Establishing an Artificial Pathway for De Novo ... - ACS Publications

Research Article pubs.acs.org/synthbio

Establishing an Artificial Pathway for De Novo Biosynthesis of Vanillyl Alcohol in Escherichia coli Zhenya Chen,†,‡ Xiaolin Shen,†,‡ Jian Wang,§ Jia Wang,†,‡ Ruihua Zhang,§ Justin Forrest Rey,§ Qipeng Yuan,*,†,‡ and Yajun Yan*,§ †

State Key Laboratory of Chemical Resource Engineering, Beijing University of Chemical Technology, Beijing 100029, China Beijing Advanced Innovation Center for Soft Matter Science and Engineering, Beijing University of Chemical Technology, Beijing 100029, China § College of Engineering, The University of Georgia, Athens, Georgia 30602, United States ‡

S Supporting Information *

ABSTRACT: Vanillyl alcohol is a phenolic alcohol and is used as a flavoring agent in foods and beverages. In this paper, we propose a novel artificial pathway for microbial production of vanillyl alcohol from simple carbon sources. The pathway extends from 4-hydroxybenzoic acid (4-HBA), and needs only three heterologous enzymes, p-hydroxybenzoate hydroxylase (PobA), carboxylic acid reductase (CAR) and caffeate O-methyltransferase (COMT). First, we examined the promiscuous activity of COMT toward 3,4-dihydroxybenzyl alcohol and found a kcat value of 0.097 s−1. Meanwhile, 499.36 mg/L vanillyl alcohol was produced by COMT in vivo catalysis when fed with 1000 mg/L 3,4-dihydroxybenzyl alcohol. In the following experiment, de novo biosynthesis of vanillyl alcohol was carried out and 240.69 mg/L vanillyl alcohol was produced via modular optimization of pathway genes. This work was to date the first achievement for microbial production of vanillyl alcohol. Additionally, the present study demonstrates the application of enzyme promiscuity of COMT in the design of an artificial pathway for the production of high-value methylated aromatic compounds. KEYWORDS: vanillyl alcohol, aromatic compounds, shikimate pathway, caffeate O-methyltransferase, enzyme promiscuity, microbial synthesis

V

conditions, and low yields. By addressing these limitations, microbial-based biosynthesis can be an appealing alternative approach for vanillyl alcohol production. Microbial-based metabolic engineering is a powerful biotechnological platform, and has been considered as an eco-friendly approach for production of many high-value compounds from simple carbon sources,9 such as amino acids,10,11 flavonoids,12−14 fatty acids,15 phenylpropanoic acids,16 terpenoids,17,18 coumarins,19,20 monolignols,21 isoprenes,22 alkanes,23 and other biofuels.24,25 So far, only the biosynthesis of vanillin, which is a precursor of vanillyl alcohol and also known as an active component of Gastrodia elata

anillyl alcohol (4-hydroxy-3-methoxybenzyl alcohol), a widely used flavoring agent, is a natural phenolic compound, existing in several diverse plants, such as Gastrodia elata Blume1−3 and Vanilla planifolia.4 Vanillyl alcohol displays a variety of biological activities. For instance, the alcohol exhibits 65% antioxidant activity by β-carotene-linoleate assay and 90% by DPPH assay,4 significant antiangiogenic activity in the chick chorioallantoic membrane (CAM), anti-inflammatory activity and antinociceptive activity in mice,5 inhibition of cell growth in food spoilage yeasts,6 and antiasthmatic activity in guinea pig.7 In addition, vanillyl alcohol possesses anticonvulsive and free radical scavenging activities in ferric chlorideinduced epileptic seizures in Sprague−Dawley rats.8 So far, the main approach for vanillyl alcohol production is via direct extraction from various plants. However, these approaches are limited by the supply of raw materials, harsh reaction © 2017 American Chemical Society

Received: April 21, 2017 Published: June 6, 2017 1784

DOI: 10.1021/acssynbio.7b00129 ACS Synth. Biol. 2017, 6, 1784−1792

Research Article

ACS Synthetic Biology

Figure 1. Novel biosynthetic pathway of vanillyl alcohol production. Black-colored arrows indicate the native pathways in E. coli; blue-colored arrow indicates the heterologous steps; PEP, phosphoenolpyruvate; E4P, D-erythrose 4-phosphate; DAHP, 3-deoxy-D-arabinoheptulosonate 7-phosphate; 4-HBA, 4-hydroxybenzoic acid; PpsA, phosphoenolpyruvate synthetase; TktA, transketolase; AroG, 2-dehydro-3-deoxyphosphoheptonate aldolase; AroL, shikimate kinase II; UbiC, chorismate lyase; PobA, p-hydroxybenzoate hydroxylase; CAR, carboxylic acid reductase; Sfp, the CAR maturation factor phosphopantetheinyl transferase; ADHs, alcohol dehydrogenases; COMT, caffeate O-methyltransferase.

were introduced into E. coli. Combined with two endogenous enzymes, chorismate lyase (UbiC) and alcohol dehydrogenase (ADH), E. coli achieves the vanillyl alcohol de novo biosynthesis. COMT, with caffeic acid and caffeyl alcohol as native substrates, was tested due to substrate similarity to catalyze 3,4dihydroxybenzyl alcohol for vanillyl alcohol production. Remarkably, the results of the in vitro enzyme assay and in vivo whole-cell bioconversion experiment of COMT indicated this attempt achieved the desired effect. COMT (Km = 0.52 ± 0.04 mM, kcat = 0.097 ± 0.002 s−1, with 3,4-dihydroxybenzyl alcohol as substrate) could produce 499.36 ± 43.75 mg/L vanillyl alcohol in vivo when fed with 1000 mg/L 3,4dihydroxybenzyl alcohol. Hence, COMT was used to substitute the catechol O-methyltransferase, which has low activity.26,29 On the basis of that, expression of the above-mentioned nonnatural pathway (Figure 1) in E. coli enabled generation of 66.94 ± 9.14 mg/L vanillyl alcohol from simple carbon sources. Further, modular optimization of pathway genes enhanced vanillyl alcohol production to 240.69 ± 22.20 mg/L, which is the highest titer of microbial-based vanillyl alcohol achieved so far.

