|Home | About | Journals | Submit | Contact Us | Français|
Several minerals, such as hydroxyapatite and β-tricalcium phosphate, have been incorporated into bioresorbable polyester bone scaffolds to increase the osteoconductivity both in vitro and in vivo. More soluble forms of calcium phosphate that release calcium and phosphate ions have been postulated as factors that increase osteoblast differentiation and mineralization. Recently, a zirconia-hybridized pyrophosphate-stabilized amorphous calcium phosphate (Zr-ACP) has been synthesized allowing controlled release of calcium and phosphate ions. When incorporated into bioresorbable scaffolds, Zr-ACP has the potential to induce osteoconductivity. In this study, 80–90% (w/v) porous poly(DL-LActic-co-glycolic acid) (PLGA) scaffolds were formed by thermal phase separation from dioxane while incorporating Zr-ACP. Scanning electron microscopy revealed a highly porous structure with a pore size ranging from a few μm to about 100 μm, smaller than we had hoped for. Zr-ACP particles were evenly dispersed in the composite structure and incorporated into the pore walls. The amorphous structure of the Zr-ACP was maintained during composite fabrication, as found by X-ray diffraction. Composite scaffolds had larger compressive yield strengths and moduli compared to pure polymer scaffolds. These initial efforts demonstrate that PLGA/Zr-ACP composites can be formed in ways that ultimately serve as promising bone scaffolds in tissue engineering.
Tissue-engineered scaffolds are promising alternatives to autografts and allografts, albeit that both of which have their limitations as bone replacements . Consequently, many synthetic scaffold materials are currently being investigated to regenerate bone tissue. Among these materials are biodegradable polymers (i.e., polylactic and polyglycolic acids) and calcium-phosphate-based materials (i.e., hydroxyapatite and β-tricalcium phosphate) due to their good biocompatibility and osteoconductivity, respectively.
Synthetic biodegradable polymers, such as poly(DL-lactic-co-glycolic acid) (PLGA), have been used for scaffolds to regenerate bone [2–5]. PLGA has good biocompatibility, the ability to metabolize into benign oligomers and a useful range of mechanical properties and degradation rates (depending on the co-polymer ratio) . PLGA scaffolds with open porosity have been fabricated that allow for cell seeding, attachment, growth and extracellular matrix production in vitro  and are remodeled when implanted into bone defects . A porosity ranging from 75 to 90% is considered preferable for PLGA bone tissue scaffolds [4, 5] and an ideal range of pore diameters from 150–710 μm has been suggested . Porous poly-meric scaffolds have been fabricated by many methods, including solvent evaporation , compression molding , gas foaming [8, 9], solution casting/salt leaching [10, 11], supercritical fluid foaming [12–16] and phase separation [17–20].
Calcium phosphate ceramics have also been extensively investigated for use as bone scaffolds and bone replacements [21–23]. Most calcium phosphate biomaterials are polycrystalline ceramics characterized by a high biocompatibility and the ability to undergo osseointegration, with varying degrees of resorbablility . Synthetic hydroxyapatite (HAP) and β-tricalcium phosphate (β-TCP) are the two most commonly used calcium phosphate materials for bone replacements due to their high osteoconductivity. Amorphous calcium phosphates have been used only sparingly as bone replacements, mainly because of rapid and uncontrolled dissolution [24, 25]. Clinical reference has been made to the use of casein-phosphopeptide-stabilized amorphous calcium phosphates under the tradename Recaldent™, which have been added to remineralizing pastes and chewing gums to retard mineral erosion [26, 27]. Recently, amorphous calcium phosphate has been stabilized by pyrophosphate ions ( ) and hybridized with zirconium (Zr-ACP) to retard its dissolution and subsequent dynamic conversion to HAP in aqueous environments [28, 29]. In addition, Zr-ACP allows for more controlled release of calcium and phosphate ions; factors that have been postulated to increase osteoblast differentiation and mineralization in vitro [30–33]. Our own observations have shown that Zr-ACP exposure to MC3T3-E1 cells is marked by transient increases in Ca2+ and ions and pH that coincide with transient increases in alkaline phosphatase activity and a significant effect on osteopontin synthesis . Incorporation of Zr-ACP could enhance graft osteoinductivity and add mechanical stability to the polymer scaffold simultaneously. One laudable goal would be to resolve ways to incorporate these structures in a composite form.
In this study, porous PLGA/Zr-ACP and PLGA/HAP composite foams (5% or 10% (w/v) polymer/solvent with 25 wt% or 50 wt% calcium phosphates) were fabricated using a previously reported thermal phase-inversion technique . The goal was to form porous PLGA structures with evenly distributed Zr-ACP fillers included, inhibiting Zr-ACP conversion to HAP during fabrication. X-ray diffraction determined the Zr-ACP crystallinity after composite fabrication. Pore size, pore distribution and mechanical properties of composites were also determined.
