- Email: [email protected]
polymer composite scaffolds for bone tissue engineering
ARTICLE IN PRESS
Biomaterials 25 (2004) 4749–4757
Structure and properties of nano-hydroxyapatite/polymer composite scaffolds for bone tissue engineering Guobao Weia, Peter X. Maa,b,c,* a Department of Biomedical Engineering, University of Michigan, Ann Arbor, MI 48109-2009, USA Department of Biologic and Materials Sciences, University of Michigan, Ann Arbor, MI 48109-1078, USA c Macromolecular Science and Engineering Center, University of Michigan, Ann Arbor, MI 48109-1055, USA b
Received 9 June 2003; accepted 11 November 2003
Abstract To better mimic the mineral component and the microstructure of natural bone, novel nano-hydroxyapatite (NHAP)/polymer composite scaffolds with high porosity and well-controlled pore architectures were prepared using thermally induced phase separation (TIPS) techniques. The morphologies, mechanical properties and protein adsorption capacities of the composite scaffolds were investigated. The high porosity (90% and above) was easily achieved and the pore size was adjusted by varying phase separation parameters. The NHAP particles were dispersed in the pore walls of the scaffolds and bound to the polymer very well. NHAP/polymer scaffolds prepared using pure solvent system had a regular anisotropic but open 3D pore structure similar to plain polymer scaffolds while micro-hydroxyapatite (MHAP)/polymer scaffolds had a random irregular pore structure. The introduction of HAP greatly increased the mechanical properties and improved the protein adsorption capacity. In a dioxane/water mixture solvent system, NHAP-incorporated poly(l-lactic acid) (PLLA) scaffolds developed a fibrous morphology which in turn increased the protein adsorption three fold over non fibrous scaffolds. The results suggest that the newly developed NHAP/polymer composite scaffolds may serve as an excellent 3D substrate for cell attachment and migration in bone tissue engineering. r 2003 Elsevier Ltd. All rights reserved. Keywords: Nano; Hydroxyapatite; Phase separation; Scaffold; Protein adsorption
1. Introduction There is a growing need for bone regeneration due to various clinical bone diseases such as bone infections, bone tumors and bone loss by trauma [1]. Current therapies for bone defects include autografts, allografts, xenografts and other artificial substitutes such as metals, synthetic cements and bioceramics [2,3]. However, these substitutes are far from ideal and each has its specific problems and limitations. For example, autografts are associated with donor shortage and donor site morbidity whereas allografts and xenografts have the risk of disease transmission and immune response [4], and synthetic materials wear and do not behave like true bone. As a result, recent research has been devoted to bone tissue engineering in which a 3D porous scaffold is loaded with specific living cells and/or tissue-inducing *Tel.: +1-734-764-2209; fax: +1-734-647-2110. E-mail address: [email protected] (P.X. Ma). 0142-9612/$ - see front matter r 2003 Elsevier Ltd. All rights reserved. doi:10.1016/j.biomaterials.2003.12.005
factors to launch a tissue regeneration or replacement in a natural way [5,6]. In the tissue engineering approach, the temporary 3D scaffold serves an important role in the manipulation of the functions of osteoblasts and a central role in the guidance of new bone formation into desired shapes. In principle, a biodegradable matrix with sufficient mechanical strength, optimized architecture and suitable degradation rate, which could finally be replaced by newly formed bone, is most desirable. The scaffolding materials for bone tissue engineering should also be osteoconductive so that osteoprogenitor cells can adhere and migrate on the scaffolds, differentiate, and finally form new bone [7,8]. Biodegradable polymers, mainly polyesters such as poly(lactic acid) (PLA), poly(glycolic acid) (PGA) and their copolymers (PLGA), have been widely used to develop porous 3-D scaffolds using various fabrication techniques [8–14]. These synthetic materials have been demonstrated to be biocompatible and degrade into non-toxic components with a
ARTICLE IN PRESS 4750
G. Wei, P.X. Ma / Biomaterials 25 (2004) 4749–4757
controllable degradation rate in vivo [15]. Another major class of biomaterials for bone repair is ceramics such as hydroxyapatite (HAP) and tricalcium phosphate (TCP) [16,17]. Being similar to the mineral component of natural bone, they showed good osteoconductivity and bone bonding ability [18]. However, the main limitation for the use of HAP ceramics was their inherent brittleness and difficulty for processing [19]. To combine the osteoconductivity of calcium phosphates and good biodegradability of polyesters, polymer/ceramic composite scaffolds have been developed for bone tissue engineering either by direct mixing or by a biomimetic approach [7,8,20]. Compared to plain polymer scaffolds in which neo tissue matrix was formed only in the surface layer (o240 mm) [21], the composite scaffolds supported cells growth and neo tissue formation throughout the scaffold including in the very center of the scaffold [6]. Polymer/ceramic composite scaffolds mimic the natural bone to some extent. Natural bone is composed of inorganic compound (mainly partially carbonated HAP on the nanometer scale) and organic compounds (mainly collagen). The nanometer size of the inorganic component (mainly bone-like apatite) in natural bone is considered to be important for the mechanical properties of the bone [22]. Recent research in this field also suggested that better osteoconductivity would be achieved if synthetic HAP could resemble bone minerals more in composition, size and morphology [23,24]. In addition, nano-sized HAP may have other special properties due to its small size and huge specific surface area. Webster et al. [25,26] have shown significant increase in protein adsorption and osteoblast adhesion on the nano-sized ceramic materials compared to traditional micron-sized ceramic materials. In previous works, the Ma group has developed a variety of scaffolds using thermally induced phase separation (TIPS) [8–10]. Both pore structure and pore wall morphology can be controlled by phase separation parameters. They have also demonstrated that the addition of micron-sized hydroxyapatite (MHAP) increase the adsorption of proteins and extracellular matrix (ECM) components [27]. The cell seeding uniformity into the scaffold was improved substantially due to the introduction of osteoconductive hydroxyapatite [6]. The MHAP/PLLA composite scaffolds were shown to suppress apoptosis of osteoblasts by a possible mechanism of enhanced adsorption of serum proteins such as vitronectin and fibronectin onto the scaffolds [27, 28]. In this paper, we further mimicked the size scale of hydroxyapatite in natural bone and aim to fabricate novel and improved composite scaffolds. The pore structure, pore wall morphology, mechanical properties and protein adsorption capacity were systematically investigated.
