Elucidating the Molecular Mechanisms of Neuropsychiatric Symptoms in Alzheimer's Disease

摘要 阿尔茨海默病（Alzheimer's disease，AD）是痴呆症最常见的病因，也是最常见的神经退行性疾病， 这种疾病在全球范围内流行，在美国每8个65岁以上的人中就有1个受到影响。在AD中，约40-60% 个体受精神病性症状（AD+P）的影响，这与更快的认知能力有关。 下降，残疾、死亡率和照顾者负担增加，导致疾病不成比例地严重 负担最近的研究表明AD+P风险的遗传基础，但AD+P的分子基础在很大程度上仍然是 这阻碍了对适当治疗和新疗法的探索。在本提案中，我们 通过系统的生成、整合， 跨两个脑区和四个细胞的转录和表观基因组表型的实验解剖 类型(1)我们分析了192例尸检中的单细胞RNA-seq和细胞类型特异性H3 K27 ac ChIP-seq。 大脑样本，每个样本在两个区域中，横跨患有精神病的AD患者，没有精神病的AD患者， 无AD的精神分裂症患者和对照个体。(2)我们将得到的数据集与遗传 信息和GWAS数据来预测AD+P的驱动基因、区域、变体和途径， 用于因果关系、中介分析和遗传贝叶斯精细分析的最先进的机器学习方法， 映射. (3)我们用我们的计算预测来指导分子的系统解剖， 在iPSC系中使用模块化和可编程CRISPR-Cas9方法来支持AD+P， 调节调节元件、基因和等位基因，并测量产生的分子和细胞表型 在细胞自主和非自主表型中。如果成功的话，这个雄心勃勃的计划有可能 提供关于AD+P和/或P-AD中精神病性症状发展的第一个机制见解， 揭示功能性风险变异和治疗干预的靶基因，可能会影响临床 管理，以减轻与这些疾病相关的个人和社会负担。

Abstract Alzheimer's disease (AD) is the most common cause of dementia, and the most common neurodegenerative disease worldwide, affecting 1 in 8 individuals over 65 years old in the US. Within AD, approximately 40-60% individuals are affected by psychotic symptoms (AD+P), which are associated with more rapid cognitive decline, greater disability, mortality and caregiver burden, resulting in a disproportionately large disease burden. Recent studies indicate a genetic basis for AD+P risk, but the molecular basis of AD+P remains largely uncharacterized, hindering the search for appropriate treatments and novel therapeutics. In this proposal, we seek to systematically dissect the mechanistic basis of AD+P by systematic generation, integration, and experimental dissection of transcriptional and epigenomic phenotypes across two brain regions and four cell types. (1) We profile single-cell RNA-seq and cell-type specific H3K27ac ChIP-seq across 192 post-mortem brain samples, each in two regions across AD patients with psychosis, AD patients with no psychosis, schizophrenia patients with no AD, and control individuals. (2) We integrate the resulting datasets with genetic information and GWAS data to predict driver genes, regions, variants, and pathways underlying AD+P using state-of-the-art machine learning methods for causality, mediation analysis, and genetic Bayesian fine- mapping. (3) We use our computational predictions to guide a systematic dissection of the molecular underpinnings of AD+P using a modular and programmable CRISPR-Cas9 methodology in iPSC lines to modulate regulatory elements, genes and alleles, and measure the resulting molecular and cellular phenotypes in cell-autonomous and non-autonomous phenotypes. If successful, this ambitious proposal has the potential to provide the first mechanistic insights on the development of psychotic symptoms in AD+P and/or P-AD, reveal functional risk variants and target genes for therapeutic intervention that will likely influence clinical management in order to alleviate the personal and societal burden associated with these disorders.