AbstractParkinson’s disease (PD) is a common neurodegenerative disorder diagnosed by the presence of bradykinesia, in combination with either rest tremor, rigidity, or both. Additionally, many non-motor symptoms (NMS) such as dementia, constipation, anosmia, depression, and sleep disorders are common. The clinical presentation of PD is heterogeneous, and increasing evidence suggests PD is rather a syndrome of parkinsonism with varying aetiologies. One of the most debilitating features of PD is the high prevalence of cognitive impairment, which has been linked to pathology spreading to limbic and cortical regions. A definite PD diagnosis can only be established post-mortem, where the loss of dopaminergic neurons and detection of protein aggregates containing α-synuclein called Lewy Bodies (LB) confirm the diagnosis. There is moreover evidence that the periphery is affected in PD. Clinically, many NMS often involve peripheral systems, including sensory and autonomic dysfunction, and systemic organelle and immune dysfunction are implicated in PD. Biomarkers are lacking in PD, both for early accurate diagnosis, and to differentiate PD endophenotypes, such as which patients will develop dementia. Previous plasma biomarker discovery studies in Alzheimer’s disease (AD) have shown promising result using mass spectrometry.
The aim of this thesis was to attempt to develop a pipeline approach for biomarker discovery and verification, in order to find novel plasma biomarkers for PD with and without cognitive impairment. A large-scale untargeted mass spectrometry experiment was performed using plasma from PD patients with varying degrees of cognitive impairment and healthy control (HC). The data was analysed for diagnosis, disease severity, cognitive impairment, and cognitive decline, and a list of candidate biomarkers was generated. The most robust peptides were selected for a targeted mass spectrometric experiment to verify the candidate biomarkers. Neurofilament light chain (NfL), arguably the most well studied PD biomarker today, was quantified in plasma as a benchmark. Finally, a validation study was performed for complement factors using immunoassays, as they are the most implicated differentially expressed protein group in PD plasma.
2260 proteins were quantified in at least half the study participants. Over 50 plasma proteins were differentially expressed in PD, and the most implicated pathway was the complement and coagulation pathway. 17 proteins had an absolute Cohen’s d effect size larger than 0.6. A single protein could differentiate PD from HC with a 76% accuracy, whereas a panel of 10 protein increased the diagnostic accuracy to 90%. Additionally, many candidate biomarkers were found for severity of cognitive impairment and longitudinal conversion to cognitive impairment and dementia. 20 proteins were significant both for level of cognitive impairment and cognitive decline. The most robust and reproducible peptides were selected for a targeted proteomic verification study. In total 22 peptides from 12 proteins were quantified, and 6 proteins correlated well with the discovery study. APOC3, C9, and TGFBI were successfully verified for PD diagnosis, and SAA1 correlated with motor and cognitive severity. Additionally high plasma LUM and APOA4 were both associated with dementia conversion for PD patients. NfL levels were elevated in PD plasma, and were pronounced in the cognitively impaired patients, and heavily affected by older age. Comparatively, some of the novel PD biomarkers identified by mass spectrometry are arguably better candidate biomarkers for PD diagnosis and PD cognition. An attempt was made to validate complement system biomarkers, but no differences were found between PD and HC, however, a decrease in C1q and C3 was found in atypical parkinsonian disorders.
In conclusion, the discovery study generated many novel plasma biomarker candidates for both PD diagnosis and cognition. The pipeline was successful in verifying several plasma proteins discovered in the untargeted PD biomarker study. Many novel biomarkers from the discovery study, and some from the verification study, showed superior biomarker performance to NfL. However, a limited number of proteins had robust enough peptides to be verified with the targeted mass spectrometric method used, and future studies should attempt using other techniques to verify the remaining candidate biomarkers.
|Date of Award||1 May 2023|
|Supervisor||Per Svenningsson (Supervisor), Latha Velayudhan (Supervisor) & Abdul Hye (Supervisor)|