Abstract
BACKGROUND: The phenolic compound tyrosol is widely used in the pharmaceutical industry, owing to its beneficial effects on human health and its use as a precursor for key pharmaceuticals, including β(1)-receptor blockers. Tyrosol can be found in olive oil, but despite its natural biosynthesis in plants, low extraction efficiencies render microbial production a more viable alternative. RESULTS: Here, we engineered the L-tyrosine overproducing Corynebacterium glutamicum strain AROM3 for the de novo production of tyrosol. Two routes were established and compared: one via 4-OH-phenylpyruvate as intermediate and the other via tyramine. We initially expected the first route to require heterologous expression of a prephenate dehydrogenase gene, given that C. glutamicum lacks this enzymatic function. However, heterologous expression of ARO10 from Saccharomyces cerevisiae (ARO10(Sc)), which encodes a phenylpyruvate decarboxylase, was sufficient to establish tyrosol production in strain AROM3. We identified that 4-OH-phenylpyruvate is synthesized from L-tyrosine by native aminotransferases, which is subsequently decarboxylated by Aro10(Sc), and reduced to tyrosol by native alcohol dehydrogenases, leading to a titer of 9.4 ± 1.1 mM (1.30 ± 0.15 g/L). We identified the furfural dehydrogenase FudC as major enzyme involved in this pathway, as its gene deletion reduced tyrosol production by 75%. Given the instability of 4-OH-phenylpyruvate, the synthesis of tyrosol via the stable intermediate tyramine was pursued via the second route. Decarboxylation of L-tyrosine followed by oxidative deamination was accomplished by overexpression of the L-tyrosine decarboxylase gene tdc from Levilactobacillus brevis (tdc(Lb)) and the tyramine oxidase gene tyo from Kocuria rhizophila (tyo(Kr)). Using this route, tyrosol production was increased by 44% compared to the route via 4-OH-phenylpyruvate. With a division of labor approach by co-cultivating L-tyrosine producing strains that either express tdc(Lb) or tyo(Kr), the highest titer of 14.1 ± 0.3 mM (1.95 ± 0.04 g/L) was achieved. CONCLUSIONS: This study demonstrates the potential of endotoxin-free C. glutamicum as production host for the L-tyrosine-derived product tyrosol. Due to its L-arogenate pathway for L-tyrosine synthesis, the unstable 4-OH-phenylpyruvate could be excluded as intermediate in the Tdc-Tyo pathway, outcompeting the most often utilized production route via phenylpyruvate decarboxylases.