National Center for Advancing Translational Sciences (NCATS), Maryland, United States
Abstract: While chemical screens remain at the forefront of drug discovery, advances in functional genomics have positioned CRISPR screens as powerful tools for understanding disease biology to better target it. In order be successful, pooled CRISPR screens rely on a well-defined and scalable biological model, efficient editing activity as well as reproducible and robust readouts. Furthermore, perturbation type and approach should be dictated by their feasibility in the relevant biological model. By satisfying those conditions and using a highly scalable and efficient model of inducible pluripotent stem cell (IPSC)-derived neurons, the I3N model, we have applied a pooled CRISPR interference (CRISPRi) approach to tackle two conditions resulting in neuronal degeneration and loss: vincristine-induced neuronal cell death and Niemann Pick Disease Type C (NPC) neuronal degeneration. Vincristine is a chemotherapeutic alkaloid used to treat various cancers. It is known to interfere with microtubule formation and exhibits dose limiting neurological side effects. Using a whole genome pooled positive selection CRISPRi screen, we uncovered a transcriptional network underlying vincristine’s toxicity in neurons. Validation experiments demonstrated that knockdown of single genes in that network prevents vincristine induced neuronal cell death. On the other hand, NPC is a lysosomal storage disorder driven by NPC1 gene mutations, resulting in aberrant cholesterol accumulation, which is the underlying cause of neurodegeneration. Using CRISPR-edited I3N that model Niemann Pick Disease Type C (NPC), we applied a pooled focused library CRISPRi phenotypic screen approach with an optimized fluorescence activated cell sorting (FACS) protocol against fluorescently conjugated perfringolysin O (PFO), a toxin known to bind unesterified cholesterol, and analyzed low and high PFO labeled neurons. Subsequently, we delineated several targetable pathways whose repression by CRISPRi reverses excessive cholesterol in NPC-/- neurons, restoring them to a wild-type phenotype. By comparing those pathways with a parallel whole genome NPC-/- synthetic lethality screen in the same neurons, it was possible to pinpoint specific targets whose repression by CRISPRi decreased aberrant cholesterol accumulation while maintaining neuronal survival in NPC deficient neurons. Taken together, these approaches underscore the importance of harnessing the power of available functional genomics tools and assay design to advance target discovery.