Automated Assembly and High Throughput Screening of Vascularized Human Organoids from Normal or Tumorigenic Tissues

Avner Ehrlich ¹ Mohammad Ghosheh ¹ Aaron Cohen ¹ Konstantinos Ioannidis ¹ Muneef Ayyash ¹ Idit Goldfracht ² Orian Leitersdof ¹ Merav Cohen ¹ Christopher Hubert ³ Lior Gepstein ² Yaakov Nahmias ¹

¹Department of Cell & Developmental Biology, Alexander Grass Center for Bioengineering, The Hebrew University of Jerusalem, Israel
²Sohnis Research Laboratory for Cardiac Electrophysiology and Regenerative Medicine, The Rappaport Faculty of Medicine and Research Institute, Technion, Israel
³Lerner Research Institute, Cleveland Clinic, USA

Vascularized human organoids mimic the microphysiological complexity of native tissues, showing many aspects of normal organization and function. However, current methods of producing organoids result in a random distribution of aggregates leading to significant variability between cultures. Data is also limited to microscopy and end-point analysis, further limiting organoid utility in screening applications.

Here, we report a robotic platform capable of robustly producing hundreds of identical human liver, cardiac and kidney organoids, as well as patient-derived mini-tumors, in a 384 well plate format. Our vascular organoids range from 400 to 800 µm in diameter, showing less than 10% variability between wells. In parallel, we developed a table-top bioanalyzer (DynamiX™) that permit continuous real-time tracking of metabolic function over days in culture, using tissue-embedded microsensors.

We demonstrate E6/E7^LOW-derived human liver organoids, showing complex metabolic zonation, as unique a model for non-alcoholic fatty liver disease (NAFLD). Human iPSC-derived cardiac organoid, with spontaneous, synchronized beating, as a kinetic model of ischemia-reperfusion injury. And, E6/E7^LOW-derived human proximal tubule organoids, as a model of drug-induced nephrotoxicity.

Finally, we will display the expansion of patient-derived glioblastoma multiforme (GBM) vascularized mini-tumors containing cancer stem cells (CSCs) sub-populations and in vivo-like hypoxic tumor core. Using the DynamiX™ platform we screen hundreds of FDA approved drugs in 2 different patient-derived models distinguishing on the fly between drugs that only damage the tumor mass (False Positive) to slow working drugs that are missed in most screens (False Negatives) but that specifically damage the tumor persistent cell population.