Dynamic quantitative description of the metabolic capabilities of Streptomyces clavuligerus for clavulanic acid production

Clavulanic acid (CA) is a β-lactam antibiotic with potent inhibitory activity against β-lactamase enzymes, which are responsible for the antibiotic resistance phenomenon in several pathogenic bacteria. CA is a secondary metabolite of pharmaceutical and industrial interest naturally produced by the filamentous Gram-positive bacterium Streptomyces clavuligerus (S. clavuligerus) under limited nutritional conditions. CA has limited availability in the market, and its high cost is a consequence of the complexity of the production process. It is determined mainly by the low titers of CA obtained in submerged cultivations of S. clavuligerus and the difficulties associated with the down-stream process. Several authors have applied experimental approaches to study the influence of some variables, especially those of nutritional nature, on CA accumulation. Nevertheless, the effect of variables relevant to the bioprocess operation, as the reactor hydrodynamics and shear stress conditions, has not been well explored. Fluxomic approaches have been recently applied to S. clavuligerus aimed to improve the understanding of its metabolism. Such approaches were mostly developed under steady-state assumptions, resulting in a limited comprehension of the metabolism under the dynamic conditions of batch and fed-batch processes. In this thesis, the reconstruction of a new and enhanced genome-scale model of S. clavuligerus, the successful combination of experimental studies in the shake flask and bioreactor scales coupled with constraint-based modeling in pseudo-steady (flux balance analysis) and dynamic (dynamic flux balance analysis) conditions were used as strategies for studying the metabolic response of S. clavuligerus to environmental and nutritional perturbations in connection with CA biosynthesis. Experimental studies in stirred tank and 2-D rocking-motion bioreactors provided valuable information on the strain's metabolic response to environmental conditions, especially regarding the effect of shear forces. Moreover, the experimental data obtained allowed to test different in silico scenarios by using constraint-based modeling with a new and enhanced reconstruction of a genome-scale metabolic network of S. clavuligerus, aimed to understand the carbon fluxes distribution during the different environmental conditions attained during the cultivations. The use of constrained-based modeling under pseudo-steady state conditions (Flux Balance Analysis, FBA) and dynamic conditions (Dynamic Flux Balance Analysis, DFBA) allowed to explain the role of primary metabolism and revealed the dynamics of intracellular carbon fluxes distribution during CA biosynthesis. Furthermore, the in silico simulation of metabolic scenarios and experimental testing showed that fed-batch operation with glutamate supplementation is a favorable condition for increasing the CA production. Potential genetic engineering targets were identified and evaluated in silico, aiming to improve the CA titers in S. clavuligerus cultures. This is the first work considering the cultivation dynamics on experimental and in silico studies of S. clavuligerus metabolism.