Blume, has been achieved in different microbes. Hansen et al. designed an artificial pathway for vanillin production from glucose in Schizosaccharomyces pombe, which was extended from 3-dehydroshikimate (3-DHS), an intermediate in the shikimate pathway. To convert 3-DHS into vanillin, three heterologous enzymes were introduced into the host, 3dehydroshikimate dehydratase from Podospora pauciseta, carboxylic acid reductase (CAR) from Nocardia genus, and catechol O-methyltransferase from Homo sapiens. Unfortunately, the vanillin titer was only 65 mg/L.26 Subsequently, in order to improve the vanillin production and decrease its toxicity, Brochado et al. employed an additional glycosyltransferase into the vanillin-producing S. cerevisiae strain, resulting in 500 mg/L vanillin β-D-glucoside production in batch cultivation via in silico metabolic engineering strategy.27 Notably, using this biosynthetic pathway only accumulated trace amounts of vanillin. Recently, an alternative route was constructed by mimicking a natural pathway of plants to achieve vanillin production from different carbon sources in E. coli. In this route, five enzymes, tyrosine ammonia-lyase (TAL), 4-coumarate 3-hydroxylase (C3H), caffeate O-methyltransferase (COMT), trans-feruloyl-CoA synthetase (FCS), and enoyl-CoA hydratase/aldolase (ECH), were arranged accordingly to convert L-tyrosine to vanillin. Grafting this artificial pathway into a tyrosine-overproducing strain only enabled the host to produce 97.2 mg/L vanillin from L-tyrosine, 19.3 mg/L from glucose, 13.3 mg/L from xylose and 24.7 mg/L from glycerol.28 Although the biosynthesis of vanillin was achieved, the titer of vanillin was low due to the long pathway, low catalytic activities of pathway enzymes, and the instability of vanillin. To overcome these issues and achieve vanillyl alcohol production from simple carbon sources, a novel biosynthetic pathway was assembled in the present study (Figure 1). Notably, this artificial pathway extends from 4-hydroxybenzoic acid (4-HBA), an endogenous compound in E. coli generated from the shikimate pathway. Four heterologous proteins, phydroxybenzoate hydroxylase (PobA), CAR, the CAR maturation factor phosphopantetheinyl transferase (Sfp) and COMT,

■

RESULTS AND DISCUSSION

Design of a Novel Biosynthetic Pathway for Vanillyl Alcohol Production. Vanillyl alcohol is a widely occurring aromatic metabolite, and the main approach for its production is extraction from a diverse set of plants. In consideration of the limitation of the plant resources and harsh reaction conditions of the extraction processes, biosynthetic vanillyl alcohol is essential to substitute for the naturally occurring vanillyl alcohol. Therefore, we proposed an artificial pathway for de novo production of vanillyl alcohol (Figure 1). In this pathway, 4-HBA is catalyzed by PobA from Pseudomonas aeruginosa, CAR from Mycobacterium marinum, endogenous ADHs and COMT from Arabidopsis thaliana sequentially in order to produce the end-product, vanillyl alcohol. 4-HBA is an endogenous compound in E. coli generated from the shikimate pathway. PobA, p-hydroxybenzoate hydroxylase, catalyzes the first step of hydroxylating 4-HBA into 3,4-dihydroxybenzoic 1785


Research Article


Figure 2. Catalytic reactions and SDS-PAGE of COMT. (A) The catalytic reactions of COMT toward different substrates. (B) SDS-PAGE of COMT. The black-colored arrow directs the band of COMT and lane M demonstrates the protein molecular weight marker.

acid. CAR (carboxylic acid reductase), when coupled with its activator Sfp (CAR maturation factor phosphopantetheinyl transferase), demonstrates catalytic versatility toward benzoic acids and fatty acids for generation of corresponding aldehydes. 30−32 Thus, we inferred CAR would permit conversion of 3,4-dihydroxybenzoic acid into 3,4-dihydroxybenzaldehyde, which was not stable and would be further converted into 3,4-dihydroxybenzyl alcohol by endogenous ADHs. The final step is methylation of 3,4-dihydroxybenzyl alcohol into vanillyl alcohol, which might be uncertain for lack of efficient methyltransferase. Given the low activity of the characterized catechol O-methyltransferase from Homo sapiens,26,29 COMT from A. thaliana was tested here to convert 3,4dihydroxybenzyl alcohol into vanillyl alcohol because of its strong activity in methylating caffeyl alcohol into coniferyl alcohol21 and substrate similarity between caffeyl alcohol and 3,4-dihydroxybenzyl alcohol (Figure 2A). Assembly of the above heterologous genes in E. coli would enable the establishment of a novel artificial pathway for the biosynthesis of vanillyl alcohol from renewable carbon sources. Enzymatic Activity Assay of COMT. To examine our assumption that COMT is able to catalyze 3,4-dihydroxybenzyl alcohol into vanillyl alcohol, the specific activity and kinetic parameters of purified COMT toward 3,4-dihydroxybenzyl alcohol were determined. The plasmid pET-COMT was transferred into E. coli BL21 Star (DE3), resulting in strain CZY10, used for expressing COMT with an N-terminal multihistidine tag. After expression, COMT was purified to homogeneity, as verified by SDS-PAGE analysis (Figure 2B). Two phenolic compounds, 3,4-dihydroxybenzyl alcohol and 3,4-dihydroxybenzoic acid, similar to its native substrate caffeyl alcohol and caffeic acid, respectively, were used as substrates to test the capability of COMT at adding the methyl group to the 3-hydroxyl group. As the results shown in Table 1, when using

3,4-dihydroxybenzyl alcohol as the substrate, COMT has a specific activity of 0.140 ± 0.003 μmol/min/mg protein, which is 4-fold higher than that of 3,4-dihydroxybenzoic acid (0.031 ± 0.001 μmol/min/mg protein). HPLC analysis of the reaction product confirmed the generation of vanillyl alcohol, suggesting that COMT could indeed add a methyl group to the 3-hydroxyl group of 3,4-dihydroxybenzyl alcohol. Unexpectedly, HPLC analysis of the reaction product from 3,4-dihydroxybenzoic acid confirmed the production of only isovanillic acid, indicating COMT methylated at the 4-hydroxyl group of 3,4-dihydroxybenzoic acid (Figure 2A). The capability of methylation was consistent with COMT toward native substrate, caffeic acid.33,34 Differently, the isoferulic acid, formed by methylation at the 4-hydroxyl group of caffeic acid, only occupied less than 5% of the methylated products.33,34 These results suggested that COMT preferred 3,4-dihydroxybenzyl alcohol over 3,4dihydroxybenzoic acid as the substrate, and the catalytic mechanisms toward the two substrates might be different. To further investigate the kinetic properties of COMT, kinetic parameters toward 3,4-dihydroxybenzoic acid and 3,4dihydroxybenzyl alcohol were measured. As shown in Table 2 Table 2. Kinetic Parameters of COMT toward 3,4Dihydroxybenzoic Acid and 3,4-Dihydroxybenzyl Alcohola substrate 3,4-Dihydroxybenzoic acid 3,4-Dihydroxybenzyl alcohol a

a

substrate

specific activity (μmol/min/mg protein) 0.031 ± 0.001 0.140 ± 0.003

kcat (s−1)

kcat/Km (mM−1 s−1)