2. Materials and methods 2.1. Materials Poly(l-lactic acid) (PLLA) with an inherent viscosity of 1.4–1.8 dl/g was purchased from Boehringer Ingelheim (Ingelheim, Germany) and was used as received. PLGA85 (Medisorbs, LA:GA=85:15) with an inherent viscosity of 0.6 was obtained from Alkermes Inc. (Wilmington, OH). NHAP were purchased from Berkeley Advanced Biomaterials Inc. (San Leandro, CA). Conventional HAP particles (MHAP, particle size on the micrometer scale), tetrahydrofuran (THF), dioxane and benzene were obtained from aldrich chemical company (Milwaukee, WI). 2.2. Fabrication method NHAP/PLLA composite scaffolds were fabricated using a phase separation technique similar to that reported previously [8]. Briefly, the HAP nanocrystal powder was dispersed in a solvent by sonication for 30 s at 15 W (Virsonic 100, Cardiner, NY). After that, PLLA was dissolved in the NHAP suspended solvent at about 60 C to make homogeneous solutions with desired concentrations. 2.5 ml polymer/NHAP mixture was added into a Teflon vial, sonicated again and then transferred into a freezer at a preset temperature to induce solid–liquid or liquid–liquid phase separation. The solidified mixture was maintained at that temperature overnight and then transferred into a freeze-drying container which was maintained at a temperature between 5 C and 10 C by salt–ice bath for freezedrying. The samples were freeze-dried at 0.5 mmHg for 7 days to remove solvent. The final composition of the composite foam was determined by the concentration of the polymer solution and NHAP content in the mixture. 2.3. SEM characterization The porous morphologies of the scaffold were examined with scanning electron microscopy (SEM) (S-3200N, Hitachi, Japan) at 20 or 15 kV. The samples were cut by a razor blade or fractured after being frozen in liquid nitrogen for several minutes and then were coated with gold for 200 s using a sputter coater (DeskII, Denton Vacuum Inc.). 2.4. Porosity The melting behavior of the matrices was characterized with a differential scanning calorimeter (DSC-7, PerkinElmer, Norwalk, CT) as detailed previously [9]. The degree of crystallinity, Xc, of a sample was calculated as 0 Xc ¼ DHm =DHm ;
ARTICLE IN PRESS G. Wei, P.X. Ma / Biomaterials 25 (2004) 4749–4757
where DHm was the measured enthalpy of melting and 0 DHm was the enthalpy of melting for 100% crystalline 0 polymer. For PLLA, DHm ¼ 203:4 J=g: The estimated density and porosity of the matrix were obtained as follows. The diameter and height of the matrix were measured to calculate the volume. The mass of the matrix was measured with an analytical balance. The density was calculated from the volume and mass. The porosity, e; was calculated from the measured overall densities Dm of the matrix and the skeletal density Ds. For a composite scaffold, the skeletal density was determined using the densities of the polymer and HAP powder. The porosity was given by Ds Dm e¼ ; Ds where Ds was calculated from the following formula: 1 ; Ds ¼ ð1 Xh Þ=Dp þ Xh =Dh where Dh is the density of the HAP powder with a value of 3.16 g/ml and Xh is the percentage of HAP in the composite while Dp is the density of the polymer. Dp was calculated from 1 Dp ¼ ; ð1 Xc Þ=Da þ Xc =Dc where Xc was the degree of crystallinity of the polymer. The density of amorphous PLLA (Da) is 1.248 g/ml and the density of 100% crystalline PLLA (Dc) is 1.290 g/ml. 2.5. Mechanical properties The compressive mechanical properties of the scaffolds were tested using a MTS Synergie 200 mechanical tester (MTS systems Co. Eden Prairie, Minnesota). The scaffolds have dimensions of about 16 mm in diameter and 3 mm in thickness. A crosshead speed of 0.5 mm/ min was used. The compressive modulus was defined as the initial linear modulus. At least five specimens were tested for each sample. The averages and standard deviations were graphed. A two-tail Student’s t-test (assuming equal variances) was performed to determine the statistical significance (po0.05) of the differences in mechanical properties. 2.6. Protein adsorption Protein adsorption was performed by incubating the scaffolds in phosphate buffered saline (PBS, 0.1 m, pH=7.4) containing 2.5% or 5% fetal bovine serum (FBS). The disk-like specimens with dimensions of 15.5– 16.5 mm in diameter and 370.1 mm in thickness were used. Before incubation in the medium containing proteins, the specimens were pretreated by ethanol for 30 min and then washed by PBS three times overnight under gentle shaking. Specimens were then put into 24-
4751
well culture plates (one specimen for each well) and 1.5 ml FBS/PBS solution was added into each well. The scaffolds were incubated at 37 C for a chosen incubation time. The concentration of the protein in the FBS/ PBS solution was then quantified with a commercial protein assay kit, BCA (Pierce, Rockford, IL), using bovine serum albumin (BSA) standards. The amount of absorbed proteins was determined by subtracting the amount of proteins left in the FBS/PBS solution after adsorption from the amount of proteins in control FBS/ PBS solution (without specimen) under the same incubation conditions.
3. Results 3.1. Porosity NHAP/polymer composite scaffolds with high porosity have been fabricated using thermally induced phase separation techniques (Tables 1 and 2). Two solvent systems were used to obtain composite scaffolds with different pore structures. One is a mono-solvent system of pure dioxane or benzene and the other, a solvent/nonsolvent mixture system of dioxane and water in different proportions. The densities and porosities of scaffolds are listed in Tables 1 and 2. Incorporation of nano-sized HAP only decreased the porosity slightly. This was especially true when NHAP content was low. Regardless of detailed morphology and composition, all scaffolds have a high porosity of at least 89%, which was considered to be beneficial for cell ingrowth and survival. The solvent system did not have much effect on the density and porosity. The quenching temperature did not affect the porosity significantly either (p=0.12). 3.2. Morphology The morphology and microstructure of the scaffolds were examined using SEM (Fig. 1): micrographs of PLLA scaffolds (a,b), MHAP/PLLA (c,d) and NHAP/ PLLA (e–h) composite scaffolds fabricated with pure
Table 1 Densities and porosities of NHAP/PLLA composite scaffolds prepared from dioxane solutions, phase separated at 18 C Scaffolds
PLLA:HAP (w/w)
Dp (g/cm3)
Ds (g/cm3)
Porosity (%)
PLLA PLLA–NHAP PLLA–NHAP PLLA–NHAP PLLA–NHAP PLLA–MHAP
100:0 90:10 70:30 50:50 30:70 50:50
1.2577 1.3383 1.5394 1.7993 2.1737 1.7993
0.08812 0.09708 0.1176 0.1475 0.2304 0.1505
93.0 92.8 92.3 91.8 89.4 91.6
ARTICLE IN PRESS G. Wei, P.X. Ma / Biomaterials 25 (2004) 4749–4757
4752
Table 2 Porosities of NHAP/PLLA scaffolds fabricated using dioxane/water mixture solvent system and phase separated at various temperatures Scaffolds
PLLA:HAP (w/w)
Dioxane:H2O (v/v)
Quenching temp. ( C)
Porosity (%)
PLLA–NHAP PLLA–NHAP PLLA–NHAP PLLA–NHAP PLLA–NHAP PLLA–NHAP PLLA–NHAP
70:30 70:30 70:30 70:30 70:30 70:30 70:30
100:0 95:5 90:10 87:13 87:13 87:13 87:13
18 18 18 4 18 70 196 (LN2)
92.3 91.6 93.1 94.5 93.1 92.8 92.5
dioxane or benzene. NHAP/PLLA composite scaffolds maintained a regular internal ladder-like pore structure similar to plain PLLA scaffolds, a typical morphology formed by solid–liquid phase separation [8,10]. By controlling the heat transfer direction, NHAP/PLLA scaffolds with tubular pore architecture were also obtained using benzene as a solvent. Nano-sized HAP particles were distributed within the pore walls of the scaffolds and no large aggregates appeared in pores. In MHAP/PLLA scaffolds, the HAP platelets with size ranging from 10 to 50 mm were found randomly distributed in the PLLA matrix. Some were embedded in the pore wall and some others were piled together between pores or in the pores. The pore size of all these scaffolds ranged from 50 to 100 mm. NHAP/PLGA85 composite scaffolds prepared with pure dioxane had leaf-like morphology and well-interconnected pore structures (Fig. 2). Very different morphologies appeared in composite scaffolds fabricated with dioxane and water mixture solvent system (Fig. 3). The proportion of water in the solvent systems had a significant effect on the pore size and overall morphology of the scaffold. Even a small amount of water (less than 5%) in the solvent system significantly reduced the pore size from 100 mm to about 10 mm and random pore structure replaced the regular ladder-like pore structure (Figs. 3a, b and c). Another obvious change was the pore wall surface texture. Unlike relative solid and smooth pore wall in NHAP/ PLLA scaffold prepared from mono-solvent system, fibrous and loose network became the characteristics of the pore structure of the scaffolds prepared from the mixed solvent system. When the water volume was 13% in the mixture solvent, the scaffolds consisted of three ranges of pore sizes. Large macropores were developed in a range of 200–500 mm while much smaller micropores (tens of microns) between the pore walls of the large pores. The pore wall surface was not smooth, and appeared as loose fibrous structure with several microns in between. NHAP particles were entrapped in the fibrous network of the porous structures and had little effect on the changes in morphology with different water proportions since plain PLLA scaffolds had the same morphological changes (data not shown).
