Bioinformatics-predicted Alzheimer's risk factors and their role in synaptic dysfunction

University Of Oklahoma Hlth Sciences Ctr, Oklahoma City OK

Investigators

Linked publications, trials & patents

Abstract

Abstract Alzheimer?s disease (AD) is the leading cause of dementia in the elderly that affects more than 5 million patients in the USA. Despite tremendous efforts to understand the pathophysiology of AD, no new therapeutic drug was approved in the last 10 years. A paradigm shift is needed in the approach to find effective treatment and cure of AD. Recent data demonstrate that synaptic failure is an essential cause of cognitive decline during the early phase of AD. Synapse loss in AD brain samples is the best correlate with the severity of cognitive impairment. The exact cellular process leading to such dramatic synapse loss in AD is still elusive. Technical advances allow now the detailed analysis of proteomics and mRNA sequencing and thus provide vast databases. Although multiple hits were described in genome wide association studies including novel genetic coding and non-coding regions, researchers are currently at the point they need to understand the molecular biology behind these genetic variations and how they impact AD. However, it is not clear how to best prioritize them for further study. This task is hindered by the fact that the physiological functions for the overwhelming majority of genes remain elusive as never tested in details. Here we propose a new innovative approach using bioinformatics, machine learning and synaptic functional assays to predict and analyze new pathways in the etiology of AD. Our iterative approach is based on prioritizing candidate genes that begins with previously validated software to predict AD-related phenotypes and is combined with a novel meta-analytic approach to predicting transcriptional changes with age. The candidate gene- phenotype list will be experimentally validated, and then this experimental data will in-turn be used to further refine the priority scores to be maximally sensitive to AD-related phenotypes. We expect that by combining transcriptional network behavior with literature-based relationships between genes, AD and AD-related phenotypes, we can determine the most likely phenotypic consequence of the identified variants linked to a specific gene. We also hypothesize that a subset of these variants encoding synaptic genes will cause synaptic dysfunction, which will be directly measured in cultured neurons and animal models of AD. We will test our hypothesis in two Specific Aims. The proposed research has the potential for significant impact as it will provide the AD field with a novel analytical tool to better predict risk factors and also will test for the first time genes or coding regions for their effect on synaptic transmission and AD pathology.