CFG - Pooled Screening

Examples screens 

A number of examples for screens performed at CFG:

• Protein-coding genes: Genome-wide KO, inhibition, activation
• lncRNAs: Inhibition and activation of 20,000 lncRNAs
• Non-coding elements: CRISPR-inhibition on open chromatin regions
• RNA targeting: Cas13d screen for targetable loci in SARS-CoV2
• Synthetic interactions: Cas12a dual guide screen, 112,000 guide combos
• 3D models: Genome-wide CRISPR-KO screen in pancreatic organoids
• Primary cells: miRNA-targeting screen in primary keratinocytes
• Perturb-Seq: Knock-out 50 genes in primary epithelial cells

Creation of Cas expressing cell lines

Verified and tested Cas-expressing cell lines are crucial for a successful screen. We use our own constructs expressing Cas-enzymes and a fluorescent marker, to create pools of cells stably expressing these enzymes, either by lentiviral or piggybac-based delivery. The fluorescent marker allows sorting of high expressors, and regular monitoring of the cell line, which is often needed due to silencing of integrated constructs.

Off-the-shelf and custom-made libraries

CFG clones all libraries in house to enable barcoding of guides. We have the following barcoded libraries ready for use:

• Human and mouse genome-wide KO library, 4 guides per gene
• Human genome-wide CRISPR-inhibition library, 5 guides per gene
• Human genome-wide CRISPR-activation library, 4 guides per gene
• Human chromatin modifiers (around 1,200 genes), 4 guides per gene

For custom libraries, we work together with the client in designing a suitable library, purchase the oligo pools, clone, sequence verify and package into lentivirus.

Lentiviral library delivery

Most commonly, the lentiviral library is then transduced into Cas expressing cells at a low multiplicity of infection, and infected cells are selected either by puromycin or fluorescence. This assures that most cells will contain precisely one integrated guide. In this way, a mutagenized cell population is established, where each cell has one perturbation that is flagged by the guide the cell contains.

Screens with phenotypic separation

This mutagenized cell population is then subjected to a selection pressure such as time, a drug, an infectious agent, etc (commonly live/dead screens, proliferation/arrest screens). Cells that cope better with the selection pressure will enrich, and cells that cannot handle the pressure will deplete from the population, and so will the guides these cells contain. Alternatively, a FACS sortable phenotype can be used to separate different phenotypes from each other (cell morphology, fluorescence readouts). In FACS, cells containing guides that promote the sorted phenotype will be enriched, and cells containing guides that repress the sorted phenotype will deplete. CFG performs all steps downstream of phenotypic selection – genomic DNA isolation, amplification of the guide cassettes, NGS library preparation, submission to NGI, and data analysis.

Screens with single cell transcriptomic readout

For smaller screens (up to around 500 guides), single cell RNA-Seq can be used to read out both the guide and the transcriptome of single cells (Perturb-Seq, CROP-Seq). CFG has a seamless pipeline with the SciLifeLab Eukaryotic Single Cell Genomics (ESCG) unit located at SciLifeLab Campus Solna. CFG supports all steps from library- and strategy design to a library-transduced cell population; helps with performing the desired treatment, and FACS sorts if needed. Viable cells are then delivered to ESCG, who perform the single cell transcriptomics and data analysis.

Scanning mutagenesis screens with base editors

In a pooled scanning mutagenesis screen, a guide library is created that tiles a protein or several proteins of interest. Cas9 fused to base-editing enzymes is then used to introduce point mutations at the guide binding site. Thus, a cell pool expressing a large variety of mutated versions of the protein(s) of interest is created. The location of any mutations conferring a fitness benefit in the assay of choice can then be narrowed down by determining enriched guides. The mutation itself is then read out by amplicon sequencing. The method can be used to study drug-target interactions, protein-protein interactions, for epitope-mapping, etc. CFG performs all steps downstream of phenotypic selection – genomic DNA isolation, amplification of the guide cassettes and of mutated amplicons, NGS library preparation, submission to NGI, and data analysis.

Screens with optical phenotyping – in development

Due to the requirement of physical separation of phenotypes, complex or subtle phenotypes cannot be easily interrogated in conventional pooled screens. Screens with single cell transcriptomic readout overcome this limitation, but are too expensive if performed at scale. We are thus implementing guide readout directly on a microscopy slide in collaboration with Mats Nilsson’s lab at SU and the SciLifeLab ISS unit. With this method, cells can be phenotyped by microscopy, followed by determining the guide present in each cell by padlock-based rolling circle amplification and microscopic deconvolution of padlock barcodes by detecting rolling circle products with cycles of hybridization and stripping of fluorescently labelled deconvolution oligo libraries.