Density–property relationships in mineralized collagen–glycosaminoglycan scaffolds
Introduction
Scaffolds for tissue regeneration are defined as: “three-dimensional open-cell porous structures synthesized from either natural or synthetic polymers which have the potential to support attachment, migration and multiplication of living cells” [1]. Although unproven, a widely believed design paradigm for scaffolds is that mimicking the composition of the natural tissue as closely as possible improves the capacity for regeneration [2]. The ability of a scaffold to regenerate tissue depends on its pore size, pore shape, porosity, biodegradability and mechanical properties. The average pore diameter must be large enough for cells to migrate through the pores yet small enough to retain an appropriate specific surface area for sufficient cell binding. For example, pore sizes in excess of 100 μm are optimal for bone growth [2], [3], [4]. Equiaxed pore shape and homogeneity are optimum for uniform cell adhesion and distribution of extracellular matrix proteins. Scaffolds must have large enough porosity (generally greater than 90%) and interconnectivity for effective transfer of cells and metabolites [5]. The degradation rate of the scaffold has to be roughly equal to the regeneration rate of the tissue. Furthermore, cells have been observed to be sensitive to the mechanical properties of the scaffold, which in turn affects the overall construct bioactivity [6].
Ideally, scaffolds should be similar to their natural counterparts in terms of chemical composition and physical structure. For this reason, natural polymers such as collagen are of major interest. To this end, collagen–glycosaminoglycan (CG) scaffolds have been developed and used clinically for skin regeneration and experimentally for nerve regeneration over the past three decades [7], [8], [9], [10], [11], [12], [13], [14], [15], [16], [17], [18], [19]. Composite scaffolds of collagen or gelatin with ceramics (e.g. hydroxyapatite and tricalcium phosphate), i.e. mineralized CG (MCG) scaffolds, have been developed to regenerate hard tissues such as bone [20], [21], [22], [23]. The most recent fabrication technique improves upon this mineralization process by forming a triple co-precipitate of mineral, collagen and glycosaminoglycan, without using a titrant, by controlling the molarity of the reactant acid and molar ratios of the different calcium sources [24], [25], [26], [27], [28], [29], [30]. Due to the in situ co-precipitation of the mineral phase, calcium phosphate crystals form within the collagen fibers, resulting in a more uniformly mineralized scaffold. Freeze-drying is then used to fabricate porous scaffolds from the triple co-precipitated slurry. These MCG scaffolds have regenerated subchondral bone at 16 weeks in a 4 mm diameter and 6 mm deep defect site at the knee joint in a goat model [31].
Extensive microstructural and mechanical characterization of CG and of MCG scaffolds of varying mineral content has been reported by Harley et al. [32] and by Kanungo et al. [30], respectively. Critical mechanical properties of scaffolds include elastic modulus, E∗, compressive crushing strength, σ∗, and compressive crushing strain, ε∗. The mechanical properties of different MCG scaffolds (along with the triple co-precipitated scaffolds) have been compared in the literature [30]. The triple co-precipitated mineralized scaffolds have relative densities (the density of the cellular solid, ρ∗, divided by that of the solid from which it is made, ρs) of roughly 0.03–0.04; that of trabecular bone varies from 0.05 to 0.60 [33], [34]. The mechanical properties of human compact and trabecular bone, along with 50 wt.% MCG scaffold (with a relative density, ρ/ρs, of 0.04), are listed in Table 1.
It is critical that the scaffold should have sufficient stiffness and strength to maintain its shape and size during surgical procedures such as implantation and to enhance bone in-growth while preventing encroachment of non-osseous tissue and competing cell types after implantation [35]. The optimal requirements for the above properties vary depending on the defect site and there are no established optimal magnitudes of the mechanical properties for bone scaffolds [36]. The current triple co-precipitated MCG scaffold (with a ρ/ρs of 0.04) can be crushed by hard thumb pressure. Hence, it is critical to improve the mechanical properties of MCG scaffolds such that they can be functionally suitable for bone regeneration. The mechanical properties (E∗ and σ∗) of the scaffold depend on those of the solid (Es and σfs) they are made from as well as the relative density of the scaffold, (ρ∗/ρs) [5], [30], [32], [37]. The overall properties of the scaffolds can be improved by either improving the properties of the solid it is made from or by increasing the relative density of the scaffold. Previous attempts to increase the mechanical properties of the scaffold by increasing the mineral content led to scaffolds with poorer mechanical properties due to the introduction of defects [30]. Our previous attempts to improve the mechanical properties by increasing the volume fraction of the components of the slurry have not been successful due to the difficulty in mixing the viscous slurry at higher volume fractions of the mineral, collagen and GAG [32]. In this paper we describe a new technique to improve the mechanical properties by increasing the relative density of the scaffold by a vacuum filtration technique.
