Improvement of R-1-3-butanediol production by engineered Escherichia coli

Improvement of (R )-1,3-butanediol production by engineered Escherichia coliNaoya Kataoka,1Alisa S.Vangnai,2,3Takahisa Tajima,1Yutaka Nakashimada,1and Junichi Kato 1,*Department of Molecular Biotechnology,Graduate School of Advanced Sciences of Matter,Hiroshima University,Hiroshima 739-8530,Japan,1Department of Biochemistry,Faculty of Science,Chulalongkorn University,Bangkok 10330,Thailand,2and Center of Excellence for Environmental and HazardousWaste Management (EHWM),Chulalongkorn University,Bangkok 10330,Thailand 3Received 30October 2012;accepted 30November 2012Available online 4January 2013(R )-1,3-Butanediol is a valuable chemical extensively used as a key intermediate for the synthesis of pharmaceuticals and several industrial compounds.Despite its high demand,the production has been restricted from multi-step chemical production,petrochemical substrate requirement and a non-existence natural synthesis pathway from renewable biomass.In this study,an arti ficial synthesis route was genetically engineered in Escherichia coli MG1655lacI q to produce 1,3-butanediol from glucose.The selection of heterologous genes from several organisms and activity level of their corresponding gene products were demonstrated to be the key element for the product formation.Improvement of fermentation aeration and the application of fed-batch system signi ficantly enhanced 1,3-butanediol production.Under the optimized conditions,1,3-butanediol was produced up to 9.05g/l (100.4mM)with 98.5±0.2%enantiomeric excess (%ee )of (R )-1,3-butanediol.This is the highest yield of 1,3-butanediol produced from glucose with the highest optical purity by the recombinant strain reported thus far.Ó2012,The Society for Biotechnology,Japan.All rights reserved.[Key words:1,3-Butanediol;Engineered Escherichia coli ;Synthetic biology;Metabolic engineering;Bioproduction](R )-1,3-Butanediol (R13BD)is a non-natural,optical active alcohol,which has been used as a building block for the production of industrial chemicals including pheromones,fragrances,insecti-cides,and as a key chiral intermediate for the synthesis of penem and carbapenem b -lactam antibiotics (1).Because b -lactam anti-biotics are the most used antibacterial agents in clinical practice worldwide,the demand for R13BD has been drastically increased and,as a consequence,the production method of R13BD has been intensively studied (2,3).So far,1,3-butanediol (13BD)has been synthesized as a racemic mixture of R -and S -forms,mainly from petroleum-based chemicals such as a prochiral precursor,4-hydroxy-2-butanone (4H2B),or its racemic and enantiomeric compound,through either multiple steps of chemical reaction (4,5),or biological means including microbial reduction by native strains or genetically-modi fied strains (1,6)and lipase-catalyzed non-aqueous reaction (7).Nevertheless,upon a current shortage of fossil resources and petrochemical supplies,the bio-based production of 13BD,especially with a high optical purity of R13BD,from renew-able biomass becomes a more attractive alternative.Since R13BD is a non-natural synthon and a naturally-occurring metabolic pathway for its production has not yet been known;therefore,R13BD production through biotechnological route is strictly relied on a development of a recombinant bacterial strain using genetic engineering and metabolic engineering approaches.The only synthetic metabolic route of R13BD production from glucose bya recombinant Escherichia coli strain HB101was recently reported to give the highest yield of 1.028g/l (11.4mM)with a moderate 86.6%ee (Okabayashi,T.,Nakajima,T.,and Yamamoto,H.,Japanese patent 2009-251276,2009).Despite the achievement of previous synthesis strategies,the current drawbacks,which include non-renewable substrate requirement,multiple reaction steps,moderate optical purity and insuf ficient productivity,restrict the industrial production of R13BD.In this study,an effective synthetic biological