A scale-down mini-bioreactor for the acceleration of data-driven bioprocess optimisation in cell therapy.

Abstract

Cell therapies have demonstrated great potential for treating a broad range of diseases where conventional treatments have failed. However, long development times and sub-optimal processing conditions often hinder their clinical translation. To efficiently develop optimal bioprocesses, a large number of experiments are required, making the screening process lengthy, costly, and resource-intensive. To address these challenges, we present a modular and 3D-printed scaled-down mini-bioreactor that enables parallelization of stirred cell cultures. In addition, the bioreactor system is coupled to real-time monitoring of critical parameters within the cell culture environment, offering the ability to generate multiple time-series data required for artificial intelligence-driven bioprocess development. In this study, a sequential screening design was employed, enabling the efficient evaluation of different combinations of bioprocess parameters (initial cell inoculum, cell-to-microcarriers surface area ratio, and rotation speed). This strategy facilitated rapid, cost-effective, and efficient convergence toward the optimal process conditions. Furthermore, the integrated sensor system demonstrated the feasibility of implementing a soft-sensing framework using metabolic indicators (dissolved oxygen, pH, glucose, and lactate) to non-invasively and non-destructively estimate cell number and gain insights into culture dynamics. Following dynamic expansion in the mini-bioreactor, several analyses were performed to confirm and assess the stemness and multipotency of the cells, which successfully underwent osteogenic, chondrogenic, and adipogenic differentiation.