Section snippets
Fabrication of mineralized collagen–glycosaminoglycan suspension
A mineralized CG suspension (50 wt.% mineral) was fabricated using microfibrillar, type I collagen isolated from bovine achilles tendon (Sigma–Aldrich Chemical Co, St. Louis, MO), chondroitin-6-sulfate (GAG) isolated from shark cartilage (Sigma–Aldrich), phosphoric acid (H3PO4) (EMD Chemicals Inc., Gibbstown, NJ), calcium nitrate (Ca(NO3)2·4H2O) and calcium hydroxide (Ca(OH)2) (Sigma–Aldrich). The suspension was prepared by combining collagen (0.019 wt.%), GAG (0.002 wt.%), calcium nitrate
Microstructural characterization
The measured overall scaffold dry densities (ρ∗) for the MCG scaffolds are listed in Table 2. The starting densities of the co-precipitate in the slurry were less than the dry scaffold densities (e.g. for the 1× scaffold, 0.042 g ml−1 and 0.076 g cm−3 are the slurry and dry densities, respectively) partly because some of the solvent was soaked into the collagen reducing the solidified solvent volume (and hence increasing the dry density) [49], [50], [51]. The relative densities can be calculated
Discussion
We were able to increase substantially the Young’s modulus and the crushing strength of the mineralized scaffold in both the dry and hydrated state by increasing the relative density by a factor of 3. The denser scaffolds, 2× and 3×, in the dry state sustained hard thumb pressure; this is a critical criterion for scaffold implantation at a defect site. Cross-linking the scaffolds with EDAC further increased the properties of the hydrated, but not the dry scaffolds. We were able to achieve an
Conclusions
50 wt.% MCG scaffolds with four different relative densities (0.045, 0.098, 0.137 and 0.187) were fabricated via a three step process: (i) titrant-free triple co-precipitation of the slurry; (ii) vacuum filtering the slurry to the desired density; and (iii) freeze-drying the slurry to obtain the dry scaffold. The MCG scaffolds had an open-cell pore structure with both walls and struts, and interconnected pores. While we have not definitively demonstrated by cell adhesion, migration,
Conflicts of interest statement
Lorna J. Gibson has a financial interest in Orthomimetics, a start-up firm that resulted from a previous collaboration on a similar mineralized collagen scaffold. However, the authors did not receive any financial support from Orthomimetics for this project.
Acknowledgments
Funding for this project was provided by the National Science Foundation, Grant No. CMS-0408259. The authors are grateful to Prof. Ioannis Yannas and Prof. Simona Socrate in the MIT Department of Mechanical Engineering, Prof. Elazer Edelman in the Harvard-MIT Division of Health Sciences & Technology, Mr. Philip Seifert in the CBSET Department of Pathology, Mr. Alan Schwartzman in the MIT Department of Materials Science and Engineering, Mr. Kaustuv DeBiswas in the MIT Department of Architecture,
References (65)
- et al.
Regeneration of trabecular bone using porous ceramics
Curr Opin Solid State Mater Sci
(2003) - et al.
Bone scaffolds: the role of mechanical stability and instrumentation
Injury-Int J Care Injured
(2005) - et al.
Cell movement is guided by the rigidity of the substrate
Biophys J
(2000) - et al.
Collagen–GAG substrate enhances the quality of nerve regeneration through collagen tubes up to level of autograft
Exp Neurol
(1998) - et al.
Early peripheral nerve healing in collagen and silicone tube implants: myofibroblasts and the cellular response
Biomaterials
(1998) Preparation and characterization of porous beta-tricalcium phosphate collagen composites with an integrated structure
Biomaterials
(2005)Preparation and histological evaluation of biomimetic three-dimensional hydroxyapatite/chitosan-gelatin network composite scaffolds
Biomaterials
(2002)- et al.
Characterization of mineralized collagen–glycosaminoglycan scaffolds for bone regeneration
Acta Biomater
(2008) - et al.
Mechanical characterization of collagen–glycosaminoglycan scaffolds
Acta Biomater
(2007) - et al.
Optimal design and fabrication of scaffolds to mimic tissue properties and satisfy biological constraints
Biomaterials
(2002)