Tiny-SEM, Technex Lab Co., Ltd., Japan) before and after thermolysis of the PLGA particles at 600 ◦ C for 3 h. Elemental analyses of the surfaces of the set substrates were performed using an SEM apparatus (JSM-6060LV, JEOL Ltd., Japan) equipped with an energy-dispersive X-ray spectrometer (EDS). 2.7. Degradation of PLGA microparticles incorporated into set CPCs To investigate the degradability of the PLGA particles incorporated into the set emulsions, the samples were immersed in 10 mL of phosphate-buffered saline (PBS) at 37 ◦ C and 95% humidity for 1, 2, 3, 4, 5 and 7 weeks. The medium was changed every 7 days. After a given time period, the samples were immersed in isopropanol for 1 day and then dried under reduced pressure for 1 day. The weights and porosities of the samples were measured. 2.8. In vitro cell-culture studies Mouse osteoblastic cells (MC3T3-E1) were used for the cell culture experiments because they are a well-characterized osteoblast-like cell line that can serve as a model for endogenous osteoblasts [24,25]. The cells were maintained in culture medium (␣-modified minimum essential medium (␣-MEM), Gibco, Invitrogen Corporation, USA) that contained 10% fetal bovine serum (FBS) and 1% antibiotic–antimycotic. The cells were routinely cultured in tissue-culture polystyrene (TCPS) flasks (Asahi Glass Co., Ltd.,

Tokyo, Japan) at 37 ◦ C in a humidified atmosphere of air and 5% CO2 . After treatment with trypsin-EDTA (Gibco, Invitrogen Corporation, USA), the MC3T3-E1 cells were harvested. The prepared emulsions were directly injected into a 24-well plate and then left undisturbed at 37 ◦ C and 95% humidity for 7 days to allow the emulsions to set. After 7 days, the samples were sterilized under UV light for 12 h and pre-conditioned for 3 days in ␣-MEM prior to cell culture. The suspension of MC3T3-E1 cells (1 mL at 1 × 104 cell/mL) in ␣-MEM was introduced into the 24well tissue-culture plate (Thermo Scientific, Rochester, USA) and incubated in culture medium at 37 ◦ C and 5% CO2 . To maintain physiological pH, the medium was changed every 12 h. After cultivation for 24 h, the density of adherent cells was measured using an LDH assay [26]. For morphological observation of adherent cells, the MC3T3 cells were seeded on the cements and cultured for 3 days in a CO2 incubator at 95% humidity. After the medium was aspirated, the cements were rinsed three times with PBS and then placed in a 2.5-vol% glutaraldehyde solution to fix the adherent cells on the cement. The cement was repeatedly rinsed with dH2 O, freeze-dried with liquid nitrogen, sputtered with gold, and examined using a Tiny-SEM. Cellular invasion into the set CPCs was also investigated. Prior to cell cultivation, set CPCs were heat-treated at 600 ◦ C to induce thermolysis of the organic phase. MC3T3-E1 cells (1 mL at 1 × 104 cell/mL) were seeded on the set CPCs, which were prepared as described above, and cultivated for 7 days. The cells that adhered to the CPCs were stained using the LIVE/DEAD® Viability/Cytotoxicity Kit (Molecular Probes, OR, USA). The CPCs were then cut perpendicularly, and cross-sectional images were obtained using a fluorescence microscope.

3. Results and discussion 3.1. Morphologies of ˛-TCP and PLGA microparticles Fig. 1 presents SEM images of milled ␣-TCP and PLGA microparticles. The majority of the milled ␣-TCP particles were less than 10 ␮m in size, and they had an angular shape. By contrast, the PLGA particles had a spherical shape, and the size of the PLGA particles was relatively larger than that of the ␣-TCP particles. XPS surface analysis of the PLGA microparticles indicated that the surfaces of the particles were covered with PHP-Na (data not shown). The morphologies of polymeric microparticles prepared using the w/o/w method have been shown to be critically dependent on the formulation conditions, such as the concentration of polymer [27], the shear conditions [28,29] and the stability of the primary w/o emulsion [30]. In the current study, the use of PHP-Na as a stabilizer in

Fig. 1. SEM images of ␣-TCP and PLGA microparticles.

