Multimodal phenotyping of synaptic damage in Alzheimer’s disease : translational perspective with focus on quantitative EEG

Abstract

Alzheimer’s disease (AD) is a progressive neurodegenerative disorder and the most common form of dementia. Accumulation of AD-associated pathology in the brain may begin a decade or more before the appearance of the first symptoms of the disease. The pathological-clinical “continuum of AD” therefore encompasses time between the initial neuropathological changes and symptoms of advanced disease. Besides cognitively healthy individuals at risk, it includes subjects with subjective cognitive decline (SCD), mild cognitive impairment (MCI) and eventually dementia when the severity of cognitive impairment affects patients’ ability to carry out everyday activities. Timely detection of the disease would therefore recognize patients that are at risk for future cognitive deterioration and provide time window for the prevention and novel therapeutical interventions.

Accumulating evidence suggests that degeneration and dysfunction of brain neuronal connections, i.e. synapses, is one of the earliest and best proxies of cognitive deficits in patients along AD continuum. Human electroencephalography (EEG) is a non-invasive and widely available diagnostic method that records real-time large-scale synaptic activity. The commonly used method in research settings is quantitative EEG (qEEG) analysis that provides objective information on EEG recorded at the level of the scalp. Quantitative EEG analysis unravels complex EEG signal and adds relevant information on its spectral components (frequency domain), temporal dynamics (time domain) and topographic estimates (space domain) of brain cortical activity.

The general aim of the present thesis was to characterize different aspects of synaptic degeneration in AD, with the focus on qEEG and its relationship to both conventional and novel synaptic markers. In study I, global qEEG measures of power and synchronization were found to correlate with conventional cerebrospinal fluid (CSF) biomarkers of Aβ and tau pathology in patients diagnosed with SCD, MCI and AD, linking the markers of AD pathology to the generalized EEG slowing and reduced brain connectivity in fast frequency bands. In study II, qEEG analysis in the time domain (EEG microstates) revealed alterations in the organization and dynamics of large-scale brain networks in memory clinic patients compared to healthy elderly controls. In study III, topographical qEEG analysis of brain functional connectivity was associated with regionspecific cortical glucose hypometabolism ([18F]Fluorodeoxyglucose positron-emission tomography) in MCI and AD patients. Study IV provided evidence that qEEG measures of global power and synchronization correlate with CSF levels of synaptic marker neurogranin, both modalities being in combination independent predictors of progression to AD dementia in MCI patients. Study V and associated preliminary study introduced in the thesis assessed the translational potential of CSF neurogranin and qEEG as well as their direct relationship to AD neuropathology in App knock-in mouse models of AD. In study V, changes in CSF neurogranin levels and their relationship to conventional CSF markers in App knock-in mice corresponded to the pattern observed in clinical AD cohorts. These findings highlighted the potential use of mouse CSF biomarkers as well as App knock-in mouse models for translational investigation of synaptic dysfunction due to AD. In general, the results of the thesis invite for further clinical validation of multimodal synaptic markers in the context of early AD diagnosis, prognosis, and treatment monitoring in individual patients.