Type Ii Diabetic Mice Urine Metabolomics Research

Type II diabetes mellitus (T2DM) is a metabolic disturbance of multiple aetiology that characterised by chronic hyperglycemia with disturbances of carbohydrate, fat, and protein metabolism resulting from defects in insulin secretion, insulin action, or both. It tends to increase rapidly on a global scale, especially in most developing countries and becomes a public problem that severely impairs people's health. If patients of T2DM are not given timely treatment, serious complicating diseases would be caused. Establishing good animal models is the basis of studying T2DM. Suitable animal models establishment has great significance in researches about pathogenesis of T2DM and those complicating diseases, and can help us take preventative measures.Metabolomics is a subject that conducts high throughout detections and data processing, analyzes the indexes of studied groups and studies the dynamic metabolites change of tissue and cell system or the whole living organisms. The subject mainly focuses on the endogenous metabolism, hereditary variation, environmental change and characters or infectious after materials entering into the metabolic system. As another new branch of science after genomics and proteomics, metabolomics has been widely used in disease diagnosis, drug effect evaluation, gene function, drug mechanism of action, seeking biomarkers and so on. Two main difficult and attractive problems that exist in researches about metabolomics are as follows:(1)how to identify the structure of metabolites and their quantities accurately; (2)how to mine useful information from the complex metabolic systems and explain them reasonably. For example, how should we combine the research results of metabolomics with pathological pharmacology effectively, and integrate them with clinical intervention. Aiming at these two problems, corresponding research was carried on in this thesis. Firstly, GCMS combined with effective trimethyl silylanized derivation, was employed to establish a stable and reliable method for extracting and qualitative and quantitative investigation of metabolites. Then chemometrics methods, such as multivariate curve resolution and pattern recognition were adopted to analyze the urine metabolic profiles of a T2DM mice model (C57BL/6J/AMPKa2-KO/KK-ay mice group). Three dimensional discriminant models for the mentioned animal model based on different strain, gender and stage were created, and biomarkers were screened out.The contents in this paper mainly includes several respects as followings:(1) GC/MS technique coupled with universal derivatization protocols were used to analyze the metabolites profiling of urea from 8 cases of normal C57BL/6J mice,8 cases of KK-ay mice and 78 cases of AMPKα2-KO mice. With the help of heuristic evolving latent projections (HELP) and selective ion analysis (SIA), overlapping peaks of acquired data were effectively handled.78 endogenous metabolites were identified and and quantified.(2) Adopting the univariate statistic t-test to analyze the remarkable metabolism differs between C57BL/6J and AMPKa2-KO mice, and Principal Component Analysis (PCA) was employed to present the efficiency of clustering.(3)With the employment of pattern recognition, a PLS-LDA model was basically established for discriminant analysis between females and males AMPKa2-KO mice. A newly proposed method that can be used to screen variables-Subwindow Permutation Analysis (SPA) coupled with PLS-LDA, were applied to maximize the separation between the male and female classes, and pick out the key compound or compounds that play important roles in the process of discriminating.(4)Based on The MCTree approach, which aims to solve the variable selection problems in multi-class metabolomics data by establishment of a large number of decision tree models, obvious and complex trajectory from the urinary GCMS data of the AMPKa2-KO mice were revealed and separation trends that related to age were maximized. All 8 metabolites that selected as biomarkers can be briefly concluded to have great relationship to energy metabolism such as lipids and carbohydrate metabolism. So our work can provide useful information for Study on the Pathogenesis of type 2 diabetes mellitus (T2DM).