1.73 ± 0.09

0.021 ± 0.0007

0.012

0.52 ± 0.04

0.097 ± 0.002

0.19

Two independent experiments were conducted to generate the data.

and Figure S1, COMT has a Km of 1.73 ± 0.09 mM and a kcat of 0.021 ± 0.0007 s−1 toward 3,4-dihydroxybenzoic acid, with isovanillic acid as the only product. Notably, COMT has a 3fold lower Km value toward 3,4-dihydroxybenzyl alcohol (0.52 ± 0.04 mM) (Figure S2), when compared with the Km value toward 3,4-dihydroxybenzoic acid, with vanillyl alcohol as the product, indicating that COMT possesses about 3-fold higher substrate affinity toward 3,4-dihydroxybenzyl alcohol. The kcat of COMT toward 3,4-dihydroxybenzyl alcohol was measured as 0.097 ± 0.002 s−1, a 5-fold higher value compared with that of

Table 1. Specific Activities of COMT toward 3,4Dihydroxybenzoic Acid and 3,4-Dihydroxybenzyl Alcohola 3,4-Dihydroxybenzoic acid 3,4-Dihydroxybenzyl alcohol

Km (mM)

Two independent experiments were conducted to generate the data. 1786


Research Article


Figure 3. Production of vanillyl alcohol from different substrates and initial in vivo activities of COMT toward different substrates. The substrates (a final concentration of 1000 mg/L) were supplemented to the cell cultures at 5.5 h. For (A) and (B), strain CZY11 was used, and 3,4dihydroxybenzoic acid and 3,4-dihydroxybenzyl alcohol were fed into medium, respectively. (C) The initial in vivo activities of COMT toward 3,4dihydroxybenzoic acid and 3,4-dihydroxybenzyl alcohol. For (D), strain CZY12 was used and 3,4-dihydroxybenzoic acid was fed into medium. For (E), strain CZY13 was used and 3,4-dihydroxybenzoic acid was fed into medium. For (F), strain CZY14 was used and 3,4-dihydroxybenzoic acid was fed into medium. Three independent experiments were conducted to generate the data.

raised with the increase of cell density in the first 24 h (Figure 3A). During this period, the isovanillic acid titer increased rapidly between 12 and 24 h, and reached 157.56 ± 0.54 mg/L at 24 h with an OD600 value of 8.21 ± 0.10. Meanwhile, only 2.31 ± 0.04 mg/L vanillic acid was accumulated. During the next 12 h, we observed the decrease of isovanillic acid and vanillic acid production and the OD600 decreased slightly to 7.81 ± 0.08. The results suggest that COMT is able to convert 3,4-dihydroxybenzoic acid into isovanillic acid in vivo, consistent with in vitro assay results. In addition, even though vanillic acid was not detected with in vitro enzyme assay, COMT can convert a trace amount of 3,4-dihydroxybenzoic acid into vanillic acid. As a comparison, to achieve the vanillyl alcohol biosynthesis in strain CZY11, 1000 mg/L 3,4-dihydroxybenzyl alcohol was fed into the medium at 5.5 h. As Figure 3B shows, the titer of vanillyl alcohol increased stably between 5.5 and 12 h with the trend of cell growth. Although the cells stopped growing, the vanillyl alcohol titer raised to 499.36 ± 43.75 mg/L in the subsequent 12 h. A trace amount of isovanillyl alcohol (13.67 ± 2.04 mg/L) was also detected. Compared with 157.56 ± 0.54 mg/L isovanillic acid production by CZY11 when fed with 1000 mg/L 3,4-dihydroxybenzoic acid, a 3-fold higher vanillyl alcohol titer was achieved when feeding the similar amount of 3,4-dihydroxybenzyl alcohol into cultures. The titer of vanillyl alcohol and isovanillyl alcohol decreased when the cultivation extended to 36 h. Similar trends were also observed in production of isovanillic acid and vanillic acid (Figure 3A), probably due to the automatic or enzymatic degradation of products. Additionally, the feeding experiments indicated COMT has the capability of methylating 3,4-dihydroxybenzyl alcohol into vanillyl alcohol in vivo as it does in vitro. Specifically, even we were not able to detect the isovanillyl

COMT toward 3,4-dihydroxybenzoic acid. This comparison illustrated that the activity of COMT toward 3,4-dihydroxybenzyl alcohol was 5-fold higher than COMT toward 3,4dihydroxybenzoic acid. The kcat/Km of COMT toward 3,4dihydroxybenzyl alcohol (0.19 mM−1 s−1) was 16-fold higher than toward 3,4-dihydroxybenzoic acid (0.012 mM−1 s−1). Overall, COMT demonstrated higher specificity and catalytic activity toward 3,4-dihydroxybenzyl alcohol than toward 3,4dihydroxybenzoic acid. Bioconversion of 3,4-Dihydroxybenzoic Acid or 3,4Dihydroxybenzyl Alcohol into Corresponding Products. To investigate the applicability of COMT for microbial production of vanillyl alcohol, whole-cell bioconversion experiments were carried out to test its in vivo catalytic efficiency. First, we incubated BW25113 (F′) with various concentrations of vanillyl alcohol (0, 1, 3, and 5 g/L) to test its toxicity. As the results shown in Figure S3, when cultivating BW25113 (F′) in the 1 g/L vanillyl alcohol medium, the OD600 value had no significant difference compared with the value of the strain cultivated in 0 g/L vanillyl alcohol medium (as control), indicating less than 1 g/L vanillyl alcohol had a negligible effect on the cell growth. Meanwhile, when treating the strains with 3 g/L or 5 g/L of vanillyl alcohol, the OD600 values slightly decreased to 5.84 ± 0.19 and 4.97 ± 0.04, respectively, compared with the value of control (6.19 ± 0.15) after 24 h cultivation. Interestingly, the OD600 values increased to comparable values to the control from 24 to 48 h. These results suggested that more than 1 g/L vanillyl alcohol inhibited the cell growth at the beginning of incubation, and as cultivation continued, this inhibition effect can be relieved. Afterward, strain CZY11, containing plasmid pZE-COMT, was used for conducting these feeding experiments. When fed with 1000 mg/L 3,4-dihydroxybenzoic acid, the isovanillic acid titer 1787


Research Article


Figure 4. Plasmid constructs for de novo production of vanillyl alcohol, including plasmids pZE-CUP, pCS-CS, pZE-CUP-APTA, pZE-C-U, pSAPobA and pZE-CU-APTA.

