On the improvement of bacterial growth on complementary substrates by partial segregration in the broth

Abstract

Previous studies have shown that when two species compete for the same substrate, a controlled degree of segregation in a bioreactor favours growth of the desired species. That concept has been extended to the growth of bacteria on a mixture of two substrates for one of which the organism has a natural preference. However, using the example of Escherichia coli cultivated in a medium containing glucose and fumarate, it has been shown that, while the physiological preference for glucose is maintained, incomplete macromixing promotes growth and enables better utilisation of both substrates. While the preferred substrate (glucose) is consumed ®rst in a perfectly mixed reactor, imperfect mixing allows both substrates to be utilised simultaneously, at different rates. By dividing the broth conceptually into an inoculated region and a non-inoculated region, it is seen that cell growth is favoured by an initial period of good mixing followed by poor mixing in the non-inoculated region.

Optimal mixing to improve the performance of batch and continuous fermentations with recombinant Escherichia coli

“…For D 2 / D 1 = 0.001, the average recombinant cell concentration peaked at 4.5 h, while for larger ratios the fermentation required longer to attain the final (best) state; at D 2 / D 1 = 1.0 the maximum was attained after 5.25 h. Variations in the average substrate concentration, average mass fraction of plasmid‐bearing cells and average plasmid loss probability (Figs 8,9 and 10) with the ratio D 2 / D 1 were also consistent with these observations. The superior performance seen here for D 1 ≠ D 2 , ie a departure from perfect mixing, agrees with similar observations reported earlier7–9 for other recombinant fermentations and strengthens the general validity of exploiting controlled imperfect mixing as a tool to increase the fermentation efficiency of large‐scale bioreactors.…”

“…Secondly, computations showed that a dilution rate of 10 h −1 was large enough to provide a good approximation of perfect mixing; larger rates produced only marginal changes in the CFSTB performance. However, this value of 10 h −1 may differ from one fermentation system to another and indeed in some cases, nearly perfect mixing may be attained at smaller values of D 1 or D 2 7–9…”

“…The determination and maintenance of conditions that favor these requirements depends on the analysis of the fermentation through a good model that complements as well as acts as a precursor to experimental design and exploration. Many investigators1–10 have quantitatively analyzed different fermentation systems through models that have considered the frequency of generation of plasmid‐free daughter cells from plasmid‐carrying mother cells and the specific growth rate difference between plasmid‐bearing and plasmid‐free cells.…”