Study on dynamic modelling of metabolic flux analysis and application in metabolic-related disease model

Abstract

In recent years, the concepts and theories of systems biology have developed rapidly. Systems biology is an approach to integrating the knowledge of each level and each component. Among these levels, fluxome is an increasingly important component in addition to well-known genomes, transcriptomes, proteomes, and metabolomes. The study of this system has produced a new omics approach, metabolic fluxomics. The focus is on determining of the direction and value of the metabolic flux within the target system. The information is important for characterizing the physiological characteristics of biological systems and assessing genetic and environmental impacts on cells. The main research tool of metabolic fluxomics is metabolic flux analysis (MFA). MFA, a key technology in bioinformatics, is an effective way of analyzing the entire metabolic system by measuring fluxes. Many existing MFA approaches are based on differential equations, which are complicated to solve mathematically. Therefore, MFA requires some simple approaches to further investigate metabolism. In this thesis, a Continuous-time Markov chain approach was proposed to MFA, called MMFA approach, and transformed the MFA problem into a set of quadratic equations by analyzing the transition probability of each carbon atom in the entire metabolic system. Unlike other methods, MMFA analyzes the metabolic model only through the transition probability. This approach is generic and could be applied to any metabolic system if all the reaction mechanisms in the system are known. The MMFA approach was applied to the pentose phosphate pathway and the results were compared to several chemical reaction equilibrium constants from early experiments. In order to apply the approach to larger and more complex central metabolic processes, such as pentose phosphate pathway, glycolysis and tricarboxylic acid (TCA) cycle, the Monte Carlo method was used to solve the problem that not all reaction rate constants k can be available. The results indicated that continuous-time Markov chain metabolic flux analysis can be applied to more biological systems using the Monte Carlo method. Furthermore, this research continued to apply MFA to biological models related to metabolic diseases. The insulin signaling pathway was selected as the model. In regulated cells, insulin receptors are on the cell surface. The receptors, in combination with insulin, lead to a series of changes in the cells (signal transduction pathway). One of the reasons that lead to Type_2 diabetes mellitus is insulin resistance. The drugs do not have a guaranteed effect because the genes themselves have several steady states, while the insulin can work in some steady states but not in all of them. When insulin is injected, once the steady states reached by the genes are similar to those resisted by insulin, it stops functioning. The steady states here refer to the stable steady point with the minimum potential energy in a single cell, which may be different from the definition in biological theory. Multiple homeostasis problems are taken into consideration in the dynamic process, while multiple homeostasis in the insulin receptor is proposed to reveal the cause of insulin resistance in Type 2 diabetes through the reaction dynamic model. In addition, meta-analysis was applied to gene expression during heart regeneration in Danio rerio. Danio rerio (zebrafish) is a well-known model vertebrate owing to its ability to regenerate its organs, and this dissertation aimed to generate a global profile of genes and transcripts participating in zebrafish heart regeneration. The discovery of its heart regeneration ability and the preliminary revelation of the underlying molecular and cellular mechanisms are considered a breakthrough in the field of organ regeneration in recent years. However, the molecular mechanisms underlying heart regeneration at the transcript level are unclear. Using statistical analysis of several RNA-seq data from the published literature, the expression data of more than 30,000 genes in zebrafish, during its heart regeneration, were obtained. Gene clustering and mathematical statistics were used to analyze the expression data from different aspects. The results showed a more accurate gene differential expression profiles during zebrafish heart regeneration and obtained the corresponding biological processes by go-cluster analysis. These data might aid in guiding future research on heart regeneration of zebrafish. Our analysis, which is comprehensive and reliable, revealed several genes that have not been linked to the heart regeneration process and might be useful for rebuilding cardiac function in zebrafish. This research contributes an innovative and generic approach to metabolic flux analysis and reveals some meaningful mechanisms in disease models.