alcohol with in vitro enzyme assay; but, COMT can convert a small amount of 3,4-dihydroxybenzyl alcohol to isovanillyl alcohol in vivo. Furthermore, we explored the initial in vivo activities of COMT toward these two substrates. As shown in Figure 3C, when feeding 3,4-dihydroxybenzoic acid, the initial in vivo activity of COMT was 3.30 ± 0.80 μM/h/OD. This value was calculated based on the isovanillic acid titer at 9 h in the conversion experiment, since the product was not observed at 6.5 h. As a comparison, the initial in vivo activity of COMT, when feeding 3,4-dihydroxybenzyl alcohol, was 11.56 ± 2.18 μM/h/OD, approximately 3.5 fold higher. In order to test the efficiency of CAR, we carried out the conversion experiment with strain CZY12, containing plasmid pCS-CS. As the results shown in Figure 3D, strain CZY12 completely consumed 1000 mg/L 3,4-dihydroxybenzoic acid generating 821.43 ± 6.79 mg/L 3,4-dihydroxybenzyl alcohol in 24 h and the cell density gradually increased throughout the cultivation, reaching 6.19 ± 0.2 at 36 h. These results suggest that CAR and its activator Sfp, coupled with endogenous ADHs, can effectively convert 3,4-dihydroxybenzoic acid into 3,4-dihydroxybenzyl alcohol in vivo. To further enhance the conversion efficiency of CZY12, we employed the use of another strain, CZY13, with the expression of an alcohol dehydrogenase (ADH6), to conduct the feeding experiments. As a result, 812.08 ± 51.48 mg/L 3,4-dihydroxybenzyl alcohol was produced from 1000 mg/L 3,4-dihydroxybenzoic acid in 24 h (Figure 3E), which was similar to strain CZY12. Cells grew continuously and the OD600 was 5.52 ± 0.16 at 36 h, lower than that of strain CZY12. These results indicate that overexpression of alcohol dehydrogenase did not improve conversion efficiency and instead caused growth stress to the host. Most notably, endogenous ADHs were sufficient to reduce generated aldehydes to alcohols, as our previous study reported.35 Upon the basis of the above conclusions that CAR and endogenous ADHs could efficiently convert 3,4-dihydroxybenzoic acid into 3,4-dihydroxybenzyl alcohol and COMT could efficiently convert 3,4-dihydroxybenzyl alcohol into vanillyl alcohol in vivo, we tested the efficiency of the downstream pathway of the vanillyl alcohol de novo biosynthesis. To achieve

this goal, the whole pathway (Figure 1) was split at 3,4dihydroxybenzoic acid into both upstream and downstream pathways. The downstream pathway plasmids pZE-COMT and pCS-CS were cotransferred into E. coli BW25113 (F′), generating strain CZY14. Conversion experiments were carried out to examine the capability of the downstream pathway. When 1000 mg/L 3,4-dihydroxybenzoic acid was fed to the cultures at 5.5 h, both cell growth and vanillyl alcohol titer had an increased trend during the first 24 h (Figure 3F). In the next 12 h, the trend of cell growth was opposite to that of the vanillyl alcohol titer. The titer of vanillyl alcohol increased to 210.17 ± 19.44 mg/L at 36 h, while the OD600 decreased to 6.45 ± 0.15 and the 3,4-dihydroxybenzyl alcohol titer also decreased mainly because of the conversion of 3,4-dihydroxybenzyl alcohol to vanillyl alcohol by COMT. Compared with CZY11, CZY14 with COMT, CAR and Sfp coexpression produced 2.4-fold lower vanillyl alcohol concentrations. Moreover, we observed 820.05 ± 18.04 mg/L 3,4-dihydroxybenzyl alcohol production at 12 h and 4.38 ± 0.99 mg/L isovanillyl alcohol accumulation at 36 h, indicating that CAR and endogenous ADHs can efficiently convert 3,4-dihydroxybenzoic acid into 3,4-dihydroxybenzyl alcohol, resulting in only trace amounts of byproduct isovanillyl alcohol production. Overall, the results of the downstream pathway bioconversion experiment demonstrated that COMT, CAR and Sfp, accompanied by endogenous ADHs, can effectively catalyze 3,4-dihydroxybenzoic acid into vanillyl alcohol, proving that they have potential to be used for vanillyl alcohol de novo biosynthesis. De Novo Production of Vanillyl Alcohol. We combined the upstream and downstream pathways to achieve vanillyl alcohol de novo production (Figure 1). To enhance chorismate conversion into 4-HBA, chorismate lyase (UbiC) with high catalytic activity32,36 was overexpressed. For the purpose of achieving vanillyl alcohol biosynthesis from simple carbon sources, plasmid pZE-CUP was created (Figure 4) and cotransferred with plasmid pCS-CS into E. coli BW25113 (F′), generating strain CZY15. Shake flask fermentation with CZY15 allowed production of vanillyl alcohol. As shown in 1788


Research Article


Figure 5. Microbial production of vanillyl alcohol. For (A), strain CZY15 was used. For (B), strain CZY16 was used. For (C), strain CZY17 was used. For (D), strain CZY18 was used. Three independent experiments were conducted to generate the data.

the titer of the intermediate, 3,4-dihydroxybenzyl alcohol, reached 282.53 ± 5.06 mg/L at the same time point. However, with further overexpression of aroL, ppsA, tktA and aroGfbr, CZY18 could only produce 83.88 ± 5.77 mg/L vanillyl alcohol, while the titer of 3,4-dihydroxybenzyl alcohol did not have an obvious improvement (Figure 5D). This was most likely due to the stress of imbalanced gene expression in the host. These results suggest that modular optimization of COMT, UbiC and PobA was an efficient approach and contributed to the titer enhancement of vanillyl alcohol.

Figure 5A, vanillyl alcohol was produced with a stable increase during the whole cultivation process. The cell density raised in the first 36 h, synchronizing with the increase of vanillyl alcohol titer. At the end of the fermentation, the vanillyl alcohol titer reached 66.94 ± 9.14 mg/L with an OD600 of 7.83 ± 0.16, with a trace amount of isovanillyl alcohol produced. Likewise, 149.65 ± 14.42 mg/L 3,4-dihydroxybenzyl alcohol was accumulated. To increase the precursor supply, we overexpressed aroL, ppsA, tktA and aroGfbr in order to boost the availability of 3,4dihydroxybenzyl alcohol. Strain CZY16 containing plasmids pZE-CUP-APTA and pCS-CS (Figure 4) was used for shake flask fermentation to demonstrate this purpose. As shown in Figure 5B, within 36 h, the 3,4-dihydroxybenzyl alcohol titer was enhanced to 314.97 ± 4.45 mg/L and the vanillyl alcohol titer increased to 118.08 ± 19.60 mg/L as well, which was 1.8fold higher than that of CZY15. Overall, introducing this artificial pathway into E. coli enabled the production of vanillyl alcohol from simple carbon sources, and the titer was enhanced by redirecting more carbon flux into the vanillyl alcohol pathway. Enhancement of Vanillyl Alcohol Production via Modular Optimization. Strain CZY15 containing plasmids pZE-CUP and pCS-CS produced limited vanillyl alcohol, possibly because the COMT, ubiC and pobA were not optimally expressed in E. coli. To modulate expression levels of both enzymes, we put the COMT and ubiC on the high-copy number plasmid under two separated operons. Additionally, in light of the high activity of PobA, pobA was fixed on the lowcopy number plasmid to balance the whole metabolic pathway. Thus, the plasmids pZE-C-U and pSA-PobA were constructed and introduced into E. coli BW25113 (F′) with pCS-CS, resulting in strain CZY17. Shake flask fermentation with strain CZY17 allowed production of vanillyl alcohol with a maximum titer of 240.69 ± 22.20 mg/L at 36 h (Figure 5C), which represents a 3.6-fold increase when compared to strain CZY15. Additionally, after modular optimization, cells grew faster and

