Chemostat culture characterization of Escherichia coli mutant strains metabolically engineered for...

ARTICLE IN PRESS

1096-7176/$ - se

doi:10.1016/j.ym

�CorrespondUniversity, Hou

E-mail addr

Metabolic Engineering 7 (2005) 337–352

www.elsevier.com/locate/ymben

Chemostat culture characterization of Escherichia coli mutant strainsmetabolically engineered for aerobic succinate production: A study ofthe modified metabolic network based on metabolite profile, enzyme

activity, and gene expression profile

Henry Lina, George N. Bennettb, Ka-Yiu Sana,c,�

aDepartment of Bioengineering, Rice University, Houston, Texas, USAbDepartment of Biochemistry and Cell Biology, Rice University, Houston, Texas, USA

cDepartment of Chemical Engineering, Rice University, Houston, Texas, USA

Received 21 February 2005; received in revised form 2 June 2005; accepted 13 June 2005

Available online 11 August 2005

Abstract

Various Escherichia coli mutant strains designed for succinate production under aerobic conditions were characterized in

chemostat. The metabolite profiles, enzyme activities, and gene expression profiles were studied to better understand the metabolic

network operating in these mutant strains. The most efficient succinate producing mutant strain HL27659k was able to achieve a

succinate yield of 0.91mol/mol glucose at a dilution rate of 0.1/h. This strain has the five following mutations: sdhAB, (ackA-pta),

poxB, iclR, and ptsG. Four other strains involved in this study were HL2765k, HL276k, HL2761k, and HL51276k. Strain HL2765k

has mutations in sdhAB, (ackA-pta), poxB and iclR, strain HL276k has mutations in sdhAB, (ackA-pta) and poxB, strain HL2761k

has mutations in sdhAB, (ackA-pta), poxB and icd, and strain HL51276k has mutations in iclR, icd, sdhAB, (ackA-pta) and poxB.

Enzyme activity data showed strain HL27659k has substantially higher citrate synthase and malate dehydrogenase activities than

the other four strains. The data also showed that only iclR mutation strains exhibited isocitrate lyase and malate synthase activities.

Gene expression profiles also complemented the studies of enzyme activity and metabolites from chemostat cultures. The results

showed that the succinate synthesis pathways engineered in strain HL27659k were highly efficient, yielding succinate as the only

major product produced under aerobic conditions. Strain HL27659k was the only strain without pyruvate accumulation, and its

acetate production was the least among all the mutant strains examined.

r 2005 Elsevier Inc. All rights reserved.

Keywords: Escherichia coli; Metabolic engineering; Succinate production; Aerobic fermentation; Real-time RT-PCR; Chemostat; Enzyme activity

1. Introduction

The production of succinate has been an area ofrecent interest due to its value as a precursor to variouscommodity chemicals used in industries like food,pharmaceutical, detergent, and polymer (Zeikus, 1980;Zeikus et al., 1999). Metabolic engineering to enhance

e front matter r 2005 Elsevier Inc. All rights reserved.

ben.2005.06.002

ing author. Department of Chemical Engineering, Rice

ston, Texas, USA.

ess: [email protected] (K.-Y. San).

succinate production in bacteria has the potential tosignificantly improve the economics of the succinatemarket, especially when coupled with the use of renew-able carbohydrates (Schilling, 1995). Examples includethe high yield production of succinic acid from woodhydrolysate, whey, or glyercol by Anaerobiospirillum

succiniciproducens through improvement of its fermen-tation conditions (Lee et al., 2000, 2001, 2003b;Samuelov et al., 1999). Other organisms that alsohave the innate capability for high yield succinicacid production include the facultative anaerobe

www.elsevier.com/locate/ymben

ARTICLE IN PRESSH. Lin et al. / Metabolic Engineering 7 (2005) 337–352338

Actinobacillus succinogenes (Van der Werf et al., 1997),and the organism from bovine rumen Mannheimia

succiniciproducens, which had been shown to be ableto produce succinic acid from whey and corn steepliquor (Lee et al., 2002, 2003a).