3.3. Mechanical properties The compressive modulus of NHAP/PLLA scaffolds increased with NHAP content (Fig. 4). The plain PLLA scaffolds prepared using solid–liquid phase separation techniques had a compressive modulus of 4.3 MPa. The addition of NHAP into the scaffold improved the mechanical properties. The modulus increased significantly when the NHAP proportion reached 30% of the composite and reached 8.3 MPa when the ratio of NHAP to PLLA was 50:50. The solvent system had a large effect on the mechanical properties of the porous scaffolds (Fig. 5). The introduction of a non-solvent component (in this case, water) in the polymer/HAP/ solvent mixture likely induced liquid–liquid phase separation instead of solid–liquid phase separation. Since regular and orientated pore architectures were replaced by random pore structures in scaffolds fabricated using mixed solvent system, the compressive modulus decreased significantly. However, the modulus recovered substantially when dioxane/water (87:13, v/v) system was used. This might have resulted from the uniform structure of the scaffolds (Figs. 3 d–f). 3.4. Protein adsorption NHAP/PLLA composite scaffolds with different amounts of NHAP were incubated in FBS/PBS solution to investigate protein adsorption. For all NHAP/PLLA composite scaffolds, the protein adsorption reached equilibrium in 25 h and there was no significant increase of protein adsorption during further incubation (Fig. 6a). The addition of NHAP increased the protein adsorption. NHAP/PLLA 50:50 and NHAP/PLLA 70:30 composite scaffolds absorbed 2.4- and 3.2-fold, respectively, more serum proteins than plain PLLA scaffold (Fig. 6b). The size of HAP particles also affected protein adsorption onto the scaffolds containing a high percentage of HAP. Serum protein adsorption was significantly greater on NHAP/PLLA scaffolds than on MHAP/PLLA scaffolds when the HAP/PLLA was 50:50 or higher (po0.05, Fig. 6c). However, when the ratio of HAP in the composite was lower than 50%, the differences in the protein adsorption between NHAP
ARTICLE IN PRESS G. Wei, P.X. Ma / Biomaterials 25 (2004) 4749–4757
4753
Fig. 1. SEM micrographs of plain PLLA, NHAP/PLLA and MHAP/PLLA scaffolds fabricated from 5% dioxane (a–e) and benzene (f,g) solution. (a,b) Pure PLLA scaffold, 50, 400; (c,d) MHAP/PLLA 50:50 scaffold, 100, 500; (e,f) NHAP/PLLA 50:50 scaffold, 100, 1000; (g) Tubular NHAP/PLLA scaffold, cross-section, 200; (h) Tubular NHAP/PLLA scaffold, longitudinal section, 100.
ARTICLE IN PRESS 4754
G. Wei, P.X. Ma / Biomaterials 25 (2004) 4749–4757
Fig. 2. SEM micrographs of NHAP/PLGA85 (30:70) scaffolds fabricated from 10% dioxane solution quenched at 70 C. (a) 150 and (b) 500.
and MHAP incorporated scaffolds were not statistically significant. With the dramatic changes in morphology of scaffolds prepared using different solvent systems, the protein adsorption also changed greatly. Scaffolds with fibrous texture absorbed about four times greater amount of protein than those with smooth pore wall surface. These results suggested that the protein adsorption on the scaffolds could be regulated by the content of HAP, its size scale, and the pore wall morphology of the scaffolds.
4. Discussion The present studies described a phase separation technique to fabricate polymer/NHAP-composite scaffolds with high porosity and controlled pore architecture. Different solvent systems were used to obtain scaffolds with different microarchitectures and properties. When dioxane was used alone, the porous structure resulted from a solid–liquid phase separation of the polymer solution. During quenching, the solvent crystallized and the polymer was expelled from the solvent crystallization front. Solvent crystals became pores after subsequent sublimation. The characteristics of the pores were determined by the morphologies of solvent crystals under such quenching conditions. The temperature gradient along the solvent crystallization direction led to anisotropic structure. The introduction of HAP particles into the polymer solution perturbed the solvent crystallization to some extent and thereby made the pore structure more irregular and isotropic. As expected, MHAP/PLLA composite scaffold exhibited an isotropic and irregular pore structure. The perturbation by NHAP particles, however, was small even in high proportion up to 50% due to their nanometer size scale and uniform distribution. As a result, NHAP/PLLA composite scaffolds maintained the main characteristic
pore architecture of solid–liquid phase separation which was anisotropic and regular. In contrast to NHAP/ PLLA, the regular anisotropic pore structure was obtained only when the HAP content was very low in MHAP/PLLA scaffolds, e.g., MHAP/PLLAo10/90 (images not shown). In this case, low content of HAP did not affect the solvent crystallization significantly enough to alter the pore structure. When mixed solvent (dioxane/water) was used to prepare 5% PLLA solutions, a liquid–liquid phase separation was induced and this process dominated the final pore structure of the system [29], resulting in random and interconnected pore architecture. The pore size was also much smaller. With the increase of water content in the system, larger quench depth allowed the system to continue on a coarsening process to further reduce the interfacial free energy between two phases separated in earlier stage [30]. The coarsening effect might have resulted in the formation of macropores in addition to the micropores (Figs. 3d–f). It is therefore possible to optimize various phase separation parameters to achieve a macroporous structure in composite scaffold. The mechanism of fibrous texture formation is not fully understood. Since PLLA is a semicrystalline polymer, the crystallization in the polymer-rich phase may present a possible explanation for the formation of the fibrous structure [9]. Most mammalian cells are anchorage-dependent cells and they need a biocompatible substrate for attachment, migration and differentiation to form new tissues. Recent research demonstrated that cell adhesion and survival could be modulated by protein pre-adsorption on the substrate [25–28]. Therefore, protein adsorption is of importance in evaluating a scaffold for tissue engineering. In this paper, we have shown that the addition of hydroxyapatite particles into the composite scaffolds greatly increased protein adsorption. In addition to the high affinity of hydroxyapatite itself for
ARTICLE IN PRESS G. Wei, P.X. Ma / Biomaterials 25 (2004) 4749–4757
4755
Fig. 3. SEM micrographs of NHAP/PLLA (30:70) scaffolds fabricated using dioxane/water mixture solvents. (a) Dioxane:water=95:5, 500. (b,c) Dioxane:water=90:10, 500, 8000. (d,e,f) Dioxane:water=87:13, 45, 500, 10000.
proteins, incorporation of HAP also altered the pore surface morphology of the composite scaffolds and may have made them more suitable for protein adsorption. Furthermore, NHAP/PLLA scaffold absorbed significantly more proteins than MHAP/PLLA at high HAP content but not at low HAP content. It was likely that most particles were embedded in the polymer pore walls and the effect of HAP size was not evident when the HAP content was low in the composite. With the increase of HAP content, more and more particles were exposed on the surfaces of the pore walls. This phenomenon was most evident when the scaffold
exhibited fibrous morphology. Fibrous scaffolds had much greater surface to volume ratio than scaffolds with solid pore walls, which might have further increased the protein adsorption capacity.