■

CONCLUSION In this study, we constructed a novel pathway for de novo biosynthesis of vanillyl alcohol. This artificial pathway only needs three heterologous enzymes: PobA, CAR and COMT. Compared with the native pathway for production of the precursor (vanillin) in plants,28 our designed pathway for vanillyl alcohol production was more efficient due to less reaction steps and high activity of pathway enzymes. First, we adopted the concept of enzyme promiscuity to validate the activity of COMT toward 3,4-dihydroxybenzyl alcohol via in vitro enzyme assay. As a result, COMT has a kcat value of 0.097 ± 0.002 s−1 toward 3,4-dihydroxybenzyl alcohol. Meanwhile, 499.36 mg/L vanillyl alcohol was produced by COMT in vivo catalysis when fed with 1000 mg/L 3,4-dihydroxybenzyl alcohol. Our following feeding experiment was conducted to examine the conversion efficiency of CAR toward 3,4dihydroxybenzoic acid and observed 821.43 ± 6.79 mg/L 3,4-dihydroxybenzyl alcohol was produced from 1000 mg/L 3,4-dihydroxybenzoic acid. Afterward, we grafted the artificial pathway into E. coli and observed 66.94 ± 9.14 mg/L vanillyl alcohol was produced from simple carbon sources. In addition, we employed a simplified method of modular optimization to balance the whole metabolic pathway enzyme expression. After using this approach, the vanillyl alcohol titer was enhanced to 240.69 ± 1789


Research Article

ACS Synthetic Biology 22.20 mg/L. In conclusion, we established a novel biosynthetic pathway to achieve vanillyl alcohol production and validated the promiscuity of COMT. However, the activity of COMT limited the production of vanillyl alcohol. Recently, protein engineering for target enzyme modification has been known as an efficient and economical approach for improving the activity of rate-limiting enzymes in the metabolic engineering field.37−41 To address the issue of COMT activity, future work may focus on improving the activity via a protein engineering approach.

Table 3. Plasmids and Strains Used in This Study

METHODS Media, Strains and Plasmids. Luria−Bertani (LB) medium containing 10 g/L NaCl, 10 g/L tryptone and 5 g/L yeast extract, was used for cell inoculation, propagation and protein expression. Modified M9 (M9Y) medium containing 10 g glycerol, 2.5 g glucose, 6 g Na2HPO4, 0.5 g NaCl, 3 g KH2PO4, 1 g NH4Cl, 2 mmol MgSO4, 0.1 mmol CaCl2 and 5 g yeast extract per liter was used for feeding experiments and de novo production of vanillyl alcohol. When needed, ampicillin, kanamycin and chloramphenicol were added to the medium to the final concentration of 100, 50, and 34 μg/mL, respectively. E. coli XL1-Blue was used for plasmid construction and propagation, while E. coli BL21 Star (DE3) was used for COMT expression and purification. E. coli BW25113 (F′) was used for feeding experiments and de novo biosynthesis of vanillyl alcohol. Plasmids pZE12-luc, pCS27 and pSA74 which are high-, medium-, and low-copy number plasmids, respectively, were used for pathway construction. Plasmid pETDuet-1 was used for COMT expression and purification. The details of the strains and plasmids, used in this study, were included in Table 3. DNA Manipulation. Plasmids pCS-APTA and pZE-COMT were constructed in our previous studies.19,21 In order to measure the in vitro activity of COMT, pET-COMT was constructed by inserting gene COMT, amplified by PCR from pZE-COMT, to pETDuet-1 using BamHI and HindIII. Plasmids pZE-COMT, pCS-CS and pZE-ADH6 were used for feeding experiments. To create plasmid pCS-CS, car encoding carboxylic acid reductase and sfp encoding CAR maturation factor phosphopantetheinyl transferase were amplified from Mycobacterium marinum and Bacillus subtilis, respectively and cloned into pCS27 using KpnI, NdeI and BamHI. Gene encoding ADH6 was amplified from Saccharomyces cerevisiae genome DNA and cloned into pZE-luc using KpnI and XbaI, generating plasmids pZE-ADH6. To achieve de novo production of vanillyl alcohol, other plasmids pSA-PobA, pZE-CUP, pZE-CU, pZE-C-U, pZE-CUP-APTA and pZE-CUAPTA were created. To construct pSA-PobA, gene pobA, encoding p-hydroxybenzoate hydroxylase, was amplified from Pseudomonas aeruginosa genome and cloned into pSA74 using KpnI and HindIII. Genes COMT amplified from pZE-COMT, ubiC amplified from E. coli BL21 Star (DE3) genome, and pobA amplified from Pseudomonas aeruginosa genome were cloned into pZE-luc using KpnI, PstI, SphI and XbaI, resulting in pZECUP. Genes COMT and ubiC were cloned into pZE-luc using KpnI, PstI and XbaI, generating pZE-CU. The expressing cassette PLlacO1-APTA was amplified from pCS-APTA and inserted into pZE-CUP and pZE-CU between SpeI and SacI, yielding plasmids pZE-CUP-APTA and pZE-CU-APTA. Plasmid pZE-C-U was constructed by inserting the expressing cassette PLlacO1-UbiC to pZE-COMT using SpeI and SacI. In Vitro COMT Enzyme Assay. E. coli BL21 Star (DE3) carrying pET-COMT (CZY10) was preinoculated in 3 mL LB

pZE-COMT pZE-ADH6

plasmids and strains Plasmids pETDuet-1 pZE12-luc pCS27 pSA74 pET-COMT

■

pZE-CUP

pZE-CU pZE-C-U pCS-APTA pZE-CUP-APTA pZE-CU-APTA pCS-CS pSA-PobA Strains XL1-Blue ZΔM15 Tn10(Tet r)] Stratagene BL21Star (DE3) BW25113 (F′)