production route of 13BD from glucose was successfully constructed,which consists of phaA (encoding 3-ketothiolase),phaB (encoding NAD(P)H-dependent acetoacetyl-CoA reductase)from Ralstonia eutropha NBRC 102504and bld (encoding butyraldehyde dehydrogenase)from Clostridium saccharoperbutylacetonicum ATCC 27012,and expressed in E.coli MG1655lacI q .The gene selection,which leads to the expression of functional enzymes and the product formation,was demonstrated to be the key element.With the optimized fermentation conditions,this recombinant whole-cell biocatalyst could produce 13BD up to 9.05g/l (100.4mM)with 98.5Æ0.2%enantiomeric excess (%ee )of R13BD.To the best of our knowledge,this is the highest yield of 13BD production with the highest optical purity by the recombinant strain reported so far.MATERIALS AND METHODSBacterial strains and plasmid construction Table 1and Table S1show the strains,plasmids and primers used in this study,respectively.The cloning components were as described:restriction enzymes (Toyobo,Osaka,Japan),ligase (Ligation High ver.2,Toyobo),In-fusion HD cloning kit (Clontech,CA,USA),and DNA polymerase (KOD DNA polymerase,Toyobo).The derivative E.coli MG1655*Corresponding author.Tel.:þ81824247757;fax:þ81824247047.E-mail address:jun@hiroshima-u.ac.jp (J.Kato)./locate/jbioscJournal of Bioscience and BioengineeringVOL.115No.5,475e 480,20131389-1723/$e see front matter Ó2012,The Society for Biotechnology,Japan.All rights reserved./10.1016/j.jbiosc.2012.11.025lacI q host strain(MG1655lacI q)was generated by introducing lacI q from NEB5-alpha F0I q(New England BioLabs,MA,USA)into E.coli MG1655by conjugation.The backbone plasmid,pHZK(8),which was previously modified to contain the ribosome binding site(AAAGGAGAGATTCAC),was constructed to generate pNK as follows.The fragment of P A1lacO-1::MCS rrnB T1T2was amplified from pKK223-3 (9)using PRf and PRr primers.The backbone pHZK was restriction digested with Eco RI and Hin dIII to remove its multiple cloning site(MCS)and lacZ gene and to be linearized.The P A1lacO-1::MCS rrnB T1T2amplified product of0.6kb was then directionally cloned into pHZK by In-fusion cloning kit generating a4.8-kb pNK.A9.4-kb pNK1(P A1lacO-1::phaA adhE phaB)was constructed to have three consecutive structural genes using the following recognition sites:phaA(Eco RI-Xho I),adhE(Xho I-Bam HI),and phaB(Bam HI-Sal I).pha A and phaB genes were amplified from genomic DNA of R.eutropha NBRC102504using PAf-PAr and PBf-PBr primers,respectively.A1.2-kb phaA and a0.8-kb phaB PCR products were digested with Eco RI-Xho I and Bam HI-Sal I,respectively.adhE gene was amplified from genomic DNA of Clostridium acetobutylicum ATCC824using ADf and ADr primers. The2.6-kb PCR product was digested with Xho I and Bam HI.The fragments were then cloned into pNK digested with Eco RI and Sal I.A9.5-kb pNK2(P A1lacO-1::phaA adhE hbd)was constructed to have three consecutive structural genes by replacing phaB with hbd gene in pNK1.hbd gene was amplified from genomic DNA of C.acetobutylicum ATCC824using HBf and HBr primers.The0.9-kb PCR product was digested with Bam HI and Sal I.The fragments were then cloned into pNK1digested with Bam HI and Sal I.A8.2-kb pNK3(P A1lacO-1::phaA bld phaB)was constructed to have three consecutive structural genes by replacing adhE with bld gene in pNK1.bld gene was amplified from genomic DNA of C.saccharoperbutylacetonicum ATCC27012using BLf and BLr primers.The1.4-kb PCR product was digested with Xho I and Bgl II.The fragments were then cloned into pNK1digested with Xho I and Bam HI.A8.3-kb pNK4(P A1lacO-1::phaA bld hbd)was constructed to have three consec-utive structural genes by replacing phaB with hbd gene in pNK3.hbd gene was amplified from genomic DNA of C.acetobutylicum ATCC824using HBf and HBr primers.The0.9-kb PCR product was digested with Bam HI and Sal I.The fragments were then cloned into pNK3digested with Bam HI and Sal I.Cultivation medium and fermentation conditions for13BD production The Luria e Bertani(LB)medium