Diagnosis of Clostridium difficile infection using an UPLC-MS based metabolomics method.

Introduction: The fecal metabolome of Clostridium difficile (CD) infection is far from being understood, particularly its non-volatile organic compounds. The drawbacks of current tests used to diagnose CD infection hinder their application.

Objective: The aims of this study were to find new characteristic fecal metabolites of CD infection and develop a metabolomics model for the diagnosis of CD infection.

Methods: Ultra-performance liquid chromatography-mass spectrometry (UPLC-MS) was used to characterize the fecal metabolome of CD positive and negative diarrhea and healthy control stool samples.

Results: Diarrhea and healthy control samples showed distinct clusters in the principal components analysis score plot, and CD positive group and CD negative group demonstrated clearer separation in a partial least squares discriminate analysis model. The relative abundance of sphingosine, chenodeoxycholic acid, phenylalanine, lysophosphatidylcholine (C16:0), and propylene glycol stearate was higher, and the relative abundance of fatty amide, glycochenodeoxycholic acid, tyrosine, linoleyl carnitine, and sphingomyelin was lower in CD positive diarrhea groups, than in the CD negative group. A linear discriminant analysis model based on capsiamide, dihydrosphingosine, and glycochenodeoxycholic acid was further constructed to identify CD infection in diarrhea. The leave-one-out cross-validation accuracy and area under receiver operating characteristic curve for the training set/external validation set were 90.00/78.57%, and 0.900/0.7917 respectively.

Conclusions: Compared with other hospital-onset diarrhea, CD diarrhea has distinct fecal metabolome characteristics. Our UPLC-MS metabolomics model might be useful tool for diagnosing CD diarrhea.