Towards the Development of Artificial Intelligence for Morphological, Biochemical and Cytopathological Characterization of Neurodegenerative Synucleinopathies
Year: 2024
Type: dissertation
Abstract: Neurodegenerative diseases are progressive and debilitating conditions affecting millions of people worldwide. Since misfolded proteins are associated with neuronal degeneration and clinical symptoms, neurodegenerative diseases are also called proteinopathies. Synucleinopathies belong to the group of neurodegenerative diseases that are characterized by the loss of neurons and deposition of misfolded α-synuclein (αSyn) protein in various cell types of the nervous system. Synucleinopathies comprise Lewy Body Disorders (LBD) and multiple system atrophy (MSA). Based on the theory that misfolded αSyn spreads cell-to-cell, the seeding potential, which is the ability of misfolded proteins to act as a template to convert normal proteins into the pathological form, can be evaluated using seed amplification assay (SAA). This is a biochemical amplification technique developed to detect minute amounts of pathological protein in real-time. Recent evidence shows that in addition to morphological differences under the microscope, seeding behavior differs between Synucleinopathies, individuals, and even brain regions within the same individual. We hypothesized that the cell-type specific accumulation of αSyn might be an important aspect contributing to the different seeding behaviors. Therefore, our aim was to: i) establish a protein extraction method using formalin-fixed paraffin-embedded (FFPE) brain tissue to confirm disease-specific αSyn seeding as observed in frozen tissue; ii) link various cytopathology (i.e., cell-type specific accumulation of αSyn) in LBD and MSA with αSyn seeding using the established FFPE-SAA; and iii) evaluate whether Synucleinopathies can be further subclassified by morphological features using artificial intelligence (AI)-based machine learning algorithms. Upon confirming the preservation of disease-specific αSyn seeding in FFPE brain tissue, small regions with mostly neuronal or glial αSyn inclusions were identified on FFPE sections under the microscope. Proteins from these precise areas were extracted using our established method for SAA-mediated evaluation of αSyn seeding. AI-based analysis provided insight into the different seeding behaviors of cell-specific pathological αSyn. Moreover, further stratification and aging-related difference within MSA were found by AI-based measurement of morphological features. Our results will open new perspectives for understanding whether cellular pathological differences seen in histological evaluations contribute to seeding molecular behavior observed through SAA. Moreover, novel subclassification of Synucleinopathies will provide additional insight into the biological relevance - potential difference in response to therapeutics between patients with different subtypes of the disease.
Source: TSpace (University of Toronto)
Author: Kim, Ain
Institution: University of Toronto
Language: English
Subfield: Clinical Neurology
Field: Medicine
Domain: Health Sciences
Open Access status: green
OpenAlex ID: W7133077142
