Cui Yang ,Xianzhong Chen ,*,Junzhuang Chang ,Lihua Zhang ,Wei Xu ,Wei Shen ,You Fan

1 Key Laboratory of Carbohydrate Chemistry&Biotechnology,Ministry of Education,Jiangnan University,Wuxi 214122,China

2 School of Biotechnology,Jiangnan University,Wuxi 214122,China

Keywords:Tyrosol Escherichia coli Phenylpyruvate decarboxylase Gene knockout Codon optimization

A B S T R A C T Tyrosol is a pharmacologically active phenolic compound widely used in the medicine and chemical industries.Traditional methods of plant extraction are complicated and chemical synthesis of tyrosolis not commercially viable.In this study,a recombinant Escherichia coli strain was constructed by overexpressing the phenylpyruvate decarboxylase ARO10 from Saccharomyces cerevisiae,which could produce tyrosol from glucose.Furthermore,genes encoding key enzymes from the competing phenylalanine and tyrosine synthesis pathways and the repression protein TyrR were eliminated,and the resulting engineered strain generated 3.57 mmol·L-1 tyrosol from glucose.More significantly,codon optimization of ARO10 increased expression and tyrosol titer.Using the novel engineered strain expressing codon-optimized AR10 in shake- flask culture,8.72 mmol·L-1 tyrosol was obtained after 48 h.Optimization of the induction conditions improved tyrosol production to 9.53 mmol·L-1(1316.3 mg·L-1).A higher titer of tyrosol was achieved by reconstruction of tyrosol synthetic pathway in E.coli.

1.Introduction

Tyrosol(2-(4-hydroxyphenyl)ethanol)is a pharmacologically active phenolic compound that is widely distributed in nature.Tyrosol and its derivatives,hydroxytyrosol and salidroside,have antioxidant activity and cardioprotective effects[1].Recent research found that tyrosol can protect streptozotoc in-induced diabetic rats through altered glycoprotein components;further,this study can be extrapolated to humans[2].Tyrosol also has taste-sharpening effects that play an important role in the taste of alcoholic beverages,particularly in sake[3]and wine[4].Arecent discovery suggests that differences in taste,astringency and bitterness,characteristic of tyrosol,might derive from the tyrosol level in sake[3].Therefore,tyrosol has recently gained considerable attention as a pharmacological chemical in industry and as a taste-sharpening compound in the taste and functionalities of alcohol beverages.

With the discovery of more and more health benefits offered by tyrosol,increasing effort has been expended to increase the yield of tyrosol.Because the cost of extracting tyrosol from nature sources is very high,tyrosol used for industrial purposes is often synthesized chemically[5].However,the chemical synthesis of tyrosol is plagued by purification problems and very low overall yield.Tyrosol occurs naturally in olive oil[6],wine[1],sake[7],and some planttissues.The tyrosolcontent of virgin olive oil ranges from 40 to 180 mg·(kg oil)-1[5].However,extracting tyrosol from olive oil is challenging because tyrosol colocalizes with a large number of other phenolic substances in olive oil[5,8].There are two pathways for tyrosolsynthesis in microbes and plants[8].One is the Ehrlich pathway,in which tyrosine is converted into tyrosol as follows:(1)transamination of tyrosine by aminotransferase to form 4-hydroxyphenylpyruvate(4HPP),(2)decarboxylation of 4HPP by pyruvate decarboxylase to form 4-hydroxyphenylacetaldehyde(4HPAAL),and(3)reduction of 4HPAAL by alcohol dehydrogenase(ADH)to form tyrosol(Fig.1)[8–10].Another pathway involves the conversion of L-tyrosine into tyramine by tyrosine decarboxylase and transformation of the resulting tyramine into tyrosol by the consecutive actions of tyramine oxidase and alcohol dehydrogenase(ADH)[9,11,12].Recently,microbial production of tyrosol has been explored using engineered Escherichia coli strains[8].Initial attempts at microbial production yielded a low tyrosol titer,so a metabolic engineering strategy is needed to increase tyrosol production to the level needed for use on an industrial scale.

In this study,we constructed a tyrosol-producing E.coli strain and used metabolic engineering to improve product titer.After codon usage and induction conditions were optimized,a yield of 9.53 mmol·L-1(1316.3 mg·L-1)was obtained.This lays the foundation for the industrialization of microbial tyrosol synthesis.