Various metabolic engineering strategies have beenapplied to improve succinate production in Escherichia

coli (Chatterjee et al., 2001; Gokarn et al., 1998, 2000,2001; Goldberg et al., 1983; Hong and Lee, 2001;Sanchez et al., 2005a, b; Stols and Donnelly, 1997;Vemuri et al., 2002a; Wang et al., 1998). Because E. coli,naturally, only produces succinate under anaerobicconditions in minimal quantities (Clark, 1989), numer-ous genetic modifications have been performed on E.

coli to enhance succinate production. Genetic engineer-ing coupled with optimization of production conditionshas shown promising results for large-scale productionof succinate from E. coli. This makes succinateproduction in E. coli competitive with that of otherorganisms like A. succiniciproducens. A geneticallyimproved E. coli mutant strain (AFP111/pTrc99A-pyc)grown anaerobically in optimized fed batch conditionswas shown to achieve succinate production of 99.2 g/Lwith yield of 110% and productivity of 1.3 g/L h(Vemuri et al., 2002b). Because of anaerobic fermenta-tion process, disadvantages include poor biomassgeneration, slow carbon throughput, and, these

Fig. 1. Genetic engineering of glycolysis, TCA cycle, and glyoxylate bypass i

2005b). 1 is icd knockout, 2 is sdhAB knockout, 5 is iclR knockout, 6 is pox

slow product formation, E. coli was also geneti-cally engineered to produce succinate under aerobicconditions (Lin et al., 2005a, b). Aerobic fed batchculture of a genetically modified E. coli mutant strain(HL27659k(pKK313)) was shown to produce 58.3 g/Lof succinate with yield of 94% and productivity of 1.1 g/L h (Lin et al., 2005c).

In this study, several E. coli mutant strains con-structed by Lin et al. during the development of aerobicsuccinate production systems were selected for char-acterization to further understand their metabolicfunctions, as a result of multiple pathway inactivations.This is important for understanding how the pathwaymanipulations in aerobic central metabolism affectedthe metabolic network and enabled aerobic succinateproduction in E. coli. These mutant strains werecharacterized in chemostat culture for their metaboliteprofiles, enzyme activities, and gene expression profiles.Enzyme activities and gene expression patterns wereexamined for those involved in pathways that affect theaerobic succinate production system. Combining geneexpression profiles with enzyme activity and metaboliteprofiles provides a holistic approach for better under-standing the connections between genes, proteins, andmetabolites.

The mutations in the mutant strains were strategicallycreated during the development process to enable

n the development of aerobic succinate production systems (Lin et al.,

B knockout, 7 is ackA-pta knockout, and 9 is ptsG knockout.

ARTICLE IN PRESS

Fig. 2. Aerobic succinate production platform, the two-route system

with the glyoxylate cycle and the oxidative branch of the TCA cycle

(mutant strain HL27659k).

H. Lin et al. / Metabolic Engineering 7 (2005) 337–352 339

aerobic succinate production in E. coli. Based onpathway modeling and optimization, the maximumtheoretical succinate yield that can be achieved in E.

coli under aerobic conditions is 1.0mol of succinate permole of glucose consumed (Lin et al., 2005a, b).Following the design for optimal succinate productionunder aerobic conditions, genetic manipulation wascarried out in E. coli. As a result, several engineered E.

coli mutant strains were able to achieve the maximumtheoretical succinate yield of 1.0mol/mol glucose underaerobic conditions in batch reactors (Lin et al., 2005b).Although the maximum theoretical succinate yield wasachieved by these strains, they still varied significantly insuccinate productivity (Lin et al., 2005b). One of themutant E. coli strains, though, exhibited significantlyhigher succinate productivity than the other strains,while still reaching the maximum succinate yield of1.0mol/mol glucose. This strain, HL27659k(pKK313),mentioned earlier was shown to have high capacity forsuccinate production in aerobic fed batch cultures.Before strain HL27659k(pKK313) was developed, thislevel of succinate production under aerobic conditionswas never possible in E. coli.

