Entkopplung von Wachstum und Überproduktion von Chemikalien in Escherichia coli

Metabolic engineering enables the construction of microbial strains that can effectively overproduce chemicals. However, overproduction strains use substrates not only for the product formation but also for growth. Thus, there is a trade-off between product formation and cell growth that can result in sub-optimal production performance. A solution is to decouple growth and overproduction in two-stage bioprocesses. After a first stage, in which cells grow without product formation, growth is stopped in the second stage while production is induced. There are two key challenges in the creation of two-stage bioprocesses: (1) the microbial metabolism and cell growth needs to be dynamically controlled to achieve a transition between the two stages, and (2) metabolic activity and production rates are typically low in non-growing cells, which needs to be adjusted by genetic engineering. For both challenges, precise knowledge about metabolite concentrations in the engineered strains is critical to guide the design efforts. Therefore, quantitative and high-throughput mass spectrometry-based metabolomics methods are developed that enable metabolite measurements in large numbers of engineered strains. One of the fastest methods is flow-injection mass spectrometry. In the here presented work, we study aspects of the dynamic control of metabolism and growth, metabolic activity under growth arrest, and flow-injection mass spectrometry. Studying all of these aspects is important to improve our understanding of how to decouple growth and overproduction of chemicals. Chapter 1 provides a general introduction to metabolic engineering, mass spectrometry-based metabolomics, and twostage bioprocesses. Chapter 2 is a short review of metabolic networks. Flow-injection mass spectrometry (FI-MS) is a metabolomics method that can detect hundreds of metabolites with measurement times in the second scale. However, FI-MS does not rely on chromatographic separation of metabolites prior to analysis. Since all metabolites arrive simultaneously at the mass spectrometer, this could lead to negative effects like in-source modifications and false-positive annotations. With Chapter 3, we provide a systematic study of in-source modifications during FI-MS. Key in our analysis was the use of 160 authentic metabolite standards added to a metabolite extract sam-ple. A network approach and information about metabolite fragmentation identified abundant in-source modifications and showed that even sequential modification events occur. Our analysis approach could explain a large fraction of these modifications. The here presented data are a valuable resource and can be helpful to avoid false-positive metabolite annotations. Temperature-sensitive proteins carry amino acid substitutions rendering them active at low temperatures. Yet, at higher temperatures, at which the wild type proteins are still active, temperature-sensitive proteins are inactive. Here, we study temperaturesensitive proteins as a tool for metabolic engineering to dynamically control cell growth and metabolism. A goal was to use temperature-sensitive mutants to decouple growth and overproduction of chemicals and create two-stage bioprocesses. Since the identification of temperature-sensitive mutants can be challenging, a focus of our work was also on the development of high-throughput approaches to find temperature-sensitive mutants. With Chapter 4,we present a high-throughput method to enrich temperature-sensitive mutants of a single essential gene in Escherichia coli. The method coupled a TIMER protein-based single cell growth rate reporter with fluorescence activated cell sorting. This allowed us to screen millions of cells and enrich temperature-sensitive mutants of argininosuccinate synthetase ArgG. We showed that temperature-sensitive ArgG functions as a metabolic valve that allows for gradual control of growth by temperature. At the same time, it also allows for the overproduction of citrulline, which is the substrate of the ArgG-catalysed reaction. Using temperature-sensitive ArgG, we achieved a two-stage bioprocess that, within 45 h, produced 3 g/L citrulline on a 1 L-bioreactor scale. We follow up on the study of temperature-sensitivity as tool for metabolic engineering and describe an approach to generate and identify temperature-sensitive mutants in many different genes (Chapter 5). We adapted a barcoded CRISPR/Cas9 genome editing method and used a custom design approach to create a pooled E. coli strain library with 15,120 members. Each strain carried a mutation causing a single amino acid substitutions in one of 346 essential proteins. In competitive fitness assays at two temperatures, we tracked the abundance of single strains in the pooled strain library by deep sequencing of plasmid-borne barcodes. This allowed us to identify 1,045 temperature-sensitive candidate strains. After isolating a subset of 92 strains, we validated the function of 42 temperature-sensitive enzymes as metabolic valves by FI-MS. As final step, we applied seven temperature-sensitive strains in the two-stage overproduction of chemicals. A promising approach to achieve high metabolic activity under growth arrest is en-forced ATP wasting. With Chapter 6, we provide a study on enforced ATP wasting in E. coli. Overexpression of ATPase resulted in strongly increased glucose uptake rates in anaerobic conditions under nitrogen starvation. Fermentation products accumulated rapidly until glucose was depleted from the medium. Following up on our study on enforced ATP wasting, we analysed how different levels of ATPase overexpression affected energy metabolism in E. coli (Chapter 7). Increasing ATPase levels also increased glucose uptake rates up to a critical point. Increasing the expression levels beyond this critical point resulted in a sharp decrease in the glucose uptake rate below the rate of a wild type strain. We showed that this effect is caused by an enzyme in upper glycolysis: phosphofructokinase, which has ATP as substrate and is allosterically activated by ADP. These findings contribute to a better understanding of E. coli energy metabolism. They also show how effective enforced ATP wasting is at increasing metabolic activity in growth arrested cells making it a very powerful tool in metabolic engineering.