Lysophosphatidylcholine and amide as metabolites for detecting alzheimer disease using ultrahigh-performance liquid chromatography-quadrupole time-of-flight mass spectrometry-based metabonomics

Abstract

Alzheimer disease (AD) can be diagnosed by clinical and neuropsychologic tests and at autopsy, but there are no simple effective diagnostic methods for detecting biomarkers in patients at early stages of cognitive impairment. Early metabolic alterations that may facilitate AD diagnosis have not been thoroughly explored. We applied a nontargeted metabonomic approach using ultrahigh-performance liquid chromatography-quadrupole time-of-flight mass spectrometry to analyze serum and urine samples from 46 patients with AD and 36 healthy controls. Metabolite profiles were processed using multivariate analysis to identify potential metabolites, which were further confirmed using tandem mass spectrometry. Ultrahigh-performance liquid chromatography mass spectrometry methods were additionally used to quantify potentially important biomarkers. Independent samples were then selected to validate the identified biomarkers. There was a clear separation between healthy controls and AD patients; AD patient samples had disordered amino acid and phospholipid metabolism and dysregulated palmitic amide. Receiver operator characteristic curve and quantification suggested that palmitic amide, lysophosphatidylcholine (LysoPC, 18:0), LysoPC(18:2), L-glutamine, and 5-L-glutamylglycine were the optimal metabolites. In addition, areas under the curve from the palmitic amide, LysoPC(18:2), and 5-L-glutamylglycine in the validation study were 0.714, 0.996, and 0.734, respectively. These data elucidate the metabolic alterations associated with AD and suggest new biomarkers for AD diagnosis, thereby permitting early intervention designed to prevent disease progression.