Dissertation

The Dynamics of the XlnR Regulon of Aspergillus Niger : a Systems Biology Approach

Omony, Jimmy

2012

Overview

In the past few decades there has been tremendous progress in the fields of Bioinformatics and Systems Biology, particularly in studying biological networks. The development of high through-put data acquisition techniques has been vital in achieving this progress. Generally, the underlying mechanisms of gene regulation constitute an important process in living cells. This process controls transmission of information encoded in the DNA sequence to proteins through the central dogma of molecular biology. The study of biological networks helps improve our understanding of complex molecular processes. The xylanolytic XlnR regulon of A. niger encodes several enzymes, which are responsible for various processes in the cell. For instance, some of these enzymes are involved in the degradation of the polysaccharide xylan and cellulose. This thesis provides a detailed look into modeling regulon dynamics and target gene regulation mechanisms. Here, the XlnR regulon of A. niger is used as a model sample system to address challenges, which are associated with the process of network dynamics and experimental design. The XlnR regulon dynamics are studied in Chapter 2, and the models are evaluated on real experimental data in Chapter 6. Primarily the models were based on qualitative prior information of the regulation mechanisms for the target genes. This thesis also highlights the need to take strategic experimental designs into consideration when studying biological networks by advocating the deployment of time course experiments. It is through integration of theory and experiments that interesting challenges on networks such as the XlnR regulon can be addressed. The application of optimal and strategic experimental design in time course experiments to study biological networks is still at its infancy. Only a few studies have paid attention to this issue in an attempt to improve the information content in their data sets. The benefits of using induction signals for studying biological networks are illustrated in this thesis. To obtain meaningful results, various aspects of experimental design were studied. There is much to gain from the use of sensitivity functions, the Fisher Information Matrix (FIM), and the E-modified criterion in the design of time course experiments using mathematical models (Chapter 3). For instance, the extent to which the individual kinetic parameters influence the measured mRNA levels (gene expression) can be assessed by using sensitivity functions. This underlines the need for smart network perturbation techniques and data sampling strategies, both of which are crucial for improving the confidence intervals on the parameters. A comparison of transcription in strains of A. niger was performed in Chapters 4 and 5. These studies provided information on both the qualitative and quantitative roles of the carbon catabolite repressor protein (CreA). The response of xylan backbone-degrading enzymes to different D-xylose induction concentrations was investigated. The expression of accessory enzyme-encoding genes was found to be favored by using a high D-xylose concentration. The extent to which transcription varies between the wild type and mutant strains after D-xylose induction was assessed. The potential of D-xylose to induce expression of cellulase-encoding genes (target genes) was assessed. High Dxylose concentrations are preferable for inducing the gene expression of enzymes in the pentose metabolic pathway. The absence of a functional CreA positively influences expression of genes encoding xylan degrading enzymes, independent of D-xylose concentration. The xlnR gene was found to be constitutively expressed. Without performing time course experiments with frequent sampling and high data resolution, there is a risk of missing out on the time dynamics of the expression of the target genes. The role of CreA in the regulation of transcription in the A. nidulans, A. oryzae and A. niger species of Aspergillus has been widely studied. It mediates repression of the target genes and it was also found that mutation of CreA significantly increases transcription in the target genes of the XlnR regulon following low D-xylose induction concentrations. Basing on the work in Chapter 6, we see that mechanistic modeling with differential equations is a powerful way to study networks such as the XlnR regulon. The XlnR regulon was perturbed to generate transcription data from both in silico and wet-lab qPCR experiments. Network induction is a good way to study the information content of genomic data sets. Compared to the low Dxylose induction, transcription was found to be lower following high D-xylose induction. Exceptions were found in a very few number of genes, which did not exhibit significant differences in transcription under the two conditions. Using nonlinear differential equations, the regulatory mechanisms of the XlnR regulon were modeled and validated on qPCR data. The strategic experimental designs led to new insights on how to maximize the transcription data information content. It was shown that differences exist both between the responses of the different target genes over time. In Chapter 6, the quantitative information of CreA proved very useful for modifying the earlier-on proposed models in Chapters 2 and 3, and to evaluate the new models. Induction with D-xylose, particularly at high concentrations, yields a diversity of transcription profiles. To understand such transcription dynamics, it is crucial to understand the roles of the individual network components e.g. genes, and transcription factors or any other signaling molecules. The explanatory power of the models was significantly improved by incorporating these components into the modeling. Getting reliable inferences from experimental data sets is essential for the study of biological networks. However, the validity of results depends on a number of processes, such as the proper planning and execution of the experimental procedure of data collection and analysis. Fortunately, numerous analytical tools are available to facilitate such studies. The work in this thesis provides a basis for future research with regards to network. This includes issues such as: i) optimal network perturbation and design of time course experiments, ii) analysis of time constants and transcription dynamic responses, iii) data sampling strategies, and iv) evaluation of candidate model structures and their appropriateness in describing the transcription dynamics from measured mRNA time course data.

• Publisher: ProQuest Dissertations & Theses
• Subject: Bioinformatics / Biology / Carbon / Cellulase / Dehydrogenases / Design / Engineering / Fungi / Gene expression / Genetics / Mathematics / Medical research / Metabolism / Morphology / Ordinary differential equations / Parameter estimation / Reverse engineering / Signal transduction / System theory
• ISBN: 9798515235680