Non-Invasive Infrared Spectroscopic Detection of Parkinson's Disease Related Spectral Signatures

Abstract

This dissertation investigates the reliability boundaries of optical photothermal infrared spectroscopy (O-PTIR) for semi-quantitative β-sheet analysis and its application to the detection of Parkinson’s disease (PD)-related skin biomarkers. The work is divided into two sub-projects.

Sub-project I used atomic force microscopy–infrared spectroscopy (AFM-IR) as a reference technique to systematically evaluate the conditions under which O-PTIR can provide reliable semi-quantitative analysis of β-sheet content. The experimental samples included Aβ amyloid fibrils with three morphologies (Thick, Thin, and Twist) and two single-chain recombinant amyloid proteins (SCRAP0 and SCRAP3). The results showed that the parallel β-sheet peak near 1630 cm−1 was the only component with consistent semi-quantitative reliability, with coefficient of variation (CV) values between 4% and 13%, whereas the CV values of the other structural components generally exceeded 30%. AFM-IR analysis further demonstrated that the β-sheet ratio was independent of fibril morphology, supporting the use of normalized peak area ratios as a robust structural parameter. When the structural homogeneity of the sample was sufficiently high (O-PTIR CV < 15%), the deviation between O-PTIR measurements and independently validated circular dichroism (CD) results was approximately 1.7%, demonstrating the semi-quantitative capability of O-PTIR under controlled conditions.

Sub-project II independently investigated ATR-FTIR spectral signatures of Parkinson’s disease-related changes in A53T transgenic mouse skin. Principal component analysis (PCA) combined with t-tests identified three PD-sensitive wavenumbers: 1647 cm−1 (amide I band, protein conformational changes), 2871 cm−1 (lipid C–H stretching), and 3356 cm−1 (skin hydration state). The clearest separation between WT and PD groups was observed in forelimb skin from 12-month-old mice. The 1647 cm−1 signal reflects a broad conformational shift in skin proteins rather than an isolated β-sheet fibril signal, indicating that the skin matrix introduces additional complexity not present in in vitro systems. Based on these target wavenumbers, commercially available quantum cascade laser (QCL) systems were systematically evaluated for potential wearable sensor integration.

Overall, this study established a reliability assessment framework for semi-quantitative O-PTIR analysis and provided both a methodological basis and target wavenumbers for the future development of photothermal infrared skin sensors for Parkinson’s disease detection.

Popular Abstract

Non-Invasive Infrared Spectroscopic Detection of Parkinson's Disease Related Spectral Signatures

For most patients with Parkinson’s disease, diagnosis occurs only after neuronal degeneration has already begun, when tremor and motor slowing become apparent. Earlier detection could improve prognosis, yet current diagnosis still relies largely on clinical observation. To detect the disease at an earlier molecular stage, researchers are increasingly turning to a familiar tool in physics: light.

At the molecular level, Parkinson’s disease is associated with the misfolding of key proteins, which lose their normal structures and instead aggregate into highly ordered β-sheet–rich fibrillar assemblies. These aggregates accumulate inside neurons and ultimately contribute to neurodegeneration. Conventional biochemical methods often require protein labeling, purification, or fixation during sample preparation, all of which may alter the native structure of the protein.

Infrared spectroscopy provides a possible solution. Chemical bonds vibrate at characteristic frequencies and absorb infrared light at corresponding wavenumbers, producing unique “vibrational fingerprints.” However, the spatial resolution of conventional infrared spectroscopy is limited by diffraction to the micrometre scale, far too coarse to resolve nanoscale protein aggregates inside neurons.

Recent advances have begun to overcome this limitation. Optical photothermal infrared (O-PTIR) microscopy achieves submicron spatial resolution by detecting infrared absorption through the photothermal effect using a visible probe beam, without labeling or physical contact. Yet because the signal is also affected by local thermal properties of the sample, an important question remains: to what extent can O-PTIR reliably reflect the true β-sheet content of a sample?

To address this, the study used structurally defined amyloid model systems together with Atomic force microscopy infrared spectroscopy as a nanoscale reference technique. The results show that only the parallel β-sheet component near 1630 cm⁻¹ exhibits consistent semi-quantitative reliability, and that this reliability depends critically on sample structural homogeneity.

Beyond its analytical capability, O-PTIR offers another practical advantage: its optical architecture does not require an interferometer or moving optical components. Replacing bulky tunable infrared sources with fixed-wavelength quantum cascade lasers (QCL) could reduce the system to a compact point-detection device targeting selected wavenumbers, thereby enabling wearable integration. However, because skin infrared signals arise from a complex mixture of proteins, lipids, and water, identifying target wavenumbers that reliably distinguish disease states remains a major challenge.

To investigate this, the study analysed skin infrared spectra from a Parkinson’s disease mouse model and identified disease-sensitive wavenumbers, most notably at 1647 cm⁻¹. However, this signal reflects broader conformational changes across the skin proteome rather than an isolated β-sheet contribution, indicating that the in vitro quantitative framework cannot be directly transferred to skin measurements. Nevertheless, these wavenumbers provide a spectroscopic basis for QCL selection, and a proof-of-concept prototype was constructed as a first step toward a wearable sensor.

• author: Sun, Mingxue
• supervisor: Anna-Lena Sahlberg; Oxana Klementieva; Su Kar Yan
• organization: Department of Physics; Combustion Physics
• course: FYSM64 20261
• year: 2026
• type: H2 - Master's Degree (Two Years)
• subject: Physics and Astronomy
• keywords: Optical photothermal infrared spectroscopy (O-PTIR), Atomic force microscopy infrared spectroscopy (AFM-IR), Amyloid β-sheet structure, Semi-quantitative reliability, Parkinson's disease biomarkers
• language: English
• id: 9240691
• date added to LUP: 2026-06-28 11:23:37
• date last changed: 2026-06-28 11:23:37