Determining lineage decisions and gene regulatory networks governing the generation of key progenitor cell types during early human brain development

点击翻译按钮获取中文摘要

Abstract The long term goal of this proposal is to quantitatively understand how gene regulatory networks (GRNs) generate the diversity of cell types during the development of the human brain. The focus of this proposal is to determine how key progenitor cell types that are uniquely enriched in humans are generated. Such an understanding is essential for uncovering the mechanisms of human developmental diseases. There are three challenges to achieving this goal: 1. Ethical issues in working with developing human tissue, 2. Computational and experimental techniques to determine the sequence of progenitor cell states and state transitions that give rise to the diversity of cell types, 3. the difficultly in building quantitative models of the gene regulatory networks in the absence of data to determine the thousands of biochemical constants. The approach of the proposal is to build the necessary computational, mathematical and experimental framework to overcome these challenges. To recapitulate early human brain development, the proposal will employ an in vitro human embryonic stem cell differentiation system. To obtain snapshots of the underlying gene regulatory network, high throughput single cell sequencing will be employed to obtain transcriptional profiles of thousands of single cells during the course of development. The challenge of inferring the sequence of cell states and cell state transitions will be overcome through a novel statistical method to obtain a joint probability distribution of the cell states, sequence of transitions and a key set of genes whose dynamics reflect these states and transitions. The inferences will be tested by mapping to in vivo data and using viral lineage tracing. The origins of forebrain and outer radial glial cells (oRG) progenitors uniquely enriched in the developing human forebrain will thus be determined. The challenge of building predictive models will be overcome by using methods from theoretical physics and ensemble modeling from statistics to build models that make probabilistic predictions. By using the available data as constraints on the model, the framework will extract joint probability distributions of all the parameters of the model. These distribution functions will then be used to produce probabilistic predictions about the responses of the underlying GRNs to perturbations. High probability predictions will be tested experimentally by perturbing gene expression and signaling during early brain development and the model will be iteratively improved. The success of this proposal will result in the first quantitative model of the gene regulatory network controlling the generation of forebrain and the oRG progenitor cells. If achieved, this work therefore would represent a major insight into the molecular and cellular events that give rise to the disproportionately gyrated human brain.