2.Materials and Methods

2.1.Genetic material,microbial hosts and plasmids

Fig.1.The engineered tyrosol pathway in E.coli.AroG,3-deoxy-D-arabino-heptulosonate-7-phosphate synthase;TyrR,autorepressor;DAHP,3-deoxy-arabino-heptulonate-7-phosphate;TyrA,chorismate mutase or prephenate dehydrogenase;PheA,chorismate mutase and prephenate dehydratase;4HPP,4-hydroxyphenylpyruvate;TyrB,aromatic amino acid transaminase;ARO10,phenylpyruvate decarboxylase;4HPAAL,4-hydroxyphenyacetaldehyde;FeaB,phenylacetaldehyde dehydrogenase;ADH,alcohol dehydrogenase.Dashed arrows indicate feedback inhibitions.Tandem dashed arrows indicate multiple enzyme reactions.Red cross bars denote the disruption of indicated genes.

The bacterial strains and plasmids used in this study are listed in Table 1.E.coli JM109 was used as the host for plasmid construction and E.coli MG1655 and its derivatives were used for tyrosol production.The plasmids pKK223-3-ARO10 and pKK223-3-ARO10*were used as expression plasmids.The plasmids pKD46,pKD13 and pCP20(Table 1),which are involved in the λ Red recombinase expression system[13]and were previously stored in our lab,were used in this study.

2.2.Pathway and plasmid construction

The primers used to amplify DNA fragments of targeted genes are listed in Table 2.The artificial tyrosol production pathway constructed in MG1655,is shown in Fig.1.The plasmids constructed during this study were confirmed by restriction enzyme digestion(Table 1).All restriction enzymes and DNA ligase were purchased from TaKaRa(Dalian,China).

Table 1Strains and plasmids used in this study

The endogenous E.coli phenylacetaldehyde dehydrogenase gene feaB(Gene ID:945933),the chorismate mutase and prephenate dehydratase gene pheA(Gene ID:947081),the tyrosine aminotransferase gene tyrB(Gene ID:948563),and the aromatic amino acid biosynthesis and transport regulon transcriptional regulator gene tyrR(Gene ID:945879)were successively knocked out of the E.coli MG1655 chromosome by using the classical λ Red homologous recombination method[14].These knockouts yielded strains CMGF,CMGP,CMGB,and CMGR,respectively.

Deletion of the feaB gene is described as an example of the procedure.The feaB fragment knockout expression cassette FRT-Km-FRT,which contains a kanamycin resistance marker flanked by arms homologous with feaB,was amplified using primers NfeaBU and NfeaBD.Km resistant strains were selected after using theλRed homologous recombination method.Insertion of the FRT-Km-FRT cassette was confirmed using colony-direct PCR with Ex Taq DNA polymerase and locusspecific primers YfeaBU and YfeaBD(Table 2),according to methods detailed in previous studies[15].The FLP recombination target,the(FRT)-flanked antibiotic resistance gene used for selection,was eliminated by induction of the pCP20 vector,which is a temperature-conditional plasmid that expresses FLP recombinase from a heat-inducible promoter.The resulting strain,CMGF,was verified using primers YfeaBU and YfeaBD.The other genes(pheA,tyrB,and tyrR)were deleted using similar strategies.

The wild-type phenylpyruvate decarboxylase gene ARO10 was amplified from Saccharomyces cerevisiae EBY100(Invitrogen)using primers ARO10U and ARO10D(Table 2).To improve ARO10 expression,optimization was performed by replacing codons predicted to be less frequently used in E.coli with more favored codons according to the JCat[16].The wild-type ARO10 gene and optimized gene ARO10*(Supplemental material S1)were inserted into plasmid pKK223-3 at the Eco RI and Hin dIII sites and generated pKK223-3-ARO10 and pKK223-3-ARO10*,respectively.The strong promoter tac was used in the regulation of expression of desired gene.

Table 2Oligonucleotides used in this study

2.3.Culturing of the recombinant E.coli strains

All strains were cultivated in LB broth medium(10 g·L-1tryptone,5 g·L-1yeast extract,10 g·L-1NaCl)and M9Y medium,which contained 1×M9 minimal salts(Na2HPO4·12H2O,17.1 g·L-1;KH2PO4,3.0 g·L-1;NaCl,0.5 g·L-1;NH4Cl,1.0 g·L-1)and 2%(w/v)glucose,and was supplemented with 0.025%(w/v)yeast extract,5 mmol·L-1MgSO4,and 100 mg·ml-1ampicillin.

