Date of Award


Document Type


Degree Name

Doctor of Philosophy (PhD)


Biomedical Sciences



Research Advisor

Francesca-Fang Liao, Ph.D.


William E. Armstrong, Ph.D. Melloni Cook, Ph.D. Michael P. McDonald, Ph.D. Rennolds S. Ostrom, Ph.D. Kazuko Sakata, Ph.D.


Alzheimer's disease, Heat shock inhibitor, Synapse protection


The excessive accumulation of amyloid peptides (Aβ) represents one major pathological hallmark of Alzheimer’s disease (AD), which is most notably characterized by synaptic dysfunction. Strategies targeting heat shock protein 90 (Hsp90) inhibition have been widely investigated in the treatment of cancer for over two decades. Its application in the treatment of neurodegenerative diseases however, has emerged more recently in the last decade. The role of the Hsp90 chaperone in clearing misfolded protein aggregates has been well established (in vitro only), but its function in synaptic activity remains elusive.

In our study, we utilized a widely used Hsp90 inhibitor, 17-AAG (17-allylamino-17- desmethoxygeldanamycin), to show that 17-AAG not only induces a heat shock-like response,

but also regulates (likely at the transcriptional level as modeled by qRT-PCR) select proteins enriched in the synapse, such as Post-synaptic density 95 (PSD95), synapsin I, and brain-derived neurotrophic factor (BDNF), among others. Confocal imaging demonstrated rescued dendritic spines with 17-AAG treatment after Aβ challenge in neuronal culture models. The functionality of these changes in the synapse was further confirmed by memory improvement in contextual fear conditioning tests. This work implied a potential strategy in using Hsp90 inhibitors to treat AD. Nevertheless, it must be noted that 17-AAG has limited feasibility in therapy due to its poor CNS-permeability.

In order to test the hypothesis inspired and encouraged by the initial 17-AAG experimental results, we instead used a similar compound with favorable CNS-permeability. OS47720, a radicicol derivative Hsp90 inhibitor, produced comparable effects to those of 17- AAG in inducing heat shock-like responses and promoting synaptic protein expression. Electrophysiology studies on coherence demonstrated enhancement of neuronal activity within the cortical hippocampal network. Most importantly, chronic 0S47720 administration for three months and six months in early and middle symptomatic AD mouse model (Tg2576) both rescued memory deficits without obvious systemic toxicity. OS47720, as an Hsp90 inhibitor, was the first reported compound to display therapeutic effects after long-term treatment in an AD animal model. Post-mortem studies revealed elevation in protein expression of heat shock proteins, synaptic proteins, and BDNF following treatment, without affecting Aβ load and neuroinflammation. Downregulation of heat shock factor 1 (HSF1) by either a pharmacological inhibitor or a virus-delivered shRNA abolished OS47720’s effects on memory improvement against Aβ toxicity. Overexpressing HSF1 partially improved memory. Together with the notion that HSF1-regulated BDNF mRNA, this implicated an HSF1-mediated mechanism at the transcriptional level of synaptic genes. Importantly, these observable effects bypassed several important signaling pathways, including the survival Protein kinase B (PKB or AKT) pathway which is an underlying mechanism of most neuroprotective drugs.

These findings revealed a novel function of Hsp90 inhibition in regulating synaptic gene expression and synaptic activity, further supporting the potential of using Hsp90 inhibitors in treating neurodegenerative diseases such as AD.




Included in

Neurosciences Commons