Identification Of Serum Protein Biomarkers By Proteomic Technique For Diagnosis And Screening Of Behcet’s Disease With Uveitis

Objective To establish a serum diagnostic pattern for BD with uveitis and identify potential biomarkers related to uveitic activity by proteomic technologyMethods Human serum proteomic spectra were generated by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF-MS) combined with Proteinchip array. A set of spectra, derived from analyzing sera from49patients with BD,31patients with VKH and48age-and sex-matched healthy volunteers, was used to train and develop a decision tree model with a machine learning algorithm called decision boosting. A blinded testing set, including13patients with BD and11healthy people, was used to determine the accuracy of the model. Combined with magnetic bead, the obtained samples (31patients with active BD,31patients with inactive BD) were subjected to Tricine-SDS-PAGE. Protein spots were visualized with the "Coomassie brilliant blue" staining. Differently expressed proteins were subsequently identified by (MALDI-TOF/TOF-MS). Enzyme-linked immunosorbent assay (ELISA) and immunoprecipitation were performed to validate the results of Peptide Mass Fingerprint (PMF).Results The diagnostic pattern with a panel of6potential protein biomarkers of mass-to-charge (m/z) ratio could accurately recognize41of49patients with BD,27of31patients with VKH and44of48healthy control. A specificity of90.91%and a sensitivity of84.62%were obtained in blinded testing. Expression levels of six of the protein spots in active BD were significantly different from those in inactive BD. Four of them elevated in active phase while the rest two decreased. Mass spectrometric protein identification revealed that the three protein spots corresponded to two proteins: Apolipoprotein A1(ApoA1) and Serum Amyloid A (SAA). ELISA and immunoprecipitation verified these results.Conclusions The preliminary data suggest a potential application of MALDI-TOF MS combined with protein array as an effective technology profiling serum proteome, and with pattern analysis, a diagnostic model comprising of six potential biomarkers, is feasible to differentiate individuals with BD from VKH and healthy controls efficiently. These results indicate that useful serum biomarkers for active BD can be identified by MALDI-TOF-MS combined with the use of magnetic beads and Tricine-SDS-PAGE..