By using stoichiometric-based modeling, we identified a number of genetic modifications that improved the intracellular availability of malonyl-CoA and thus, indirectly, the production of flavanones from phenylpropanoic acid precursors in recombinant E.coli. While traditional metabolic engineering approaches were used to construct overexpressions and deletion, CRISPRi based on dCas9 was used for downregulating a number of gene targets. Such interventions resulted in production titers that exceeded 100 mg/lt. In a similar approach, we applied metabolic pathway balancing for the biosynthesis of flavan-3-ols, resulting in an E.coli recombinant strain producing flavan-3-ols at over 900 mg/lt final titers.
The two strains were tested for their cross compatibility in a co-culture system. Such a co-culture system would allow the conversion of phenylpropanoic acid precursors to flavan-3ols. Optimization involved four different parameters: initial inoculation cell ratios, induction points, temperatures of growth and two different carbon sources using an empirical scaled-Gaussian model. These experiments resulted in a maximum titer of 40.7±0.1 mg/L, a 65% increase over the highest titer measured prior to computational optimization.
In a final example, we will demonstrate the use of such a poly-culture approach in the case of de novo anthocyanin biosynthesis from glucose. A similar modelling approach was used in order to further optimize fermentation.