was used for culture cultivation during cloning and as a pre-culture medium for fermentation.Antibiotic was added when appropriated(ampicillin,100m g/ml,and tetracycline10m g/ml).For batch fermentation,a time course analysis using resting cell reaction was carried out.The resting cells were prepared by inoculating an overnight,LB-grown cell as an inoculum(1%,v/v)into50-ml LB medium in a300-mlflask with baffles and cultured at28 C on a rotary shaker(150strokes/min).When cell optical density (OD600)reached0.6e0.8,isopropyl-b-D-thiogalacto-pyranoside(IPTG)was added for induction to thefinal concentration ranging from0.01to0.5mM.After4-h induction,cells were collected by centrifugation at6000rpm,4 C,and resuspended to thefinal OD600of10in the production M9medium(6.8g Na2HPO4,3.0g KH2PO4,0.5g NaCl,1.0g NH4Cl,2mM MgSO4,0.1mM CaCl2per litter)containing3%(w/v)glucose,0.1M HEPES(pH7.4)and the appropriate antibiotic.The fermentation was carried out at37 C on a reciprocal shaker at150 strokes/min for12h or48h as indicated prior to the quantitative determination of13BD formation and glucose consumption.Effect of aeration on13BD production was examined by transferring the resting-cell suspension to a sealed 10-ml glass tube or a200-mlflask with baffles with various volume ratios between the culture suspension to a headspace remaining in the container(1:0.6, 1:1.5,1:4and1:7).During the fermentation,the pH was generally maintained above6.0by addition of5N NaOH,except when effect of pH on13BD production was examined.Fed-batch fermentation conditions were similar to that above mentioned protocol except that the fermentation period was extended to120h and glucose was fed at30g/l at20h and50h after the initiation of reaction.Determination of enzyme activities To determine in vitro activity of each enzyme involved in the synthetic pathway for13BD production,cell-free extract was prepared as followed.The resting cells of E.coli MG1655lacI q harboring pNK1,pNK2, pNK3or pNK4was prepared as described above,harvested,and resuspended in a sonication buffer(50mM Tris e HCl buffer and2mM1,4-dithiothreitol,pH7.5). Cells were disrupted by sonication(Digital Sonifier450,Branson Ultrasonics Corporation,CT,USA)for20min with2min pulse,on ice.After removal of cell debris,the supernatant was used as cell-free extract.Protein concentration was measured with a Bio-Rad protein assay kit.All enzymatic assays are spectrophotometrical-based assays(DU800,Beckman Coulter Inc.,CA,USA), which were conducted in the reaction mixture of0.2ml at30 C under aerobic conditions.The results were reported as specific activity(unit per mg of protein) of each enzyme.The thiolysis activity of3-ketothiolase(phaA gene product)was measured by monitoring the disappearance of acetoacetyl-CoA at303nm in the reaction,which consisted of100mM Tris e HCl(pH8.0),10mM MgCl2,1mM1,4-dithiothreitol, 50m M acetoacetyl-CoA,0.2mM CoA,and cell-free extract.The control reaction was conducted without Co-A.One unit was defined as the activity for1m mole of substrate consumption per min(ε303of14,000MÀ1cmÀ1)(10).Both hbd and phaB genes encode NAD(P)H-dependent acetoacetyl-CoA reduc-tase,which catalyzes the conversion of acetoacetyl-CoA to3-hydroxybutyryl-CoA. The enzyme activity was assayed by adding cell-free extract to an assay mixture containing100mM MOPS buffer(pH7.0),1mM1,4-dithiothreitol,0.1mM ace-toacetyl-CoA,and0.15mM NAD(P)H.The decrease of NAD(P)H was monitored at 340nm.One unit was defined as the activity for1m mole NAD(P)H consumed per min(ε340¼6220MÀ1cmÀ1)(11).Analytical techniques13BD and all metabolites were analyzed using a normal phase high performance liquid chromatography(HPLC)with an ion-exclusion column(RSpak KC-811,8.0mm IDÂ300mm L,Shodex,Tokyo,Japan) using0.1%phosphate as a mobile phase at aflow rate of0.7ml minÀ1,at60 C13BD was acetylated(12)prior to check its optical purity.The enantiomeric excess(%ee) was examined using a reverse phase HPLC equipped with a UV detector(at220nn). The separation was performed at40 