Metabolic analysis of shikimic acid producing Escherichia coli

Abstract: Shikimic acid production using recombinant E.coli has been developed during the past 15 years due to the increased demand of this product, primarily as a starting material for the anti-influensa medicine Tamiflu?. A detailed metabolic analysis of a shikimic acid producing E.coli strain was made in the current work in order to increase the understanding of shikimic acid production and identify means of improving the process. The strain used (called W3110.shik1) was modified for shikimate production by deletion of aroL, and by overexpression of aroFFBR (on a plasmid). Physiological characterizations were made using techniques such as metabolite analysis, stoichiometric network analysis, transcriptome analysis, and in situ 2D-fluorometry at different modes of cultivation, and compered to a wild-type strain (W3110). A high yield of shikimic acid requires both a high flux into the pathway and minimum by-product formation. In carbon-limited chemostat cultivation the total flux into the pathway was almost 2 times higher than the one obtained in phosphorus-limited chemostat cultivation (yields on glucose: ~0.2 vs. 0.1 c-mole/c-mole), but at the expense of increased formation of DHS, DHQ and quinic acid. Typical yields carbon- vs. phosphorus-limitation: shikimate: 0.02 vs. 0.06, DHS 0.03-0.12 (increased with increased growth rate) vs. 0.03, DHQ: 0.02 vs. 0.01, quinic acid: 0.02 vs. 0.01. The batch and fed-batch conditions studied showed similar characteristics as their chemostat dittos. A theory to explain the mechanisms of by-product formation, based on intracellular equilibration, is discussed and compared to the hydroaromatic equilibration theory. Gene expression patterns of the two growth conditions showed differences in agreement with observed changes in total flux and by-product formation. Under carbon-limitation an upregulation of aroF, aroG, talA and talB agreed with the larger total flux into the pathway, and an upregulation of the genes ydiB, ydiN and aroD may explain the increased by-product formation. A slight starvation of the aromatic amino acids (identified by an induction of the trp and tyr-operons) under carbon-limitation, which was relieved under phosphorus limitation, was observed. The relief could possibly be explained by a simultaneous increased expression of aroK and aroA, thus enabling a higher flux downstream of shikimate. Metabolic flux analysis showed that the flux into the pathway was mostly dependent on the relationship between consumption and production of erythrose-4-phosphate. In addition, a higher flux connected to overflow metabolism was observed in the central metabolism under phosphorus-limitation. These fluxes were generally lower under carbon-limitation. 2D-fluorometry was used to monitor strain stability in chemostat cultivation. It was shown that strain stability was not maintained for more than about 25 generations, even in selective medium. This method provided additional information about metabolic changes coupled to adaptation mechanisms. For example a change in fluorescence probably coupled to pyridoxal-5-phosphate dependent proteins and FAD-related compounds/proteins was observed.