5. Conclusion The present study investigated new NHAP/polymer composite scaffolds developed using thermally induced phase separation techniques. Nano-sized HAP particles were successfully incorporated into the PLLA porous
ARTICLE IN PRESS 4756
G. Wei, P.X. Ma / Biomaterials 25 (2004) 4749–4757
scaffolds. The incorporation of NHAP improved the mechanical properties and protein adsorption of the composite scaffolds while maintaining high porosity and
Fig. 4. Effects of NHAP content on the compressive modulus of NHAP/PLLA scaffolds fabricated from 5% dioxane solutions % po0.05, wp=0.29, n=5.
Fig. 5. Effects of mixture solvent composition and quenching temperature on the compressive moduli of NHAP/PLLA scaffolds fabricated from 5% dioxane/water solvent systems.
Fig. 6. Protein adsorption on porous composite scaffolds incubated in FBS/PBS solution. (a) as a function of time, 2.5% FBS/PBS, labels indicate the w/w ratios of NHAP to PLLA; (b) as a function of NHAP content, 2.5% FBS; (c) Comparison of MHAP and NHAP, 5.0% % FBS/PBS. po0.05, wp=0.71, n=3. (d) Comparison of NHAP/PLLA (30:70) scaffolds prepared using different solvent systems, 2.5% % FBS/PBS. po0.05, n=3.
suitable microarchitecture. These results suggest that the newly developed NHAP/polymer composite scaffolds may be superior for bone tissue engineering.
ARTICLE IN PRESS G. Wei, P.X. Ma / Biomaterials 25 (2004) 4749–4757
References [1] Braddock M, Houston P, Campbell C, Ashcroft P. Born again bone: tissue engineering for bone repair. News Physiol Sci 2001;16:208–13. [2] Goldberg VM, Stevenson S. Natural history of autografts and allografts. Clin Orthop 1987;225:7–16. [3] Costantino PD, Friedman CD. Synthetic bone graft substitutes. Otolaryngol Clin North Am 1994;27:1037–74. [4] Bonfiglio M, Jeter WS. Immunological responses to bone. Clin Orthop 1972;87:19–27. [5] Langer R, Vacanti JP. Tissue engineering. Science 1993;260:920–6. [6] Ma PX, Zhang R, Xiao G, Franceschi R. Engineering new bone tissue in vitro on highly porous poly(alpha-hydroxyl acids)/ hydroxyapatite composite scaffolds. J Biomed Mater Res 2001;54:284–93. [7] Zhang R, Ma PX. Porous poly(l-lactic acid)/apatite composites created by biomimetic process. J Biomed Mater Res 1999;45:285–93. [8] Zhang R, Ma PX. Poly(alpha-hydroxyl acids)/hydroxyapatite porous composites for bone tissue engineering. I. Preparation and morphology. J Biomed Mater Res 1999;44:446–55. [9] Ma PX, Zhang R. Synthetic nano-scale fibrous extracellular matrix. J Biomed Mater Res 1999;46:60–72. [10] Ma PX, Zhang R. Micro-tubular architecture of biodegradable polymer scaffolds. J Biomed Mater Res 2001;56:469–77. [11] Mikos AG, Thorsen AJ, Czerwonka LA, Bao Y, Langer R, Winslow DN, Vacanti JP. Preparation and characterization of poly(l-lactic acid) foams. Polymer 1994;35:1068–77. [12] Whang K, Thomas CH, Healy KE, Nuber G. A novel method to fabricate bioabsorbable scaffolds. Polymer 1995;36:837–42. [13] Mooney DJ, Baldwin DF, Suh NP, Vacanti JP, Langer R. Novel approach to fabricate porous sponges of poly(d,l-lactic-coglycolic acid) without the use of organic solvents. Biomaterials 1996;17:1417–22. [14] Borden M, Attawia M, Laurencin CT. The sintered microsphere matrix for bone tissue engineering: in vitro osteoconductivity studies. J Biomed Mater Res 2002;61:421–9. [15] Visscher GE, Robison RL, Maulding HV, Fong JW, Pearson JE, Argentieri GJ. Biodegradation of and tissue reaction to 50:50 poly(DL-lactide-co-glycolide) microcapsules. J Biomed Mater Res 1985;19:349–65. [16] Li SH, De Wijn JR, Layrolle P, de Groot K. Synthesis of macroporous hydroxyapatite scaffolds for bone tissue engineering. J Biomed Mater Res 2002;61:109–20.
4757
[17] Flautre B, Descamps M, Delecourt C, Blary MC, Hardouin P. Porous HA ceramic for bone replacement: role of the pores and interconnections—experimental study in the rabbit. J Mater SciMater Med 2001;12:679–82. [18] LeGeros RZ. Properties of osteoconductive biomaterials: calcium phosphates. Clin Orthop 2002;395:81–98. [19] Cooke FW. Ceramics in orthopedic surgery. Clin Orthop 1992;276:135–46. [20] Thomson RC, Yaszemski MJ, Powers JM, Mikos AG. Hydroxyapatite fiber reinforced poly(alpha-hydroxy ester) foams for bone regeneration. Biomaterials 1998;19:1935–43. [21] Ishaug SL, Crane GM, Miller MJ, Yasko AW, Yaszemski MJ, Mikos AG. Bone formation by three-dimensional stromal osteoblast culture in biodegradable polymer scaffolds. J Biomed Mater Res 1997;36:17–28. [22] Rho JY, Kuhn-Spearing L, Zioupos P. Mechanical properties and the hierarchical structure of bone. Med Eng Phys 1998;20: 92–102. [23] Du C, Cui FZ, Zhu XD, de Groot K. Three-dimensional nanoHAp/collagen matrix loading with osteogenic cells in organ culture. J Biomed Mater Res 1999;44:407–15. [24] Du C, Cui FZ, Feng QL, Zhu XD, de Groot K. Tissue response to nano-hydroxyapatite/collagen composite implants in marrow cavity. J Biomed Mater Res 1998;42:540–8. [25] Webster TJ, Ergun C, Doremus RH, Siegel RW, Bizios R. Specific proteins mediate enhanced osteoblast adhesion on nanophase ceramics. J Biomed Mater Res 2000;51:475–83. [26] Webster TJ, Siegel RW, Bizios R. Osteoblast adhesion on nanophase ceramics. Biomaterials 1999;20:1221–7. [27] Woo KM, Zhang R, Deng H, Ma PX. Protein-mediated osteoblast survival and migration on biodegradable polymer/hydroxyapatite composite scaffolds. Transaction of the 28th Annual Meeting of the Society for Biomaterial, vol. 605, 2002. [28] Woo KM, Wei G, Ma PX. Enhancement of fibronectin- and vitronectin-adsorption to polymer/hydroxyapatite scaffolds suppresses the apoptosis of osteoblasts. J Bone Miner Res 2002;17(Suppl 1):M49. [29] Zhang R, Ma PX. Processing of polymer scaffolds: phase separation. In: Atala A, Lanza R, editors. Methods of tissue engineering. San Diego, CA: Academic Press; 2001. p. 715–24. [30] Nam YS, Park TG. Porous biodegradable polymeric scaffolds prepared by thermally induced phase separation. J Biomed Mater Res 1999;47:8–17.
