Background: The emergent field of 3D bioprinting has the potential to overcome hypoxia and lack of immunoprotection, two major limitations of islet transplantation in encapsulation systems. When transitioning tissue engineered constructs from bench to bedside, several design parameters must be considered, including tissue geometry, islet density, and oxygen tension in different transplantation sites. To address this challenging multifactorial optimization process, we have designed an in vitro flow device that can be used to evaluate bioprinted tissue performance in vitro under different flow, oxygen and tissue geometry conditions. The aim of this work was to assess the function and viability of single 3D bioprinted core-shell fibres containing pseudo-islets or human islets in a novel perfusion device designed to accommodate bioprinted tissues.

Methods: The perfusion device was designed using computer-aided design modelling. Computational fluid dynamics (CFD) was used to simulate flow and compute input parameters that would ensure laminar, uniform flow. Pseudo-islets were formed after aggregation of MIN6 cells in AggrewellTM culture plates during 48h.  Pseudo-islets and human islets were 3D bioprinted in alginate using an RX1 bioprinter (AspectBiosystems, Vancouver, CA). Free or encapsulated pseudo-islets or human islets were cultured in static conditions as controls. Pseudo-islet function was determined using a glucose stimulation insulin secretion (GSIS) assay in static or perfused conditions. Viability of human islets was evaluated using Calcein AM/Ethidium Homodimer staining.

Results: Based on our computational models, by setting the inlet flow speed to 1.0 cm/s, we can achieve physiological flow velocities within the device. This inlet flow speed is also expected to generate smooth, undisturbed streamlines and laminar flow. After a 24h culture in the perfusion device, we detected increased insulin secretion of pseudo-islets in fibres in response to high glucose stimulation (ratio of secreted to total insulin of 0.27% after 15 min at high glucose vs 0.13% after 15 min at low glucose). Human islets bioprinted in fibres showed higher cell viability compared to free human islets after a 48h culture (95% of viability in fibres cultured in the perfusion device compared to 80% viability for free islets). As these results were obtained from a single human pancreas donor, further studies are needed to assess the reproducibility and statistical significance of these observations.

Conclusion: These promising preliminary results suggest that (1) the flow device we designed can be used to evaluate the performance of 3D bioprinted pancreatic tissue and (2) pseudo-islets and human islets can be safely bioprinted in core-shell fibres. The platform can be used to streamline the characterization and optimize the configuration in vitro of promising artificial tissues at human-scale.