Y. Iwasaki et al. / Colloids and Surfaces B: Biointerfaces 126 (2015) 394–400

397

Fig. 2. Optical macroscopic and microscopic images of particle-stabilized emulsions prepared under various conditions.

the external aqueous phase resulted in the formation of spherical microparticles. The concentration of PHP-Na and the shear rate for mixing were optimized to obtain PLGA particles of several dozen micrometers in diameter.

3.2. Preparation of emulsion stabilized with ˛-TCP and PLGA microparticles The surface wettability, size and concentration of the solid particles used as stabilizers regulate the formation of particlestabilized emulsions [14]. Fujii and co-workers [31] recently reported the preparation of o/w emulsions using hydroxyapatite (HAp, Ca10 (PO4 )6 (OH)2 ) nanoparticles as a particulate emulsifier in the absence of any molecular surfactant. HAp nanoparticles can be applied for the formation of emulsions of various oils in aqueous media. We therefore considered that ␣-TCP might also serve as a suitable stabilizer for the creation of an o/w emulsion. In the current study, various vegetable oils were tested for the preparation of ␣-TCP-stabilized emulsions. The results of these tests are presented in supporting Fig. S2. In all cases, o/w emulsions were formed, as confirmed by the drop test [30]. No clear correlation was observed between the composition of the oil and emulsion formation. However, vegetable oils with higher surface tensions [32–34] could more effectively form stable emulsions. Among the investigated oils, castor and sunflower seed oils were good candidates for emulsion formation. Castor oil was subsequently selected as the oil phase because castor oil is already used as a therapeutic agent. Additionally, no significant adverse effects of castor oil administration have been noted, as reported by Irwin [35]. The milled ␣-TCP particles were well adsorbed at the interface of the water and the castor oil, and stable o/w emulsions were formed. Optical micrographs of particle-stabilized emulsions prepared using various concentrations of ␣-TCP are presented in supporting Fig. S3. The oil droplet size of the emulsions decreased from several hundreds of micrometers to several tens of micrometers as the concentration of solid particles increased; thus, a larger oil–water interfacial area can be stabilized at higher concentrations of solid particles.

The optimal concentration of ␣-TCP microparticles for emulsion formation was found to be 20–40 wt%. Fig. 2 presents macroscopic and microscopic images of emulsions stabilized by ␣-TCP particles (CPC-P0) or ␣-TCP/PLGA mixed particles (CPC-P10, CPC-P20). The preparation conditions for each emulsion are summarized in Table 1. Even with the addition of PLGA microparticles alone, the formation of an o/w emulsion was clearly observed. Although we hypothesized that the oil droplet size would decrease as a result of adding PLGA microparticles, the oil droplet size of the emulsion did not remarkably decrease after PLGA microparticles were added because the diameter of the microparticles was considerably larger than that of the ␣-TCP microparticles, as shown in Fig. 1. The smaller ␣-TCP microparticles predominantly controlled the droplet sizes of the emulsions. Supporting Fig. S4(A) and (B) present phase-contrast and fluorescence micrographs, respectively, of emulsions stabilized by ␣-TCP/PLGA mixed microparticles (CPC-P20). The PLGA particles were previously stained with rhodamine 6G. The surfaces of the emulsions could be observed using not only an optical microscope but also a fluorescence microscope. This result indicated that the PLGA particles also acted as a stabilizer for the formation of particle-stabilized emulsions, as shown in supporting Fig. S4(C). The surface of the PLGA microparticles was considered to be hydrophilic because of its coverage by PHP-Na, making the particles suitable for the formation of o/w emulsions. The expression of CPC-P20 from a plastic syringe was performed, as shown in supporting Fig. S5. The emulsion was paste-like in nature, and its shape was preserved after expression. Furthermore, the emulsion was easy to handle, making it appropriate for filling bone defects with irregular shapes. Table 1 Preparation conditions for self-setting emulsions.

CPC-P0 CPC-P10 CPC-P20

␣-TCP (wt%)

PLGA (wt%)

H2 O (wt%)

Castor oil (wt%)

30 30 30

0 10 20

35 30 25

35 30 25

398

Y. Iwasaki et al. / Colloids and Surfaces B: Biointerfaces 126 (2015) 394–400

Fig. 3. XRD patterns reflecting the phase evolution during the setting reactions of the emulsions.