Biomaterials 25 (2004) 4749–4757
Structure and properties of nano-hydroxyapatite/polymer composite scaffolds for bone tissue engineering Guobao Weia, Peter X. Maa,b,c,* a Department of Biomedical Engineering, University of Michigan, Ann Arbor, MI 48109-2009, USA Department of Biologic and Materials Sciences, University of Michigan, Ann Arbor, MI 48109-1078, USA c Macromolecular Science and Engineering Center, University of Michigan, Ann Arbor, MI 48109-1055, USA b
Received 9 June 2003; accepted 11 November 2003
Abstract To better mimic the mineral component and the microstructure of natural bone, novel nano-hydroxyapatite (NHAP)/polymer composite scaffolds with high porosity and well-controlled pore architectures were prepared using thermally induced phase separation (TIPS) techniques. The morphologies, mechanical properties and protein adsorption capacities of the composite scaffolds were investigated. The high porosity (90% and above) was easily achieved and the pore size was adjusted by varying phase separation parameters. The NHAP particles were dispersed in the pore walls of the scaffolds and bound to the polymer very well. NHAP/polymer scaffolds prepared using pure solvent system had a regular anisotropic but open 3D pore structure similar to plain polymer scaffolds while micro-hydroxyapatite (MHAP)/polymer scaffolds had a random irregular pore structure. The introduction of HAP greatly increased the mechanical properties and improved the protein adsorption capacity. In a dioxane/water mixture solvent system, NHAP-incorporated poly(l-lactic acid) (PLLA) scaffolds developed a fibrous morphology which in turn increased the protein adsorption three fold over non fibrous scaffolds. The results suggest that the newly developed NHAP/polymer composite scaffolds may serve as an excellent 3D substrate for cell attachment and migration in bone tissue engineering. r 2003 Elsevier Ltd. All rights reserved. Keywords: Nano; Hydroxyapatite; Phase separation; Scaffold; Protein adsorption
1. Introduction There is a growing need for bone regeneration due to various clinical bone diseases such as bone infections, bone tumors and bone loss by trauma [1]. Current therapies for bone defects include autografts, allografts, xenografts and other artificial substitutes such as metals, synthetic cements and bioceramics [2,3]. However, these substitutes are far from ideal and each has its specific problems and limitations. For example, autografts are associated with donor shortage and donor site morbidity whereas allografts and xenografts have the risk of disease transmission and immune response [4], and synthetic materials wear and do not behave like true bone. As a result, recent research has been devoted to bone tissue engineering in which a 3D porous scaffold is loaded with specific living cells and/or tissue-inducing *Tel.: +1-734-764-2209; fax: +1-734-647-2110. E-mail address: [email protected] (P.X. Ma). 0142-9612/$ - see front matter r 2003 Elsevier Ltd. All rights reserved. doi:10.1016/j.biomaterials.2003.12.005
factors to launch a tissue regeneration or replacement in a natural way [5,6]. In the tissue engineering approach, the temporary 3D scaffold serves an important role in the manipulation of the functions of osteoblasts and a central role in the guidance of new bone formation into desired shapes. In principle, a biodegradable matrix with sufficient mechanical strength, optimized architecture and suitable degradation rate, which could finally be replaced by newly formed bone, is most desirable. The scaffolding materials for bone tissue engineering should also be osteoconductive so that osteoprogenitor cells can adhere and migrate on the scaffolds, differentiate, and finally form new bone [7,8]. Biodegradable polymers, mainly polyesters such as poly(lactic acid) (PLA), poly(glycolic acid) (PGA) and their copolymers (PLGA), have been widely used to develop porous 3-D scaffolds using various fabrication techniques [8–14]. These synthetic materials have been demonstrated to be biocompatible and degrade into non-toxic components with a
ARTICLE IN PRESS 4750
G. Wei, P.X. Ma / Biomaterials 25 (2004) 4749–4757
controllable degradation rate in vivo [15]. Another major class of biomaterials for bone repair is ceramics such as hydroxyapatite (HAP) and tricalcium phosphate (TCP) [16,17]. Being similar to the mineral component of natural bone, they showed good osteoconductivity and bone bonding ability [18]. However, the main limitation for the use of HAP ceramics was their inherent brittleness and difficulty for processing [19]. To combine the osteoconductivity of calcium phosphates and good biodegradability of polyesters, polymer/ceramic composite scaffolds have been developed for bone tissue engineering either by direct mixing or by a biomimetic approach [7,8,20]. Compared to plain polymer scaffolds in which neo tissue matrix was formed only in the surface layer (o240 mm) [21], the composite scaffolds supported cells growth and neo tissue formation throughout the scaffold including in the very center of the scaffold [6]. Polymer/ceramic composite scaffolds mimic the natural bone to some extent. Natural bone is composed of inorganic compound (mainly partially carbonated HAP on the nanometer scale) and organic compounds (mainly collagen). The nanometer size of the inorganic component (mainly bone-like apatite) in natural bone is considered to be important for the mechanical properties of the bone [22]. Recent research in this field also suggested that better osteoconductivity would be achieved if synthetic HAP could resemble bone minerals more in composition, size and morphology [23,24]. In addition, nano-sized HAP may have other special properties due to its small size and huge specific surface area. Webster et al. [25,26] have shown significant increase in protein adsorption and osteoblast adhesion on the nano-sized ceramic materials compared to traditional micron-sized ceramic materials. In previous works, the Ma group has developed a variety of scaffolds using thermally induced phase separation (TIPS) [8–10]. Both pore structure and pore wall morphology can be controlled by phase separation parameters. They have also demonstrated that the addition of micron-sized hydroxyapatite (MHAP) increase the adsorption of proteins and extracellular matrix (ECM) components [27]. The cell seeding uniformity into the scaffold was improved substantially due to the introduction of osteoconductive hydroxyapatite [6]. The MHAP/PLLA composite scaffolds were shown to suppress apoptosis of osteoblasts by a possible mechanism of enhanced adsorption of serum proteins such as vitronectin and fibronectin onto the scaffolds [27, 28]. In this paper, we further mimicked the size scale of hydroxyapatite in natural bone and aim to fabricate novel and improved composite scaffolds. The pore structure, pore wall morphology, mechanical properties and protein adsorption capacity were systematically investigated.
