Combine Genomics and Symptoms Data Driven Models to Discover Synergistic Combinatory Therapies for Alzheimer's Disease

点击翻译按钮获取中文摘要

Project Summary In 2018, an estimated 5.7 million people have Alzheimer's Disease (AD) or a related dementia in the U.S., with related healthcare costs of ~$277 billion1. However, there is no cure yet for AD. One major challenge is that the complicated pathogenesis of AD remains unclear, though >42 genes/loci have been associated with AD2,3. These genes are not actionable or druggable yet for AD management2. Over 240 drugs were tested in clinical trials but no new drugs have been approved for AD since 20031,4. The failure of these drugs is likely, in part, due to the limited efficacy of single agents to treat AD that is a genetically complex, multifactorial disease2, i.e., robust molecular signaling crosstalks among multi-pathways2,5,6,7,8,9, as well as complicated niche factors, e.g., oxidative stress10,11,12,13, and inflammation14,15,16, leading to neuron de-generation. Therefore, combination therapies eliminating these niche factors, and disrupting the dysfunctional signaling pathways and cross-talks, can be more effective than single agents for in AD patients. The goal of this study is to fill the gap of accelerating repositioning of combination therapies for AD using following novel genomics and symptoms data-driven models seamlessly integrating well designed iPSC Aβ AD models. The Washington University Charles F. and Joanne Knight Alzheimer's Disease Research Center (Knight-ADRC), has generated comprehensive omics data for a large group of AD samples. We propose to (in Aim 1) uncover core signaling pathways and crosstalks of ApoE4 genotype-specific AD subtypes via a novel signaling convergence network model, and consequently to discover synergistic Signaling Network Disruption drug combinations (SNDdc) via novel drug prediction models integrating heterogenous pharmacogenomics datasets. On the other hand, we propose to (in Aim 2) discover potential Neuron Protective drug combinations (NPdc) using electronic health records (EHR), available in BJC HealthCare system (includes 14 academic and community hospitals in Missouri and Illinois), of patients with brain injury diseases, especially Traumatic Brain Injury (TBI), via a novel high-order poly-pharmacy efficacy and safety model. We hypothesize that acute brain damage in TBI will share the aforementioned key AD-related niche factors. Also because TBI patients often require multiple drugs daily (high-order poly-pharmacy use), we propose that TBI provides an appropriate model to study synergy and interactions of combination therapies that can ameliorate acute brain injury, and thus suggest potentially neuron protective combinations. Combinations in SNDdc and NPdc provide candidates for novel and effective AD treatment. To filter the false positives, we will (in Aim 3) utilize pooled CRISPR functional genomics and iPSC neurodegeneration model to identify key signaling genes, and validate combination therapies with ApoE4 genotype-specific iPSC Aβ models. Our new models represent a potential breakthrough in AD combination therapies discovery.