A single colony of the indicated E.coli strain was used to inoculate 20 ml liquid LB medium containing appropriate antibiotics and allowed to grow overnight at 37°C.The overnight culture was then diluted 1:100 with 50 ml fresh LB medium and incubated in a gyratory shaker incubatorat37 °C and 200 r·min-1.When the absorbance of the culture(measured at 600 nm)reached approximately 0.6–0.8,isopropyl β-D-thiogalactopyranoside(IPTG)was added to the medium to reach a final concentration of 0.6 mmol·L-1.After the IPTG was added,the culture was incubated at 30°C for 8 h to induce recombinant protein expression.Subsequently,the cells were collected by centrifugation,resuspended in 50 ml M9Y medium,and cultured at 30°C for 48 h for the extraction of compounds and further analysis.The shake- flask experiments were conducted in triplicate.

2.4.Analysis of tyrosol and byproducts from the recombinant E.coli strains

The tyrosol and byproduct,the pyruvic acid,the lactic acid and the acetic acid,content of the fermentation medium were analyzed using high-performance liquid chromatography(HPLC)with diode array detection or using HPLC with a mass spectrometric(MS)detector(HPLC–MS).

For the analysis of tyrosol:HPLC analysis was performed with an Agilent 1260 system(Agilent Technologies Inc.,Santa Clara,CA)equipped with an Agela Innoval C18 column(4.6 mm×250 mm;particle size,5 mm)and a diode array spectrophotometer(Agilent Technologies).A 10 μl sample of the fermentation supernatant was applied to the column,which was eluted at room temperature with a mobile phase containing 80%solvent A(0.1%methanoic acid in H2O)and 20%solvent B(methanol)at a flow rate of 1 ml·min-1.The products were detected at 276 nm.Under these conditions,the retention time of a tyrosol standard was 10.28 min.To quantify tyrosol in the culture medium,calibration curves were generated with a series of known concentrations of the tyrosol standard dissolved in culture medium.The R2coefficients for the calibration curves were＞0.999.

For the analysis of byproducts:HPLC analysis was performed with an Agilent1260 system(Agilent Technologies Inc.,Santa Clara,CA)equipped with a Bio-Rad Aminex HPX-87H column(300 mm×7.8 mm)and the detector used for the detection was RID-10A.A 10 μl sample of the fermentation supernatant was applied to the column,which was eluted at room temperature with a mobile phase of 5 mmol·L-1H2SO4, flow rate of 0.5 ml·min-1,and column temperature of 65 °C.Under these conditions,the retention time of the pyruvic acid standard was 11.18 min;the retention time of the lactic acid standard was 15.89 min;and the retention time of the acetic acid standard was 17.12 min.To quantify byproducts in the culture medium,calibration curves were generated with a series of known concentrations of the byproduct standard dissolved in culture medium.The R2coefficients for the calibration curves were＞0.999.

HPLC–MS analysis was performed with a Waters Acquity UPLC system that included a BEH C18 column(2.1 mm× 150 mm × 1.7 μm),a Waters Acquity photodiode array detector,and a Waters MALDI Synapt Q-TOF MS detector equipped with an electrospray ionization probe.Samples were analyzed at 45°C using a gradient elution protocol at a flow rate of 0.3 ml·min-1as follows.For the first 0.1 min,the mobile phase consisted of 10%solvent A(acetonitrile)and 90%solvent B(0.1%formic acid in H2O).From 0.1 to 5 min,a linear gradient from 10%A and 90%B to 20%B was used;from 5 to 7 min,a linear gradient from 80%A and 20%B to 100%A was used;from 7 to 7.1 min a linear gradient from 100%A to 10%A and 90%B was used;and from 7.1 to 10 min,the mobile phase was maintained at 10%A and 90%B.The products were detected at 276 nm and identified using MS.

3.Results

3.1.Creation of a tyrosol biosynthesis pathway in E.coli MG1655

Neither E.coli MG1655 nor strain CMG0 could synthesize detectable amounts of tyrosol(data not shown).The tyrosol biosynthesis pathway created in E.coli MG1655(Fig.1)begins with 4HPP,the critical precursor of tyrosine in the aromatic amino acid biosynthesis pathway.To construct a biorthogonal pathway catalyzing the conversion of 4HPP to tyrosol,the S.cerevisiae ARO10 gene,which encodes phenylpyruvate decarboxylase,was inserted into pKK223-3 and the resulting plasmid was inserted into E.coli MG1655 to form strain CMGA.The CMGA strain could produce 0.05 mmol·L-1tyrosol(Fig.2).The result showed that overexpression of ARO10 combined with endogenous ADHs could convert 4HPP into tyrosol using glucose as substrate.