CZY10 CZY11 CZY12 CZY13 CZY14 CZY15 CZY16 CZY17 CZY18

description

source

pT7, PBR322 ori, Ampr PLlacO1, colE ori, luc, Ampr PLlacO1, P15A ori, Kanr PLlacO1, pSC101 ori, Cmr pETDuet-1 containing COMT from Arabidopsis thaliana pZE12-luc containing COMT pZE12-luc containing ADH6 from Saccharomyces cerevisiae pZE12-luc containing ubiC from E. coli, COMT from A. thaliana, and pobA from Pseudomonas aeruginosa pZE12-luc containing ubiC and COMT pZE12-luc containing ubiC and COMT, two operons pCS27 containing aroL, ppsA, tktA and aroGfbr from E. coli pZE12-luc containing PLlacO1-CUP and PLlacO1-APTA pZE12-luc containing PLlacO1-CU and PLlacO1-APTA pCS27 containing car from Mycobacterium marinum and sfp from Bacillus subtilis pSA74 containing pobA recA1 endA1 gyrA96 thi-1 hsdR17 supE44 relA1 lac [F′ proAB lacIqZΔM15 Tn10 (Tetr)] q ZΔM15 Tn10(Tet r)] F− ompT hsdSB (rB−mB−) gal dcm (DE3) rrnBT14 ΔlacZWJ16 hsdR514 ΔaraBADAH33 ΔrhaBADLD78 F′ [traD36 proAB lacIqZΔM15 Tn10(Tetr)] BL21Star (DE3) with pET-COMT BW25113 (F′) with pZE-COMT BW25113 (F′) with pCS-CS BW25113 (F′) with pCS-CS and pZE-ADH6 BW25113 (F′) with pZE-COMT and pCS-CS BW25113 (F′) with pZE-CUP and pCS-CS BW25113 (F′) with pZE-CUP-APTA and pCS-CS BW25113 (F′) with pZE-C-U, pCS-CS and pSA-PobA BW25113 (F′) with pZE-CU-APTA, pCS-CS and pSA-PobA

Novagen 19 19 19

This study 21

This study This study

This study This study 19

This study This study This study This study Stratagene

Invitrogen 43

This This This This This This This

study study study study study study study

This study This study

medium containing ampicillin and grown aerobically at 37 °C, respectively. After 12 h, 500 μL of preinoculum was transferred into 50 mL of fresh LB containing ampicillin and cultured until OD600 reached around 0.6 at 37 °C, and then induced overnight with 0.5 mM IPTG at 30 °C. Cells were then harvested and lysed by a beads beater. The recombinant proteins with an N-terminal multihistidine tag were purified using His-Spin protein miniprep kit (ZYMO RESEARCH).42 Pierce BCA Protein Assay Kit (Thermo Scientific) was used for measuring the protein concentrations. The COMT assays were carried out by mimicking catechol O-methyltransferase activity assay, described by Kunjapur et al.29 A 1 mL reaction system contained 100 mM Tris-HCl (pH 8.0), 5 mM MgCl2, 2 mM Sadenosyl-L-methionine tosylate, 0.5 μM purified COMT and 0−2000 μM substrate (3,4-dihydroxybenzoic acid or 3,4dihydroxybenzyl alcohol). The reactions were conducted at 30 °C for 30 min (for 3,4-dihydroxybenzoic acid) and 15 min (for 3,4-dihydroxybenzyl alcohol) and terminated by adding 10 μL 100% HCl. The reaction rates of COMT were calculated 1790


Research Article


2 min and 5% solvent B for an additional 5 min. 3,4Dihydroxybenzoic acid, 3,4-dihydroxybenzyl alcohol, vanillic acid, vanillyl alcohol, isovanillic acid and isovanillyl alcohol were quantified based on their peak areas at specific wavelengths (260 nm for 3,4-dihydroxybenzoic acid, vanillic acid and isovanillic acid, 280 nm for 3,4-dihydroxybenzyl alcohol, vanillyl alcohol and isovanillyl alcohol).

according to the product formation, which were measured by HPLC. The kinetic parameters were estimated with OriginPro8.5 through nonlinear regression of the Michaelis−Menten equation. Toxicity Test. Single colonies of E. coli BW25113 (F′) were preinoculated into 3 mL of LB medium and cultured overnight at 37 °C. 200 μL overnight cultures were inoculated into 20 mL M9Y medium containing 0, 1, 3 and 5 g/L vanillyl alcohol, respectively. The cultures were cultivated at 37 °C for 48 h. Samples were collected at 0, 2, 4, 6, 8, 10, 12, 24, 36 and 48 h and the cell growth was confirmed by measuring OD600. Feeding Experiments. In order to conduct in vivo conversion experiments, E. coli BW25113 (F′)43 was transformed with plasmid pZE-COMT, generating strain CZY11, E. coli BW25113 (F′) was transformed with plasmid pCS-CS, generating strain CZY12, E. coli BW25113 (F′) was cotransformed with plasmids pCS-CS and pZE-ADH6, generating strain CZY13, and E. coli BW25113 (F′) was cotransformed with plasmids pZE-COMT and pCS-CS, generating strain CZY14. Single colonies were preinoculated into 3 mL of LB medium containing ampicillin and cultured overnight at 37 °C. 200 μL overnight cultures were inoculated into 20 mL M9Y medium containing ampicillin. The cultures were cultivated at 37 °C until OD600 reached 0.6 and then induced with IPTG (a final concentration of 0.5 mM) at 30 °C. After 3 h induction, for strain CZY11, 1000 mg/L 3,4dihydroxybenzoic acid or 3,4-dihydroxybenzyl alcohol was fed into the cultures. For strains CZY12, CZY13 and CZY14, 3,4dihydroxybenzoic acid was fed into the cultures at a final concentration of 1000 mg/L. Samples were collected at the time when substrates were added (5.5 h), 9, 12, 24, and 36 h; cell growth was confirmed by measuring OD600 and the products and intermediates were analyzed by HPLC. Additional samples were taken at 6.5 h to calculate the initial in vivo activity of COMT. De Novo Production of Vanillyl Alcohol. E. coli BW25113 (F′) containing plasmids pZE-CUP and pCS-CS (CZY15), E. coli BW25113 (F′) containing plasmids pZECUP-APTA and pCS-CS (CZY16), E. coli BW25113 (F′) containing plasmids pZE-C-U, pCS-CS and pSA-PobA (CZY17), and E. coli BW25113 (F′) containing plasmids pZE-CU-APTA, pCS-CS and pSA-PobA (CZY18) were used for de novo biosynthesis of vanillyl alcohol. Transformants were preinoculated in 3 mL LB overnight and then 200 μL samples were inoculated into 20 mL M9Y medium containing suitable antibiotics and 0.5 mM IPTG. The cultures were cultivated at 30 °C and samples were collected every 12 h until 48 h. OD600 values were measured and the concentrations of the products were analyzed by HPLC. HPLC Analysis. 3,4-Dihydroxybenzoic acid and vanillic acid were purchased from Alfa Aesar. Isovanillic acid and isovanillyl alcohol were purchased from Sigma-Aldrich. 3,4-Dihydroxybenzyl alcohol and vanillyl alcohol were purchased from VWR and TCI AMERICA, respectively. These six compounds all have over 95% purities and were used as standards. HPLC (Dionex Ultimate 3000), equipped with a reverse phase ZORBAX SB-C18 column and an Ultimate 3000 Photodiode Array Detector, was used for analysis and quantification of standards and samples. The column temperature was set to 28 °C. Flowing phase contains solvent A (water with 0.1% formic acid) and solvent B (100% methanol) with a flow rate of 1 mL/ min. The following gradients were used: 5% to 50% solvent B for 20 min, 100% solvent B for 2 min, 100% to 5% solvent B for

