Ultrasound-assisted biofabrication and bioprinting of preferentially aligned three-dimensional cellular constructs

Abstract

A critical consideration in tissue engineering is to recapitulate the microstructural organization of native tissues that is essential to their function. Scaffold-based techniques have focused on achieving this via the contact guidance principle wherein topographical cues offered by scaffold fibers direct migration and orientation of cells to govern subsequent cell-secreted extracellular matrix organization. Alternatively, approaches based on acoustophoretic, electrophoretic, photophoretic, magnetophoretic, and chemotactic principles are being investigated in the biofabrication domain to direct patterning of cells within bioink constructs. This work describes a new acoustophoretic three-dimensional (3D) biofabrication approach that utilizes radiation forces generated by superimposing ultrasonic bulk acoustic waves (BAW) to preferentially organize cellular arrays within single and multi-layered hydrogel constructs. Using multiphysics modeling and experimental design, we have characterized the effects of process parameters including ultrasound frequency (0.71, 1, 1.5, 2 MHz), signal voltage amplitude (100, 200 mVpp), bioink viscosity (5, 70 cP), and actuation duration (10, 20 min) on the alignment characteristics, viability and metabolic activity of human adipose-derived stem cells (hASC) suspended in alginate. Results show that the spacing between adjacent cellular arrays decreased with increasing frequency (p < 0.001), while the width of the arrays decreased with increasing frequency and amplitude (p < 0.05), and upon lowering the bioink viscosity (p < 0.01) or increasing actuation duration (p < 0.01). Corresponding to the computational results wherein estimated acoustic radiation forces demonstrated a linear relationship with amplitude and a nonlinear relationship with frequency, the interaction of moderate frequencies at high amplitudes resulted in viscous perturbations, ultimately affecting the hASC viability (p < 0.01). For each combination of frequency and amplitude at the extremities of the tested range, the hASC metabolic activity did not change over 4 d, but the activity of the low frequency-high amplitude treatment was lower than that of the high frequency-low amplitude treatment at day 4 (p < 0.01). In addition to this process-structure characterization, we have also demonstrated the 3D bioprinting of a multi-layered medial knee meniscus construct featuring physiologically-relevant circumferential organization of viable hASC. This work contributes to the advancement of scalable biomimetic tissue manufacturing science and technology.

Characterizing the Process Physics of Ultrasound-Assisted Bioprinting

“…3e) indicated that the experimentally measured T construct after 30 min (38.9 °C) was close to the computational model estimate (37 °C). The slightly higher experimental T construct could be attributed to fluid streaming-induced perturbations observed for the 1.5 MHz transducers 40 , which may have caused a mixing of the high temperature fluid near the transducer with the low temperature fluid in the construct region.…”

“…This could be attributed to the viscous streams at the initiation of each excitation burst, wherein the PBS is scattered away from the transducer every 1 s. For each construct, the top layer exhibits a greater degree of 3D alignment, while the bottom layer exhibits a greater degree of alignment near the top. This could be attributed to a greater compliance of the transducer surface near its top region as compared to its bottom 40 . This non-uniform compliance could also explain the increased tilting of cellular arrays in the top layer.…”

“…The image for the side view of the 1.5 MHz–64 Vpp construct is not shown here because the large amount of perturbations, large radiation forces, and exposure to high temperatures for 60 min lowered the cell viability significantly, thereby lowering the neutral red fluorescence below detectable levels. It should also be noted that some artifacts in the constructs and images resulted from slicing procedures, non-homogenous crosslinking, and fluid streaming-induced perturbation effects, which were more pronounced for layers at greater z-heights 40 .

Figure 6( a ) Macroscopic and (b) microscopic fluorescence images of neutral red-stained MG63 cellular alignment within bilayered constructs as viewed from the top and across a side section. The images of 1.5 MHz–64 Vpp group have not been shown because the alignment was not visible due to low fluorescence (resulting from a poor cell viability) and heavy perturbations.

…”

“…This might not be the case when a photo-crosslinkable hydrogel such as GelMA is used, wherein uninterrupted increase in temperature may lead to reduction in viability. This effect would be further exacerbated by streaming-induced perturbations and temperature shocks at higher duration or voltage amplitude of excitation 40 . However, since the viscosity of the liquid GelMA at 37 °C is lower than that of the alginate used in this study 56 , comparatively lower pressure amplitudes can be used to induce alignment.…”

Mechanical Considerations of Bioprinted Tissue

“…In a different application, but still relevant to bioprinting, computational force modeling of the tip movement inside gels with different viscoelastic properties was used to design and create channels within the material (Barua et al, 2020). Similarly, a combination of physical modeling of acoustic waves and engineering solutions was used to pattern cells within a bioprinted cartilage precursor via ultrasound (Chansoria et al, 2019).…”