Premium
Lipid metabolism dysfunction in progranulin mutation carriers: Unbiased metabolomics reveals strong relationship to clinical status in FTLD
Author(s) -
Vandevrede Lawren,
Rojas Julio C.,
Wang Ping,
Heuer Hilary W.,
Karydas Anna M.,
Ljubenkov Peter A.,
Boeve Bradley F.,
Rosen Howard J.,
Boxer Adam L.
Publication year - 2020
Publication title -
alzheimer's and dementia
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 6.713
H-Index - 118
eISSN - 1552-5279
pISSN - 1552-5260
DOI - 10.1002/alz.046594
Subject(s) - medicine , metabolomics , frontotemporal lobar degeneration , population , disease , endocrinology , biology , genetics , bioinformatics , frontotemporal dementia , dementia , environmental health
Abstract Background Heterozygous progranulin ( GRN) mutations are an important cause of familial Frontotemporal Lobar Degeneration (f‐FTLD) associated with TDP‐43 pathology. The cellular role of progranulin remains unclear, but alterations in lipid metabolism and lysosomal function have been suggested. Better understanding the pathophysiology of progranulin deficiency is critical as several clinical trials are underway in this population, and fluid biomarkers are urgently needed to assess pharmacodynamic effects. Therefore, we conducted an unbiased metabolomic analysis in GRN mutation carriers and family member controls, and identified at least two classes of lipids altered in symptomatic GRN carriers. Methods Matched plasma and CSF samples were obtained from GRN mutation carriers (46 plasma, 36 CSF) and age‐matched blood relatives (36 plasma, 18 CSF). Untargeted metabolite detection was conducted via LC‐MS/MS in both plasma (1389 compounds) and CSF (508 compounds) by Metabolon. To identify important compounds in differentiating these groups, we conducted an unbiased analysis between family controls, asymptomatic carriers, and symptomatic carriers using univariate one‐way ANOVA with post‐hoc Fisher’s LSD corrected for false discovery rate (FDR) and partial least squares ‐ discriminant analysis (PLS‐DA, Metaboanalyst). For compounds identified by this unbiased approach, correlation with measures of disease severity was analyzed via linear regression, controlling for age and gender. Results In plasma, 22 compounds were identified as significant in discriminating between groups via ANOVA (FDR<0.05), and 17/22 were also identified by PLS‐DA (VIP>2.4). Of these, 21/22 compounds were lipids; 11 were sphingomyelins. Comparatively, in the relatively lipid‐poor CSF samples, 6/15 compounds identified by PLS‐DA (VIP>2.0) were phosphatidylcholines. Pathway impact analysis revealed a large impact from the GRN mutation on sphingolipids in plasma and glycerophospholipids in CSF. In secondary analyses, sphingomyelins were elevated in plasma from symptomatic carriers, while phosphatidylcholines were decreased in CSF. Secondary analyses showed these compounds were correlated with clinical measures of disease severity, including MoCA and CDR+NACC‐FTLD. Conclusion Increased plasma concentration of sphingomyelin and decreased CSF phosphatidylcholines are both associated with worse clinical status in FTLD‐ GRN . These findings suggest that impaired lipid metabolism could play a pathogenic role in FTLD‐ GRN . Targeted lipidomic measures in plasma and CSF may potentially reveal useful clinical trial biomarkers.