Profiling of schizophrenia-associated serum peptides by MALDI-TOF-MS

Schizophrenia is a psychiatric condition characterized by poor prognosis and severe symptoms that decrease the quality of life of patients. It is therefore important to develop rapid and reliable methods for early diagnosis. To achieve this aim, identification of accurate biomarkers will promote research into the mechanisms of schizophrenia and the design of effective diagnosis tools. Discriminative peptides in the serum of participants (277 schizophrenia patients and 294 healthy controls) were detected using the matrix-assisted laser desorption ionization time-of-flight tandem mass spectrometry (MALDI-TOF-MS). The diagnostic model for schizophrenia was validated using the ClinProTool software. Discrimination models were tested in the training set (200 schizophrenia patients and 200 healthy controls), and the robustness of the models was evaluated using the independent test set (77 cases and 94 controls). A total of 77 peaks were significantly different between schizophrenia patients and healthy controls with a signal-to-noise ratio of over 5. Among them, 35 peptides were down-regulated and 42 peptides were up-regulated in schizophrenia patients. With a cross-validation rate of 92.7% and a recognition capability rate of 96.5%, the supervised neural network comprising 25 discriminative peaks was found to be the most efficient model for schizophrenia diagnosis (sensitivity = 96.10%, specificity = 94.68%). Peptides at m/z = 2022.18 corresponded to complement C3f, and peptides at m/z = 1020.89, 1351.02, 1466.1 were identified as fragments of fibrinopeptide A. These two peptides may be potential biomarkers for schizophrenia in the Chinese population. The serum peptide levels present a potential clinical diagnostic tool for schizophrenia.