3.3. Characterization of self-setting particle-stabilized emulsions Fig. 3 presents the XRD patterns of porous substrates as functions of the ␣-TCP hydration time at 37 ◦ C. The XRD results indicated that the hydrolysis of ␣-TCP into CDHA was slightly slowed by the addition of PLGA particles. Although the phase of ␣-TCP was completely transformed into an apatite phase after 24 h in the absence of PLGA particles, a small signal at 31◦ (2), corresponding to ␣TCP, was still observed when PLGA particles were added. Ginebra [36] and co-workers have reported the effects of polymeric additives bearing carboxyl groups on the hydrolysis of ␣-TCP into CDHA. These additives can interact with the calcium present in the calcium phosphate particles, bonding to them through the carboxylic groups and creating an adsorbed layer on both the ␣-TCP and the HAp particles, which can hinder the progress of the reaction. In the current system, PHP-Na was used as a stabilizer for the PLGA particles. We have previously reported that polyphosphoester ionomers can interact with the calcium present in HAp substrates [37]. In the current study, the lower rate of ␣-TCP hydrolysis is attributable to the chelate-bonding ability of the PHP-Na covering the PLGA particles. Fig. 4 presents the compressive-strength evolution of the three investigated cements as a function of the storage time under humidified conditions. Although the hydrolysis rate of ␣-TCP was slightly slowed by the addition of PLGA microparticles, the hardening rate was similar for every sample, and the compressive strengths of the cements plateaued after 24 h. Furthermore, the compressive strengths of the set substrates increased with an increase in the amount of PLGA particles. This trend can be

Fig. 4. Compressive strengths of CPCs vs. reaction time. (): PCP-P0; ( P10;(䊉): PCP-P20 (SD, n = 3).

): PCP-

attributed to the composition of the solid particles and to the interaction between the PHP-Na coating the PLGA particles and the calcium present in the CDHA substrate. Supporting Fig. S6 presents the distributions of the elements, namely, calcium (Ca) and carbon (C), with respect to the surface morphologies of the set CPCs. Calcium was homogeneously distributed in every cement specimen. By contrast, localization of carbon along the pores was observed for CPC-P10 and CPCP20, which contained PLGA particles. This carbon signal was not detected for CPC-P0, which was prepared without PLGA particles. This result indicated that the pores in the set CPCs were generated by the emulsion templates. Fig. 5 presents the changes in the weights of the cements during soaking in PBS. The weight of CPC-P0 remained stable for 4 weeks. By contrast, CPC-P10 and CPC-P20 gradually became lighter. This weight loss can be attributed to the degradation of the PLGA microparticles because the PLGA particles used in the current study completely degraded after 4 weeks, as shown in supporting Fig. S7. Moreover, the porosities of CPC-P10 and CPC-P20 increased remarkably after soaking in PBS, as shown in supporting Fig. S8. Good correlations between the weight loss and the porosity gain of the cements were observed. Fig. 6 presents SEM images of the set CPCs before and after the thermolysis of the PLGA particles at 600 ◦ C. The pore structure of the ␣-TCP substrate did not change upon thermal treatment at 600 ◦ C, and isolated spherical pores were observed. By contrast, interconnected pores formed in the cements prepared with PLGA particles after thermolysis. Notably, an interconnected

Fig. 5. Changes in the weights of set CPCs after immersion in PBS. (): PCP-P0; ( ): PCP-P10; (䊉): PCP-P20 (SD, n = 3).

Y. Iwasaki et al. / Colloids and Surfaces B: Biointerfaces 126 (2015) 394–400

399

Fig. 6. SEM images of set CPCs before and after thermolysis. The bars represent 200 ␮m.

macroporous structure was observed as the concentration of PLGA particles was increased. 3.4. Cell-culture studies on CPCs The densities of adherent cells are presented in supporting Fig. S9. The density of adherent cells on the set CPC generated from only ␣-TCP was less than that on TCPS. By contrast, the adhesion of cells on the set CPC generated from ␣-TCP and PLGA microparticles was similar to that on TCPS. The cell compatibility of the CPCs was improved by the addition of PLGA. It is well known that cell adhesion on material surfaces strongly depends on the surface properties, such as chemistry, wettability, zeta potential, topology, and stiffness [38–41]. Although the exact factors determining the cell compatibility of composite cements have not yet been clarified,

