Biosynthesis of fraxetin from three different substrates using engineered Escherichia coli

Fraxetin, which is a simple coumarin, is a phytochemical present in medicinal plants, such as Fraxinus rhynchophylla, and Cortex Fraxini. In plants, it serves as a controller of iron homeostasis. The health-enhancing activities of fraxetin, such as anticancer, neuroprotective and antibacterial activities, are known. Scopoletin 8-hydroxylase (S8H) is a key enzyme involved in the synthesis of fraxetin from scopoletin. Scopoletin can be synthesized either from esculetin by O-methylation or from ferulic acid by feruloyl CoA 6′-hydroxylase (F6′H) and 4-coumaric acid CoA ligase (4CL). To enable fraxetin synthesis, the fraxetin biosynthesis pathway was introduced into Escherichia coli. Three distinct routes, from ferulic acid, esculetin, and scopoletin, were designed for the synthesis of fraxetin. In the first approach, E. coli strain harboring S8H was used and found to synthesize 84.8 μM fraxetin from 100 μM scopoletin. Two E. coli strains were used for the other two approaches because these approaches required at least two enzymatic reactions. Through this approach, 41.4 μM fraxetin was synthesized from 100 μM esculetin, while 33.3 μM fraxetin was synthesized from 100 μM ferulic acid.

Fraxetin belongs to the family of simple coumarins and is synthesized from scopoletin by the hydroxylation of its carbon at position 8. Fraxetin is involved in iron metabolism in plants [4,13]. Similar to other phytochemicals, fraxetin was found to exert beneficial effects in humans. These included antitumor [10,14], neuroprotective [15], antihyperglycemic [16], and anti-inflammatory [17] effects.
Since the metabolic pathway responsible for the synthesis of simple coumarins is well established, these compounds have been synthesized in E. coli. Scopoletin, esculetin, umbelliferone, skimming (umbelliferone 7-O-glucoside), and herniarin (7-O-methyl umbelliferone) were synthesized in E. coli [18,19]. For the fraxetin synthesis process in E. coli, two routes were postulated (Fig. 1). The first route involved the synthesis of esculetin from glucose, followed by the conversion of esculetin into fraxetin by 7-O-methylation and 8-hydroxylation. The second route started with the synthesis of scopoletin from ferulic acid, followed by 8-hydroxylation. In the present study, one coumarin, namely fraxetin, was synthesized using E. coli via these two routes.