C on a chiral column(Chiralcel OB-H,4.6mm IDÂ250mm,Daicel Chem.Co.,Osaka,Japan)using a hexane:isopropanol mixture (19:1)as a mobile phase at aflow rate of1ml minÀ1.The quantitative analysis was conducted using a calibration curve generated from a standard racemic13BD(TCI, Tokyo,Japan).RESULTS AND DISCUSSIONDesign of the artificial pathway for13BD synthesis and expression of the gene products R-1,3-Butanediol is a non-natural synthon,of which its naturally-occurring productionTABLE1.Strains and plasmids used in this study.Strain/Plasmid Relevant characteristics SourceStrainsEscherichia coliNEB5-alpha F0I q F0proAþBþlacI q D(lacZ)M15zzf::Tn10(Tet R)/fhuA2D(argF-lacZ)U169phoA glnV44F80D(lacZ)M15gyrA96recA1relA1endA1thi-1hsdR17New England Biolabs,USAMG1655F-l-ilvG-rfb-50rph-1NBRCMG1655lacI q F0[proAB lacI q Z D M15Tn10(Tet R)]l-ilvG-rfb-50rph-1;source of atoB This studyMG-NK1MG1655lacI q/pNK1This studyMG-NK2MG1655lacI q/pNK2This studyMG-NK3MG1655lacI q/pNK3This studyMG-NK4MG1655lacI q/pNK4This studyRalstonia eutropha NBRC102504Source of phaA and phaB NBRCClostridium acetobutylicum ATCC824Source of hbd and adhE ATCCC.saccharoper-butylacetonicum ATCC27012Source of bld ATCCPlasmidspKK223-3Source of MCS;rrnB T1T29pHZK p15A ori;Amp r;Kan r8pNK p15A ori;Amp r;Kan r;P A1lac O-1::MCS;rrnB T1T2This studypNK1p15A ori;Amp r;Kan r;P A1lac O-1::phaA(RE)-adhE(CA)-phaB(RE);rrnB T1T2This studypNK2p15A ori;Amp r;Kan r;P A1lac O-1::phaA(RE)-adhE(CA)-hbd(CA);rrnB T1T2This studypNK3p15A ori;Amp r;Kan r;P A1lac O-1::phaA(RE)-bld(CS)-phaB(RE);rrnB T1T2This studypNK4p15A ori;Amp r;Kan r;P A1lac O-1::phaA(RE)-bld(CS)-hbd(CA);rrnB T1T2This studyNBRC,NITE Biological Resource Center,Japan;ATCC,American Type Culture Collection;RE,Ralstonia eutropha NBRC102504;CA,Clostridium acetobutylicum ATCC824;CS, C.saccharoperbutylacetonicum ATCC27012.476KATAOKA ET AL.J.B IOSCI.B IOENG.,pathway from biomass substrate has not yet been known.In this study,a synthetic genetic route was designed and constructed from foreign genes to produce 13BD from the most commonly used biomass for industrial production,glucose,and expressed in a readily genetically-manipulated host,E.coli MG1655lacI q .The success of production of industrial chemicals through synthetic metabolic route is generally depending on type of gene and activity of the corresponding gene product in the selected host;therefore,the selection of proper gene or its homologs from several organ-isms would be one of the most important keys to increase productivity,titer and yield (13,14).The overall synthesis pathway of 13BD from glucose requires two main metabolic parts.The first element includes the generation of central metabolic intermediates,i.e.pyruvate and acetyl-CoA,and the formation of reducing equivalents.The second element encompasses theconversion pathway of acetyl-CoA to 13BD,which requires the following enzymes:3-ketothiolase,acetoacetyl-CoA reductase,and alcohol/aldehyde dehydrogenase (Fig.1).So far,the only arti ficial production pathway of 13BD from glucose was reported to consist of phbA (b -ketothiolase-encoding gene),phbB (acetoacetyl-CoA reductase-encoding gene)from R.eutropha DSM 531,and adhE (aldehyde/alcohol dehydrogenase gene derived from C.acetobutylicum )(Okabayashi,T.,Nakajima,T.,and Yamamoto,H.,Japanese patent 2009-251276,2009).In this study,in order to enhance the productivity and,perhaps,the optical purity,the metabolic route for 13BD formation was redesigned using different sources of gene and gene combination.The following genes and its homolog were selected from several sources (Table 1)to encode the essential enzymes,used in the construction of the plasmids,pNK1to pNK4(Table 1),before transforming into E.coli MG1655lacI q to create the recombinant strains,MG-NK1to MG-NK4:phaA or atoB for 3-ketothiolase,phaB or hbd for NAD(P)H-dependent acetoacetyl-CoA reductase,and adhE or bld for alcohol/aldehyde dehydrogenase or butylraldehyde dehydrogenase,respectively.Since the overexpression of atoB constructed in the plasmid