PLGA particles covered with PHP-Na could be expected to constitute a good substrate for cell adhesion. Indeed, good adhesion of osteoblastic cells has been reported on polyphosphoesters [42,43]. The morphologies of cells cultivated on the set CPCs for 3 days in the present study are presented in supporting Fig. S10. On every substrate, each of the adherent MC3T3 cells adopted a spindle shape. There was no difference among the set CPCs with respect to the shape of the adherent cells. The cells were distributed not only on the surfaces of the CPCs but also inside the pore structures. Finally, we examined the in vitro invasion of MC3T3 cells into the set CPCs. Cross-sectional images are presented in Fig. 7. After cultivation for one week, the cells were stained using the LIVE/DEAD® assay kit. All cells that invaded a set CPC were stained with calcein, but not with EthD-1. The viability of all cells was preserved in the set CPCs. Adherent cells were observed on the surface of the

Fig. 7. Fluorescence micrographs illustrating the invasion of MC3T3-E1 cells into set CPCs after 7 days of cultivation. The bars represent 200 ␮m.

400

Y. Iwasaki et al. / Colloids and Surfaces B: Biointerfaces 126 (2015) 394–400

set CPC prepared from only ␣-TCP. It appeared to be difficult for the adherent cells to migrate into the CPC because of its non-connected pore structure. By contrast, significant cellular invasion into the CPC prepared from ␣-TCP and PLGA microparticles was observed. The invasion depth increased with an increase in the concentration of PLGA microparticles. 4. Conclusions In the current study, self-setting CPCs that generated interconnected macroporous substrates were fabricated using o/w emulsion templates stabilized with ␣-TCP and PLGA microparticles. The sizes of the pores and the degree of porosity could be optimized by changing the microparticle concentrations because the oil droplet size of the emulsion was determined by the microparticles. The biodegradable PLGA microparticles resulted in the formation of an interconnected macroporous structure in the set cements. Extensive cellular invasion was observed in the CPCs that possessed an interconnected macroporous structure. Self-setting CPCs generated using an ␣-TCP/PLGA microparticle-stabilized emulsion template demonstrate good potential for use as matrices in hardtissue engineering. Acknowledgments This work was supported by a Kansai University Grant-in-Aid for Progress of Research in a Graduate Course, 2014 and by JSPS KAKENHI Grant Number 24500532 from the Ministry of Education, Culture, Sports, Science, and Technology of Japan. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.colsurfb. 2014.12.003. References [1] [2] [3] [4]

W.E. Brown, L.C. Chow, J. Dent. Res. 62 (1983) 672. S.V. Dorozhkin, J. Mater. Sci. 43 (2008) 3028. M. Bohner, Eur. Cell Mater. 20 (2010) 1. M.P. Ginebra, E. Fernández, F.C.M. Driessens, J.A. Planell, J. Am. Ceram. Soc. 82 (1999) 2808.