Production and analysis of metabolites
For the synthesis of fraxetin from scopoletin, an overnight culture of an E. coli transformant containing pG-OsS8H, pG-AtS8H, pC-AtS8H, pE-AtS8H, or pR-AtS8H was inoculated into fresh LB medium containing 50 μg/mL ampicillin and grown at 37 °C until the OD 600 reached 0.8; following this, isopropyl β-D-1thiogalactopyranoside (IPTG) was added to the medium at a final concentration of 0.1 mM or 1 mM and incubated at 18 °C for 16 h. Cells were harvested and the cell concentration was adjusted to an OD 600 of 3.0. The cells were resuspended in M9 medium containing 2% glucose, ampicillin (50 μg/mL), and either 0.  in a test tube. A total concentration of 100 μM of the substrate (esculetin, isoscopoletin, scopoletin, or scoparone) was added, and the resulting culture was incubated at 30 °C for 24 h. An E. coli transformant containing the pG-AtS8H construct was employed to determine the substrate (scopoletin) concentration. The cell concentration was adjusted to an OD 600 of 3.0. The substrate was added to the appropriate M9 medium at 0.1, 0.2, 0.3, or 0.5 mM. The reaction culture was incubated at 30 °C for 24 h.
The E. coli transformant harboring ROMT9 was used to methylate esculetin to scopoletin, isoscopoletin, and scopoletin. Three reaction products were purified using thin layer chromatography (silica gel 60 F254, Millipore). A mixture of benzene and ethyl acetate (3:1) was used as a solvent. The E. coli transformant harboring POMT9 was used to synthesize scopoletin from esculetin. The methylation reaction using E. coli was carried out as described by Kim et al. [20].
E. coli harboring AtS8H or OsS8H was tested for the conversion of esculetin, isoscopoletin, scopoletin, and scoparone by the administration of each compound. E. coli harboring OsS8H did not convert any coumarins used. However, for E. coli harboring AtS8H, scopoletin and isoscopoletin were converted into novel compounds that had retention times different from those of the corresponding substrates (Fig. 2). Other substrates , and esculetin (c); the reaction product was analyzed. d-f denote standard scopoletin, isoscopoletin, and esculetin, respectively. P1, reaction product from scopoletin; P2, reaction product from isoscopoletin; S1, scopletin; S2, isoscopoletin; S3, esculetin (esculetin and scoparone) did not generate any new product. The molecular mass of the products from scopoletin and isoscopoletin was 207.937 Da, which is the molecular mass obtained if hydroxylation occurs. S8H utilized one methylated esculetins (scopoletin and isoscopoletin) as a substrate and did not utilize dimethylated (scoparone) or unmethylated esculetin. Scopoletin was a better substrate than isoscopoletin; 84.8% of scopoletin was converted, as opposed to the conversion of only 55% isoscopoletin. To determine the structure of the biotransformation product from scopoletin, the reaction product was purified, and its structure was analyzed using proton NMR. The reaction product was determined to be fraxetin (see
A two-step reaction was conducted using two E. coli transformants to augment the final yield of fraxetin. The first reaction was carried out using E. coli harboring POMT7. Approximately 56.9 μM scopoletin was synthesized from 100 μM esculetin (Fig. 3a). Further incubation did not result in the conversion of more esculetin. Thereafter, the culture filtrate from the first reaction was combined with E. coli harboring AtS8H. Approximately 41.4 μM fraxetin was synthesized from 56.9 μM scopoletin (Fig. 3b), indicating that there was approximately 72.7% conversion from the synthesized fraxetin. E. coli harboring POMT7 was administered esculetin (S1), following which the reaction product was analyzed. P1 denotes the reaction product from scopoletin. b Synthesis of fraxetin from scopoletin using E. coli harboring AtS8H. The culture filtrate from E. coli harboring POMT7 was administered to the E. coli harboring AtS8H, and the reaction product was analyzed Fraxetin was successfully synthesized from esculetin by a two-step reaction. The final yield of fraxetin synthesized from esculetin was lower than that from scopoletin; moreover, the conversion rate of scopoletin into fraxetin in the two-step reaction was lower than that seen for the direct conversion. This could possibly be attributed to the metabolite(s) in the first step inhibiting the second reaction. It was attempted herein to synthesize fraxetin from esculetin using an E. coli transformant harboring both POMT7 and AtS8H. Only 3.4 μM fraxetin and 17.2 μM scopoletin were synthesized from 100 μM esculetin. Next, fraxetin was synthesized from ferulic acid. Three enzymatic reactions are required for this. Previously, scopoletin was successfully synthesized from ferulic acid using E. coli harboring IbF6′H2 and Os4CL. It was reasoned that introducing AtS8H into E. coli harboring IbF6′H2 and Os4CL could result in the synthesis of fraxetin from ferulic acid. Three genes (IbF6′H, Os4CL, and AtS8H) were introduced into E. coli, and the resulting transformant was administered ferulic acid. The E. coli transformant synthesized fraxetin from ferulic acid. To optimize fraxetin synthesis, several initial ferulic acid concentrations (100, 200, 300, and 500 μM) were tested. The synthesis of fraxetin was optimal at 100 μM of initial ferulic acid, and approximately 33.3 μM fraxetin was synthesized (Fig. 4). Unreacted ferulic acid and scopoletin were accumulated at the higher concentrations of ferulic acid.
Fraxetin was synthesized from three different substrates (scopoletin, esculetin, and ferulic acid). As shown in Fig. 1, more enzymes are required when fraxetin is synthesized from ferulic acid or esculetin than when it is synthesized from scopoletin. Consequently, the final yield of fraxetin was higher (84.8 μM) when it was synthesized from scopoletin (100 μM). Its yield was decreased when synthesis was carried out from esculetin (41.4 μM) or ferulic acid (33.3 μM).
An attempt was made to synthesize fraxetin from esculetin or ferulic acid. One E. coli transformant harboring both POMT7 and AtS8H synthesized a lower amount of fraxetin from esculetin than the other two E. coli transformants, each of which conducted one reaction. However, fraxetin was successfully synthesized from ferulic acid using one E. coli transformant harboring three genes (Os4CL, IbF6′H, and AtS8H). Esculetin may compete with scopoletin for AtS8H. In the E. coli transformant harboring both POMT7 and AtS8H, esculetin served as a substrate for POMT7 and an inhibitor of AtS8H. Therefore, following the synthesis of scopoletin by POMT7, AtS8H could not utilize scopoletin because it was inhibited by esculetin. Conversely, when two independent E. coli transformants were used, more scopoletin synthesized by the first E. coli transformant harboring POMT7 was present in the medium and was converted into fraxetin by the second E. coli transformant harboring AtS8H. When fraxetin was synthesized from ferulic acid, only scopoletin was synthesized; therefore, it was possible to synthesize fraxetin using E. coli harboring Os4CL, IbF6′H, and AtS8H.  Fig. 4 Synthesis of fraxetin from ferulic acid using E. coli harboring Os4CL, IbF6′H, and AtS8H. S1, ferulic acid; P1, scopoletin; P2, fraxetin