Ferric particle-assisted LDI-MS platform for metabolic fingerprinting of diabetic retinopathy.

Abstract

To explore the metabolic fingerprints of diabetic retinopathy (DR) in individuals with type 2 diabetes using a newly-developed laser desorption/ionization mass spectrometry (LDI-MS) platform assisted by ferric particles. Metabolic fingerprinting was performed using aferric particle-assisted LDI-MS platform. A nested population-based case-control study was performed on 216 DR cases and 216 control individuals with type 2 diabetes. DR cases and control individuals with type 2 diabetes were comparable for a list of clinical factors. The newly-developed LDI-MS platform allowed us to draw theblueprint of plasma metabolic fingerprints from participants with and without DR. The neural network afforded diagnostic performance with an average area under curve value of 0.928 for discovery cohort and 0.905 forvalidation cohort (95 % confidence interval: 0.902-0.954 and 0.845-0.965, respectively). Tandem MS and Fourier transform ion cyclotron resonance MS with ultrahigh resolution identified seven specific metabolites that were significantly associated with DR in fully adjusted models. Ofthese metabolites, dihydrobiopterin, phosphoserine, N-arachidonoylglycine, and 3-methylhistamine levels in plasma were first reported to show the associations. This work advances the design of metabolic analysis for DR and holds the potential to promise as an efficient tool for clinical management of DR.