Shikimic acid is one of several industrially interesting chiral starting materials formed in the aromatic amino acid pathway of plants and microorganisms. In this study, the physiology of a shikimic acid producing strain of Escherichia coli (derived from W3110) deleted in aroL (shikimic acid kinase II gene), was compared to that of a corresponding control strain (W3110) under carbon- and phosphate-limited conditions. For the shikimic acid producing strain (referred to as W3110.shik1), phosphate limitation resulted in a higher yield of shikimic acid (0.059 +/- 0.012 vs. 0.024 +/- 0.005 c-mol/c-mol) and a lower yield of by-products from the shikimate pathway, when compared to carbon-limited condition. The yield of the by-product 3-dehydroshikimic acid (DHS) decreased from 0.076 +/- 0.028 to 0.022 +/- 0.001 c-mol/c-mol. Several other by-products were only detected under carbon-limited conditions. The latter group included 3-dehydroquinic acid (0.021 +/- 0.021 c-mol/c-mol), quinic acid (0.012 +/- 0.005 c-mol/c-mol), and gallic acid (0.002 +/- 0.001 c-mol/c-mol). For both strains, more acetate was produced under phosphate than the carbon-limited case. Considerable cell lysis was found for both strains but was higher for W3110.shik1, and increased for both strains under phosphate limitation. The advantages of the latter condition in terms of an increased shikimic acid yield was thus counteracted by an increased cell lysis, which may make downstream processing more difficult.

High amounts of outer membrane (OM) components were released in glucose-limited fed-batch (GLFB) cultures at 37 degrees C at specific growth rates approaching 0.05 h(-1). Endotoxin analyses from a 20 degrees C GLFB culture gave similar results. An alternative fermentation technique, the temperature-limited fed-batch (TLFB) technique, reduced the endotoxin concentration in a culture with a biomass concentration of 30 g l(-1) from the 850 mg l(-1) in traditional GLFB cultures to about 20 mg l(-1). The TLFB technique uses the temperature to regulate the dissolved oxygen tension, while all substrate components are unregulated. It appears to be severe glucose limitation that triggers the extensive release of endotoxins rather than a low growth rate. Furthermore, it is not the low temperature that stabilizes the OM when using the TLFB technique. Simulations and experimental data show that this technique results in the same biomass productivity as the GLFB technique.

No abstract