Yeast Cell Factory-Platform for the Screening and the Industrial Production of Flavonoids and other Phenolic Compounds: PhD thesis by Beata Joanna Lehka

Flavonoids are secondary plant metabolites derived from the phenylpropanoid pathway. These bioactive compounds are of great commercial interest due to their varied properties, such as anti-oxidative, anti-tumor and/or antibacterial. However, industrial production of flavonoids based on purification from plants or on their organic synthesis can be problematic. On the other hand, the so called bio-industrial production of a variety of compounds in microorganisms is gaining great popularity. The objective of this study is to produce naringenin and some of its derivatives in Saccharomyces cerevisiae as a model for industrial production of flavonoids.
By combining a balanced heterologous expression of (phenylpropanoid) naringenin biosynthetic pathway genes and the optimisation of yeast metabolism we developed a strain producing 430 mg/L of naringenin from glucose. In this set up naringenin was produced from phenylalanine by action of a phenylalanine ammonia lyase (PAL) in a strain overproducing phenylalanine and tyrosine. As tyrosine can also be a precursor for naringenin production via tyrosine ammonia lyase (TAL), we evaluated in vivo the activity of several TAL enzymes in S. cerevisiae. Most of the ammonia lyases have affinity to both phenylalanine and tyrosine, with a strong preference for phenylalanine. Some TALs have been described before but these either have very low overall activity on tyrosine or are also capable of using phenylalanine as substrate. Here we identified a novel TAL, an enzyme from Aeromonas salmonicida with relatively high activity towards tyrosine and no activity towards phenylalanine.
Production of flavonoids and stilbenoids in S. cerevisiae is challenging, partially due to carbon loss towards phloretic acid by the action of an unknown endogenous reductase. Through the screening of 26 putative double bond reductase knockout strains we identified Tsc13 as the enzyme responsible for the reduction of p-coumaroyl-CoA to phloretic acid. Since Tsc13 is an essential enzyme, a combination of homology modelling and site-saturation mutagenesis was used to identify residues that would impair its activity towards p-coumaroyl-CoA without compromising cell growth. Several mutations leading to lower phloretic acid production were found and discussed. As an alternative approach we complemented the deletion of the TSC13 gene with homologues from various plants. Genes from Arabidopsis thaliana, Gossypium hirsutum and Malus domestica were able to sustain cell progression while not resulting in p-coumaroyl-CoA reduction to phloretic acid. These findings should be of great value for the industrial production of p-coumaric acid-derived molecules in S. cerevisiae.
As naringenin is a precursor for other commercially relevant flavonoids we developed a platform for the production of a library of flavonoid derivatives and screened them for antibacterial properties. Seven different yeast strains producing flavonoids (naringenin, kaempferol, dihydrokaempferol, apigenin and afzelechin), a stilbenoid (resveratrol) and a dihydrochalcone (phloretin) were used as a base for the introduction of a library of decorating enzymes known to be active towards some of these compounds. Cariogenic bacteria were selected as a model for the screening of the combinatorial library for antibacterial potential. Although no compounds with significant antibacterial properties were identified, the method for the construction of flavonoid libraries that is presented here could serve as a base for future screenings.