Marion Nikolas Gaeve, Fraunhofer IPA, Stuttgart

For the past decades, cell based assays and animal models have represented the gold standard test systems for fundamental scientific research and pre-clinical validation of novel therapeutic agents. Despite their crucial role in these fields, both systems harbor individual disadvantages in the establishment, maintenance, application and relevance and are often only partially adequate for implementation.

Organoids represent a novel, uprising test system by combining the advantages of both standards. The microspheres typically consist of human pluripotent stem cells or patient-derived tumor cells developed into specific organ-like tissues within days to weeks. The cells naturally aggregate and grow in 3D formation, and the resulting organoids can be subsequently used, e.g. for high-quality research or drug screening. Organoids resemble a complex environment similar to human in-vivo tissues with superior species and physiological relevance compared to conservative 2D cell based assays or animal models.

Despite the increasing interest in the new test system, production of organoids is very limited by complex protocols, high production costs, long cultivation times and low yields. After formation, the resulting spheres may vary dramatically in their capability of representing a functional organoid. Early identification and separation of suitable microspheres can increase the productivity while reducing the overall production costs, but isolation of functional organoids is a tedious and work-intensive process. Modern single entity isolation technologies can detect, isolate and deposit a large variety of microbodies for further processing. However, commercially available products differ from each other in many ways, including general principles for detection and deposition, and bear individual advantages and disadvantages in terms of precision, scalability, throughput, and induced cellular stress.

The Fraunhofer Institute for Automation and Manufacturing Engineering IPA in Stuttgart, Germany, aims to reduce these limitations by developing automated technological solutions for organoid handling and production. To overcome the challenges of efficient organoid sorting, we developed an AI-driven technology to analyze and isolate single spheroids from suspension in a non-harmful, high-throughput and automated approach. The spheres are applied into a microfluidic setup, investigated via light microscopy, classified by the machine learning algorithm and transferred into a target vessel according to their morphological properties. Initial experiments resulted in robust recognition and separation of organoids from suspension. The system can be used to improve development and validation of new standard protocols for organoid differentiation into any functional tissue. In addition, increased production of functional organoids with already well-established protocols contributes to serving the growing demand for the new model system for drug target research and other applications worldwide.