2. Materials and methods 2.1. Materials Poly(l-lactic acid) (PLLA) with an inherent viscosity of 1.4–1.8 dl/g was purchased from Boehringer Ingelheim (Ingelheim, Germany) and was used as received. PLGA85 (Medisorbs, LA:GA=85:15) with an inherent viscosity of 0.6 was obtained from Alkermes Inc. (Wilmington, OH). NHAP were purchased from Berkeley Advanced Biomaterials Inc. (San Leandro, CA). Conventional HAP particles (MHAP, particle size on the micrometer scale), tetrahydrofuran (THF), dioxane and benzene were obtained from aldrich chemical company (Milwaukee, WI). 2.2. Fabrication method NHAP/PLLA composite scaffolds were fabricated using a phase separation technique similar to that reported previously [8]. Briefly, the HAP nanocrystal powder was dispersed in a solvent by sonication for 30 s at 15 W (Virsonic 100, Cardiner, NY). After that, PLLA was dissolved in the NHAP suspended solvent at about 60 C to make homogeneous solutions with desired concentrations. 2.5 ml polymer/NHAP mixture was added into a Teflon vial, sonicated again and then transferred into a freezer at a preset temperature to induce solid–liquid or liquid–liquid phase separation. The solidified mixture was maintained at that temperature overnight and then transferred into a freeze-drying container which was maintained at a temperature between 5 C and 10 C by salt–ice bath for freezedrying. The samples were freeze-dried at 0.5 mmHg for 7 days to remove solvent. The final composition of the composite foam was determined by the concentration of the polymer solution and NHAP content in the mixture. 2.3. SEM characterization The porous morphologies of the scaffold were examined with scanning electron microscopy (SEM) (S-3200N, Hitachi, Japan) at 20 or 15 kV. The samples were cut by a razor blade or fractured after being frozen in liquid nitrogen for several minutes and then were coated with gold for 200 s using a sputter coater (DeskII, Denton Vacuum Inc.). 2.4. Porosity The melting behavior of the matrices was characterized with a differential scanning calorimeter (DSC-7, PerkinElmer, Norwalk, CT) as detailed previously [9]. The degree of crystallinity, Xc, of a sample was calculated as 0 Xc ¼ DHm =DHm ;
ARTICLE IN PRESS G. Wei, P.X. Ma / Biomaterials 25 (2004) 4749–4757
where DHm was the measured enthalpy of melting and 0 DHm was the enthalpy of melting for 100% crystalline 0 polymer. For PLLA, DHm ¼ 203:4 J=g: The estimated density and porosity of the matrix were obtained as follows. The diameter and height of the matrix were measured to calculate the volume. The mass of the matrix was measured with an analytical balance. The density was calculated from the volume and mass. The porosity, e; was calculated from the measured overall densities Dm of the matrix and the skeletal density Ds. For a composite scaffold, the skeletal density was determined using the densities of the polymer and HAP powder. The porosity was given by Ds Dm e¼ ; Ds where Ds was calculated from the following formula: 1 ; Ds ¼ ð1 Xh Þ=Dp þ Xh =Dh where Dh is the density of the HAP powder with a value of 3.16 g/ml and Xh is the percentage of HAP in the composite while Dp is the density of the polymer. Dp was calculated from 1 Dp ¼ ; ð1 Xc Þ=Da þ Xc =Dc where Xc was the degree of crystallinity of the polymer. The density of amorphous PLLA (Da) is 1.248 g/ml and the density of 100% crystalline PLLA (Dc) is 1.290 g/ml. 2.5. Mechanical properties The compressive mechanical properties of the scaffolds were tested using a MTS Synergie 200 mechanical tester (MTS systems Co. Eden Prairie, Minnesota). The scaffolds have dimensions of about 16 mm in diameter and 3 mm in thickness. A crosshead speed of 0.5 mm/ min was used. The compressive modulus was defined as the initial linear modulus. At least five specimens were tested for each sample. The averages and standard deviations were graphed. A two-tail Student’s t-test (assuming equal variances) was performed to determine the statistical significance (po0.05) of the differences in mechanical properties. 2.6. Protein adsorption Protein adsorption was performed by incubating the scaffolds in phosphate buffered saline (PBS, 0.1 m, pH=7.4) containing 2.5% or 5% fetal bovine serum (FBS). The disk-like specimens with dimensions of 15.5– 16.5 mm in diameter and 370.1 mm in thickness were used. Before incubation in the medium containing proteins, the specimens were pretreated by ethanol for 30 min and then washed by PBS three times overnight under gentle shaking. Specimens were then put into 24-
4751
well culture plates (one specimen for each well) and 1.5 ml FBS/PBS solution was added into each well. The scaffolds were incubated at 37 C for a chosen incubation time. The concentration of the protein in the FBS/ PBS solution was then quantified with a commercial protein assay kit, BCA (Pierce, Rockford, IL), using bovine serum albumin (BSA) standards. The amount of absorbed proteins was determined by subtracting the amount of proteins left in the FBS/PBS solution after adsorption from the amount of proteins in control FBS/ PBS solution (without specimen) under the same incubation conditions.
3. Results 3.1. Porosity NHAP/polymer composite scaffolds with high porosity have been fabricated using thermally induced phase separation techniques (Tables 1 and 2). Two solvent systems were used to obtain composite scaffolds with different pore structures. One is a mono-solvent system of pure dioxane or benzene and the other, a solvent/nonsolvent mixture system of dioxane and water in different proportions. The densities and porosities of scaffolds are listed in Tables 1 and 2. Incorporation of nano-sized HAP only decreased the porosity slightly. This was especially true when NHAP content was low. Regardless of detailed morphology and composition, all scaffolds have a high porosity of at least 89%, which was considered to be beneficial for cell ingrowth and survival. The solvent system did not have much effect on the density and porosity. The quenching temperature did not affect the porosity significantly either (p=0.12). 3.2. Morphology The morphology and microstructure of the scaffolds were examined using SEM (Fig. 1): micrographs of PLLA scaffolds (a,b), MHAP/PLLA (c,d) and NHAP/ PLLA (e–h) composite scaffolds fabricated with pure
Table 1 Densities and porosities of NHAP/PLLA composite scaffolds prepared from dioxane solutions, phase separated at 18 C Scaffolds
PLLA:HAP (w/w)
Dp (g/cm3)
Ds (g/cm3)
Porosity (%)
PLLA PLLA–NHAP PLLA–NHAP PLLA–NHAP PLLA–NHAP PLLA–MHAP
100:0 90:10 70:30 50:50 30:70 50:50
1.2577 1.3383 1.5394 1.7993 2.1737 1.7993
0.08812 0.09708 0.1176 0.1475 0.2304 0.1505
93.0 92.8 92.3 91.8 89.4 91.6
ARTICLE IN PRESS G. Wei, P.X. Ma / Biomaterials 25 (2004) 4749–4757
4752
Table 2 Porosities of NHAP/PLLA scaffolds fabricated using dioxane/water mixture solvent system and phase separated at various temperatures Scaffolds
PLLA:HAP (w/w)
Dioxane:H2O (v/v)
Quenching temp. ( C)
Porosity (%)
PLLA–NHAP PLLA–NHAP PLLA–NHAP PLLA–NHAP PLLA–NHAP PLLA–NHAP PLLA–NHAP
70:30 70:30 70:30 70:30 70:30 70:30 70:30
100:0 95:5 90:10 87:13 87:13 87:13 87:13
18 18 18 4 18 70 196 (LN2)
92.3 91.6 93.1 94.5 93.1 92.8 92.5
dioxane or benzene. NHAP/PLLA composite scaffolds maintained a regular internal ladder-like pore structure similar to plain PLLA scaffolds, a typical morphology formed by solid–liquid phase separation [8,10]. By controlling the heat transfer direction, NHAP/PLLA scaffolds with tubular pore architecture were also obtained using benzene as a solvent. Nano-sized HAP particles were distributed within the pore walls of the scaffolds and no large aggregates appeared in pores. In MHAP/PLLA scaffolds, the HAP platelets with size ranging from 10 to 50 mm were found randomly distributed in the PLLA matrix. Some were embedded in the pore wall and some others were piled together between pores or in the pores. The pore size of all these scaffolds ranged from 50 to 100 mm. NHAP/PLGA85 composite scaffolds prepared with pure dioxane had leaf-like morphology and well-interconnected pore structures (Fig. 2). Very different morphologies appeared in composite scaffolds fabricated with dioxane and water mixture solvent system (Fig. 3). The proportion of water in the solvent systems had a significant effect on the pore size and overall morphology of the scaffold. Even a small amount of water (less than 5%) in the solvent system significantly reduced the pore size from 100 mm to about 10 mm and random pore structure replaced the regular ladder-like pore structure (Figs. 3a, b and c). Another obvious change was the pore wall surface texture. Unlike relative solid and smooth pore wall in NHAP/ PLLA scaffold prepared from mono-solvent system, fibrous and loose network became the characteristics of the pore structure of the scaffolds prepared from the mixed solvent system. When the water volume was 13% in the mixture solvent, the scaffolds consisted of three ranges of pore sizes. Large macropores were developed in a range of 200–500 mm while much smaller micropores (tens of microns) between the pore walls of the large pores. The pore wall surface was not smooth, and appeared as loose fibrous structure with several microns in between. NHAP particles were entrapped in the fibrous network of the porous structures and had little effect on the changes in morphology with different water proportions since plain PLLA scaffolds had the same morphological changes (data not shown).
