PROJECT SUMMARY
Alzheimer’s disease (AD) is a fatal disease that currently afflicts almost six million Americans. With an
aging population, we risk a public health crisis by 2050, unless effective treatments are identified. Despite
extensive research, there are currently no drugs that slow or alter the course of disease. AD is defined by the
presence of ß-amyloid (Aß) plaques and intraneuronal tau inclusions called neurofibrillary tangles (NFTs) in
the brain. Drug candidates that reduce Aß plaques have not, yet, been shown to have clinical benefit, and
growing data suggests that it may be more important to target NFTs over Aß plaques to prevent cognitive
decline. Recently, the macroautophagy-lysosomal pathway of protein degradation has emerged as a
compelling target for reducing pathogenic tau in the brain. Our hypothesis is that increasing the rate of tau
degradation will reduce tau levels and stop, or greatly slow, the rate of tau aggregation. We recently
discovered a novel pathway to accomplish this objective. Inhibiting the farnesylation of Rhes, a GTPase
protein in the Ras family, activates the lysosome and results in the selective degradation of pathological
tau. Confirmation of the therapeutic hypothesis was achieved by administering a farnesyltransferase inhibitor
(FTI) in a mouse model of tauopathy, which reduced tau pathology and attenuated behavioral abnormalities in
the mice.
Known FTIs are not suitable for human development as CNS drugs. Optimized for cancer indications,
they are efficiently pumped out of the brain by efflux proteins. We propose a three-pronged approach to identify
chemical matter that can reach pharmacologically significant and dose-proportional brain levels. For two of the
known inhibitors, L-778,123 and lonafarnib, we will make strategic changes to the structures, eliminating
functional groups that serve as recognition substrates for the efflux pumps. Concurrently, we will initiate a high-
throughput screen of a chemical library with chemical properties consistent with known CNS drugs. As a third
step, we will engage in a multi-million compound artificial intelligence-based virtual screen with AtomWise to
identify novel FTIs. By generating x-ray co-crystal structures of the most promising hits and using computer-
aided drug design, we plan to accelerate the process of hit validation, lead discovery, and optimization to
identify small molecule drug candidates. We will advance inhibitors to an in vivo pharmacodynamic model
and select compounds with linear pharmacokinetic/pharmacodynamic (PK/PD) relationships that can be
advanced into the clinic. Three of the top compounds will be tested for efficacy in a tauopathy animal model
using doses derived from the PK/PD relationship. Short-term studies will identify compounds that reduce all
pathogenic tau species. The most efficacious compound will be moved into long-term dosing studies to
evaluate life-span extension and reduction in NFT formation.