[5] A. Almirall, G. Larrecq, J.A. Delgado, S. Martínez, J.A. Planell, M.P. Ginebra, Biomaterials 25 (2004) 3671. [6] H. Monma, T. Kanazawa, J. Ceram. Soc. Jpn. 84 (1976) 209. [7] M.P. Ginebra, E. Fernandez, E.A.P. DeMaeyer, R.M.H. Verbeeck, M.G. Boltong, J. Ginebra, F.C.M. Driessens, J.A. Planell, J. Dent. Res. 76 (1997) 905. [8] E.B. Montufar, T. Traykova, C. Gil, I. Harr, A. Almirall, A. Aguirre, E. Engel, J.A. Planell, M.P. Ginebra, Acta Biomater. 6 (2010) 876. [9] R.P. Félix Lanao, S.C. Leeuwenburgh, J.G. Wolke, J.A. Jansen, Biomaterials 32 (2011) 8839. [10] A.R. Studart, U.T. Gonzenbach, E. Tervoort, L.J. Gauckler, J. Am. Ceram. Soc. 89 (2006) 1771. [11] M.P. Ginebra, M. Espanol, E.B. Montufar, R.A. Perez, G. Mestres, Acta Biomater. 6 (2010) 2863. [12] W. Ramsden, Proc. R. Soc. Lond. 72 (1903) 156. [13] S.U. Pickering, J. Chem. Soc. 91 (1907) 2001. [14] R. Aveyard, B.P. Binks, J.H. Clint, Adv. Colloid Interface Sci. 100–102 (2003) 503. [15] B.P. Binks, T.S. Horozov (Eds.), Colloidal Particles at Liquid Interfaces, Cambridge University Press, Cambridge, 2006. [16] B.P. Binks, Adv. Mater. 14 (2002) 1824. [17] A. Menner, V. Ikem, M. Salgueiro, M.S. Shaffer, A. Bismarck, Chem. Commun. 41 (2007) 4274. [18] Y. Iwasaki, E. Yamaguchi, Macromolecules 43 (2010) 2664. [19] R.G. Carrodeguas, S. De Aza, Acta Biomater. 7 (2011) 3536. [20] M. Bohner, Injury 31 (2000) 37. [21] H. Liao, X.F. Walboomers, W.J. Habraken, Z. Zhang, Y. Li, D.W. Grijpma, A.G. Mikos, J.G. Wolke, J.A. Jansen, Acta Biomater. 7 (2011) 1752. [22] H. Monma, S. Ueno, T. Kanazawa, J. Chem. Technol. Biotechnol. 31 (1981) 15–24. [23] C. Durucan, P.W. Brown, J. Mater. Sci. Mater. Med. 11 (2000) 365. [24] R.T. Franceschi, B.S. Iyer, Y. Cui, J. Bone Miner. Res. 9 (1994) 843. [25] D. Wang, K. Christensen, K. Chawla, G. Xiao, P.H. Krebsbach, R.T. Franceschi, J. Bone Miner. Res. 14 (1999) 893. [26] Y. Iwasaki, U. Takami, Y. Shinohara, K. Kurita, K. Akiyoshi, Biomacromolecules 8 (2007) 2788. [27] H. Jeffery, S.S. Davis, D.T. O’Hagan, Pharm. Res. 10 (1993) 362. [28] H.K. Sah, R. Toddywala, Y.W. Chien, J. Microencapsul. 12 (1995) 59. [29] C. Yana, J.H. Resaub, J. Hewetsona, M. Westa, W.L. Rilla, M. Kende, J. Control. Release 32 (1994) 231. [30] N. Nihant, C. Schugens, C. Grandfils, R. Jérôme, P. Teyssié, Pharm. Res. 11 (1994) 1479. [31] S. Fujii, M. Okada, T. Furuzono, J. Colloid Interface Sci. 315 (2007) 287. [32] M.E. Stolar, J. Soc. Cosmet. Chem. 17 (1966) 607. [33] T. Stadler, A. Fornes, M. Buteler, Bull. Insectol. 58 (2005) 57. [34] B. Esteban, J.–R. Riba, G. Baquero, R. Puig, A. Rius, Fuel 102 (2010) 231. [35] R. Irwin, Toxic Rep. Ser. 12 (1982), 1-B5. [36] M.P. Ginebra, A. Rilliard, E. Fernández, C. Elvira, J. San Román, J.A. Planell, J. Biomed. Mater. Res. 57 (2001) 113. [37] Y. Iwasaki, K. Katayama, M. Yoshida, M. Yamamoto, Y. Tabata, J. Biomater. Sci. Polym. Ed. 24 (2013) 882. [38] P. Roach, D. Eglin, K. Rohde, C.C. Perry, J. Mater. Sci. Mater. Med. 18 (2007) 1263. [39] Z. Ma, Z. Mao, C. Gao, Colloids Surf. B: Biointerfaces 60 (2007) 137. [40] W. Zheng, W. Zhang, X. Jiang, Adv. Healthc. Mater. 2 (2013) 95. [41] X. Yao, R. Peng, J. Ding, Adv. Mater. 25 (2013) 5257. [42] C. Wachiralarpphaithoon, Y. Iwasaki, K. Akiyoshi, Biomaterials 28 (2007) 984. [43] X.Z. Yang, T.M. Sun, S. Dou, J. Wu, Y.C. Wang, J. Wang, Biomacromolecules 10 (2009) 2213.