Plasma Characteristic Metabolites of Pediatric Community Acquired Pneumonia in TCM Syndrome Differentiation Syndrome Differentiation


 Background Cold syndrome CS and Heat syndrome ( H S)S), with the opposite clinical manifestations, are two main syndrome types of pediatric community acquired pneumonia ( in Traditional Chinese medicine TCM )). According to research of syndrome types among CAP children, about 2.92% of them can be identified with CS while 91.25% with HS. This study aim s t o analyze plasma metabolic profiles and find out potential biomarkers for distinguishing H S from CS.Methods A total of 296 patients and 55 healthy controls (HC) were divided into discovery group ( n =213, HS=160, CS=23, HC= 30) and validation group N=138, HS=93, CS=20, HC = Plasma metabolic profiles were detected by ultra performance liquid chromatography combined with linear ion trap quadrupole-Orbitrap mass spectrometry (UPLC/LTQ Orbitrap MS) in both positive and negative mode. Finally, plasma metabolic profiles were obtained through principal component analysis (PCA) and orthogonal partial least squares discriminant analysis (OPLS DA)DA). Differential metabolites in CS and HS were screened by counting the value of fold change (FC 1.5 or 0.667 P value ( P 0.05 ) and false discovery rate (FDR 0.05 Diagnostic accuracy of potential biomarkers was evaluated by receiver operating characteristic (ROC) curve.Results Among all metabolites 3 amino acids including alanine, phenylalanine, arginine, and 4 glycerophospholipids including lysoPC20:1, lysoPE16:0, lysoPE18:0, PE(16:0 22:6) were increased (FC 1.5 FDR 0.05) in CS versus HC . While 6 glycerophospholipids including PC(18:1 18:1), PC(20:4 20:4), lysoPE 20:4, lysoPE 18:2, lysoPE 22:6 andPE(16:0 18:2) were decreased (FC 0. 667 FDR 0.05) in HS versus HC. However, Ceramide(d18:1 24:1) was increased in HS versus HC . Differential metabolites were mainly involved in amino acid, glycerophospholipid and linoleic acid metabolism via KEGG pathway analysis P<0.05) and they all exhibited good diagnostic abilities with ROC analysis AUC 0.9 54 Mo reover, lyso PC20:1, lysoPE16:0, lysoPE18:2, lysoPE20:4, lysoPE22:6, PC(18:1_18:1), PC(20:4_20: and PC(P 16:0_22:4) may serve as potential biomarkers for distinguishing HS from CS.Conclusions Plasma metabolism of amino acids and lipids (triglyceride, glycerophospholipid and sphingomyelin were largely disturbed in CAP children with CS or HS. Among them, detection of glycerophospholipids via metabolomics can help diagnose these two syndrome types on clinic.

Background Cold syndrome CS and Heat syndrome ( H S)S), with the opposite clinical manifestations, are two main syndrome types of pediatric community acquired pneumonia ( in Traditional Chinese medicine TCM )). According to research of syndrome types among CAP children, about 2.92% of them can be identi ed with CS while 91.25% with HS. This study aim s t o analyze plasma metabolic pro les and nd out potential biomarkers for distinguishing H S from CS. Methods A total of 296 patients and 55 healthy controls (HC) were divided into discovery group ( n =213, HS=160, CS=23, HC= 30) and validation group N=138, HS=93, CS=20, HC = Plasma metabolic pro les were detected by ultra performance liquid chromatography combined with linear ion trap quadrupole-Orbitrap mass spectrometry (UPLC/LTQ Orbitrap MS) in both positive and negative mode. Finally, plasma metabolic pro les were obtained through principal component analysis (PCA) and orthogonal partial least squares discriminant analysis (OPLS DA)DA). Differential metabolites in CS and HS were screened by counting the value of fold change (FC 1.5 or 0.667 P value ( P 0.05 ) and false discovery rate (FDR 0.05 Diagnostic accuracy of potential biomarkers was evaluated by receiver operating characteristic (ROC) curve.
Conclusions Plasma metabolism of amino acids and lipids (triglyceride, glycerophospholipid and sphingomyelin were largely disturbed in CAP children with CS or HS. Among them, detection of glycerophospholipids via metabolomics can help diagnose these two syndrome types on clinic.
Full-text Due to technical limitations, full-text HTML conversion of this manuscript could not be completed. However, the manuscript can be downloaded and accessed as a PDF. Figure 1 Flow chart of children enrolled in this study. Abbreviations: CAP=community acquired pneumonia.