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acssynbio.7b00129.

■

Figures S1−S3 (PDF)

AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected]. Phone: +86-10-64437610. *E-mail: [email protected]. Phone: +1-706-542-8293. ORCID

Yajun Yan: 0000-0002-9993-3016 Author Contributions

Z.C. conceived this study, designed and conducted the experiments. X.S., J(ian).W., J(ia).W. and R.Z. participated in the research. Z.C. analyzed the data and wrote the manuscript. Q.Y. and Y.Y. directed the research. J.R. and Y.Y. revised the manuscript. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was supported by the National Natural Science Foundation of China (21406010, 21606012 and 21636001), the Programme of Introducing Talents of Discipline to Universities (“111” project, B13005), the Program for Changjiang Scholars and Innovative Research Team in Universities in China (No. IRT13045), the Academic Leader of Beijing Polytechnic (DTR201601), and the Key Project of Beijing Polytechnic (YZK028). We also acknowledge the College of Engineering, The University of Georgia, Athens and the International Joint Graduate Training Program of Beijing University of Chemical Technology.

■

REFERENCES

(1) Zhao, Y., Cao, Q., Xiang, Y., and Hu, Z. (1999) Identification and determination of active components in Gastrodia elata Bl. by capillary electrophoresis. J. Chromatogr. A 849, 277−283. (2) Liu, C., Liu, M., and Zhu, P. (2002) Determination of gastrodin, p-hydroxybenzyl alcohol, vanillyl alcohol, p-hydroxylbenzaldehyde and vanillin in tall gastrodia tuber by high-performance liquid chromatography. Chromatographia 55, 317−320. (3) Ong, E. S., Heng, M. Y., Tan, S. N., Yong, H., Wan, J., Koh, H., Teo, C. C., and Hew, C. S. (2007) Determination of gastrodin and vanillyl alcohol in Gastrodia elata Blume by pressurized liquid extraction at room temperature. J. Sep. Sci. 30, 2130−2137. (4) Shyamala, B., Naidu, M. M., Sulochanamma, G., and Srinivas, P. (2007) Studies on the antioxidant activities of natural vanilla extract and its constituent compounds through in vitro models. J. Agric. Food Chem. 55, 7738−7743. (5) Jung, H., Song, Y., Lim, C., and Park, E. (2008) Anti-angiogenic, anti-inflammatory and anti-nociceptive activities of vanillyl alcohol. Arch. Pharmacal Res. 31, 1275−1279.

1791


Research Article

ACS Synthetic Biology (6) Fitzgerald, D. J., Stratford, M., and Narbad, A. (2003) Analysis of the inhibition of food spoilage yeasts by vanillin. Int. J. Food Microbiol. 86, 113−122. (7) Jang, Y. W., Lee, J. Y., and Kim, C. J. (2010) Anti-asthmatic activity of phenolic compounds from the roots of Gastrodia elata Bl. Int. Immunopharmacol. 10, 147−154. (8) Hsieh, C., Chang, C., Chiang, S., Li, T., Tang, N., Pon, C., Hsieh, C., and Lin, J. (2000) Anticonvulsive and free radical scavenging activities of vanillyl alcohol in ferric chloride-induced epileptic seizures in Sprague-Dawley rats. Life Sci. 67, 1185−1195. (9) Sun, X., Shen, X., Jain, R., Lin, Y., Wang, J., Sun, J., Wang, J., Yan, Y., and Yuan, Q. (2015) Synthesis of chemicals by metabolic engineering of microbes. Chem. Soc. Rev. 44, 3760−3785. (10) Becker, J., Zelder, O., Häfner, S., Schröder, H., and Wittmann, C. (2011) From zero to hero-Design-based systems metabolic engineering of Corynebacterium glutamicum for L-lysine production. Metab. Eng. 13, 159−168. (11) Zhang, K., Li, H., Cho, K. M., and Liao, J. C. (2010) Expanding metabolism for total biosynthesis of the nonnatural amino acid Lhomoalanine. Proc. Natl. Acad. Sci. U. S. A. 107, 6234−6239. (12) Leonard, E., Yan, Y., and Koffas, M. A. (2006) Functional expression of a P450 flavonoid hydroxylase for the biosynthesis of plant-specific hydroxylated flavonols in Escherichia coli. Metab. Eng. 8, 172−181. (13) Santos, C. N. S., Koffas, M., and Stephanopoulos, G. (2011) Optimization of a heterologous pathway for the production of flavonoids from glucose. Metab. Eng. 13, 392−400. (14) Wang, J., Guleria, S., Koffas, M. A., and Yan, Y. (2016) Microbial production of value-added nutraceuticals. Curr. Opin. Biotechnol. 37, 97−104. (15) Xu, P., Gu, Q., Wang, W., Wong, L., Bower, A. G., Collins, C. H., and Koffas, M. A. (2013) Modular optimization of multi-gene pathways for fatty acids production in E. coli. Nat. Commun. 4, 1409. (16) Sun, J., Lin, Y., Shen, X., Jain, R., Sun, X., Yuan, Q., and Yan, Y. (2016) Aerobic biosynthesis of hydrocinnamic acids in Escherichia coli with a strictly oxygen-sensitive enoate reductase. Metab. Eng. 35, 75− 82. (17) Leonard, E., Ajikumar, P. K., Thayer, K., Xiao, W. H., Mo, J. D., Tidor, B., Stephanopoulos, G., and Prather, K. L. (2010) Combining metabolic and protein engineering of a terpenoid biosynthetic pathway for overproduction and selectivity control. Proc. Natl. Acad. Sci. U. S. A. 107, 13654−13659. (18) Paddon, C. J., Westfall, P. J., Pitera, D., Benjamin, K., Fisher, K., McPhee, D., Leavell, M., Tai, A., Main, A., and Eng, D. (2013) Highlevel semi-synthetic production of the potent antimalarial artemisinin. Nature 496, 528−532. (19) Lin, Y., Shen, X., Yuan, Q., and Yan, Y. (2013) Microbial biosynthesis of the anticoagulant precursor 4-hydroxycoumarin. Nat. Commun., DOI: 10.1038/ncomms3603. (20) Lin, Y., Sun, X., Yuan, Q., and Yan, Y. (2013) Combinatorial biosynthesis of plant-specific coumarins in bacteria. Metab. Eng. 18, 69−77. (21) Chen, Z., Sun, X., Li, Y., Yan, Y., and Yuan, Q. (2017) Metabolic engineering of Escherichia coli for microbial synthesis of monolignols. Metab. Eng. 39, 102−109. (22) Alper, H., Miyaoku, K., and Stephanopoulos, G. (2005) Construction of lycopene-overproducing E. coli strains by combining systematic and combinatorial gene knockout targets. Nat. Biotechnol. 23, 612−616. (23) Choi, Y. J., and Lee, S. Y. (2013) Microbial production of shortchain alkanes. Nature 502, 571−574. (24) Yim, H., Haselbeck, R., Niu, W., Pujol-Baxley, C., Burgard, A., Boldt, J., Khandurina, J., Trawick, J. D., Osterhout, R. E., and Stephen, R. (2011) Metabolic engineering of Escherichia coli for direct production of 1, 4-butanediol. Nat. Chem. Biol. 7, 445−452. (25) Zhang, K., Sawaya, M. R., Eisenberg, D. S., and Liao, J. C. (2008) Expanding metabolism for biosynthesis of nonnatural alcohols. Proc. Natl. Acad. Sci. U. S. A. 105, 20653−20658.

