Skip to main navigation Skip to search Skip to main content

Alternative Treatment Strategy and Validation of a Novel Brain Target for Deep Brain Stimulation in Alzheimer’s Disease and Epilepsy

Student thesis: Doctoral Thesis

Abstract

Alzheimer’s disease (AD) and epilepsy are two prevalent neurological disorders that often co-occur and share pathophysiological mechanisms. Current pharmacotherapies are limited by drug resistance and adverse events, and existing deep brain stimulation (DBS) targets (fornix for AD; anterior thalamic nucleus for epilepsy) have significant drawbacks associated with psychological side effects. To address these unmet needs, we investigated the pedunculopontine nucleus (PPN) - a brainstem structure containing cholinergic, glutamatergic and GABAergic neurons that projects to memory related (thalamus, basal forebrain, retrosplenial cortex (RSC)) and seizure related (primary motor cortex (MOp), substantia nigra pars reticulata (SNr), anterior thalamic nucleus (ATN)) regions - as a novel DBS target.

In AD studies, we first confirmed that PPN projects and innervate into the hippocampus, and other memory-related brain regions including nucleus basalis of Meynert (NBM), RSC, prefrontal cortex (PFC), and thalamus. anterograde tracing. Next, we employed APP/PS1 transgenic and streptozotocin induced sporadic mouse models. Chronic low frequency (60 Hz) deep brain stimulation of pedunculopontine nucleus (PPN DBS) significantly improved memory performance in both models, as assessed by Y maze, novel object recognition, or Barnes maze tasks. However, no significant reduction in amyloid-beta (Aβ) plaque burden or soluble Aβ levels was observed in aged APP/PS1 mice, and the vast majority of neuroinflammatory and microglial markers remained unchanged, with the exception of Lpl, Csf1 and Hrh1, which highly link to neuronal hyperactivity and apoptosis. In the STZ model, PPN-DBS restored the expression of immediate early genes (Arc, JunB). In 9 month old APP/PS1 mice, PPN-DBS normalized the aberrant expression of multiple ionotropic glutamate receptor subunits (Grin1, Grin2a, Grin2d, Gria1, Grik3, Grid1) and immediate early genes (c-Fos, JunB) in the hippocampus. However, 15-monthold- APP/PS1 mice showed no aberrant alteration in neuronal number, synaptic integrity, apoptosis and BDNF/TrkB signaling components in comparison with WT control, indicating that this mouse model may not be suitable to study excitotoxicity. Overall, the data suggested that cognitive benefits are not mediated by direct anti-Aβ or broad anti-inflammatory actions, but rather by normalizing hippocampal hyperexcitability through restoration of glutamatergic and activity-dependent gene expression.

In epilepsy studies, we first confirmed that PPN projects to the MOp, SNr, and ATN using AAV based anterograde tracing. In the acute pentylenetetrazole (PTZ) model (65 mg/kg), systematic parameter optimization identified 60 Hz and 100 µA as the most effective combination, significantly prolonging myoclonic and generalized tonic clonic (GTC) latencies, and most importantly, remarkably reducing GTC duration from ~24 s to ~3 s, and improving survival. This optimized protocol also suppressed seizures - delay the seizure progression, prolong the seizure onset, reduce the tonic-clonic duration in chronic PTZ kindling (35 mg/kg) and intrahippocampal kainic acid (KA) chronic models. Given the PPN-DBS successfully suppress the seizure demonstrated in the video recording, we then found PPN-DBS also suppressed cortical EEG and PPN power spectral density, normalized its specific frequency bands - delta and theta, and reduced epileptiform discharges. Mechanistically, PPN DBS restored the epilepsy induced imbalance of excitatory/inhibitory markers (vGLUT1/GAD67) in the PPN. In gene level analysis, we found downregulated seizure associated immediate early genes (c-Fos, FosB), and normalized AMPA receptor subunit expression (Gria1 gene and Gria1/Gria2 ratio) coupled with glutamate transporter levels (Eaat2, Eaat3) in the cortex and PPN. At circuit level, we first verified that PTZ kindling mouse model present the significant decrease in AAV signal of the PPN-downstream brain targets – MOp, SNr and ATN, indicating these regions were affected in epilepsy. Finally, the local field potential (LFP) analysis revealed that MOp and SNr were overactive in acute PTZ mice, and PPN-DBS can successfully normalize them.

In conclusion, chronic low-frequency (60 Hz) PPN-DBS exerts potent therapeutic effects in mouse models of both AD and epilepsy. In AD, the cognitive benefits are not mediated by direct anti-amyloid or broad anti-inflammatory actions, but rather by the normalization of hippocampal hyperexcitability - specifically through restoration of glutamatergic receptor subunit expression and activity-dependent immediate early genes. In epilepsy, the same optimized stimulation protocol robustly suppresses acute and chronic seizures, normalizes cortical and PPN electrophysiological activity, and restores the excitatory/inhibitory balance together with AMPA receptor and glutamate transporter expression. These findings establish PPN-DBS as a promising dual indication neuromodulation strategy that targets shared mechanisms of network hyperexcitability, offering a strong preclinical foundation for future clinical translation in drug resistant epilepsy and AD.
Date of Award24 Jun 2026
Original languageEnglish
Awarding Institution
  • City University of Hong Kong
SupervisorChi Him Eddie MA (Supervisor)

Cite this

'