(268c) Mechanical Properties and Permeability of Collagen-Gag Scaffolds for Tissue Engineering: Cellular Solids Modeling and Experimental Results

The mechanical properties of 2D substrates and 3D scaffolds have been observed to affect cell migration and contractile behavior. The permeability of tissue engineering scaffolds controls diffusion-based metabolite and waste transport to and from the scaffold and influences the final solution pressure distribution in the scaffold. Both of these parameters can significantly influence cell behavior and overall scaffold bioactivity. To perform quantitative cell behavior studies in a scaffold, detailed mechanical and structural analysis is required to understand the local cell environment.

Here we have measured the microstructural, mechanical, and permeability properties of a series of collagen-glycosaminoglycan (CG) scaffolds fabricated via freeze-drying: uniform scaffolds with homogeneous pore structure and equiaxed pores, constant composition and relative density (R_d, 0.6%), but with distinct pore sizes (151, 121, 110, 96 µm) have been produced [1,2]. The scaffold microstructure is characteristic of an open-cell foam [1]; cellular solids theory suggests that the scaffold Young's modulus (E^*) and compressive strength (σ_el^*) depend on scaffold R_d and the Young's modulus of the scaffold material (E_s) [3]. After fabrication, all scaffolds were crosslinked via dehydrothermal (DHT) crosslinking (105^oC, 24hrs, <50mTorr) [1]; two intensities of carbodiimide (EDAC) crosslinking were also used to increase scaffold stiffness relative to DHT independent of pore microstructure [4]. Mechanical characterization was performed on dry and hydrated CG scaffolds with pore sizes ranging between 96 and 151 µm and a constant R_d of 0.6% [2]; mechanical tests were also performed on a series of scaffolds with a range of R_d (0.6, 0.9, 1.2, 1.8%). The modulus of individual scaffold struts (E_s) was measured via AFM, allowing comparison of cellular solids predictions and experimental results of scaffold mechanical properties.

The CG scaffolds were found to be mechanically isotropic. E^* and σ_el^* of the dry scaffolds (DHT, R_d 0.6%) was 30,000 ± 3900 Pa, and 5150 ± 530 Pa, respectively.  E^* and σ_el^* of the hydrated scaffolds (DHT, R_d 0.6%) was 208 ± 41 Pa and 21 ± 8
Pa. Both the E^* and σ_el^* increased with the degree of crosslinking (2.0 and 7.2x stiffness of DHT using two distinct EDAC intensities). E^* and σ_el^* were found to be independent of mean pore size, but dependent on scaffold relative density (E α R_d¹, R_d: 0.6 ? 1.8%).

The Young's modulus of the dry individual scaffold strut  (DHT crosslinked, R_d 0.6%) was measured experimentally via AFM to be 760 ± 35.4 MPa. Comparison between the normalized modulus (E^*/E_s) and collapse strength (σ_el^*/E_s) with scaffold relative density was performed and results were correlated with cellular solids model predictions of scaffold mechanical properties. Note that cellular solids models for open cell foams indicate a modulus dependence on the square of R_d; the observed best-fit linear dependence is likely due to structural heterogeneities in the higher density scaffolds used in this experiment.

A cellular solids description of scaffold permeability (K) in terms of scaffold mean pore size (d), relative density (R_d), and applied strain (ε) has been developed:

K = C₁ · d² · (1 ? ε)² · (1 ? R_d)^3/2

Fig. 1. Experimentally measured (solid bars) vs. cellular solids model predicted (striped bars) permeability as a function of mean pore size and compressive strain.

This model has been utilized successfully to model the permeability of four uniform CG scaffold variants (Mean pore size: 96 ? 151 µm) under a variety of compressive strains (0 ? 40%), K: 0.2E-10 ? 1.4E-10 m⁴/Ns [5]. The excellent comparison between experimentally measured and cellular solids model predicted scaffold permeability suggests that cellular solids modeling techniques can be used as a predictive model of scaffold permeability for many different scaffold architectures under a variety of physiological loading conditions.

  References

[1] F. J. O'Brien, B. A. Harley, I. V. Yannas and L. J. Gibson, Biomaterials 25, 1077-1086, 2004.

[2] F.J. O'Brien, B.A. Harley, I.V. Yannas, and L.J. Gibson, Biomaterials, 26(4):433-441, 2005.

[3] L. J. Gibson and M. F. Ashby, Cellular Solids: Structure and Properties. 2^nd Ed.,
Cambridge Univ. Press, 1997.

[4] C.R. Lee, A.J. Grodzinsky and M. Spector, Biomaterials, 22:3145-3154, 2002.

[5] F. O'Brien, B. Harley, M. Waller,
I. Yannas, L. Gibson, P. Prendergast, Submitted, Technol. Health Care, 2005

Keywords: scaffolds, mechanical properties, modeling, permeability