Abstract: Human induced pluripotent stem cells (hiPSCs) possess the capacity to differentiate into a variety of cell types and hold great promise for clinical applications and disease modeling. However, the efficiency of differentiation into specific lineages, particularly endodermal cells, varies among cell lines, a phenomenon referred to as differentiation propensity. This variability poses a significant challenge for the effective utilization, research, and development of hiPSCs and their derivatives, necessitating the optimization of differentiation methods for each cell line and the selection of cell lines with consistent differentiation capabilities toward desired cell types. Despite these challenges, factors that can overcome the differentiation propensity remain to be comprehensively explored. Using a genome-wide CRISPR-KO screen for definitive endoderm (DE) differentiation and inhibitors targeting hit genes, we discovered that a combinatorial treatment with DOT1L and JNK inhibitors substantially enhanced the efficiency of DE cell differentiation. This improvement was also confirmed in several hiPSC and hESC lines. In addition, DE cells induced by these inhibitors demonstrated the capacity of differentiating into DE derivates, including pancreatic and lung lineages. Mechanistic analysis of DOT1L in DE differentiation revealed that DOT1L suppressed DE differentiation by maintaining the expression of BCOR, which was the 2nd ranked hit gene in our CRISPR screen. Disruption of BCOR similarly enhanced DE differentiation, achieving levels comparable to those observed in DOT1L-KO hiPSCs. Additionally, we found that the expression level of BCOR negatively correlated with DE differentiation efficiency and that variability in BCOR expression was associated with genetic mutations. In summary, we developed a novel DE differentiation method applicable to a wide range of cell lines by identifying differentiation-blocking genes with a genome-wide CRISPR screen. This approach has the potential to be applied to each differentiation stage across various lineages, leading to the development of more robust differentiation protocols.