3.3. Mechanical properties The compressive modulus of NHAP/PLLA scaffolds increased with NHAP content (Fig. 4). The plain PLLA scaffolds prepared using solid–liquid phase separation techniques had a compressive modulus of 4.3 MPa. The addition of NHAP into the scaffold improved the mechanical properties. The modulus increased significantly when the NHAP proportion reached 30% of the composite and reached 8.3 MPa when the ratio of NHAP to PLLA was 50:50. The solvent system had a large effect on the mechanical properties of the porous scaffolds (Fig. 5). The introduction of a non-solvent component (in this case, water) in the polymer/HAP/ solvent mixture likely induced liquid–liquid phase separation instead of solid–liquid phase separation. Since regular and orientated pore architectures were replaced by random pore structures in scaffolds fabricated using mixed solvent system, the compressive modulus decreased significantly. However, the modulus recovered substantially when dioxane/water (87:13, v/v) system was used. This might have resulted from the uniform structure of the scaffolds (Figs. 3 d–f). 3.4. Protein adsorption NHAP/PLLA composite scaffolds with different amounts of NHAP were incubated in FBS/PBS solution to investigate protein adsorption. For all NHAP/PLLA composite scaffolds, the protein adsorption reached equilibrium in 25 h and there was no significant increase of protein adsorption during further incubation (Fig. 6a). The addition of NHAP increased the protein adsorption. NHAP/PLLA 50:50 and NHAP/PLLA 70:30 composite scaffolds absorbed 2.4- and 3.2-fold, respectively, more serum proteins than plain PLLA scaffold (Fig. 6b). The size of HAP particles also affected protein adsorption onto the scaffolds containing a high percentage of HAP. Serum protein adsorption was significantly greater on NHAP/PLLA scaffolds than on MHAP/PLLA scaffolds when the HAP/PLLA was 50:50 or higher (po0.05, Fig. 6c). However, when the ratio of HAP in the composite was lower than 50%, the differences in the protein adsorption between NHAP
ARTICLE IN PRESS G. Wei, P.X. Ma / Biomaterials 25 (2004) 4749–4757
4753
Fig. 1. SEM micrographs of plain PLLA, NHAP/PLLA and MHAP/PLLA scaffolds fabricated from 5% dioxane (a–e) and benzene (f,g) solution. (a,b) Pure PLLA scaffold, 50, 400; (c,d) MHAP/PLLA 50:50 scaffold, 100, 500; (e,f) NHAP/PLLA 50:50 scaffold, 100, 1000; (g) Tubular NHAP/PLLA scaffold, cross-section, 200; (h) Tubular NHAP/PLLA scaffold, longitudinal section, 100.
ARTICLE IN PRESS 4754
G. Wei, P.X. Ma / Biomaterials 25 (2004) 4749–4757
Fig. 2. SEM micrographs of NHAP/PLGA85 (30:70) scaffolds fabricated from 10% dioxane solution quenched at 70 C. (a) 150 and (b) 500.
and MHAP incorporated scaffolds were not statistically significant. With the dramatic changes in morphology of scaffolds prepared using different solvent systems, the protein adsorption also changed greatly. Scaffolds with fibrous texture absorbed about four times greater amount of protein than those with smooth pore wall surface. These results suggested that the protein adsorption on the scaffolds could be regulated by the content of HAP, its size scale, and the pore wall morphology of the scaffolds.
4. Discussion The present studies described a phase separation technique to fabricate polymer/NHAP-composite scaffolds with high porosity and controlled pore architecture. Different solvent systems were used to obtain scaffolds with different microarchitectures and properties. When dioxane was used alone, the porous structure resulted from a solid–liquid phase separation of the polymer solution. During quenching, the solvent crystallized and the polymer was expelled from the solvent crystallization front. Solvent crystals became pores after subsequent sublimation. The characteristics of the pores were determined by the morphologies of solvent crystals under such quenching conditions. The temperature gradient along the solvent crystallization direction led to anisotropic structure. The introduction of HAP particles into the polymer solution perturbed the solvent crystallization to some extent and thereby made the pore structure more irregular and isotropic. As expected, MHAP/PLLA composite scaffold exhibited an isotropic and irregular pore structure. The perturbation by NHAP particles, however, was small even in high proportion up to 50% due to their nanometer size scale and uniform distribution. As a result, NHAP/PLLA composite scaffolds maintained the main characteristic
pore architecture of solid–liquid phase separation which was anisotropic and regular. In contrast to NHAP/ PLLA, the regular anisotropic pore structure was obtained only when the HAP content was very low in MHAP/PLLA scaffolds, e.g., MHAP/PLLAo10/90 (images not shown). In this case, low content of HAP did not affect the solvent crystallization significantly enough to alter the pore structure. When mixed solvent (dioxane/water) was used to prepare 5% PLLA solutions, a liquid–liquid phase separation was induced and this process dominated the final pore structure of the system [29], resulting in random and interconnected pore architecture. The pore size was also much smaller. With the increase of water content in the system, larger quench depth allowed the system to continue on a coarsening process to further reduce the interfacial free energy between two phases separated in earlier stage [30]. The coarsening effect might have resulted in the formation of macropores in addition to the micropores (Figs. 3d–f). It is therefore possible to optimize various phase separation parameters to achieve a macroporous structure in composite scaffold. The mechanism of fibrous texture formation is not fully understood. Since PLLA is a semicrystalline polymer, the crystallization in the polymer-rich phase may present a possible explanation for the formation of the fibrous structure [9]. Most mammalian cells are anchorage-dependent cells and they need a biocompatible substrate for attachment, migration and differentiation to form new tissues. Recent research demonstrated that cell adhesion and survival could be modulated by protein pre-adsorption on the substrate [25–28]. Therefore, protein adsorption is of importance in evaluating a scaffold for tissue engineering. In this paper, we have shown that the addition of hydroxyapatite particles into the composite scaffolds greatly increased protein adsorption. In addition to the high affinity of hydroxyapatite itself for
ARTICLE IN PRESS G. Wei, P.X. Ma / Biomaterials 25 (2004) 4749–4757
4755
Fig. 3. SEM micrographs of NHAP/PLLA (30:70) scaffolds fabricated using dioxane/water mixture solvents. (a) Dioxane:water=95:5, 500. (b,c) Dioxane:water=90:10, 500, 8000. (d,e,f) Dioxane:water=87:13, 45, 500, 10000.
proteins, incorporation of HAP also altered the pore surface morphology of the composite scaffolds and may have made them more suitable for protein adsorption. Furthermore, NHAP/PLLA scaffold absorbed significantly more proteins than MHAP/PLLA at high HAP content but not at low HAP content. It was likely that most particles were embedded in the polymer pore walls and the effect of HAP size was not evident when the HAP content was low in the composite. With the increase of HAP content, more and more particles were exposed on the surfaces of the pore walls. This phenomenon was most evident when the scaffold
exhibited fibrous morphology. Fibrous scaffolds had much greater surface to volume ratio than scaffolds with solid pore walls, which might have further increased the protein adsorption capacity.
