Alpha-synuclein aggregation induces prominent cellular lipid changes as revealed by Raman spectroscopy and machine learning analysis.

Abstract

The aggregation of α-synuclein is a central neuropathological hallmark in neurodegenerative disorders known as Lewy body diseases, including Parkinson's disease and dementia with Lewy bodies. In the aggregation process, α-synuclein transitions from its native disordered/α-helical form to a β-sheet-rich structure, forming oligomers and protofibrils that accumulate into Lewy bodies, in a process that is thought to underlie neurodegeneration. Lipids are thought to play a critical role in this process by facilitating α-synuclein aggregation and contributing to cell toxicity, possibly through ceramide production. This study aimed to investigate biochemical changes associated with α-synuclein aggregation, focusing on lipid changes, using Raman spectroscopy coupled with machine learning. HEK293, Neuro2a and SH-SY5Y expressing increased levels of α-synuclein were treated with sonicated α-synuclein pre-formed fibrils, to model seeded aggregation. Raman spectroscopy, complemented by an in-house lipid spectral library, was used to monitor the aggregation process and its effects on cellular viability over 14 days. We detected α-synuclein aggregation by assessing β-sheet peaks at 1045 cm⁻¹, in cells treated with α-synuclein pre-formed fibrils, using machine learning (principal component analysis and uniform manifold approximation and projection) analysis based on Raman spectral features. Changes in lipid profiles, and especially sphingolipids, including a decrease in sphingomyelin and increase in ceramides, were observed, consistent with oxidative stress and apoptosis. Altogether, our study informs on biochemical alterations that can be considered for the design of therapeutic strategies for Parkinson's disease and related synucleinopathies.