| Objective: The purpose of this article is to study the effect of cassia aqueous extract and aurantio-obtusin on metabolites in hyperlipidemia rats based on metabolomics via a GC-MS method. Common analysis methods of metabolomics were used to find out the biomarkers of hyperlipidemia, and the association between cassia & aurantio-obtusin and the biomarkers of hyperlipidemia was discussed. The results of this study could provide experimental basis for the further development and clinic application of cassia and aurantio-obtusin.Methods:(1) An establishment of hyperlipidemia model for rats.Male Sprague-Dawley rats were randomly divided into two groups after adapting to the environment for 3 days. The normal control group received common feed, while the hyperlipidemia model group was fed on high fat diet(which contains 8.0% yolk powder, 6.0% lard, 4.0% cholesterol, 0.5% sodium cholate, 0.2% propylthiouracil and 81.3% normal feed). The biochemical indexes were detected by the blood sampling of the fundus venous plexus to confirm the modeling results 30 days later.(2) A GC-MS method was established for quantitative analysis of endogenous metabolites in rat serum.The ingredients that can be detected by GC-MS must be volatile. So we pretreated the serum sample with methoximation followed by trimethylsilyl(TMS) derivatization in order to make metabolites that are non-volatile and have large polarity can be detected. The quality control(QC) samples were produced by mixing a plurality of samples. Both the QC samples and the mixed samples were used to inspect the GC-MS methodology. Among them, the RSD value of the relative content of each standard in the mixed standard samples was used to evaluate the stability of the instrument system.(3) The screening of hyperlipidemia-related biomarkersCompared the changes of hyperlipidemia rats in vivo metabolites before and after the treatment with the aqueous extract of cassia, aurantio-obtusin and atorvastatin. Variable importance in projection and random forest algorithm were used to find out potential biomarkers associated with hyperlipidemia. Then the potential biomarkers were confirmed by T-test analysis to search for the final biomarkers. By finding out the association of biomarkers and the lipid regulating effect by the aqueous extract of cassia, aurantio-obtusin and atorvastatin, we can infer its lipid metabolism.Results:(1) Except the rats died in the process of modeling, the biochemical indexes of the alive rats have a change. The LDL-c and TC were increased significantly, suggesting that hyperlipidemia model for rats was successful established.(2) The RSD value of the most metabolites in QC samples and the relative contents of the standard samples in mixed standard samples were all less than 20%, which shows that the sample pretreatment has good repeatability, while the instrument also has good stability. So the data obtained are reliable and can be used for further data analysis. Standard substances and NIST107 & NIST05 MS library were used to give a quantitative analysis of endogenous metabolites in rat serum. Finally, we get a total of 45 metabolites, 27 were identified by standard substances, and the rest 18 were identified through retrieving in the NIST107 & NIST05 MS library, insuring the similarity were above 85%. These 45 metabolites include organic acids, fatty acids, amino acids, glucose, and cholesterol.(3) Nine potential hyperlipidemia-related biomarkers were found out by using the method with the orthogonal partial least squares discriminant analysis. They are L-valine, alanine, glycine, fumaric acid, isoleucine, serine, stearic acid, linoleic acid and cholesterol respectively. And ten potential hyperlipidemia-related biomarkers were found out by using the method with the random forest algorithm. They are alanine, glycine, L-valine, isoleucine, fumaric acid, serine, 1, 5-anhydro-d-ghlcitol, linoleic acid, stearic acid and cholesterol respectively. These biomarkers were analysed by T-test analysis, and finally we get 5 kinds of metabolites with difference, they are alanine, L-valine, isoleucine, linoleic acid and cholesterol respectively. The PLS-DA model was established with these five kinds of metabolites, which could obtain good resolution. So they are identified as the hyperlipidemia-related biomarkers.Conclusions:(1) There are some defects for GC-MS as a data acquisition of metabolomics. So we pretreated the serum sample with methoximation followed by trimethylsilyl(TMS) derivatization to improve the accuracy and stability of the system, and it became a better method service to metabolomics finally.(2) In this study, we sought to find out biomarkers associated with hyperlipidemia by two different analysis methods. Finally 5 kinds of potential hyperlipidemia-related biomarkers were obtained by using T-test analysis and PLS-DA model. Compared the changes of metabolites in vivo of the hyperlipidemia rats before and after the treatment with the aqueous extract of cassia and aurantio-obtusin by PCA and PLS-DA, we found that aurantio-obtusin has a better lipid regulating effect. |
PDF Full Text Request