pNK resulted in protein aggregation (data not shown),the gene was no longer used in further study.Therefore,phaA was the only gene remaining as 3-ketothiolase-encoding gene in all constructs.The recombinant strains were then prepared as described before the enzyme activity and 13BD formation under a semi-aerobic condition were examined (Fig.2).The activity of 3-ketothiolase (PhaA)was in comparable levels in all recombinants.The activity assay of alcohol/aldehyde dehydrogenase (AdhE)and butylralde-hyde dehydrogenase (Bld)showed inclusive result,possibly due to the instability of the enzymes under an aerobic assay conditions (15).Nevertheless,their activity could be presumed from the end-point production level of 13BD.The strain MG-NK2harboring hbd from C.acetobutylicum ATCC 824exhibited a 50-time higher ace-toacetyl-CoA reductase activity than MG-NK1carrying phaB from R.eutropha NBRC 102504,but did not yield 13BD production.This result was in agreement to the previous reports in that,in non-native pathways,a highly over-expressed gene or too high enzyme activity may cause an imbalance metabolism of the host,which may result in a reduction of product formation (15).In this case,an excessive consumption of NADH of NAD(P)H-dependent acetoace-tyl-CoA reductase (Hbd)may disturb NAD þ/NADH balance of the production pathway.Although strain MG-NK1harboring phaBandFIG.1.Schematic representation of 1,3-butanediol production pathway from glucose in the engineered E.coli MG1655lacI q .The metabolic pathway starting from acetyl-CoA consists of three enzymatic steps,of which the genes were obtained from different organisms.3-Ketothiolase (atoB from the native E.coli ,phaA from R.eutropha NBRC 102504);an NAD(P)H-dependent acetoacetyl-CoA reductase (hbd from C.acetobutylicum ATCC 824,phaB from R.eutropha NBRC 102504);alcohol/aldehyde dehydrogenase and butyraldehyde dehydrogenase (adhE from C.acetobutylicum ATCC 824and bld from C.saccharoperbutylacetonicum ATCC 27012,respectively).FIG.2.13BD production gene organization in plasmids pNK1to pNK4,and enzymatic activities and 13BD production in MG1655lacI q harboring these plasmids.Batch cultures were performed at 37 C in a 10-ml sealed-glass vial with a 1:1.6volume ratio between the culture suspension to a culture container,which was considered as a semi-aerobic condition.The 13BD production was determined after 12h of fermentation.The meaning of each symbol is as follows:no existence of the enzyme in the recombinant strain;ND,not detected.Source of gene coding for each enzyme was from Ralstonia eutropha NBRC 102504(RE),Clostridium acetobutylicum ATCC 824(CA),and Clostridium saccharoperbutylacetonicum ATCC 27012(CS).Note:The enzymatic assay of adhE and bld gene product was not possible due to enzyme instability.V OL .115,2013(R )-1,3-BUTANEDIOL PRODUCTION BY ENGINEERED E.COLI 477adhE could produce 13BD,the yield was relatively low (0.37mM)when compared to the previous report (11.4mM;1.03g/l)(Oka-bayashi,T.,Nakajima,T.,and Yamamoto,H.,Japanese patent 2009-251276,2009).Further improvement,thus,focused on the use of an alternative gene from adhE .In addition to alcohol/aldehyde dehydrogenase (AdhE)of C.acetobutylicum ATCC 824,butyraldehyde dehydrogenase (Bld)of C.saccharoperbutylacetonicum ATCC 27012was examined for itsability to catalyze the conversion of 3-hydroxybutyryl-CoA to 3-hydroxybutyraldehyde.Then,in this case,the following step of 13BD formation from 3-hydroxybutyraldehyde was catalyzed by the native enzyme of the host.The NAD(P)H-dependent acetoace-tyl-CoA reductase (PhaB)activities in MG-NK1and MG-NK3were comparable,but the production of 13BD in MG-NK3harboring bld gene reached 1.45mM,which was signi ficantly higher (4folds)(Fig.2).These results indicate the importance of gene selection and the balance of gene product activities in order to have proper expression and product formation from the synthetic production pathway in the selected recombinant host.According to the results,the recombinant strain MG-NK3harboring phaA ,phaB ,and bld was selected for further bioprocess improvement to enhance 13BD production.Batch production of 13BD The production of 13BD was initially conducted in batch fermentation.The following factors affecting batch production were investigated and optimized to enhance 13BD titer:IPTG concentration and oxygen supply.Since the genes constructed in the plasmid pNK3were under control of the IPTG-inducible promoter P A1lacO-1;therefore the appropriate level of IPTG would result in high gene expressions and,conse-quently,a high 13BD e of IPTG concentration of 0.01mM yielded the relatively highest 13BD under the conditions tested (data not shown);therefore it was selected as a proper inducer concentration for further investigation.The fact that 13BD production route is a reductive pathway and butylraldehyde dehydrogenase from C.saccharoperbutylacetonicum ATCC 27012(Bld)is an oxygen-sensitive enzyme;therefore,an2468510151:0.61:1.51:41:71,3-b u t a n e d i o l (m M )Glucose consumption rate (mM/h)Volume ratio(cell suspension :headspace)FIG.3.Batch fermentation of 1,3-butanediol by the recombinant E.coli strain MG-NK3at various aeration levels (expressed as differences in volume ratio of cell suspension to headspace).The culture medium contained glucose at 30g/l (167mM)as the initial substrate.Batch cultivations were conducted at 37 C (bars).Glucose consumption rate (circles)was determined for each fermentationcondition.40801201602000204060801001201400122436484080120160200010*******1224364845678Glucose (mM)M e t a b o l i t e s (m M )Glucose (mM)M e t a b o l i t e s (m M )1,3-b u t a n e d i o l (m M )p H1,3-b u t a n e d i o l (m M )p HTime (h)Time (h)02040608010012014012243648FIG.4.Time course of batch production of 13BD by the recombinant E.coli strain MG-NK3without (A,B,C)and with pH control (D,E,F).The batch fermentation was conducted at 37 C with a baf fled flask under aerobic condition.The actual pH (A)and the controlled pH (D)were monitored.The production of 13BD (closed squares),glucose consumption (open squares)(B,E)and the formation of other metabolites (C,F)were examined.478KATAOKA ET AL.J.B IOSCI .B IOENG .,anaerobic or a semi-aerobic condition was hypothesized to be preferable for 13BD production.Nevertheless,to produce 13BD,a suf ficient supply of acetyl-CoA and NAD(P)H is strictly required from aerobic condition.Due to this discrepancy,it was necessary to determine the suitable oxygen supply for 13BD production.The resting cell reaction was conducted at 37 C with different volume ratios between culture suspension and a headspace remaining in the container (1:0.6,1:1.5,1:4and 1:7)resulting in different aera-tion.Glucose consumption rate in each condition was determined and high consumption rate represented relatively higher oxygen supply during the production.The result (Fig.3)indicated that the production was higher with an increase of oxygen supply.13BD production with the volume ratio of 1:7markedly increased the titer to 11.0mM (0.99g/l)in batch fermentation.Although butyl-raldehyde dehydrogenase (Bld)in 13BD synthesis pathway is known as an oxygen-sensitive enzyme,it was clearly shown that aerobic condition is preferred for 13BD production in MG-NK3.The results suggest that high acetyl-CoA substrate is required from glucose conversion in order to drive the reaction direction through 13BD production pathway.These optimized conditions were subsequently applied in further investigations.Effect of pH on 13BD batch production Time course of batch fermentation was investigated for 13BD production and formation of other fermentation metabolites.In this study,to obtain high production of the target chemical,13BD,high-concentrated,resting cells were used in batch production.Without pH adjustment,the formation and accumulation of organic acids,i.e.succinate,lactate,formate,and a signi ficantly high amount of acetate,caused drastic pH change,which resulted in a decrease of glucose conversion and 13BD production (Fig.4A e C).Acetate accumulated at