Calcium nitrate tetrahydrate (Ca(NO3)2 · 4H2O), disodium hydrogenphosphate (Na2HPO4), sodium pyrophosphate (Na4P2O7 · 10H2O) and sodium hydroxide were purchased from Sigma-Aldrich (St. Louis, MO, USA). Ammonium hydroxide and acetone were purchased from Fisher (Fairlawn, NJ, USA). Zirconyl chloride (ZrOCl2) was purchased from GFS Chemicals (Columbus, OH, USA).
Zr-ACP and HAP were prepared at the American Dental Association Foundation Paffenbarger Research Center according to the method of Skrtic et al. . Briefly, for Zr-ACP, a 1.06 M solution of a Ca(NO3)2 · 4H2O was prepared using CO2-free water (in a N2 closed system), designated as solution 1. A 0.25 M ZrOCl2 solution (40 ml) was prepared and designated as solution 2. A third solution containing 1.16 M Na2HPO4 and 0.024 M Na4P2O7 · 10H2O was made using CO2-free water (mechanically stirred at 400 rpm) and mixed with 67 ml of a 1 M NaOH solution under N2 in a covered beaker to maintain a pH of 8.5–9. Solutions 1 and 2 were added simultaneously to the third under N2 and mechanically stirred at 400–500 rpm. Precipitation of the Zr-ACP was instantaneous. The solution was allowed to stabilize for 5 min, after which precipitates were filtered and washed twice with cold ammoniated water (2% ammonia) and then washed with cold acetone. Zr-ACP was then frozen in a −50°C freezer for 1 h and transferred to a VirTis bench top freeze dryer (SP Industries, Warminster, PA, USA) for 48 h at a temperature of −70°C and pressure of 5 mTorr. Further characterization of the product crystallinity, surface morphology, Ca/PO4 molar ratios and dissolution/transformation of Zr-ACP has been documented in the literature [28, 29].
PLGA/HAP and PLGA/Zr-ACP composites were fabricated using a similar thermal phase inversion technique as previously reported . Briefly, 75/25 poly(DL-lactic-co-glycolic acid) (Absorbable Polymers, Pelham, AL, USA, catalog No. 75DG0651) was dissolved in 1,4-dioxane (Sigma-Aldrich) to a 5% or 10% (w/v) solution. Samples were heated and stirred at 50°C for 2 h to obtain a homogeneous polymer solution. Equal volumes of polymer solution were placed into individual wells of a 24-well cell-culture plate (Corning, Corning, NY, USA) and Zr-ACP or HAP was added to make a 25% or 50% (w/v) mineral final concentration. Minerals were then sonicated for 20 s at 8 W and transferred to a −50°C ethanol bath for 1 h to induce solid–liquid phase separation. The samples were maintained in the freezer at 50°C overnight and then freeze dried using a VirTis bench top freeze dryer (SP Industries) for 3 days at a temperature of −70°C and pressure of 5 mTorr. The samples took the form of the well plate (nominal diameter 16 mm, the scaffolds were roughly 5 mm thick).
Scaffold morphology was observed by scanning electron microscopy (SEM). Samples were immersed in liquid nitrogen and then freeze-fractured longitudinally using a razor blade (parallel to the solidification direction). The samples were then sputtered with 20 nm Pd and observed using a LEO 1550 Field Emission SEM (Carl Zeiss SMT, Thornwood, NY, USA). Micrographs were obtained using the LEO image software.
The pore size and distribution of the scaffolds were analyzed using the public domain Scion Image program (available at http://www.scioncorp.com) following published procedures . Briefly, a 0.25 mm2 area of the SEM image was selected for further analysis (Fig. 1a). Next, contrast and brightness were adjusted to the same level for each SEM image because pore size analysis uses a grayscale image analysis (Fig. 1b). Thresholding was performed at a set value for each image (Fig. 1c) and then qualitatively compared to the original image (Fig. 1a) to validate that thresholding yielded similar contrast. If the threshold image did not match the original image, thresholding was repeated until a better match between the original and threshold image was obtained. After calibrating with a known scale, each pore was labeled and measured (Fig. 1d) and the data were exported to an Excel spreadsheet for further analysis. Pore diameter was determined by taking the average of the major and minor axes. Two dimensional pore area was determined by using the equation below:
The crystallinity of the lyophilized solids and composites was determined by X-ray diffractometry (XRD). The XRD profiles were recorded using an XDS 2000 Diffractometer (Scintag, Cupertino, CA, USA) between 15° and 70° with a CuKα source (λ = 1.54 Å) at 40 kV and 40 mA. The samples were step scanned in intervals of 0.020° at a scanning speed of 2.000°/min.
The compressive modulus and yield strength of the composites were found using an Instron 4204 equipped with a 1 kN static load cell (Instron, Canton, MA, USA) . Samples were sectioned to achieve final composite dimensions of 16 mm in diameter and 3 mm in thickness. The compression tests were done using a cross-head speed of 0.5 mm/min. Modulus was determined by the initial slope of the stress–strain curve. Yield strength was determined by using a parallel 0.02% strain offset. Three samples were tested for each condition.
Values are presented as a mean ± standard deviation. Statistical analysis was performed using Origin® 6.1 (OriginLab, Northampton, MA, USA). A one-way analysis of variance (ANOVA) with a significance level (α) of 0.05 was used to determine significance between groups.
