Quantitative Label-Free Proteomics for Discovery of Biomarkers in Cerebrospinal Fluid: Assessment of Technical and Inter-Individual Variation

• DOI: 10.1371/journal.pone.0064314
• Publication Date: 2013-05-20T21:01:19Z
• Authors: Richard J. Perrin; Jacqueline E. Payton; James P. Malone; Petra Gilmore; Alan E. Davis; Chengjie Xiong; Anne M. Fagan; R. Reid Townsend; David M. Holtzman
• Keywords: Biomarker Discovery

Background Biomarkers are required for pre-symptomatic diagnosis, treatment, and monitoring of neurodegenerative diseases such as Alzheimer's disease. Cerebrospinal fluid (CSF) is a favored source because its proteome reflects the composition brain. Ideal biomarkers have low technical inter-individual variability (subject variance) among control subjects to minimize overlaps between clinical groups. This study evaluates process multi-affinity fractionation (MAF) quantitative label-free liquid chromatography tandem mass spectrometry (LC-MS/MS) CSF biomarker discovery by (1) identifying reparable sources variability, (2) assessing subject variance residual numerous proteins, (3) testing ability segregate samples on basis desired characteristics. Methods/Results Fourteen aliquots pooled two from six cognitively normal individuals were randomized, enriched low-abundance proteins MAF, digested endoproteolytically, randomized again, analyzed nano-LC-MS. Nano-LC-MS data time m/z aligned across relative peptide quantification. Among 11,433 charge groups, 1360 relatively abundant ones annotated MS2, yielding 823 unique peptides. Analyses, including Pearson correlations LC-MS ion chromatograms, performed all pairwise sample comparisons, identified several variability: i) incomplete MAF keratins; ii) globally- or segmentally-decreased current in isolated analyses; iii) oxidized methionine-containing Exclusion these yielded 609 peptides representing 81 proteins. Most showed very coefficients variation (CV<5%) whether they quantified mean only 2 most-abundant Unsupervised clustering, using 24 selected high variance, perfect segregation individual samples. Conclusions Quantitative LC-MS/MS can measure scores with according criteria. Thus, this technique shows potential neurological diseases.