Serum metabolites associate with biofluid neurofilament light chain in Alzheimer’s disease

Abstract

AbstractBackgroundNeurodegeneration is one of the main components of A/T/N (amyloid/ tau/ neurodegeneration) classification. Recently, neurofilament light chain (NFL), a neuronal protein that is expressed in myelinated axons, has been increasingly recognized as a biomarker of neurodegeneration. In this study, we performed large‐scale association analysis of serum metabolites with NFL in Alzheimer’s disease.MethodWe performed cross‐sectional and longitudinal association analysis of serum metabolome with plasma and CSF NFL in Alzheimer’s disease (AD), after adjusting for multiple testing (N=1,317 for 140 metabolites). Subjects who had both NFL and metabolomics panel data were included. Metabolites were measured using a targeted metabolomics approach using the AbsoluteIDQ‐p180 kit.ResultWe identified 70 metabolites including acylcarnitines sphingomyelins, biogenic amines, amino acids, and glycerophospholipids as significantly associated with baseline plasma NFL, where we identified sex‐specific and APOE ε4‐specific metabolites. Of those, eight metabolites (C16:1, C18:1, SM(OH) C14:1, SM(OH) C16:1, SM C20:2, PC.aa.C28.1, PC.ae.C30.2, and PC.ae. C32.2) were also associated with baseline cerebrospinal fluid NFL and 10 metabolites (C9, C10:2, citrulline, creatinine, SDMA, PC.ae.C34.2, PC.ae.C36.2, PC.ae. C36.3, SM C20:2, and SM C16:0) were associated with change rates of plasma NFL in AD. Metabolites with significant associations were related to biological functions including amyloid precursor protein processing, neuroinflammation, nitric oxide production, and renal function. There were two significant interactions between sex and two medium‐chain acylcarnitine metabolites (C8 and C10) on plasma NFL levels.ConclusionOur present study provides insight into the association of serum‐based circulating metabolites with biofluid biomarker evidence of neurodegeneration, which is closely related to overt clinical symptoms in AD. Metabolites found in our study are related to various biological functions or associated with known risk factors of AD.