Bioinformatics Modelling and Metabolic Engineering of the Branched Chain Amino Acid Pathway for Specific Production of Microbial Biosurfactants and Biopesticides.

International audienceThe lipopeptides produced by B. subtilis have a wide range of activities. For example, mycosubtilin is a strong antifungal molecule and surfactin is one of the most powerful biosurfactants known. These two bioactive molecules can be used in many markets (cosmetics, phytosanitary, detergents, ...) and their industrial production is being developed in Europe and Asia. Both molecules are made up of a cyclic heptapeptide linked to a fatty acid chain. The isomery and the length of the fatty acid chain (FA) are responsible for the activities of the lipopeptides. In this work, the activities of the different lipopeptide isoforms were first investigated for their surfactant or antifungal properties against several pathogens, revealing the strong surfactant activity of the surfactin isoform nC14 and the strong antifungal activity of the mycosubtilin isoform anteiso-C17. Then, the modification of lipopeptides production was studied by amino acid feeding experiments. In parallel, bioinformatic and metabolic engineering strategies were performed. The knockout (KO) or overexpression of genes, leading to the specific overproduction of the most active surfactin or mycosubtilin isoforms, were predicted using computational tools that provide logical reasoning with formal models of reaction networks. In this way, different genes impacting the production of amino acids or branched-chain fatty acids or the regulation of these metabolic pathways were predicted and targeted for KO or overexpression. The influence of these deletions or these overexpression on the quantitative and qualitative lipopeptides production was analysed using HPLC, LC-MS-MS and GC-MS methods. To better understand the impact of certain deletions on the metabolism of strains and the production of their lipopeptides, two different approaches were undertaken. On the one hand, for the production of surfactin, a complete analysis of extracellular metabolites by quantitative 1H NMR was used. On the other hand, for the production of mycosubtilin, a comprehensive analysis of cellular fatty acid production was undertaken using GC-MS. The results obtained show important metabolic changes in the different mutants allowing an increase in the specific production of the nC14 isoform from 1.4 to 5.8 times and of the mycosubtilin anteiso-C17 from 1.4 to 2.8 times.These results show that increased selective synthesis of particular lipopeptide isoforms by metabolic engineering and bioinformatic-assisted synthetic biology is possible, revealing the interest of these approaches for the future development of lipopeptide-producing strains and for more targeted applications