Porous PLGA/HAP and PLGA/Zr-ACP composites were made using thermally induced solid-liquid phase separation and subsequent solvent sublimation. The original HAP particle size ranged between 0.5 and 10 μm (Fig. 2a), whereas the average Zr-ACP particle size was <1 μm (Fig. 2b). Pore size ranged from 3.3 μm to 114.7 μm for the 10% PLGA scaffold (Fig. 3a, b) and from 1.8 μm to 32.5 μm for 5% PLGA scaffolds (Fig. 3c, d) with median pore sizes of 20.9 μm and 4.3 μm, respectively (Table 1). Although the average pore size seems small for the 5% scaffolds, the cumulative pore area distribution shows that 87% of the pore area is comprised of pores with average diameters greater than 50 μm for the 10% PLGA scaffold (Fig. 4). Clearly these pores are smaller than the ideal size identified by Ishaug-Riley et al. .
PLGA/Zr-ACP 75:25 composites were made from 5% and 10% polymer solutions with pore sizes ranging from 3 to 107 μm and 4 to 154 μm, respectively (Fig. 5). As scaffolds increased in Zr-ACP content (50:50 blends as compared to 75:25 blends), average pore size decreased from 46.8 μm to 6.28 μm, respectively, and the pore morphology was more uneven (Fig. 6). All scaffolds including HAP displayed irregular pore architecture with average pore sizes ranging from 2.5 μm to 9.5 μm (Figs 7 and and8).8). One explanation for smaller and more irregular pores for the PLGA/HAP scaffolds is that the HAP particles used are larger than the Zr-ACP particles. These findings are consistent with results of Zhang et al. , who found that scaffolds filled with nano-hydroxyapatite particles have much larger pores and a more regular pore morphology than scaffolds fabricated with larger hydroxyapatite particles.
Adhesion between the Zr-ACP particles and PLGA matrix is observed (Fig. 9) and the surface of the pore walls shows dispersion and inclusion of the Zr-ACP particles, which range in size from <1 to 5 μm for a representative PLGA/Zr-ACP scaffold (Fig. 9). Zr-ACP incorporated in the pore walls could increase osteoconductivity of the composite scaffold.
Although these structures show high levels of porosity, no indication of pore interconnectivity was seen for these scaffolds using SEM. Other techniques known to raise the level of interconnectivity using common salt leaching would have led to ACP conversion which we wanted to avoid. Still it is possible that there is some level of interconnectivity although not as much as in other studies.
It is important that Zr-ACP retains its amorphous structure and does not convert to HAP during PLGA/Zr-ACP composite fabrication. The amorphous structure of Zr-ACP is known to increase dissolution rates and subsequently release more Ca2+ and ions in aqueous environments as compared to HAP . These ions are postulated to regulate osteoblastic differentiation [30, 31] and mineralization [32, 33] in vitro, but the mechanisms by which these factors promote osteogenesis remain unknown. The 50:50 PLGA/Zr-ACP composite diffraction pattern lacks discrete diffraction peaks, which is characteristic of amorphous materials (Fig. 10a). In contrast, the XRD pattern for the 50:50 PLGA/HAP composite shows a semi-crystalline structure as evidenced by discrete diffraction peaks (Fig. 10b). This confirms that the Zr-ACP retains its amorphous structure during the solid–liquid phase separation technique used here.
Compressive yield strength and modulus were found for 10% PLGA and PLGA/Zr-ACP composite scaffolds. The compressive modulus for the 50:50, 75:25 PLGA/Zr-ACP composites and PLGA foams were 51.9 ± 5.6 MPa, 35.6 ± 8.9 MPa and 14.7 ± 6.8 MPa, respectively (Fig. 11). Higher compressive yield was also obtained by adding Zr-ACP to the foam: 138 ± 13 kPa (50:50) and 78 ± 9.7 kPa (75:25) as compared to 35 ± 6.5 kPa (Fig. 11). It was shown that 5% polymer — 50:50 poly(lactic acid)/HAP composites have a compressive yield strength and modulus of about 40 kPa and 12 MPa, respectively . The rise in stiffness here was due to the larger polymer concentration (10% as compared to 5%). These results show that incorporation of Zr-ACP significantly strengthens the composite as compared to PLGA alone.
Thermal solid–liquid phase separation was used to make PLGA/Zr-ACP scaffolds. Porous structures were observed with pore sizes ranging from 3.6 μm to 154.4 μm for composites with 25% (w/v) Zr-ACP using a 10% (w/v) PLGA/dioxane mixture. SEM images showed adhesion between PLGA matrix and Zr-ACP and distribution of Zr-ACP in the pore walls. X-ray diffraction analysis confirmed that Zr-ACP remained amorphous during composite fabrication. 50:50 PLGA/Zr-ACP composites had a compressive modulus of 51.9 ± 5.6 MPa, whereas unreinforced PLGA had a modulus of only 14.7 ± 6.8 MPa. These studies ultimately show that Zr-ACP can be incorporated into porous biodegradable polymer composites for scaffolds.
The team acknowledges the support of NIH grant DE-015529 in executing this work.