Abstract

Metabolic network reconstruction is a list comprising annotated possible reactions for a given organism. It relies on genomic data and provide useful information about the organism’s ability in converting metabolites. Reconstructions can vary in scope depending on one’s objective: they can be as complete as possible, in the case of genome scale network reconstructions, or they can focus on a metabolism subset. We adopted the last approach with the bacteria Burkholderia sacchari because we were interested in the set of reactions involved in poly-3-hydroxybutyrate (P3HB) production, since the bacteria is a natural producer of it. As we have recently sequenced and automatically annotated the bacteria’s genome, our aim was to use this draft annotation to reconstruct the core metabolism of P3HB production from a carbon source, which we adopted to be glucose. Therefore, we first selected a subset of metabolic reactions to guide the curation process. This subset comprised Embden-Meyerhoff-Parnas, Entner-Doudoroff and Pentose-Phosphate pathways, the Citric Acid Cycle, reactions involved in P3HB formation and other reactions from core reconstructions of Escherichia coli and Pseudomonas putida (i.e. transhydrogenases and atp maintenance reaction). Then, we checked each one of these pre selected reactions in reference databases (i.e. KEGG and BioCyc) to obtain accurate annotated information such as reaction stoichiometry, enzyme commission number of proteins, reaction subsystem, among others. In this process, some reactions were rewritten and a few were added. Next, we evaluated the presence of each one of these accurate written reactions in the annotated genome so that we just selected the present ones to create our reconstruction, which was used in Metatool to obtain elementary modes. Finally, we ran bioreactor experiments to access which combination of elementary modes better explained the measured metabolic fluxes.