(26) Hansen, E. H., Møller, B. L., Kock, G. R., Bünner, C. M., Kristensen, C., Jensen, O. R., Okkels, F. T., Olsen, C. E., Motawia, M. S., and Hansen, J. (2009) De novo biosynthesis of vanillin in fission yeast (Schizosaccharomyces pombe) and baker’s yeast (Saccharomyces cerevisiae). Appl. Environ. Microbiol. 75, 2765−2774. (27) Brochado, A. R., Matos, C., Møller, B. L., Hansen, J., Mortensen, U. H., and Patil, K. R. (2010) Improved vanillin production in baker’s yeast through in silico design. Microb. Cell Fact. 9, 84. (28) Ni, J., Tao, F., Du, H., and Xu, P. (2015) Mimicking a natural pathway for de novo biosynthesis: natural vanillin production from accessible carbon sources. Sci. Rep., DOI: 10.1038/srep13670. (29) Kunjapur, A. M., Hyun, J. C., and Prather, K. L. (2016) Deregulation of S-adenosylmethionine biosynthesis and regeneration improves methylation in the E. coli de novo vanillin biosynthesis pathway. Microb. Cell Fact. 15, 61. (30) Akhtar, M. K., Turner, N. J., and Jones, P. R. (2013) Carboxylic acid reductase is a versatile enzyme for the conversion of fatty acids into fuels and chemical commodities. Proc. Natl. Acad. Sci. U. S. A. 110, 87−92. (31) Kallio, P., Pásztor, A., Thiel, K., Akhtar, M. K., and Jones, P. R. (2014) An engineered pathway for the biosynthesis of renewable propane. Nat. Commun. 5, 4731. (32) Bai, Y., Yin, H., Bi, H., Zhuang, Y., Liu, T., and Ma, Y. (2016) De novo biosynthesis of Gastrodin in Escherichia coli. Metab. Eng. 35, 138− 147. (33) Finkle, B. J., and Nelson, R. F. (1963) Enzyme reactions with phenolic compounds: a meta-O-methyltransferase in plants. Biochim. Biophys. Acta 78, 747−749. (34) Edwards, R., and Dixon, R. A. (1991) ) Purification and characterization of S-adenosyl-L-methionine: caffeic acid 3-O-methyltransferase from suspension cultures of alfalfa (Medicago sativa L.). Arch. Biochem. Biophys. 287, 372−379. (35) Wang, J., Shen, X., Jain, R., Wang, J., Yuan, Q., and Yan, Y. (2017) Establishing A Novel Biosynthetic Pathway for the Production of 3, 4-Dihydroxybutyric Acid from Xylose in Escherichia coli. Metab. Eng. 41, 39. (36) Pugh, S., McKenna, R., Osman, M., Thompson, B., and Nielsen, D. R. (2014) Rational engineering of a novel pathway for producing the aromatic compounds p-hydroxybenzoate, protocatechuate, and catechol in Escherichia coli. Process Biochem. 49, 1843−1850. (37) Bommareddy, R. R., Chen, Z., Rappert, S., and Zeng, A.-P. (2014) A de novo NADPH generation pathway for improving lysine production of Corynebacterium glutamicum by rational design of the coenzyme specificity of glyceraldehyde 3-phosphate dehydrogenase. Metab. Eng. 25, 30−37. (38) Chen, W., Zhang, S., Jiang, P., Yao, J., He, Y., Chen, L., Gui, X., Dong, Z., and Tang, S. (2015) Design of an ectoine-responsive AraC mutant and its application in metabolic engineering of ectoine biosynthesis. Metab. Eng. 30, 149−155. (39) Chen, Z., and Zeng, A. (2016) Protein engineering approaches to chemical biotechnology. Curr. Opin. Biotechnol. 42, 198−205. (40) Wang, X., Wang, G., Li, X., Fu, J., Chen, T., Wang, Z., and Zhao, X. (2016) Directed evolution of adenylosuccinate synthetase from Bacillus subtilis and its application in metabolic engineering. J. Biotechnol. 231, 115−121. (41) Zhang, L., Liang, Y., Wu, W., Tan, X., and Lu, X. (2016) Microbial synthesis of propane by engineering valine pathway and aldehyde-deformylating oxygenase. Biotechnol. Biofuels 9, 80. (42) Lin, Y., Sun, X., Yuan, Q., and Yan, Y. (2014) Extending shikimate pathway for the production of muconic acid and its precursor salicylic acid in Escherichia coli. Metab. Eng. 23, 62−69. (43) Atsumi, S., Hanai, T., and Liao, J. C. (2008) Non-fermentative pathways for synthesis of branched-chain higher alcohols as biofuels. Nature 451, 86−89.

1792
