Metabolomics Study Of Severe Aplastic Anemia Patients Based On Liquid Chromatography-mass Spectrometry

ObjectiveIn our study,Ultra High Performance Liquid Chromatography Quadrupole Time of Flight Mass Spectrometry was applied to observe the changes of plasma metabolomics in patients with severe aplastic anemia and,looking for potential metabolic markers.MethodsA total of 30 patients diagnosed in Hematology Department of Tianjin Medical Univesity were enrolled in this study from January 2018 to June 2018,including 12 males and 18 females.Patients were grouped according to different types of disease(severe aplastic anemia group 8 cases,immnorelated pancytopenia group 12 cases,myelodysplastic syndrome group 10 cases).The control group includes 10 cases of healthy adults whose race was same as those patients in this study,including 4 males and 6 females.One of the control samples was removed due to hemolysis.Plasma samples were collected and cryopreserved.Four groups of plasma samples were detected by Liquid chromatograph(2777C UPLC system(Waters,UK)and mass spectrometer(Xevo G2-XS QTOF(Waters,UK).Student's test and Fold-change Analysis,and PLS-DA methods were applied to analyze the differential metabolites between SAA and the control,IRP and the control,MDS and the control,SAA and IRP,SAA and MDS,IRP and MDS groups.Results1.There was a clear separation trend in the volcano plot,and PLS-DA score plot between SAA and the control groups.The PLS-DA model under positive and negative ion patterns had an explanation rate(R2)of nearly 1 and a prediction rate(Q2)of nearly 0.5,and intercept of R2 less than 0.3,intercept of Q2 less than 0.05 after Permutation test,the model was proved to be effective,and there was no over-fitting phenomenon.2.There was a separation trend in the volcano plot,and PLS-DA score polt between SAA and IRP,SAA and MDS,IRP and MDS,IRP and the control,MDS and the control groups.The PLS-DA model between IRP and the control groups underpositive and negative ion patterns had an explanation rate of nearly 1,but the prediction rate was less than 0.5,the model was invalid.The PLS-DA model in the other four groups under positive and negative ion patterns had an explanation rate of nearly 1,but the prediction rate was less than 0.5,and the R2 intercept less than 0.3and the Q2 intercept less than 0.05 were not satisfied at the same time after Permutation test,the model was proved to be invalid,and there was over-fitting phenomenon.3.Eight metabolites were significantly different between SAA group and healthy control group under the screening and identification of differential ions.Coumaric acid,L-phenylalanine and sulfate were up-regulated in SAA group,L-glutamic gamma-semialdehyde,heobromine,3a,7a-dihydroxy-5b-cholestane,Gamm-delta-Dioxovaleric acid,9,10-DHOME were down-regulated.Seven metabolites were significantly different between IRP group and healthy control group.Hypoxanthine,Pyroglutamic acid,3a,6a,7b-Trihydroxy-5b-cholanoic acid,Cholic acid,3a,6b,7bTrihydroxy-5b-cholanoic acid,Ursocholic acid were up-regulated in IRP group,3a,7adihydroxy-5b-cholestane was down-regulated.Three metabolites were significantly different between MDS group and healthy control group.Hypoxanthine,Pyroglutamic acid were up-regulated in MDS group,3a,7a-dihydroxy-5b-cholestane were downregulated.Four metabolites were significantly different between SAA group and IRP group.L-Phenylalanine were up-regulated in SAA group,N6,N6,N6-Trimethyl-Llysine,Pyroglutamic acid,9,10-DHOME were down-regulated.Eight metabolites were significantly different between SAA and MDS group.L-Phenylalanine was up-regulated in SAA group,Hypoxanthine,L-glutamic gamma-semialdehyde,N6,N6,N6-Trimethyl-L-lysine,Biliverdin,Gamma-delta-Dioxovaleric acid,Pyroglutamic acid,9,10-epoxyoctadecanoic acid were down-regulated.ConclusionsMetabolomics research methods could distinguish between SAA and healthy controls,differential metabolites were identified and it was found that L-phenylalanine increased between SAA and the control,SAA and IRP,SAA and MDS groups,however,there was no increase between IRP and the group,MDS and the group,indicating that L-phenylalanine may be the characteristic differential metabolite of SAA,providingnew strategies for the diagnosis of SAA in the future and new ideas for the exploration of the molecular mechanism of the disease.