Mosaic in Vivo Perturbation (MVP) for systematic assessment of gene function in the eye

Project Summary Genetic screens have been a mainstay of scientific research in animals for over a century contributing immensely to our understanding of molecular cell biology, animal development, and numerous diseases including cancer and neurodegeneration. Despite their success, genetic screens have been limited by several

factors. Pleiotropy, when the same gene is used in more than one process, can mask gene function in one tissue if the same gene is required at an earlier stage. Redundancy, when more than one gene is sufficient for a given function, can prevent the identification of genes with homologues with overlapping expression patterns.

Pleiotropy and redundancy are both quite common in animals precluding the functional assessment of many genes. And finally, traditional genetic screens are very labor intensive, especially in vertebrates such as zebrafish and mice, which has limited their full use. In this multi-PI grant, we will combine the expertise of Dr.

Allon Klein in technology development and computation for single cell approaches with that of Dr. Sean Megason on zebrafish genetics and developmental biology. Here we propose a novel genetic screening approach termed Mosaic in Vivo Perturbation (MVP) that can overcome the difficulties of pleiotropy,

redundancy, and scalability. MVP relies on: i) recent technological advances including single-cell RNAseq, CRISPR/Cas9 pooled perturbations; and ii) the remarkable regulative ability of animals especially fish and mammals to undergo “mosaic complementation”–where wildtype cells fill the niche of mutant cells in genetic

mosaics allowing for the formation of phenotypically normal organisms. Here we propose the development of CRISPR-based MVP libraries using transposon based expression of multiple sgRNAs in zebrafish that will allow for hundreds of cells per embryo to each receive a different defined mutation or combination of mutations

for putatively redundant genes. Given the regulative nature of zebrafish, these embryos will most often be phenotypically normal; however, the proportion of cells with a given genotype that form each cell type will depend on the fitness of each cell's genotype for that cell type (if a gene is important for forming a cell type, its

sgRNA will be underrepresented). Using scRNA-Seq we will jointly genotype (determine sgRNAs) and phenotype (determine cell type) large numbers of cells from MVP zebrafish embryos to quantify the function of all genes in the library for all measured cell types. Here we focus on the eye, because of its large size and

dissectability in zebrafish, remarkable development, and the large unmet medical challenges of eye diseases. In three largely independent aims, we will 1) optimize the MVP approach through iterative testing culminating in measuring the function of the complete TFome in the eye, 2) develop an ultra-scalable readout approach for

cell type based on our novel “capsule” technology, and 3) develop validated whole genome MVP CRISPR libraries. At the conclusion of this work, we anticipate having generated key resources of widespread utility including the MVP approach, CRISPR libraries, and a systematic catalog of gene function in the eye.