Serum Metabolomics Study For Schizophrenia Based On UPLC-QTOF/MS Platform

Schizophrenia(SZ)is a kind of severe mental illness with high incidence and poor prognosis,and it is difficult to identify at early stage of SZ,resulting in huge economic losses and psychological burden to family and patients themselves.Now,SZ has been one of major public health problems for the whole world.The mechanism of SZ is still elusive,mainly thought to be related with genetic and environmental factors and neurodevelopmental abnormalities.The occurrence,development and prognosis of SZ are a long-term process under the interaction between genetic and risk factors.Currently,the diagnosis of schizophrenia is mainly based on scale,which is subjective,lack of objective evidence and can not distinguish heterogeneity among patients.Although brain imaging techniques can display basic organic disorders which provides a convenient way for people to understand the structure and function of the brain,it is still expensive and with poor applicability.Therefore,looking for objective biomarkers can not only provide a basis for further exploring the pathogenesis of schizophrenia,but be able to provide objective indicators for early identification and diagnosis of SZ.Metabolomics is an omics method which analyzes changes in endogenous small molecule metabolite concentrations from the global or omics level.The state of the body's biochemical metabolism will change accordingly in the development and progression of disease,furthermore the concentration of metabolites can be measured and analyzed by metabolomics method,so distinguished metabolites differentially expressed in patients and healthy people can be found.Therefore,metabolomics method has a significant role in understanding the pathogenesis of complex diseases and provides a new technical method for the prevention and early diagnosis of complex diseases.In this study,ultra-high pressure liquid chromatography-quadrupole-time of flight tandem mass spectrometry(UPLC-QTOF/MS)was used to conduct non-targeted metabolomics test of serum in schizophrenia patients and healthy controls to obtain metabolic fingerprints under the modes of positive and negative ions.XCMS pretreatment was adopted to acquire standard data format that can be processed statistical analysis,then principal component analysis(principal component analysis,PCA)without supervision and partial least square discriminant analysis(partial least square discriminant analysis,PLS-DA)supervised were used to obtain metabolic classification patterns between schizophrenia patients and controls.Finally,biomarkers were selected according to variable importance in projection(VIP)and results of univariate statistical analysis.Biomarkers were identified through online database query,primary and secondary mass spectra,then the biological function and pathogenic mechanism leading to SZ of biomarkers were further explored.Results:The metabolic patterns between SZ patients and healthy controls were significantly different,and the result of classification had good performance.Carnitine,sphingosine,stearamide and PI(P-18:0/22:6(4Z,7Z,10Z,13Z,16Z,19Z))were indentified.In addition,concentrations of these biomarkers were obviously different between SZ and control groups.They were mainly involved in fatty acid metabolism and neurodevelopment process.Among the six statistical pattern recognition models,random forest and logistic discriminant models had better classification and discrimination ability.Conclusions:SZ patients and healthy controls have different metabolic characteristics,and the concentrations of metabolites and metabolic pathways in SZ patients were changed.The occurrence and development of SZ may be primarily concerned with energy metabolism changes and neurodevelopmental abnormalities.Carnitine,sphingosine,stearamide and PI(P-18:0/22:6(4Z,7Z,10Z,13Z,16Z,19Z))have potential clinical application value in the early diagnosis of schizophrenia.