Mass spectrometry on cerebrospinal fluid uncovers association of novel glycolysis biomarkers with Alzheimer’s disease in a complex clinical cohort

Abstract

AbstractBackgroundDiagnosis of Alzheimer’s disease (AD) primarily relies on cognitive assessments supported by imaging and cerebrospinal fluid (CSF) biomarkers of amyloid and tau proteins. These limited biomarkers correlate poorly with disease progression or treatment response. Broader, more comprehensive biomarker sets together with traditional ATN biomarkers could improve the diagnosis and prognosis of AD, identify contributing pathophysiologies and comorbidities, and prove useful in measuring treatment efficacy. In this study we leverage mass‐spectrometry to simultaneously quantify hundreds of proteins in CSF from a large‐scale neurology clinic cohort, all of whom had clinical indication for diagnostic lumbar puncture. We aim to define markers that can stratify and diagnose AD in this “real‐life” clinically complex cohort.MethodCSF samples from 408 patients were collected from the Massachusetts General Hospital Neurology Service following written informed consent for research biobanking and analysis of excess CSF. The patient cohort consisted of ATN‐verified cognitively‐unimpaired (n = 81), mild‐cognitive impairment from AD (n = 116), mild‐cognitive impairment from other non‐AD causes (n = 78), dementia from AD (n = 65) and dementia from other non‐AD causes (n = 31). Samples were analyzed on an Orbitrap Fusion using a Data Independent Acquisition method and raw files were searched using Scaffold‐DIA.Result1761 unique peptide sequences belonging to 516 proteins were robustly quantified by DIA in CSF. Linear regression analyses found 572 peptides to be significantly differentially abundant across all diagnoses compared to AD (padj < 0.05). Ontological analysis of these peptides indicated enrichment of glycolysis in AD, potentially indicating a shift in brain metabolism in AD. 4 peptides belonging to 2 glycolysis proteins (PKM, ALDOA) were elevated in AD compared to all non‐AD diagnoses, suggesting specificity for AD. Differential abundance of these proteins was verified through immunoassay.ConclusionHere, unbiased mass‐spectrometry on CSF highlighted novel AD specific biomarkers ALDOA and PKM, potentially indicating a metabolic shift in the brain. Further mechanistic studies should investigate dysregulation of these markers in AD and correlation with FDG‐PET.