Shotgun Lipidomics of Cerebrospinal Fluid in Neurological Disorders Gábor Balogh (Presenter) Biological Research Centre, Hungarian Academy of Sciences Presenter Bio: Received PhD in biochemistry at the Cardiff University, senior staff member of the Laboratory of Molecular Stress Biology and manage the Lipidomics Research Group in the Biological Research Center, Szeged. Research interests are development and application of lipidomics techniques and information technology solutions in the areas of lipidomics. He has broad experience in mass spectrometry of lipids, membrane and lipid biochemistry and biophysics, lipids in membrane function and signal transduction. Has 46 publications and over 1900 citations.

Authors: Mária Péter (1), Magdolna Pákáski (2), Anna Petrovics-Balog (3), Zita Oláh (2), Zsolt Datki (2), Eszter Ivitz (2), Wanda Török (3), Tibor Kovács (4), Dénes Zádori (3), János Tajti (3), Péter Klivényi (3), Ibolya Horváth (1), János Kαlmαn (2), László Vécsei (3,5), László Vígh (1), Gábor Balogh (1
(1) Biological Research Centre, Hungarian Academy of Sciences, Szeged, Hungary; (2) Department of Psychiatry, University of Szeged, Szeged, Hungary; (3) Department of Neurology, University of Szeged, Szeged, Hungary; (4) Department of Neurology, Semmelweis University, Budapest, Hungary; (5) MTA-SZTE Neuroscience Research Group, University of Szeged, Szeged, Hungary.

Biomarkers are urgently needed to improve diagnosis and prognosis of neurodegenerative diseases. Here we demonstrate that the lipid composition of cerebrospinal fluid (CSF) from Multiple Sclerosis, Alzheimer's disease (AD) and Guillain–Barré syndrome (GBS) patients shows significant signs of metabolic disturbances. In AD patients we found several lipid markers which were indicative to the severity of dementia, while the shift observed in the lipid profile of GBS patients supports the hypothesis of leakage from blood through the blood-nerve barrier. Therefore, lipidomic analysis of CSF might capable to reflect the pathophysiology of several neurological disorders.

Introduction

Cerebrospinal fluid (CSF), as it is in close interaction with brain tissue, represents an appropriate source for biomarkers that reflect processes of brain biochemistry and the pathophysiology of neurological disorders including Alzheimer's disease (AD) and Guillain–Barré syndrome (GBS). AD is a progressive neurodegenerative disorder, while GBS is a potentially life-threatening post-infection disease characterized by acute inflammatory polyneuropathy in the peripheral nervous system. Dysregulated lipid metabolism has been linked to these diseases, however, studies examining the changes in CSF lipids are limited.

In the present work, we applied high-resolution shotgun lipidomics for investigating the CSF lipidome to further clarify the role of lipids in the pathophysiology and diagnosis of AD and GBS.

Methods

CSF samples were obtained from 92 AD and 22 GBS patients and from 92 and 21 control patients, respectively. Control groups consisted of cognitively normal individuals with other neurological diseases. Mass spectrometric analyses were performed on an Orbitrap Elite instrument with robotic nanoflow direct injection [1]. Lipids were identified and quantitated by LipidXplorer and by in-house-built software. Beta-amyloid, phosphorylated tau and total tau proteins were determined by ELISA.

Results

We analyzed ca. 300 lipid species from CSF. Several lipids (e.g., docosahexaenoic acid-containing phospholipids, especially phosphatidylserine 40:6, saturated diacylglycerols, alkyl-acyl phosphatidylcholines) showed alterations distinguishing AD patients from controls and were also found to be indicative for the severity of dementia. Correlation was detected between certain lipid species and beta-amyloid or tau proteins. Artificial neural network model was able to learn and identify AD patients based on CSF lipid data. In GBS, the elevation of CSF protein content was strongly correlated with the increased lipid content conceivably due to the blood–CSF barrier damage. Hierarchical cluster analysis of GBS lipidomes may be useful in the differential diagnosis of this disease.

Conclusions & Discussion

The combination of high throughput lipidomic approach with advanced statistics and machine learning may help in the identification of potential biomarkers for the diagnosis of neurological disorders like AD or GBS, and may shed new light on their pathomechanism.

[1] Péter M, Glatz A, Gudmann P, Gombos I, Török Z, Horvαth I, Vígh L, Balogh G.PLoS One. 2017;12(3):e0173739.

Financial support: Hungarian Research and Technology Innovation Fund (KTIA_13_NAP-A-II/16); Ministry for National Economy (GINOP-2.3.2-15-2016-00001 and GINOP-2.3.2-15-2016-00006).