Abstract 158: How low can you go: Maintenance of tumoroid phenotype with a highly scalable and automation-compatible reduced-ECM suspension culture method

Tumoroid cultures, also known as cancer organoids, have been shown to maintain patient-specific mutational and gene expression profiles over the course of long-term culture better than traditional cancer lines. Despite the physiological relevance of tumoroid models, they have yet to supplant cancer lines, particularly for high-throughput screening (HTS) applications, in large part due to the relative difficulty of the culture workflow. The current gold standard tumoroid culture method involves embedding the cells into a scaffold (most commonly basement membrane extract or BME), which is highly manual, costly in time and resources, and difficult to implement in HTS workflows. We have developed a novel tumoroid culture medium, named GibcoTM OncoProTM Tumoroid Culture Medium, and method in which the addition of diluted BME to tumoroid suspension cultures preserves apical-in polarity (equivalent to that of embedded cultures) while leveraging the benefits of a suspension culture workflow. Additionally, histology for a subset of suspension cultures have been compared with xenograft tumors formed in mice and demonstrate that the suspension culture maintains similar cellular organization. Our system enables a single user to generate hundreds of millions of cells, a feat that would require hundreds of BME domes using standard tumoroid culture methods, and has been shown compatible with colorectal, lung, pancreatic, and head and neck tumoroid lines that were derived in embedded culture with various complete media formulations. Critically, our data confirm that, when paired with our novel tumoroid culture medium, suspension culture maintains patient-specific characteristics comparably to embedded culture when passaged side-by-side for multiple months. Tumoroid lines representing the four unique cancer indications evaluated maintained patient-specific mutational and gene expression profiles (both ≥ 90% correlated with the original material) over the course of culture. An evaluation of differentially expressed genes from a panel of over 20,000 human RefSeq genes showed that there was no change (greater than a two-fold increase/decrease; p-value < 0.05) in gene expression for ≥ 98% of genes, on average, between embedded and suspension culture. Following scale up in suspension culture, we also show that tumoroids can be plated using automated liquid handling techniques for downstream HTS assays. Taken together, our data indicate that this suspension culture workflow pair