< Back to previous page


Metabolic engineering and modelling of Escherichia coli for the production of succinate

Book - Dissertation

Subtitle:Metabolisch modelleren en modificeren van Escherichia coli voor de productie van succinaat
Current climate issues and the ongoing depletion of oil reserves have led to an increased attention for biobased production processes. Not only the production of bio-energy but also biochemicals have gained interest. Recent reports of the US department of energy and the GROWTH program of the European commission review a comprehensive list of chemicals that can be produced via biological processes and which may be of great importance to sustain a green chemical industry in the future. Succinate is one of those biochemicals. Today, this compound is synthesised via maleic anhydride, which is produced by a petrochemical production process. The conditions which a biological production processes have to meet to be economically viable are quite strict. Such a process has to obtain a yield of 0.88 g/g, a rate between 1.8 and 2.5 g/l/h and a titer around 80 g/l. None of the available (reported) processes reach either of these values. In most cases the rate and the titer are still a problem. To optimise succinate production via metabolic engineering, first a mutation strategy has to be developed. This strategy can then be applied to a suitable production host. The choice of this host has nowadays become less important due to the recent developments in genetic engineering and synthetic biology. These developments allow the introduction or altering of almost every cellular function. What has become important is the availability of information on the potential host and its genetic accessibility. E. coli is therefore still an excellent host for the development of production processes. Since its isolation vast amounts of information have been gathered and several biological databases are devoted to it. Moreover, almost each cellular function has been modified. However, E. coli does not naturally produce succinate in large amounts. It will have to undergo some genetic modifications to overproduce this chemical. Which modifications are needed can be uncovered in silico. A functional and comparative genomics analyses of natural producing and non-producing strains revealed which genes and reactions may influence succinate production. The optimal biochemical route towards succinate is then uncovered via stoichiometric network analysis. For this analysis, elementary flux modes was combined with partial least squares regression. Both tools resulted in the identification of optimal biochemical production routes for several substrates and allowed to evaluate how reactions that do not naturally occur in E. coli may affect the succinate yield. The transport reaction is one of the reactions that could be identified by the EFM-PLS model. E. coli possesses both succinate import as well as export proteins. However, export is normally only active under anaerobic conditions and import under aerobic conditions. Therefore, the import protein was knocked out and the export protein was expressed with an artificial promoter. These modifications led to an increased succinate yield and production rate, but also revealed alternative import proteins. An analysis of the phenotype of mutant strains in these alternative importers did however not lead to increases in succinate yield. These mutations influenced biomass yield and growth rate. A second route that was identified in the stoichiometric network analysis was the glyoxylate route. This route correlated positively with succinate production and is strongly regulated by the transcription factors ArcA and IclR. In order to gain more insight into the synergy that may exist between both regulators, knock outs in both genes were studied under chemostat and batch conditions. This analysis revealed a synergetic effect between both proteins on the biomass yield. A strain in which both arcA and iclR are knocked out showed a biomass yield that approached the maximal theoretical yield. The single knock out strains did not have such an outspoken phenotype. Finally, several mutations were introduced and evaluated for succinate production and byproduct formation. The formation of acetate was studied in detail to uncover alternative acetate formation reactions. First, the known reactions, acetate kinase, phospho-acetyltransferase and pyruvate oxidase were knocked out. This resulted in a significant decrease in acetate production but not in the total elimination. Several alternatives such as citrate lyase and acetate CoA-transferase were evaluated, but without success. The remaining acetate formation reactions could not be identified. Succinate dehydrogenase can be seen as one of the most crucial enzymes for succinate production. This enzyme converts succinate into fumarate and therefore has to be knocked out to increase production. Strains that possess a succinate dehydrogenase deletion immediately show an increased production. However, pyruvate becomes one of the main byproducts. Several enzymes influence pyruvate production. The most important enzymes in the context of succinate production are PEP carboxykinase, oxaloacetate decarboxylase, malic enzyme, PEP carboxylase, and citrate synthase. The three former reactions are gluconeogenic reactions that can form futile cycles. Deletions in these genes resulted in an increase in biomass yield due to a more energy efficient metabolism, but does not increase succinate yield. Point mutations in PEP carboxylase and citrate synthase increased the flux towards the TCA cycle. The flux ratio between the glyoxylate pathway and the reductive and oxidative TCA cycle can be influenced by these enzymes. The activity of the reductive TCA is however strongly dependent on the availability of reduced equivalents. To modulate this availability a point mutation was introduced in FNR, an anaerobic transcription factor that activates the reductive TCA and represses the electron transport chain. Although none of the developed strains are economically viable yet, many of the mutations that have been introduced show great promise for future improvements. In fact, the next steps in strain development should not be to identify new targets to modify, but rather to fine tune activities of the routes towards succinate in such a way that the theoretical yields can be approached with sufficiently high rates.
Publication year:2010