Mapping information-rich genotype-phenotype landscapes with genome-scale Perturb-seq

Maximizing CRISPRi efficacy and accessibility with dual-sgRNA libraries and optimal effectors

“…In the realm of CRISPRi effectors, our work points to a clear consensus: Zim3-dCas9 is the effector of choice, as it appears equal or superior to other effectors in every test we performed and had no downsides. We had previously measured by Perturb-seq that Zim3-dCas9 afforded median mRNA knockdown of 91.6% across 2,285 genes in RPE1 cells (Replogle et al, 2022), and here we further found that Zim3-dCas9 mediates robust knockdown across a range of cell types. Our work highlights the importance of using multiple assays to assess effector function including single-cell assays to assess cell-to-cell heterogeneity, of directly measuring knockdown instead of relying on proxies such as growth phenotypes that conflate multiple factors, and of evaluating effectors in stably transduced cells rather than in transiently transfected cells to evaluate longer-term consequences for cell viability.…”

“…Like other CRISPR approaches, CRISPRi has been paired with large-scale sgRNA libraries to conduct systematic genetic screens. Such screens have been deployed to identify essential protein-coding and non-coding genes (Gilbert et al, 2014; Haswell et al, 2021; Horlbeck et al, 2016a; Liu et al, 2017; Raffeiner et al, 2020), to map the targets of regulatory elements (Fulco et al, 2019, 2016; Gasperini et al, 2019; Kearns et al, 2015; Klann et al, 2017; Thakore et al, 2015), to identify regulators of cellular signaling and metabolism (Coukos et al, 2021; Liang et al, 2020; Luteijn et al, 2019; Semesta et al, 2020), to uncover stress response pathways in stem cell-derived neurons (Tian et al, 2021, 2019), to uncover regulators of disease-associated states in microglia and astrocytes (Dräger et al, 2022; Leng et al, 2022), to decode regulators of cytokine production in primary human T-cells (Schmidt et al, 2022), to define mechanisms of action of bioactive small molecules (Jost et al, 2017; Morgens et al, 2019; Sage et al, 2017), to identify synthetic-lethal genetic interactions in cancer cells (Du et al, 2017; Horlbeck et al, 2018), and to identify genetic determinants of complex transcriptional responses using RNA-seq readouts (Perturb-seq) (Adamson et al, 2016; Replogle et al, 2022, 2020; Tian et al, 2021, 2019), among others.…”

scPerturb: Harmonized Single-Cell Perturbation Data

“…A multivariate distance measure between sets of cells resulting from different perturbations is needed to perform effective statistics comparing perturbations. With such a distance measure, one can describe the difference or similarity between sets of cells treated with distinct perturbations to infer difference or similarity in terms of mechanism or perturbation target; shared mechanisms tend to produce similar shifts in molecular profiles (Replogle et al, 2022;Tian et al, 2021). A number of distance measures for scRNA-seq have been explored by the single-cell community in recent years, including Wasserstein distances (Chen et al, 2020), maximum mean discrepancy (Kimmel et al, 2021), neighborhood-based measures (Burkhardt et al, 2021;Dann et al, 2022), E-distance (Replogle et al, 2022), and distances in learnt autoencoder latent spaces (Lotfollahi et al, 2019).…”

Transcriptional Regulation of the Hippo Pathway: Current Understanding and Insights from Single-Cell Technologies

“…These single-cell CRISPR screen technologies can provide rich phenotypes for massive numbers of perturbations. A recent effort has expanded Perturb-seq to the genome-wide level for the first time [ 84 ]. Besides CRISPR-based gene perturbations, scRNA-seq can also be combined with the pooled ORF overexpression to enable screening for cell reprogramming [ 85 , 86 ] or coding variants of oncogenes [ 87 ].…”