Plasma Lipoproteome in Alzheimer’s Disease: a Proof-of-concept Study

Danni Li,Fangying Huang,Yingchun Zhao,Peter W. Villata,Timothy J. Griffin,Lin Zhang,Ling Li,Fang Yu

DOI: https://doi.org/10.1186/s12014-018-9207-z

2018-01-01

Clinical Proteomics

Abstract:Background: Although total plasma lipoproteome consists of proteins that have shown promises as biomarkers that can identify Alzheimer's disease (AD), effect sizes are modest. The objective of this study is to provide initial proof-of-concept that the plasma lipoproteome more likely differ between AD cases and controls when measured in individual plasma lipoprotein fractions than when measured as total in immunodepleted plasma. Methods: We first developed a targeted proteomics method based on selected reaction monitoring (SRM) and liquid chromatography and tandem mass spectrometry for measurement of 120 tryptic peptides from 79 proteins that are commonly present in plasma lipoproteins. Then in a proof-of concept case-control study of 5 AD cases and 5 sex- and age-matched controls, we applied the targeted proteomic method and performed relatively quantification of 120 tryptic peptides in plasma lipoprotein fractions (fractionated by sequential gradient ultracentrifugation) and in immunodepleted plasma (of albumin and IgG). Unadjusted p values from two-sample t-tests and overall fold change was used to evaluate a peptide relative difference between AD cases and controls, with lower p values (<0.05) or greater fold differences (> 1.05 or < 0.95) suggestive of greater peptide/protein differences. Results: Within-day and between-days technical precisions (mean %CV [SD] of all SRM transitions) of the targeted proteomic method were 3.95% (2.65) and 9.31% (5.59), respectively. Between-days technical precisions (mean % CV [SD]) of the entire plasma lipoproteomic workflow including plasma lipoprotein fractionation was 27.90% (14.61). Ten tryptic peptides that belonged to 5 proteins in plasma lipoproteins had unadjusted p values< 0.05, compared to no peptides in immunodepleted plasma. Furthermore, 27, 32, 17, and 20 tryptic peptides in VLDL, IDL, LDL and HDL, demonstrated overall peptide fold differences > 1.05 or < 0.95, compared to only 6 tryptic peptides in immunodepleted plasma. The overall comparisons, therefore, suggested greater peptide/protein differences in plasma lipoproteome when measured in individual plasma lipoproteins than as total in immunodepleted plasma. Specifically, protein complement C3's peptide IHWESASLLR, had unadjusted p values of 0.00007, 0.00012, and 0.0006 and overall 1.25, 1.17, 1.14-fold changes in VLDL, IDL, and LDL, respectively. After positive False Discovery Rate (pFDR) adjustment, the complement C3 peptide IHWESASLLR in VLDL remained statistically different (adjusted p value <0.05). Discussion: The findings may warrant future studies to investigate plasma lipoproteome when measured in individual plasma lipoprotein fractions for AD diagnosis.