high concentration in fermentation system was previously reported to be a main inhibitory substance not only to E.coli growth,but also its production ef ficiency (16,17).To mitigate the pH problem,pH was adjusted and maintained at approximately 6.0.As a consequence,13BD production and glucose conversion were not ceased and the highest 13BD titer of 34.6mM was achieved at 36-h of batch cultivation (Fig.4D,E).The fact that interval pH adjustment during the cultivation derestricted 13BD production may suggest the involvement of pH-dependent enzyme(s)in the metabolic network leading to 13BD production.In addition,it is worth noting that upon a rapid increase of 13BD production,acetate was rapidly consumed.Within a short subsequence at low level of acetate,13BD formation rate was signi ficantly slowed down.This result may suggest that acetate or other metabolites generated during the fermentation may playa role in activation of 13BD formation.Although the mechanism of this phenomenon cannot be described in this study,it was previously reported that organic acids may trigger a metabolic shift in C.saccharoperbutylacetonicum N1-4in acetone e butanol e ethanol fermentation (17).Fed-batch fermentation for 13BD production To enhance 13BD yield,a fed-batch fermentation with a pH control (Fig.5A e C)was carried out to eliminate any inhibition,which may be caused by high concentration of substrate and other metabolites rapidly generated and accumulated in a single,heavy-loaded substrate of a batch fermentation.In this study,the initial glucose feeding was at 30g/l (167mM),followed by the repeating feedings with a similar glucose concentration at 20h and 50h after the initial substrate loading period.With pH-control,fed-batch system,13BD was produced in a linear manner up to 80h of cultivation and the concentration of 100.4mM (9.05g/l)of 13BD could be achieved at 110h (Fig.5A).The optical purity (%ee )of 13BD was further determined by a chiral HPLC,quantitatively analyzed with a calibration curve with standard chemicals (Fig.S1)and revealed that 13BD produced from the synthetic pathway of recombinant E.coli strain MG-NK3has 98.5Æ0.2%ee .To the best of our knowledge,this study is the first report of the highest titer and the optical purity of 13BD accounted so far.Supplementary data related to this article can be found at /10.1016/j.jbiosc.2012.11.025.ACKNOWLEDGMENTSThis work was the collaboration of Chulalongkorn University e Hiroshima University under the Asian Core Program (ACP)and financially supported by The Japan Society for the Promotion of Science (JSPS)and the National Research Council of Thailand (NRCT)(Bilateral Project).References1.Matsuyama,A.,Yamamoto,H.,Kawada,N.,and Kobayashi,Y.:Industrial production of (R )-1,3-butanediol by new biocatalysts,J.Mol.Catal.B:Enzym.,11,513e 521(2001).2.Llarrull,L.I.,Testero,S.A.,Fisher,J.F.,and Mobashery,S.:The future of the beta-lactams,Curr.Opin.Microbiol.,13,551e 557(2010).3.Zheng,R.C.,Ge,Z.,Qiu,Z.K.,Wang,Y.S.,and Zheng,Y.G.:Asymmetric synthesis of (R )-1,3-butanediol from 4-hydroxy-2-butanone by a newly isolated strain Candida krusei ZJB-09162,Appl.Microbiol.Biotechnol.,94,969e 976(2012).rcheveque,M.,Mambu,L.,and Petit,Y.:Preparation of enantiomerically pure 1,3-butanediol from threonine,mun.,21,2295e 2300(1991).20406080100120020406080100120040801202040608010012068105791,3-b u t a n e d i o l (m M )p HGlucose (mM)M e t a b o l i t e s (m M )Cell biomass (OD 600)Time (h)Time (h)FIG.5.Time course of fed-batch production of 13BD by the recombinant E.coli strain MG-NK3with pH control.The fed-batch fermentation (A)was conducted at 37 C with a baf fled flask under aerobic condition with an initial glucose concentration of 30g/l with twice repeating feedings (open square),while the production of 13BD was examined (closed square).The system pH (open squares)and cell biomass (closed squares)were monitored (B).The formation of other metabolites (C)was also examined.V OL .115,2013(R )-1,3-BUTANEDIOL PRODUCTION BY ENGINEERED E.COLI 479。