[Screening of serum oxysterol biomarkers for colon cancer by liquid chromatography-tandem mass spectrometry]
Zhanjun Ma,Zhenguo Li,Huan Wang,Renjun Wang,Xiaofei Han
DOI: https://doi.org/10.3724/SP.J.1123.2022.01001
Abstract:Colon cancer (CC) is one of the most common malignant tumors worldwide. As there are no effective biomarkers for the early diagnosis and intervention tracking, the incidence of CC is increasing every year. Cholesterol is an important component of cell membrane, and it has been shown to be associated with CC. Oxysterol is an oxidized derivative of cholesterol, which plays an important role in many malignant tumors. In this study, liquid chromatography-tandem mass spectrometry (LC-MS/MS) was used to determine serum cholesterol and ten oxysterol metabolites related to cholesterol in CC patients and healthy controls, and qualitative and quantitative analyses were carried out. Raw data were processed and analyzed using GraphPad Prism 8.3.0 and the MetaboAnalyst 5.0 platform (https://www.metaboanalyst.ca/MetaboAnalyst/ModuleView.xhtml). To perform the independent sample t-test, it was necessary to ensure that all the sample data followed a normal distribution; therefore, the normal distribution test was performed in advance. The Mann-Whitney U test, which is a nonparametric test, was adopted for samples without a normal distribution. For the processed data, we used the statistical analysis function module of the MetaboAnalyst 5.0 platform to perform partial least-square discriminant analysis (PLS-DA) and orthogonal partial least-square discriminant analysis (OPLS-DA). Both PLS-DA and OPLS-DA are supervised discriminant analysis methods. The OPLS-DA model is based on the PLS-DA model and eliminates variables that are unrelated to the experiment. In both models, the samples from the two groups were well separated by the score plot. In the PLS-DA model, the horizontal and vertical coordinates of the score plot represent the interpretation rates of the principal components of the model. The horizontal coordinates show the differences between groups, and the vertical coordinates show the differences within groups. In addition to the score plot in the PLS-DA model, another crucial factor is variable importance in the projection (VIP). When VIP>1, the compound makes an important contribution to the model and is also used as a criterion for screening differential metabolites. Based on 10-fold cross-validation (CV) of the PLS-DA model, the performance of the model was the best when the number of components was three. To avoid overfitting of the data, three metabolic markers were selected by using not only the VIP values of metabolites of the PLS-DA model, but also the optimal compositions and K-mean clusters. The three biomarkers were 4β-hydroxycholesterol (4β-OHC), cholestane-3β,5α,6β-triol (Triol), and cholesterol. A receiver operating characteristic (ROC) curve was constructed. The area under the curve (AUC) was generally between 0.5 and 1.0. In the case of AUC>0.5, the closer the AUC is to 1, the better is the performance of the model. In this study, the area under the ROC curve constructed jointly by the three metabolic markers was 0.998, indicating that their combined ability to predict CC was strong and that the diagnostic performance was excellent. In addition, to understand the role of the three metabolic markers in the pathogenesis of CC, the genes associated with the metabolic markers were identified using GeneCards (https://www.genecards.org/). Finally, 110 genes were identified. Gene ontology (GO) enrichment and Kyoto Encyclopedia of Genes and Genomes (KEGG) were used to analyze the biological processes, metabolic pathways, and possible roles in the body. GO enrichment showed that the three markers are mainly distributed in the endoplasmic reticulum lumen and coated vesicles, and they are mainly involved in biological processes such as cholesterol metabolism, transportation, and low-density lipoprotein particle remodeling. Their molecular functions are cholesterol transfer activity and low-density lipoprotein particle receptor binding. KEGG pathway analysis showed that biomarkers are enriched in steroid biosynthesis, PPAR (peroxisome proliferator-activated receptor) signaling pathways, and ABC (ATP-binding cassette) transport pathways. The results of this study are helpful to understand the role of cholesterol and oxysterol in the pathogenesis of CC and to elucidate the pathogenesis of CC.