3.2.Engineering aromatic amino acid pathways to improve tyrosol yield

Fig.2.The influence of gene deletions on tyrosol production.All strains were derived from E.coli MG1655.Data are averages of results for three biological replicates.Error bars represent the standard deviation of three sets of parallel averages.

It has been reported that 4HPAAL is reduced to tyrosol by the endogenous ADHs in E.coli(Fig.1)[12].However,this intermediate compound could also be oxidized to 4HPA by endogenous phenylacetaldehyde dehydrogenases(FeaB or PadA)[17].Therefore,the feaB gene was deleted to evaluate its effect on the tyrosol production.We found that the engineered strain CMGFA(E.coli MG1655ΔfeaB harboring pKK223–3-ARO10)could produce 0.52 mmol·L-1tyrosol in the M9Y for 48 h,which was 10-fold the amount produced by CMGA(Fig.2).Thus,knockout of feaB had a positive effect on the accumulation of tyrosol.

Chorismate is a very important node in the biosynthesis of phenylalanine,tyrosine and tyrosine[18,19].To reduce carbon flux from chorismate toward phenylalanine,the pheA gene,which encodes chorismate mutase and prephenate dehydratase,was knocked out of the genome of CMGF to produce CMPG.After introducing pKK223-3-ARO10,the resulting engineered strain CMGPA was employed to produce tyrosol.A yield of 2.48 mmol·L-1tyrosol,which is 49.6-fold greater than that obtained with the control strain CMGA,was obtained(Fig.2).Furthermore,our observations indicated that the combined deletion of feaB and pheA resulted in an obviously synergistic effect on tyrosol synthesis.

To reduce tyrosine formation from the tyrosol precursor 4HPP,the tyrB gene was deleted from the genome of the double knockout strain of feaB and pheA CMGP to form strain CMGB.After inserting pKK223-2-ARO10,the resulting strain CMGBA yielded 3.26 mmol·L-1tyrosol(Fig.2),which was higher 31.5%than that obtained with CMGPA.In our opinion,cumulative elimination of competing pathways increased the efficiency of tyrosol production and increased its accumulation in the culture medium.

3.3.Effect of elimination of repression protein TyrR on tyrosol synthesis

In E.coli,tyrR encodes a repressor of the aromatic amino acid pathway.The activity of the TyrR protein is modulated by binding to one or more of the aromatic amino acids[20].The consensus DNA sequence for TyrR-binding sites,which are referred to as TyrR boxes,is TGTA AAN6TTTACA[20,21].TyrR exists as a dimer in solution,but in the presence of ATP and tyrosine it can restrain the transcription of tyrB,aroP,aroL-aroM,aroF-tyrA and tyrP[21,22].Therefore,it seemed reasonable to hypothesize that its presence in our system had a negative effect on the tyrosol yield.To test this hypothesis,the tyrR gene was deleted from CMGB to produce strain CMGR.After introducing pKK223-3-ARO10 to form CMGRA,induction of ARO10 overexpression and incubation in M9Y for 48 h,CMGRA could produce 3.57 mmol·L-1tyrosol(Fig.2),which was 9.5%higher than that of CMGBA(Fig.2).These results indicated that tyrR knockout had a positive effect on the accumulation of tyrosol.

Moreover,the effects of gene deletion on the byproducts and cell growth were investigated.We found that deletion of feaB,pheA and tyrR genes led to the little impact on the byproduct accumulation(Fig.3),however elimination of tyrB gene could decrease the pyruvic acid content in the final broth(48 h)(Fig.3a).More importantly,deletion of tyrB gene decreased the lactic acid accumulation significantly(Fig.3b).In our opinion,due to TyrB catalyzing 4HPP to tyrosine(Fig.1),deletion of tyrB gene could reduce tyrosine production and release the feedback repression,thereby increasing the tyrosol flux from pyruvate.Also,the deletion of the indicated genes gave the very slight effect on the acetic acid accumulation(Fig.3c).Fig.3d showed that accumulative deletion of the genes caused a defect in the growth performance.The engineered strain of CMGRA(E.coli MG1655 ΔfeaB ΔpheA ΔtyrB ΔtyrR with pKK223-3-ARO10)exhibited the lowest biomass content compared to the control strain and other engineered strains(Fig.3d).

