High-throughput single cell transcriptome analysis and CRISPR screen identify key ß cell-specific disease genes

Pancreatic endocrine cells orchestrate the precise control of blood glucose levels, but the contribution of each cell type to diabetes or obesity remains elusive. Here we used a massively parallel single-cell RNA-seq technology (Drop-Seq) to analyze the transcriptome of 26,677 pancreatic islets cells from both healthy and type II diabetic (T2D) donors. We have analyzed cell type-specific gene signatures, and detected several rare a or ß cell subpopulations with high sensitivity. We also developed RePACT, a sensitive single cell analysis algorithm to identify genes associated with rare disease causing cells, or to capture the subtle disease-relevant cellular variation. We successfully identified both common and specific signature genes of obesity and T2D with only a small number of islet samples. We also performed an unbiased genome-wide CRISPR screen and mapped these Drop-Seq signature genes to the core insulin regulatory network in ß cells. Notably, our integrative analysis discovered a ß cell-specific function of the cohesin loading complex in regulating insulin gene transcription, and a previously unrecognized role of the NuA4/Tip60 histone acetyltransferase complex in regulating insulin release. These data demonstrated that single-cell trancriptomics is necessary to dissect the heterogeneity, disease state, and functionality of islet ß cells and other cell types. Overall design: Single cell sequencing (Drop-seq) for Human Pancreatic islet from 9 individuals respectively. Including 6 Healthy donor and 3 Type II diabetes patient donor. 4 Chipseq for further validation.