BackgroundHydroxycinnamoyl anthranilates, also known as avenanthramides (avns), are a group of phenolic alkaloids with anti-inflammatory, antioxidant, anti-itch, anti-irritant, and antiatherogenic activities. Some avenanthramides (avn A–H and avn K) are conjugates of hydroxycinnamic acids (HC), including p-coumaric acid, caffeic acid, and ferulic acid, and anthranilate derivatives, including anthranilate, 4-hydroxyanthranilate, and 5-hydroxyanthranilate. Avns are primarily found in oat grain, in which they were originally designated as phytoalexins. Knowledge of the avns biosynthesis pathway has now made it possible to synthesize avns through a genetic engineering strategy, which would help to further elucidate their properties and exploit their beneficial biological activities. The aim of the present study was to synthesize natural avns in Escherichia coli to serve as a valuable resource.ResultsWe synthesized nine avns in E. coli. We first synthesized avn D from glucose in E. coli harboring tyrosine ammonia lyase (TAL), 4-coumarate:coenzyme A ligase (4CL), anthranilate N-hydroxycinnamoyl/benzoyltransferase (HCBT), and anthranilate synthase (trpEG). A trpD deletion mutant was used to increase the amount of anthranilate in E. coli. After optimizing the incubation temperature and cell density, approximately 317.2 mg/L of avn D was synthesized. Avn E and avn F were then synthesized from avn D, using either E. coli harboring HpaBC and SOMT9 or E. coli harboring HapBC alone, respectively. Avn A and avn G were synthesized by feeding 5-hydroxyanthranilate or 4-hydroxyanthranilate to E. coli harboring TAL, 4CL, and HCBT. Avn B, avn C, avn H, and avn K were synthesized from avn A or avn G, using the same approach employed for the synthesis of avn E and avn F from avn D.ConclusionsUsing different HCs, nine avns were synthesized, three of which (avn D, avn E, and avn F) were synthesized from glucose in E. coli. These diverse avns provide a strategy to synthesize both natural and unnatural avns, setting a foundation for exploring the biological activities of diverse avns.Electronic supplementary materialThe online version of this article (10.1186/s12934-018-0896-9) contains supplementary material, which is available to authorized users.

Hydroxybenzoic acids (HBAs) such as 4-hydroxybenzoic acid (4-HBA) and 3,4-dihydroxybenzoic acid (DHB; protocatechuic acid) and its ester with methanol (methylparaben [MP]) are known to have various functional biological properties, including antibacterial, anticancer, antidiabetic, antiaging, antiviral, and anti-inflammatory activities. Since these compounds are widely used in cosmetic, food, and pharmaceutical industries, the use of renewable feedstocks for the production of HBAs is an area of growing interest. In this study, we used Escherichia coli to synthesize these three hydroxybenzoic acid derivatives (4-HBA, DHB, and MP). We overexpressed ubiC in E. coli to synthesize 4-HBA from chorismate, a substrate that is produced by the shikimate pathway in E. coli. For the synthesis of DHB, an additional gene (pobA) was introduced, while hbad and EHT1 were co-expressed to synthesize MP. To supply more chorismate, we introduced the shikimate gene module construct and selected the best construct for increased yields. Using this approach, 723.5 mg/L 4-HBA, 942.0 mg/L DHB, and 347.7 mg/L MP were synthesized. Our study showed that the shikimate gene module constructs can be applicable to increase the yields of HBA derivatives in HBA-tolerant microorganisms.

BackgroundHydroxycinnamic acids (HCAs) including cinnamic acid, p-coumaric acid, caffeic acid, and ferulic acid, are C6–C3 phenolic compounds that are synthesized via the phenylpropanoid pathway. HCAs serve as precursors for the synthesis of lignins, flavonoids, anthocyanins, stilbenes and other phenolic compounds. HCAs can also be conjugated with diverse compounds including quinic acid, hydroxyl acids, and amines. Hydroxycinnamoyl (HC) amine conjugates such as N-HC tyramines and N-HC phenethylamines have been considered as potential starting materials to develop antiviral and anticancer drugs.ResultsWe synthesized N-HC tyramines and N-HC phenethylamines using three different approaches in Escherichia coli. Five N-HC phenethylamines and eight N-HC tyramines were synthesized by feeding HCAs and phenethylamine or tyramine to E. coli harboring 4CL (encoding 4-coumarate CoA:ligase) and either SHT (encoding phenethylamine N-HC transferase) or THT (encoding tyramine N-HC transferase). Also, N-(p-coumaroyl) phenethylamine and N-(p-coumaroyl) tyramine were synthesized from p-coumaric acid using E. coli harboring an additional gene, PDC (encoding phenylalanine decarboxylase) or TDC (encoding tyrosine decarboxylase). Finally, we synthesized N-(p-coumaroyl) phenethylamine and N-(p-coumaroyl) tyramine from glucose by reconstructing the metabolic pathways for their synthesis in E. coli. Productivity was maximized by optimizing the cell concentration and incubation temperature.ConclusionsWe reconstructed the metabolic pathways for synthesis of N-HC tyramines and N-HC phenethylamines by expressing several genes including 4CL, TST or SHT, PDC or TDC, and TAL (encoding tyrosine ammonia lyase) and engineering the shikimate metabolic pathway to increase endogenous tyrosine concentration in E. coli. Approximately 101.9 mg/L N-(p-coumaroyl) phenethylamine and 495.4 mg/L N-(p-coumaroyl) tyramine were synthesized from p-coumaric acid. Furthermore, 152.5 mg/L N-(p-coumaroyl) phenethylamine and 94.7 mg/L N-(p-coumaroyl) tyramine were synthesized from glucose.