The goal of this proposal is to discover drug-like retinoid X receptor (RXR) agonists that can halt and reverse the progression of Alzheimer's disease (AD). The research strategy is to target ApoE gene expression through RXR agonists;this has recently been shown to improve A? turnover, reduction in A? plaque area, and the reversal of cognitive, social and olfactory deficits. The research plan consists of two steps. The first step is computer-aided prediction of agonistic RXR binding, blood brain barrier permeability, and absence of side effects. This step relies on a novel method for pharmacophore analysis by examining the joint space of chemical compounds, targets, and chemical/biological properties. This joint space is defined using machine learning on the 3D geometry of spatial arrangement of pharmacophoric points, using attributes such as donors, acceptors, aromatic rings, and charged fragments. The second step consists of biochemical assays including competitive binding with bexarotene, and ApoE level determination in cultured mouse brain cells. A small number of drug-like compounds will be obtained through multiple iterations of the two steps of in-silico prediction and assays at the end of Phase 1. The outcome of this study will be new drug leads to potentially treat and prevent AD at different stages of cognitive decline and neurodegeneration.
Interleaving of computer-aided screening of large chemical databases through novel machine learning based models with biochemical assays will identify drug-like compounds that can clear toxic chemicals responsible for Alzheimer's Disease from brain cells with minimal side-effects. These compounds will form the starting point for the first true cure of Alzheimer's Disease by slowing and reversing its effects.
