Scaling Up Synthetic Cell Production Using Robotics and Machine Learning Toward Therapeutic Applications

Abstract

Synthetic cells (SCs), developed through bottom-up synthetic biology, hold great potential for biomedical applications, with the promise of replacing malfunctioning natural cells and treating diseases with spatiotemporal control. Currently, most SC synthesis and characterization processes are manual, limiting scalability and efficiency. In this study, an automated method is developed for large-scale production of protein-producing SCs for therapeutic applications. The optimized process, compatible with a robotic liquid handling system (LiHa), reduces production time by half. Additionally, incorporation of an automated tissue dissociator-based emulsification increases batch size 30-fold while preserving SC characteristics. To assess SC quality and protein synthesis, artificial intelligence (AI)-based image analysis is employed, allowing for automated, accurate and high-throughput SC characterization. Large-scale luciferase-expressing SCs from a single homogeneous batch are administered to mice, allowing for real-time monitoring of protein expression and reducing experimental variability. By troubleshooting several central steps in SC synthesis, it is demonstrated that automation and computerized quality control can significantly improve the process of SC synthesis for preclinical and clinical applications.