5. Conclusion The present study investigated new NHAP/polymer composite scaffolds developed using thermally induced phase separation techniques. Nano-sized HAP particles were successfully incorporated into the PLLA porous
ARTICLE IN PRESS 4756
G. Wei, P.X. Ma / Biomaterials 25 (2004) 4749–4757
scaffolds. The incorporation of NHAP improved the mechanical properties and protein adsorption of the composite scaffolds while maintaining high porosity and
Fig. 4. Effects of NHAP content on the compressive modulus of NHAP/PLLA scaffolds fabricated from 5% dioxane solutions % po0.05, wp=0.29, n=5.
Fig. 5. Effects of mixture solvent composition and quenching temperature on the compressive moduli of NHAP/PLLA scaffolds fabricated from 5% dioxane/water solvent systems.
Fig. 6. Protein adsorption on porous composite scaffolds incubated in FBS/PBS solution. (a) as a function of time, 2.5% FBS/PBS, labels indicate the w/w ratios of NHAP to PLLA; (b) as a function of NHAP content, 2.5% FBS; (c) Comparison of MHAP and NHAP, 5.0% % FBS/PBS. po0.05, wp=0.71, n=3. (d) Comparison of NHAP/PLLA (30:70) scaffolds prepared using different solvent systems, 2.5% % FBS/PBS. po0.05, n=3.
suitable microarchitecture. These results suggest that the newly developed NHAP/polymer composite scaffolds may be superior for bone tissue engineering.
ARTICLE IN PRESS G. Wei, P.X. Ma / Biomaterials 25 (2004) 4749–4757
References [1] Braddock M, Houston P, Campbell C, Ashcroft P. Born again bone: tissue engineering for bone repair. News Physiol Sci 2001;16:208–13. [2] Goldberg VM, Stevenson S. Natural history of autografts and allografts. Clin Orthop 1987;225:7–16. [3] Costantino PD, Friedman CD. Synthetic bone graft substitutes. Otolaryngol Clin North Am 1994;27:1037–74. [4] Bonfiglio M, Jeter WS. Immunological responses to bone. Clin Orthop 1972;87:19–27. [5] Langer R, Vacanti JP. Tissue engineering. Science 1993;260:920–6. [6] Ma PX, Zhang R, Xiao G, Franceschi R. Engineering new bone tissue in vitro on highly porous poly(alpha-hydroxyl acids)/ hydroxyapatite composite scaffolds. J Biomed Mater Res 2001;54:284–93. [7] Zhang R, Ma PX. Porous poly(l-lactic acid)/apatite composites created by biomimetic process. J Biomed Mater Res 1999;45:285–93. [8] Zhang R, Ma PX. Poly(alpha-hydroxyl acids)/hydroxyapatite porous composites for bone tissue engineering. I. Preparation and morphology. J Biomed Mater Res 1999;44:446–55. [9] Ma PX, Zhang R. Synthetic nano-scale fibrous extracellular matrix. J Biomed Mater Res 1999;46:60–72. [10] Ma PX, Zhang R. Micro-tubular architecture of biodegradable polymer scaffolds. J Biomed Mater Res 2001;56:469–77. [11] Mikos AG, Thorsen AJ, Czerwonka LA, Bao Y, Langer R, Winslow DN, Vacanti JP. Preparation and characterization of poly(l-lactic acid) foams. Polymer 1994;35:1068–77. [12] Whang K, Thomas CH, Healy KE, Nuber G. A novel method to fabricate bioabsorbable scaffolds. Polymer 1995;36:837–42. [13] Mooney DJ, Baldwin DF, Suh NP, Vacanti JP, Langer R. Novel approach to fabricate porous sponges of poly(d,l-lactic-coglycolic acid) without the use of organic solvents. Biomaterials 1996;17:1417–22. [14] Borden M, Attawia M, Laurencin CT. The sintered microsphere matrix for bone tissue engineering: in vitro osteoconductivity studies. J Biomed Mater Res 2002;61:421–9. [15] Visscher GE, Robison RL, Maulding HV, Fong JW, Pearson JE, Argentieri GJ. Biodegradation of and tissue reaction to 50:50 poly(DL-lactide-co-glycolide) microcapsules. J Biomed Mater Res 1985;19:349–65. [16] Li SH, De Wijn JR, Layrolle P, de Groot K. Synthesis of macroporous hydroxyapatite scaffolds for bone tissue engineering. J Biomed Mater Res 2002;61:109–20.
4757
[17] Flautre B, Descamps M, Delecourt C, Blary MC, Hardouin P. Porous HA ceramic for bone replacement: role of the pores and interconnections—experimental study in the rabbit. J Mater SciMater Med 2001;12:679–82. [18] LeGeros RZ. Properties of osteoconductive biomaterials: calcium phosphates. Clin Orthop 2002;395:81–98. [19] Cooke FW. Ceramics in orthopedic surgery. Clin Orthop 1992;276:135–46. [20] Thomson RC, Yaszemski MJ, Powers JM, Mikos AG. Hydroxyapatite fiber reinforced poly(alpha-hydroxy ester) foams for bone regeneration. Biomaterials 1998;19:1935–43. [21] Ishaug SL, Crane GM, Miller MJ, Yasko AW, Yaszemski MJ, Mikos AG. Bone formation by three-dimensional stromal osteoblast culture in biodegradable polymer scaffolds. J Biomed Mater Res 1997;36:17–28. [22] Rho JY, Kuhn-Spearing L, Zioupos P. Mechanical properties and the hierarchical structure of bone. Med Eng Phys 1998;20: 92–102. [23] Du C, Cui FZ, Zhu XD, de Groot K. Three-dimensional nanoHAp/collagen matrix loading with osteogenic cells in organ culture. J Biomed Mater Res 1999;44:407–15. [24] Du C, Cui FZ, Feng QL, Zhu XD, de Groot K. Tissue response to nano-hydroxyapatite/collagen composite implants in marrow cavity. J Biomed Mater Res 1998;42:540–8. [25] Webster TJ, Ergun C, Doremus RH, Siegel RW, Bizios R. Specific proteins mediate enhanced osteoblast adhesion on nanophase ceramics. J Biomed Mater Res 2000;51:475–83. [26] Webster TJ, Siegel RW, Bizios R. Osteoblast adhesion on nanophase ceramics. Biomaterials 1999;20:1221–7. [27] Woo KM, Zhang R, Deng H, Ma PX. Protein-mediated osteoblast survival and migration on biodegradable polymer/hydroxyapatite composite scaffolds. Transaction of the 28th Annual Meeting of the Society for Biomaterial, vol. 605, 2002. [28] Woo KM, Wei G, Ma PX. Enhancement of fibronectin- and vitronectin-adsorption to polymer/hydroxyapatite scaffolds suppresses the apoptosis of osteoblasts. J Bone Miner Res 2002;17(Suppl 1):M49. [29] Zhang R, Ma PX. Processing of polymer scaffolds: phase separation. In: Atala A, Lanza R, editors. Methods of tissue engineering. San Diego, CA: Academic Press; 2001. p. 715–24. [30] Nam YS, Park TG. Porous biodegradable polymeric scaffolds prepared by thermally induced phase separation. J Biomed Mater Res 1999;47:8–17.