3.4.Enhancement of tyrosol production by codon-optimized

The ARO10 gene was cloned from S.cerevisiae.Many studies have shown that codon preference in bacteria is significantly different than that in eukaryotes.This is particularly true for E.coli and S.cerevisiae[23].Our codon analysis showed that ARO10 possesses a number of codons that are rarely used in the E.coli genome.SDS-PAGE analysis also indicated that ARO10 was expressed at a low level(Fig.4a).Therefore,ARO10 was redesigned(optimized)to include codons preferred by E.coli,and then artificially synthesized(Supplemental material S1).The resulting gene,ARO10*,was inserted into pKK223-3 and the resulting plasmid was inserted into CMGR to generate the recombinant strain CMGRA*.SDS-PAGE analysis showed thatcodon optimization significantly improved the foreign gene expression level,compared with expression of the wild-type gene(Fig.4a).Furthermore,tyrosolproduction by the two engineered strains CMGRA and CMGRA*was compared.The tyrosol yield of CMGRA*reached 8.72 mmol·L-1,which was 2.44-fold greater than the tyrosol yield CMGRA(Fig.4b).These results demonstrate that codon optimization of ARO10 significantly improved its expression in CMGR,leading to significantly improved tyrosol production.

3.5.Optimization of the tyrosol production process

Induction conditions were optimized to improve the efficiency of tyrosol synthesis using CMGRA*.Firstly,we evaluated the effect of the growth phase of the cells being induced on tyrosol production.The results showed that tyrosol production was greater when ARO10*expression was induced after 8 h of growth(OD6000.6–0.8)than when expression was induced after 4,16,or 24 h(Fig.5a).Next,we assessed the effect of induction temperature on tyrosol production.Four different temperatures(20,25,30,and 37°C)were chosen for the induction phase.The production of tyrosol by CMGRA*was greatest(9.21 mmol·L-1)when the induction was conducted at 30°C(Fig.5b).Finally,the effect of inducer concentration on tyrosol production was investigated.IPTG was added to final concentrations of 0.1,0.2,0.4,0.6,0.8,1.0,and 1.2 mmol·L-1.Tyrosol production by CMGRA*was greatest with an IPTG concentration of 0.6 mmol·L-1(Fig.5c).As shown in Fig.5c,the production of tyrosol from glucose by CMGRA*reached 9.53 mmol·L-1under optimal culture conditions(induction with 0.6 mmol·L-1IPTG at an OD600of 0.6–0.8 and maintaining the temperature at 30 °C during the induction phase).Finally,we repeated the fermentation experiment under the optimal conditions,and assessed the relationship between tyrosol yield and the bacterial OD during the fermentation process.The experiment demonstrated that the fermentation was reproducible and the final tyrosol titer reached 9.53 mmol·L-1(1316.3 mg·L-1).Therefore,a higher titer of tyrosol was achieved by reconstruction of tyrosol synthetic pathway in E.coli.

Fig.3.The influence of gene deletions on byproduct production and the cell growth.Broth samples were taken at 24 h,36 h and 48 h and pyruvic acid(a),lactic acid(b)and acetic acid(c)were detected,respectively.Cell density was also monitored(d).Error bars represent the standard deviation of three sets of parallel averages.

Fig.4.Improvement of tyrosol production by codon-optimization of ARO10 gene.(a)SDS-PAGE analyses of ARO10*and ARO10 expression.M represents the molecular marker,lanes 1,2 and 3 indicate the strains CMG0(pKK223-3),CMGA(pKK223-3-ARO10)and CMGA*(pKK223-3-ARO10*),respectively.(b)The influence of codon optimization on tyrosol production.Strain CMGRA harbors the vector pKK223-3-ARO10;strain CMGRA*harbors the vector pKK223-3-ARO10*.ARO10*is the codon-optimized version of ARO10.Data are averages of results for three biological replicates.Error bars represent the standard deviation of three sets of parallel averages.

Fig.5.Determination of optimal culture conditions for the strain CMGRA*.(a)Effect of induction time on tyrosol titer at an OD600 of 0.6,an IPTG concentration of 0.2 mmol·L-1,and a temperature of 25 °C.(b)Effect of induction temperature on tyrosol production at an IPTG concentration of 0.2 mmol·L-1 and an induction time of 8 h.(c)Effect of IPTG concentration on tyrosol production at an OD600 of 0.6 and a temperature of 30°C.(d)Evaluation of tyrosol production under the optimal conditions using the recombinant strain CMGRA*.Data are the average experiments performed in triplicate.Error bars represent standard deviations from the mean.Data are averages of results for three biological replicates.Error bars represent the standard deviation of three sets of parallel averages.

