Biomechanics of Tissue Engineered Cartilage Produced Using 3D Printing and Bioprinting

INTRODUCTION

The prevalence and cost of arthritis accounts for more than 10% of annual healthcare costs. 3D printing and bioprinting are disruptive technologies that could revolutionize the treatment of arthritic cartilage. This study was undertaken to compare the techniques of 3D printing, bioprinting and a combination thereof in terms of shape fidelity and biomechanics.

METHODS

Chondrocytes were isolated from the knee of rabbits (n=3; 7-9 months) euthanized under an IACUC approved protocol. Cartilage was dissected, minced and digested using hyaluronidase (30 min) and collagenase (8 h) before cryopreservation (95% fetal bovine serum, 5% dimethyl sulfoxide). Chondrocytes were thawed and seeded onto devitalized porcine synoviocyte matrix coated flasks (~6000 cells/cm2). At 90% confluence, cells were trypsinized (0.25% trypsin/EDTA; Hyclone) and split between 8 groups: 1) scaffold-free sheet, 2) scaffold-free dome, 3) direct-write polycaprolactone scaffold sheet, 4) direct-write polycaprolactone scaffold dome, 5) collagen-encapsulated (LifeInk200, Advanced Biomatrix) sheet, 6) collagenen capsulated (LifeInk200, Advanced Biomatrix) dome, 7) collagen-encapsulated (LifeInk200, Advanced Biomatrix) sheet on polycaprolactone scaffold, 8) collagen encapsulated (LifeInk200, Advanced Biomatrix) dome on a direct-write polycaprolactone scaffold. Convex dome molds were 3D printed in acrylonitrile butadiene styrene (ABS, Hictop iprusa) then sterilized by incubation in bleach, neutralization in sodium thiosulphate and rinsed in sterile water before drying in a laminar flow cabinet. Direct write scaffolds were produced with sequentially smaller, then larger, grids (70, 60, 50, 40, 30, 40, 50, 60, 70 µm) in polycaprolactone (RegenHu); domes were produced by incubation over an ABS dome at 55 °C. Bioprinting was performed with chondrocytes resuspended in DMEM/F12 + 10 % FBS mixed 1:5 with collagen (LifeInk200 Advanced Biomatrix, Se3d Reb3l syringe extrusion bioprinter). Constructs were incubated in 30 ml Nalgene containers modified with a 0.2 µm sterile syringe filter for gas exchange (Fig. 1A; 5% CO2, 5% O2, 90% N2). Groups 1-4 were incubated in chondrogenic media [1] immediately, whilst bioprinted constructs (groups 5-8) were incubated in DMEM/F12 supplemented with 10 % FBS for 24 h prior to chondrogenic culture, as previous experiments determined that immediate culture in chondrogenic media resulted in extremely low chondrocyte viability. All constructs were placed on a rotary shaker at 10 rpm on day 2 and 60 rpm on day 5. Constructs were cultured for 28 days, with chondrogenic media changes every 2-3 days. At harvest, visual and tactile assessment was scored 1-5 (5=best) before freezing on dry ice. Mechanical testing: samples were thawed at room temperature in PBS. Then, each dome sample was fixed on a spherical stainless steel layer. Camera-registration (Mapping Toolbox, Biomomentum Inc, Canada) was used to superimpose a position grid (~20-30 positions per sample). Mechanical properties were mapped using a technique allowing for automated normal alignment and indentation mapping with a multiaxial mechanical tester (Mach-1v500css, Biomomentum Inc, Canada; 17N range and 0.85mN resolution on the vertical axis, 12N range and 0.60mN resolution on the horizontal axes). Perpendicular indentation was obtained using a spherical indenter (D=0.50 mm) moved perpendicularly to the surface. The cartilage was indented by 100 µm at 100 µm/s and allowed for a 60 second relaxation time. Afterwards, the spherical indenter was replaced by a needle probe to perform automated thickness mapping at each position of the same grid. Analysis: From the automated thickness mapping results, cartilage thickness was calculated at each position by the difference between the vertical position of the surface (where the load starts to increase) and that of the cartilage/bone interface (corresponding to the first inflection point in the displacement/force curve). The instantaneous modulus at each position was obtained by fitting the load-displacement curve (with corresponding thickness and effective Poisson ratio of 0.5) to an elastic model in indentation [2]. Statistical analysis was performed using ANOVA with Dunnett’s Post-Hoc analysis.

RESULTS

Table 1 and Figure 2 show that scaffold-free constructs maintained their shape fidelity more consistently than the other constructs with greater stiffness. Defects in bioprinted constructs due to bubbles or extrusion issues persisted, i.e. were not filled in by the cells, during culture. Dome shaped constructs formed on ABS molds were stiffer and thicker than those produced on direct-write scaffolds. Chondrocyte contraction of bioprinted constructs was a major problem and prevented biomechanical testing of those tissues.

DISCUSSION

Whilst several aspects of 3D printing and bioprinting were compared, many more approaches in each exist. Further optimization of the approach in each may achieve stiffer, higher fidelity constructs. What is commonly lacking in 3D bioprinting studies is the effect of long-term culture. Here, short term shape congruence was readily achieved but was lost in longer culture as the chondrocytes manipulated their environment and constructs contracted.

SIGNIFICANCE

3D printing and bioprinting have great potential, but side by side comparisons testing functional outcomes such as this are lacking. This study supports the use of scaffold-free, self-assembled, cartilage constructs over the other methods tested.

Figures