High-Throughput Genetic Interaction Sequencing in Mammalian Cells

Project Summary The goal of our R21 project is to develop a powerful platform that will be able to generate and assay millions of combinations of CRISPR/Cas9 single-guide RNAs (sgRNAs) or other genetic perturbagens. Doing so will clear the way for systematic exploration of genetic interactions in mammalian cells. There are many reasons to develop this high-throughput mammalian genetic technology. Currently there is no systematic method to unravel the assortment of genetic interactions that drive specific cancers and determine the variability of individual treatment responses. Nor is there a robust method to study gene interactions in other complex multigenic diseases such as Parkinson's. We are motivated by the tremendous advances recently made in yeast and worms that have resulted from systematic discovery of genetic interactions. Genetic interaction maps in yeast have revealed functional relationships within and between protein complexes orders of magnitude beyond those revealed by protein- protein interaction screens. Systematic screening 65,000 pairs of genes in worms led to the identification of a class of highly connected `hub' genes encoding chromatin regulators. The technologies behind these discoveries depend on several high-throughput steps, including a dependable gene knockout or knockdown method, a method to deliver two gene knockouts/knockdowns into the same cell and to monitor which cells receive which combinations, and also a reliable assay to measure relative fitness. Our major enabling technology for development of a high-throughput system for mammalian cells is the tandem-integration landing pad that allows two plasmids to be inserted next to each other at a neutral location of the genome. Each plasmid contains a DNA barcode that uniquely identifies the associated genetic perturbagen (e.g. sgRNAs). When both plasmids are integrated into the genome, the two barcodes are in close enough proximity to be sequenced together by paired-end amplicon sequencing. We have established this methodology in yeast and have shown that it can generate a library of >108 double barcoded cells via pooled sequential plasmid transformation and integration. The fitness of large double barcode libraries can then be measured using the fit-seq approach that we pioneered: pooled growth and double barcode amplicon sequencing over several time points accurately measures the relative fitness of each double barcoded cell in the pool. In yeast, Genetic interaction Sequencing (GiSeq) promises to be a cheaper and higher-throughput alternative to the commonly used synthetic-genetic array technology. In mammalian cells, GiSeq promises to be a major leap forward over existing technologies: not only will genome-scale interaction libraries become practical, but negligible work will be needed to repeat a screen in a different cells or different conditions. For this proposal, we will establish the utility of GiSeq in mammalian cells (Aims 1 and 2), and prepare reagents to perform genetic interaction screens in vivo (Aim 3).