Objectives: Schizophrenia is a serious mental illness. The serum protein biomarkers of schizophrenia were explored using isobaric tags for relative and absolute quantitation (iTRAQ) technology. The underlying function of the identified protein biomarker was also investigated.
Methods: We first collected serum samples from 12 schizophrenia patients and 12 healthy control (HC) subjects, followed by global screening with iTRAQ and tandem mass spectrometry (LC-MS/MS). In total, 691 serum proteins were detected and eight proteins, including ZYX, OSCAR, TPM4, SDPR, BST1, ARGHDB, ITIH5 and SH3BGRL3, were selected for further specific validation with enzyme-linked immunosorbent assay (ELISA) on the serum samples from 52 schizophrenia patients and 50 HC subjects.
Results: Schizophrenia patients had significantly lower serum level of BST1 and higher ITIH5 level than the HC subjects did. Using the levels of BST1, ITIH5 and OSCAR combined with machine learning algorithm, we developed a prediction model of schizophrenia with an auROC value 0.78. Moreover, in vitro cell assay confirmed that BST1 significantly repressed neutrophil infiltration through endothelial layer, highlighted the anti-inflammation nature of BST1.
Conclusions: Four novel protein markers (BST1, ITIH5, SDPR, and OSCAR) of schizophrenia were identified, and BST-1 could serve as a serum protein biomarker involved in neutrophil infiltration in schizophrenia.
Keywords: gene; iTRAQ; inflammation; proteomics; schizophrenia.