Figure 2
Metabolic pro les of plasma samples between CS, HS versus HC. A OPLS DA sc ore plots of plasma upper phase between CS versus HC. B OPLS DA score plots of plasma upper phase between HS versus HC. C OPLS DA score plots of plasma upper phase between CS versus HS. D OPLS DA score plots of plasma lower phase between CS versus HC. E OP LS DA score plots of plasma lower phase between HS versus HC. F OPLS DA score plots of plasma lower phase between CS versus HS. Elliptical area represents 95% CI of Hotelling's T squared distribution in each gure. R2Y indicates goodness of t, Q2 indicates goodness of prediction. Abbreviations: CS=cold syndrome, HS=heat syndrome, HC=healthy control , OPLS DA=orthogonal partial least squares discriminant analysis, C I=con dence interval.

Figure 3
Analysis of differential metabolites screened in CS, HS versus HC. Venn diagram of differential metabolites in CS vs . HC, HS vs. HC and CS vs. HS groups. Each cycle in Venn diagram represent s 7, 13, 15 differential metabolites screened in CS vs. HS, CS vs. HC, HS vs. HC. The overlapping area represents the same differential metabolites in different groups. Box whisker p lots showed the relative concentration of 7 differential metabolites for distinguishing CS f rom HS. X axis means grouping and Y axis means relative concentration. Whiskers are 10 90 percentile range and boxes represent interquartile range . P <0.05, P <0.0 1, *** P <0.001, P <0.0001 Abbreviations: ns=not signi cant, HS=heat syndrome, CS=cold syndrome, HC=healthy control, PC=phosphatidylcholine, PE=phosphatidylethanol amine, lysoPC=lysophosphatidylcholine, lysoPE=lysophosphatidylethanolamine. KEGG pathway analysis and metabolic pathway network of differential metabolites screened in HS and CS. KEGG pathway analysis were summarized based on 24 screened metabolites (P 0.05). Names of 5 enriched pathways were annotated beside each point (P 0.05). Metabolic pathway network was drawn by reference to screened differential metabolites and KEGG pathway analysis. Rectangle represents CS, circle represents HS, blue color represents down-regulation, red color represents up-regulation, grey triangle represents undetected or not signi cant metabolites. Name of differential metabolites screened in CS and HS were all labeled in red. Abbreviations: CAP=community acquired pneumonia, HC=healthy control, CS=cold syndrome, HS=heat syndrome, PC=phosphatidylcholine, PE=phosphatidylethanolamine, lysoPC=lysophosphatidylcholine, lysoPE=lysophosphatidylethanolamine. ROC curve for testing diagnostic signi cance of differential metabolites screened in CS and HS. A ROC curve of 13 differential metabolites screened i n CS vs. HC. 95% CI: 0.967 1, sensitivity: 96%, speci city: 97%, AUC: 0.995. B ROC curve of 15 differential metabolites screened in HS vs. HC. 95% CI: 0.906 0.985, sensitivity: 96%, speci city: 60%, AUC: 0.954. C ROC curve of 7 differential metabolites screened in CS vs. HS. 95% CI: 0.955 1, sensitivity: 98%, speci city: 83%, AUC: 0.982. Abbreviations: HC=healthy control, CS=cold syndrome, HS=heat syndrome, ROC receiver operating characteristic , AUC area under the curve, CI=con dence interval.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.