Josh Huang Sept 6, 2020 RNA-programmable cell-type targeting, editing, and therapy Abstract Systematic identification and manipulation of cell types is necessary for dissecting mechanisms of biological functions in health and disease. Although large-scale, single-cell transcriptome profiling now enables identification of all major transcription-defined cell types in many organisms, easy and systematic experimental access to all major cell populations is needed to physiologically and anatomically validate these statistical “transcriptional clusters” as cell types and, more importantly, to interrogate their roles in tissue organization and function. The difficulty of selectively manipulating cell types remains a critical barrier to such studies. Current approaches to this problem mostly rely on germline DNA engineering, which is slow and expensive and poses ethical issues, especially in humans and other primates. Cell-type transcriptional enhancers afford a non- germline approach, but their identification and validation remain effort-intensive and costly. To overcome these barriers, all of biomedical research urgently needs a novel approach to manipulate cell types in a way that is specific, easy yet comprehensive, affordable, and generalizes across organs and species, akin to CRISPR- based manipulation of genes. Here I propose to develop a paradigm-shifting technology that will enable RNA- programmable cell-type targeting and manipulation based on the fundamental biology of RNA editing. To achieve this breakthrough, I will harness a set of next-generation, multi-functional ribonucleoprotein devices, which can detect the presence of specific RNAs in somatic cells and trigger the expression of effector genes for cell visualization, monitoring, and manipulation. This method builds upon the universal RNA sensing and editing system within all metazoan cells, centered around the editing enzyme adenosine deaminase acting on RNA (ADAR). I term this method CellREADR (Cell access through RNA sensing by Endogenous ADAR). As CellREADR leverages endogenous cellular machinery and is built with a single modular RNA molecule that functions through Watson-Crick base pairing, it is highly specific, inherently programmable, fast, affordable, easy to use, and widely applicable. I propose to build and optimize CellREADR devices in cell-culture systems and validate the method in a highly complex organ - the brain - by targeting and manipulating a large set of neuronal cell types in the mouse cerebral cortex. We will extend CellREADR across species by targeting cell types in ex vivo human brain specimens, and in the macaque and avian brain. Further, we will design intersectional strategies for targeting increasingly specific cell types, and combinatorial strategies for simultaneous and differential manipulation of multiple cell types in a tissue. By linking cell-type RNA sensors to a variety of effector genes that alter cell functions, ranging from ablation to subtle ph...