(1081-B) High-throughput, AI-driven, pooled phenotypic CRISPR screening by ghost cytometry.
Monday, February 5, 2024
2:00 PM – 3:00 PM EST
Location: Exhibit Halls AB
Abstract: Current high-throughput drug screening approaches are heavily reliant on plate-based imaging using conventional high-resolution microscopy, which is multi-instrument intensive and prohibits scale-up to screening large genomic libraries. Here we present a novel drug screening approach using high-throughput machine vision-based cell sorting (Ghost Cytometry (GC)) to study the effects of drugs and genetic perturbations on cellular morphological phenotypes by flow cytometry. GC combines high-speed morphological profiling in flow with artificial intelligence (AI) to identify and sort cells exhibiting phenotypes of interest.
We show the ability of GC to identify complex immune cell phenotypes used in small and large molecule drug screening including B-cell activation, T-cell health (glycolysis level, viability, exhaustion level) and macrophage polarization. For functional genomics screening, we demonstrate application of GC in a pooled CRISPR screening workflow to identify genes regulating the nuclear translocation of nuclear factor kappa B (NF-κB) in a human leukemia monocytic cell line (THP-1). We screened a pathway-specific library of 7,290 sgRNAs targeting 729 kinase genes in a pooled format in one day. Using the embedded AI in a GC-powered device, LPS (a TLR4 agonist) stimulated Cas9-expressing THP-1 cells transduced with the kinase CRISPR library were sorted based on subcellular morphological changes and the approach identified enrichment of sgRNAs targeting known genes downstream of TLR4 signaling including MAP3K7, IRAK4, IKBKB, and IKBKG.
Here we present GC as a novel method well-suited for next-generation high-throughput drug screening. The approach provides significant advantages over traditional arrayed screens including: (1) scalability to large screening libraries (2) applicability to a diverse repertoire of complex morphological phenotypes in both adherent and suspension cells (3) and compatibility with commercial sequencing platforms to identify target gene perturbations in functional genomics screens.