CARTILAGE 3D PRINTING - 3D Bioprinting and Nanotechnology in Tissue Engineering and Regenerative Medicine - page 263

single system can print with multiple bioinks making it feasible to construct complex tissues containing
multiple cell types ( Xu et al., 2013 ). The small size of individual droplets increases the resolution of the
printed constructs and reduces the amount of bioink needed. However, the higher resolution comes at
the cost of increased printing duration and limits the application to low viscosity inks. Cell settling and
clumping can clog the microchannels and nozzles in an ink-jet system ( Chahal et al., 2012 ).
Laser-induced forward transfer (LIFT), biological laser printing (BioLP), matrix-assisted pulsed la-
ser evaporation-direct writing (MAPLE-DW), and laser-assisted bioprinting (LAB) use a laser to print
bioink, which is spread across an optically transparent “ribbon” ( Figure 12.3 D). The ribbon is suspend-
ed with the bioink orientated downward, over the print substrate or “bio-paper.” A laser then superheats
and vaporizes a thin layer of material on the underside of the ribbon, inducing a jet of bioink on the
target substrate ( Barron et al., 2004; Guillemot et al., 2010; Guillotin et al., 2010; Ringeisen et al., 2006 ).
The need for carefully prepared ribbons and the small ejection size limits the throughput of the process
and printing large multilayered 3D constructs using LIFT techniques has not yet been reported.
While the vast majority of bioprinting research involves one of the four major approaches just
mentioned, other unique techniques are being explored. For instance, patterned chips have been used
for cell placement ( Jing et al., 2011; Rosenthal et al., 2007 ). Cells are preferentially trapped in wells on
the chip, which is then overturned to drop the cells in the well pattern on another substrate. While this
technique is not immediately obvious as a 3D manufacturing method, it has shown potential for inves-
tigating cell-cell interactions and other spatially varying phenomenon. Another uncommon method for
cell placement uses microfluidic channels to deposit cells one layer at a time ( Tan and Desai, 2004 ).
Having control of cell and matrix types at the resolution of microfluidic chips has advantages for both
laboratory investigation and construct development, although the technique is limited to shapes that can
be molded in channels.
12.1.4 APPLYING 3D PRINTING FOR CARTILAGE AND BONE
Initial 3D printing-based approaches have involved forming cell-free 3D scaffolds in vitro that
were subsequently seeded with cells. Sherwood et al. ( Sherwood et al., 2002 ) used the TheriForm TM
process to bind powder particles with a liquid binder to form an osteochondral construct that was
subsequently seeded with cells. A mixture of PLGA and PLA polymer powder particles was bound
using chloroform to generate a porous scaffold. Adding tricalcium phosphate to the subchondral
region generated scaffold mechanical properties that were comparable to bone. Woodfield et al.
( (Woodfield et al., 2004 ) utilized a pressure-driven syringe to deposit molten copolymer fibers onto
a computer-controlled x - y - z table to produce scaffolds comprised of poly(ethylene glycol)-tere-
phthalate-poly(butylene terephthalate) (PEGT/PBT), into which they seeded bovine articular chon-
drocytes. The mechanical properties of these scaffolds could be altered by changing PEGT/PBT
composition and by altering the geometry and porosity. These scaffolds supported a homogeneous
cell distribution and subsequent cartilage-like tissue formation. Normal articular cartilage possess-
es an equilibrium modulus of 0.27 MPa and a dynamic stiffness of 4.10 MPa, and these engineered
scaffolds approximated these qualities with an equilibrium modulus ranging from 0.05-2.5 MPa
and a dynamic stiffness range spanning 0.16-4.33 MPa ( (Woodfield et al., 2004 ). To reconstruct the
zonal arrangements of articular cartilage, high-density pellet cultures of cells derived from different
zones were seeded into PEGT/PBT scaffolds, although conclusive evidence of zonal structure was
not provided ( Schuurman et al., 2013b ).