Computational and experimental methods for metabolic engineering: applications in Escherichia coli and Bacillus subtilis

Metabolic engineering was defined more than 25 years ago as the directed modulation of metabolic pathways, using methods of recombinant DNA technology, for the purpose of overproducing high-value compounds, such as pharmaceutical products, food additives and fuels. Given the increasing need of more sustainable processes for the production of value-added chemicals and materials from renewable resources, metabolic engineering became a powerful tool for the development of highly efficient microbial cell factories. The main innovation introduced from metabolic engineering, compared to traditional trial-and-error approaches, is the use of predictive modelling methods to study the behaviour of cellular metabolism and to guide the rational strain design. In this context, the cellular metabolism is described by the complete set of biochemical reactions that occur in the target microorganism, known as genome-scale metabolic model, and can be analyzed in terms of flux distributions, namely the reaction rates. Differently from gene expression levels or protein and metabolite concentration, the metabolic flux profiles are able to reflect the consequences of cellular component interactions. Despite a variety of in-silico modelling approaches have been developed for the study of cellular metabolism, only those requiring a limited number of readily available parameters can be successfully applied to genome-scale models. Currently, constraint-based modelling approach is the best methodology by which genome-scale models are constructed and analyzed. This approach identifies a set of allowable solutions, by the assumption of steady-state conditions and limiting the fluxes, and then finds an unique flux distribution, by an optimization problem that maximizes or minimizes a biological objective function. Several methods based on different objective functions, and therefore appropriate for specific study goals, were developed. Flux Balance Analysis is the most popular method, which determines the flux through the metabolic network that maximizes growth rate. However, in some contexts the reliability of such models in the quantitative prediction of cellular phenotypes and fluxes through biochemical reactions can be low. The integration of additional biological information in the model, e.g., genome-scale transcriptomic or proteomic profiles, has been recently proposed as an attempt to improve prediction accuracy. The last and crucial step for strain improvement is the application of genetic manipulations for the control of metabolic fluxes through recombinant DNA technologies. The perturbations, identified by the in-silico design phase, are implemented through the synthetic biology techniques for the tight control of gene expression levels, namely over-, down-expression and deletion. Synthetic biology is an emerging discipline, closely coupled with metabolic engineering field, that promotes the optimization of microorganisms using toolkits of pre-characterized regulatory elements. In particular, regulatory parts, such as promoters or ribosome binding sites, are commonly used for the over- or down-regulation of transcriptional and translational processes of target genes, respectively, whereas gene knockouts are implemented using homologous recombination or silencing the gene via the new proposed techniques. This thesis work includes both in-silico and in-vivo investigations on different metabolic engineering tools on Escherichia coli and Bacillus subtilis.