3D Printing of Bone-Templated Scaffolds

Abstract

SJ Fernando1, JA Sterling2,3, SA Guelcher1 1Vanderbilt University Medical Center, Department of Chemical and Biomolecular Engineering, 2 Veterans' Affairs Tennessee Valley Healthcare System, Nashville, 3Vanderbilt University, Department of Clinical Pharmacology Statement of Purpose: Biomimetic 3D tissue engineered systems have been proposed for investigating molecular mechanisms of disease progression and for screening drugs [1]. We have utilized 3D printing technology to investigate how the mechanical and topological properties of the bone microenvironment influence both cellular interactions and tumor progression in bone. Trabecular curvature, pore size, and pore shape have also been reported to affect the rate of new bone formation. Cells sense and respond to radii of curvature larger than themselves, and the rate of new bone formation increases with the curvature of the surface [2]. The Structure Model Index (SMI) parameter characterizes the structure of trabecular bone, with 0 representing plate-like (e.g., 2D) structures and 3 representing rod-like structures.[3] The trabecular structure largely varies between values of 0 and 3, rather than being predominantly plates or rods. Vertebral trabeculae have an SMI typically closer to a rod-like structure (SMI=3), whereas femoral trabeculae have structure closer to plates (SMI=0). We reason that 3D scaffolds replicating the mechanical and topological properties of bone at different anatomic sites will enable the development of 3D cell culture models for investigating the spatio-temporal dynamics of both bone cellular physiology and cancer progression [1]. We have developed a fabrication process in which wax templates of trabecular bone are prepared by a 3D inkjet printer and subsequently filled with reactive polyurethanes to create scaffolds with elastic modulus and SMI comparable to human bone. Methods: Human cadaver samples from the proximal tibia, proximal femur, and lumbar vertebrae were obtained from the Program in Advanced Anatomy and Simulated Skills Program at Vanderbilt. Isolated tissue samples were imaged using uCT technology from Scanco Inc. Using Scanco software the scans were converted into STL images, a format representative of the surface of bone uCT images, which specifically captures the trabecular architecture. These images were inverted using the same software to create a representation of the trabecular spacing (Tb.Sp.) The STL images of Tb.SP were then uploaded into the Solidscape Studio 3Z Printer for fabrication. The resulting wax molds were filled with a reactive two-component polyurethane (PUR) composed of lysine diisocyanate (LDI), a poly(e-caprolactone-coglycolide) triol (Mw=300 or 3000 g/mol), and iron(III) acetylacetonate (FeAA) catalyst. The polyol (900 Da) was prepared from glycerol starter and backbone comprising 70wt% ɛ-caprolactone, 20wt% glycolide, and 10wt% D,L-lactide. The mixture was cured overnight under vacuum at 80° C. The cured polyurethane structure was extracted from the wax mold using acetone solution to dissolve the wax (sulphonamide derivatives, polyester resin, and benzoate derivatives,) and dried under vacuum overnight. The resulting structures were imaged by SEM and uCT. Nanoindentation was performed on 2D films and on the 3D biomimetic PUR scaffolds. Rat bone marrow-derived stromal cells and monocytes, as well as bone-metastatic MDA-MB-231 breast cancer cells were incubated with the scaffolds under standard cell culture conditions.