Solvent based 3D printing of hydroxyapatite laden scaffolds for bone tissue engineering

摘要

Introduction: The classic tissue engineering paradigm uses any combination of cells, a biomaterial scaffold, and bioactive factors to produce tissue. 3D printing is promising biomaterial fabrication method in bone tissue engineering for its inherent ability to create porous scaffolds of arbitrary shape'11. Hydroxyapatlte (HA) is a popular material for bone tissue engineering for Its bioattractrve and osteoinductive properties. We created HA laden scaffolds with a solvent based printing technique and report novel surface architecture resulting from the technique. Materials and Methods: Our 3D printer consists of three linear actuators mounted orthogonally in an aluminum frame to move an extruder assembly along X, Y, and Z axes. Linux CNC drives the printer according to custom G-code, depressing the syringe as the extruder assembly is moved in a series of overlapping ovals Is drawn to approximate a rectangle. The next layer is printed orthogonal to the first to form a grid pattern. The origin of alternate pairs of perpendicular layers is offset by one half the spacing between lines, eliminating the long hollow columns typical of 3D printed scaffolds. Scaffolds are printed with a 1:3 (w:v) polycaprolactone (PCL):chloroform(CF) solution or a 1:2:6 (w:w:v) HA:PCL:CF solution mixed overnight in a rotary mixer. Samples were gold sputter coated and imaged with a SEM at Whitman College. Results and Discussion: We identified a variety of parameters that effect print quality. Print height, tip gauge, and the ratio of travel speed to extrusion rate strongly influenced fiber diameter and drying times. By modifying these parameters, we obtained comparable fiber diameters in samples printed with 22 gauge and 27 gauge dispensing tips. SEM revealed μm scale patterning in both HA laden and pure PCL scaffolds, presumably due to solvent evaporation. Pure PCL scaffolds show longitudinal furrows on the order of tens of μm across and μm scale surface roughness. HA laden scaffolds lack longitudinal furrows, but do have μm scale roughness. This can be explained by decreased drying time for the HA laden samples (2 days) relative to the pure PCL samples (3 weeks). Conclusion: We have successfully developed a system for solvent based printing of biomaterial scaffolds. Solvent based printing provides a convenient method for preliminary materials experimentation without the need to extrude printing filament; allows for printing of non-thermoplastic polymers; and introduces novel surface micro-architecture presumably as an effect of solvent evaporation. Because the technique is uncommon and holds promise for biomaterials manufacturing, we have prepared a detailed troubleshooting guide for other researchers. Future work will include precise determination of porosity, the effect of drying time on surface architechture, compressive strength, and scaffold cyto-compatibility.