No abstract

Curcuminoids are well-known for their therapeutic properties. However, their extraction from natural sources is environmentally unfriendly, expensive and limited by seasonal variability, highlighting the need for alternative production processes. We propose an optimized artificial biosynthetic pathway to produce curcuminoids, including curcumin, in Escherichia coli. This pathway involves six enzymes, tyrosine ammonia lyase (TAL), 4-coumarate 3-hydroxylase (C3H), caffeic acid O-methyltransferase (COMT), 4-coumarate-CoA ligase (4CL), diketide-CoA synthase (DCS), and curcumin synthase (CURS1). Curcuminoids pathway was divided in two modules, the first module included TAL, C3H and COMT and the second one 4CL, DCS and CURS1. Optimizing the first module of the pathway, from tyrosine to ferulic acid, enabled obtaining the highest ferulic acid titer reported so far (1325.1 µM). Afterward, ferulic acid was used as substrate to optimize the second module of the pathway. We achieved the highest concentration of curcumin ever reported (1529.5 µM), corresponding to a 59.4% increase. Subsequently, curcumin and other curcuminoids were produced from tyrosine (using the whole pathway) in mono-culture. The production increased comparing to a previously reported pathway that used a caffeoyl-CoA O-methyltransferase enzyme (to convert caffeoyl-CoA to feruloyl-CoA) instead of COMT (to convert caffeic to ferulic acid). Additionally, the potential of a co-culture approach was evaluated to further improve curcuminoids production by reducing cells metabolic burden. We used one E. coli strain able to convert tyrosine to ferulic acid and another able to convert the hydroxycinnamic acids produced by the first one to curcuminoids. The co-culture strategies tested led to 6.6 times increase of total curcuminoids (125.8 µM) when compared to the mono-culture system. The curcuminoids production achieved in this study corresponds to a 6817% improvement. In addition, by using an inoculation ratio of 2:1, although total curcuminoids production decreased, curcumin production was enhanced and reached 43.2 µM, corresponding to an improvement of 160% comparing to mono-culture system. To our knowledge, these values correspond to the highest titers of curcuminoids obtained to date. These results demonstrate the enormous potential of modular co-culture engineering to produce curcumin, and other curcuminoids, from tyrosine.

Curcuminoids are phenylpropanoids with high pharmaceutical potential. Herein, we report an engineered artificial pathway in Escherichia coli to produce natural curcuminoids through caffeic acid. Arabidopsis thaliana 4-coumaroyl-CoA ligase (4CL1) and Curcuma longa diketide-CoA synthase (DCS) and curcumin synthase (CURS1) were used to produce curcuminoids and 70 mg/L of curcumin was obtained from ferulic acid. Bisdemethoxycurcumin and demethoxycurcumin were also produced, but in lower concentrations, by feeding p-coumaric acid or a mixture of p-coumaric acid and ferulic acid, respectively. Additionally, curcuminoids were produced from tyrosine through the caffeic acid pathway. To produce caffeic acid, tyrosine ammonia lyase (TAL) from Rhodotorula glutinis and 4-coumarate 3-hydroxylase (C3H) from Saccharothrix espanaensis were used. Caffeoyl-CoA 3-O-methyltransferase (CCoAOMT) from Medicago sativa was used to convert caffeoyl-CoA to feruloyl-CoA.Using caffeic acid, p-coumaric acid or tyrosine as a substrate, 3.9 mg/L, 0.3 mg/L and 0.2 mg/L of curcumin were produced, respectively. This is the first time DCS and CURS1 were used in vivo to produce curcuminoids and that curcumin was produced by feeding tyrosine. We have shown that curcumin can be produced using a pathway through caffeic acid. This alternative pathway represents a step forward in the heterologous production of curcumin using E. coli.