ABSTRACT
Electroencephalographic (EEG) abnormalities in Alzheimer’s disease are poorly understood and can
provide insight into network changes accompanying disease progression and serve as non-invasive
biomarkers. This proposal is based on a novel EEG abnormality that is common very early in life in mouse
models of Alzheimer’s disease neuropathology. The abnormality consists of high frequency oscillations
(HFOs), ultrafast (>250 Hz), and very brief (~30 ms) electrical events in the EEG. Notably, HFOs are known to
occur in epilepsy but their role in Alzheimer’s disease models remains unknown.
The goals of the proposed project are: (i) to determine the role of high frequency oscillations (HFOs) as a
new type of abnormal Alzheimer’s disease-related electrical biomarker using electrode arrays, (ii) to
manipulate HFOs in real time as a method to boost or repress memory, since impaired memory is a major
symptom of Alzheimer’s disease and last (iii) to test an underlying mechanism contributing to HFOs in
Alzheimer’s disease.
In Aim 1 we will address the novel hypothesis that HFOs start as a local phenomenon in the hippocampus
and progressively expand into more superficial areas as the mice get older. To this end, we will take advantage
of a novel neurotechnology that uses flexible electrode arrays and a mouse model of cerebral amyloid
overexpression that shows robust HFOs to record longitudinally. We will also test if HFOs precede other
epileptiform abnormalities and reduced gamma power, which are commonly seen in Alzheimer’s disease.
These experiments will tell us if HFOs can be considered an early biomarker. Furthermore, we will ask whether
and how HFOs relate to the other abnormalities.
In Aim 2 we will address the role of HFOs in impaired memory and test an underlying mechanism. We will
use a novel closed-loop protocol we developed to detect and disrupt HFOs in real time to test whether
selective disruption of HFOs during sleep is beneficial to memory and Alzheimer’s disease neuropathology.
Importantly, our preliminary data suggest that the closed-loop approach improves memory in a mouse model of
Alzheimer’s disease. To address mechanism, we will test a gene therapy approach to reduce HFOs in vivo by
overexpressing a specific K+ channel and test if memory is improved. Notably, our pilot data suggest that
overexpression of this K+ channel reduces HFOs in vivo. Effects of the closed-loop approach or K+ channel
overexpression will be tested on HFOs and other epileptiform activity. Because HFOs occur during a sleep
stage where Aβ is being cleared, we will also stain for soluble Aβ and ask if it is reduced when HFOs are
disrupted.
This project will address an unmet goal for new and improved biomarkers and two novel strategies to
disrupt them using state-of-the-art methods. It also aims to provide a solid foundation for future investigation of
HFOs in diverse models to address mechanisms and circuits that HFOs utilize in Alzheimer’s disease
pathophysiology. Lastly, we expect the findings and protocols developed in this project to be relevant for a
range of brain disorders affecting memory including epilepsy.