HPLC and HPLC–MS analysis of a sample of the material produced in this experiment confirmed that the substance was tyrosol(Fig.6a and b).

4.Discussion

E.coli has become a promising host for the microbial production of a variety of valuable chemicals from renewable resources[24].Engineering E.coli for tyrosol production has been explored in recent years.Two pathways for tyrosol synthesis were evaluated:through expression of ARO10 and the indole-3-pyruvate decarboxylase gene ipdC[12,25],and through expression of tyrosine decarboxylase and tyrosine oxidase[8].More recently,we constructed a tyrosol-producing E.coli BL21(DE3)system that was able to effectively produce 4.15 mmol·L-1tyrosol from glucose[9].This engineered strain,when used as a biocatalyst,could successfully convert 10 mmol·L-1tyrosine to 8.71 mmol·L-1tyrosol,achieving a conversion rate of 87.1%.In this study,a more efficient tyrosol-producing engineered strain was constructed.This strain,CMGRA*,was able to produce 9.53 mmol·L-1(1316.3 mg·L-1)tyrosol from glucose.To the best of our knowledge,this is the tyrosol production ever reported using E.coli as the host.

The elimination of competing pathways is a useful strategy for improving bacterial production of interesting chemicals[26].In this study,we found that deleting both pheA and feaB significantly increased tyrosol production.Deleting tyrB,which encodes tyrosine aminotransferase,further increased the tyrosol yield.Together,these three deletions,which eliminate competing pathways,increased tyrosol production by 65-fold.

In contrast,deletion of tyrR,which eliminates potential repression of the tyrosol biosynthesis pathway,had a minimal effect on tyrosol production(9.1%increase).This suggests that eliminating the competing pathways lowered the concentrations of the aromatic amino acids,making repression by TyrR less of an issue.Subsequently,we assessed whether the strain CMGB was tyrosine deficient.Interestingly,the triple-deletion mutant CMGB was not tyrosine deficient;however,it was phenylalanine deficient.A plausible explanation for this is that deletion of pheA had a greater effect on phenylalanine biosynthesis than deletion of tyrB had on tyrosine biosynthesis.It seems likely that an alternative enzyme,like aspartate aminotransferase or histidine aminotransferase(AspC or HisC)may be able to partially compensate for the loss of TyrB[27].

Fig.6.(a)HPLC analysis of tyrosol production in the fermentation supernatant of recombinant strain CMGRA*.CMGRA*,E.coli MG1655(ΔfeaB ΔpheA ΔtyrB ΔtyrR)harboring plasmid pKK223-3-ARO10*;CMGRA,E.coli MG1655(ΔfeaB ΔpheA ΔtyrB ΔtyrR)harboring plasmid pKK223-3-ARO10;standard,authentic tyrosol.(b)Mass spectrum of tyrosol from the supernatant of a CMGRA*fermentation.Three biological replicates were performed and the representative data are presented.

Codon optimization of ARO10 led to a significant,2.4-fold increase in tyrosol production.Our data are clearly consistent with the well-known positive effect of codon optimization on gene expression in E.coli.In this particular instance,codon optimization also had a crucial effect on tyrosol production,suggesting that phenylpyruvate decarboxylase activity was limiting tyrosol production.Future work on pathway engineering should take into consideration the issue of codon usage.

In another study,the indole-3-pyruvate decarboxylase gene(ipdC)was introduced into the genome of E.coli BW25113(DE3),which overproduces tyrosine,and then the feaB and pheA genes were deleted to generate a recombinant tyrosol producer[25].The resulting strain produced 8.3 mmol·L-1tyrosol from 1%glucose.Although the yield of tyrosol in this study was not as high as that seen with the production system described in this report,they integrated ipdC into the bacterial genome by homologous recombination.Thus,the strain was no longer subject to the growth pressure caused by the plasmid,nor did they have to add antibiotics to prevent plasmid loss.In future studies,we will pursue the integration of key enzymes into the bacterial chromosome to avoid the use of inducers and antibiotics,which should smooth the path toward the industrialization of novel biocatalysts.

Supplementary Material

Supplementary data to this article can be found online athttps://doi.org/10.1016/j.cjche.2018.04.020.