Miniaturization and characterization of patient derived hepatocellular carcinoma tumor organoid cultures for cancer drug discovery applications

Patient derived tumor organoid (PDTO) models retain the structural, morphological, genetic, and clonal heterogeneity of the original tumors. The ability to efficiently generate, expand, and biobank PDTOs has the potential to make the clinical diversity of cancer accessible for personalized medicine assay guided therapeutic drug selection and drug discovery. We describe the miniaturization and growth in 96- and 384-well formats of a single non-tumor liver and two Hepatocellular carcinoma (HCC) organoids derived from cryopreserved PDTO cells and the application of high content imaging (HCI) to characterize the models and enhance drug sensitivity testing. Non-invasive sequentially acquired transmitted light images showed that seeding cryopreserved cells from non-tumoral and HCC PDTOs into 96- or 384-well plates in reduced growth factor Matrigel (rgf-MG) that were fed with growth medium every 3 days supported organoid growth up to 15 days. The number and sizes of organoids increased with longer times in culture. HCC PDTO's had more heterogeneous morphologies than non-tumor organoids with respect to size, shape, and optical density. Organoids cultured in rgf-MG could be stained in situ with HCI reagents without mechanical, chemical or enzymatic disruption of the hydrogel matrices and quantitative data extracted by image analysis. Hoechst and live/dead reagents provided organoid numbers and viability comparisons. HCC PDTO's stained with phalloidin or immuno-stained with α-tubulin antibodies revealed F-actin and microtubule cytoskeleton organization. HCC PDTO's stained with antibodies to signaling pathway proteins and their phosphorylation status allowed comparisons of relative expression levels and inference of pathway activation. Images of HCC PDTO's exposed to ellipticine showed that drugs penetrate Matrigel hydrogels and accumulate in organoid cells. 9-day 384-well HCC organoid cultures exhibited robust and reproducible growth signals suitable for cancer drug testing. Complimenting cell viability readouts with multiple HCI parameters including morphological features and dead cell staining improved the analysis of drug impact and enhanced the value that could be extracted from these more physiologically relevant three-dimensional HCC organoid cultures.