Dementia Proteomics

Time: Fri 2020-06-12 09.00

Location:, (English)

Subject area: Biotechnology

Doctoral student: Julia Remnestål , Science for Life Laboratory, SciLifeLab, Affinity Proteomics, Nilsson Lab

Opponent: Assoc. Prof, MD., PhD. Pieter Jelle Visser, Karolinska Institutet

Supervisor: Professor Peter Nilsson, Science for Life Laboratory, SciLifeLab, Affinity Proteomics; Anna Månberg, Science for Life Laboratory, SciLifeLab


The term dementia encompass a number of conditions arising as a consequence of tissue degeneration in the brain. This degeneration is caused by molecular events occurring on a cellular level including inflammation, defective waste disposal and accumulation of insoluble proteins and peptides. Many of these molecular events are in turn also reflected in the composition of the cerebrospinal fluid (CSF) which circulates within and around the brain. This thesis summarise five studies conducted with the aim to explore and profile CSF proteins in the context of dementia and other neurodegenerative disorders. Protein profiles were obtained by so-called suspension bead arrays (SBAs), created by coupling antibodies to color-coded microspheres, allowing detection of more than 350 CSF proteins simultaneously. The majority of the explored proteins are referred to as brain-enriched, entailing that the corresponding genes are highly expressed in brain tissue in comparison to other tissues.

In Paper I, the SBA technology was utilised to profile about 280 proteins in CSF from several neurodegenerative disorders, i.e. Alzheimer’s disease (AD), dementia with Lewy Bodies and Parkinson’s disease. Distinct differences in the CSF proteome were identified depending on site of collection (ventricular or lumbar) and time point (post mortem or ante mortem). Disease-associated profiles for the two synaptic proteins neuromodulin (GAP43) and neurogranin (NRGN) could be confirmed, in which both proteins displayed higher levels in AD compared to controls. High levels of the two proteins were furthermore observed in patients at preclinical stages of AD in two independent cohorts. To verify the identified protein profiles, parallel reaction monitoring (PRM) assays were developed for 17 proteins in Paper II, including GAP43. Eight proteins displayed concordance to data generated with SBAs and among these were GAP43, cholecystokinin, neurofilament medium chain (NF-M), leucine-rich alpha-2-glycoprotein and vascular cell adhesion protein 1. 

In Paper III, the SBA technology was again applied to characterise early dementia-related changes in the CSF proteome by comparing samples from individuals with mild cognitive impairment (MCI), controls and AD patients in two independent cohorts. The MCI individuals were moreover stratified based on CSF concentration of the core AD biomarkers Aβ42 and tau. The six proteins amphiphysin, aquaporin 4, cAMP regulated phosphoprotein 21, β-synuclein, GAP43 and NF-M did all show significant differences between sample groups in both cohorts. Further exploration of how the pathological processes preceding dementia affect the CSF proteome, was done by analysis of 104 brain-enriched proteins in CSF from asymptomatic 70 year-olds in Paper IV. Protein profiles were correlated to Aβ42, t-tau and p-tau CSF concentration, revealing a large number of proteins displaying significant correlations to tau levels. Upon dividing the asymptomatic individuals based on Aβ42 CSF pathology, some proteins showed significantly different associations in the two groups. Most of these proteins yielding interesting profiles, were plasma membrane proteins or proteins connected to synaptic vesicle transport.

While AD is the most common form of dementia, accounting for more than 60 % of all cases worldwide, frontotemporal dementia (FTD) is the most frequently occurring form of young-onset dementia. In Paper V, CSF protein profiles were explored in the context of FTD. Patients with behavioural variant FTD and primary progressive aphasia, were compared to unaffected individuals with a high risk of developing FTD. Proteomic differences between patients with FTD and the unaffected individuals were observed already at a global level, and particularly for the six proteins NF-M, neurosecretory protein VGF, neuronal pentraxin receptor, prodynorphin, transmembrane protein 132D and tenascin-R.

The disease-associated profiles identified in the presented studies provide a basis for future research within dementia proteomics. Whether the proteins identified will have the possibility to aid in clinical diagnosis, prognosis or characterisation of dementia, remains to be evaluated. Given the fortunate situation, especially in Sweden, with access to large and well characterised CSF collections, there are ample opportunities for future proteomic studies to elucidate the true potential of these proteins.