Strain HL27659k(pKK313) is the most efficient andoptimal E. coli strain created thus far for aerobicsuccinate production. Plasmid pKK313 provides theoverexpression of a mutant Sorghum vulgare phosphoe-nolpyruvate carboxylase (PEPC) to enhance carbon fluxto oxaloacetate (OAA) for the succinate synthesispathways. Background strain HL27659k has five muta-tions created in genes encoding enzymes of aerobiccentral metabolism. These five mutations are sdhAB,inactivation of succinate dehydrogenase (SDH), (ackA-pta), inactivation of acetate kinase-phosphotransacety-lase, poxB, inactivation of pyruvate oxidase, iclR,inactivation of aceBAK operon repressor, and ptsG,inactivation of glucose phosphotransferase system(Fig. 1). Naturally, under aerobic conditions, succinateis only an intermediate of the tricarboxylic acid (TCA)cycle and it is formed by succinyl-CoA synthetase andsubsequently oxidized to fumarate by SDH. Because ofthis, succinate is never detected in aerobic cultures of E.

coli. Inactivation of SDH is, therefore, essential foraccumulation of succinate under aerobic conditions. Byinactivating SDH, the TCA cycle becomes branchedwith the oxidative branch capable of producing succi-nate as a product under aerobic conditions. Inactivationof the two acetate producing pathways ackA-pta andpoxB increases the carbon flux toward the branchedTCA cycle for succinate production. Inactivation of theaceBAK operon repressor (iclR) activates the glyoxylatebypass for succinate production (Gui et al., 1996). Theglyoxylate bypass consists of two steps. The first step iscarried out by isocitrate lyase (ICL), which convertsisocitrate (C6) to succinate (C4) and glyoxylate (C2). Thesecond step is carried out by malate synthase (MS),

which condenses acetyl-CoA (C2) with glyoxylate (C2) toform malate (C4). With the inactivation of iclR andsdhAB, two routes are created for aerobic succinateproduction. One route is the oxidative branch of theTCA cycle and the second route is the glyoxylate cycle.Inactivation of ptsG increases the phosphoenolpyruvate(PEP) pool for succinate synthesis, since glucosetransport would no longer require the conversion ofPEP to pyruvate for phosphorylation. This would alsoimprove the balanced carbon metabolism by slowingdown the glucose uptake, since acetate is formed whenthere is excess influx of glucose that the cell is unable toutilize for biomass synthesis (Doelle et al., 1981; Yanget al., 1999). The five mutations in strain HL27659k(Fig. 2) together create an aerobic succinate productionplatform that can produce substantial levels of succinatewith high productivity.

Another strain that was also shown to reach themaximum theoretical succinate yield of 1.0mol/molglucose under aerobic conditions in batch reactorwas strain HL51276k(pKK313) (Lin et al., 2005b). Inbatch cultures, strain HL51276k(pKK313) and strainHL27659k(pKK313) both achieved the succinate yieldof 1.0mol/mol glucose, but strain HL51276k(pKK313)had much lower productivity than strain HL27659k(pKK313) (Lin et al., 2005b). The background strain

ARTICLE IN PRESS

Fig. 3. Aerobic succinate production platform, the glyoxylate cycle

(mutant strain HL51276k) (Lin et al., 2005b).

Table 1

List of mutant strains studied in chemostat and their mutations

Strains Mutations

sdhAB ackA-pta PoxB iclR icd ptsG

HL27659k X X X X X

HL2765k X X X X

HL276ka X X X

HL2761k X X X X

HL51276k X X X X X

aBackground control strain for all the other strains.

H. Lin et al. / Metabolic Engineering 7 (2005) 337–352340

HL51276k has five mutations, sdhAB, (ackA-pta), poxB,iclR, and icd. It differs from strain HL27659k in that itonly has the glyoxylate cycle for succinate production(Fig. 3). Because of the inactivation of isocitratedehydrogenase (icd), the oxidative branch of the TCAcycle is not active for succinate production. The TCAcycle of strain HL51276k, in essence, is reduced to theglyoxylate cycle.

Certain mutant strains constructed during the processof developing an aerobic succinate production system inE. coli were selected for chemostat characterization. Themost prominent mutant strain created for aerobicsuccinate production is strain HL27659k. Therefore,this strain was selected for further study in chemostatcultures. Four additional mutant strains were selected tocomplement strain HL27659k in understanding how thegenetic manipulations affected the metabolic functionsamong the related strains. One mutant strain HL2765k(sdhAB, (ackA-pta), poxB, and iclR), has the samemutations as strain HL27659k, except for ptsG. StrainHL2765k will provide the basis for studying the effect ofptsG inactivation on the glucose metabolism in strainHL27659k. Strain HL276k (sdhAB, (ackA-pta), poxB),has three mutations, serves as the base control for

strains HL2765k and HL27659k. Strain HL276k doesnot have the iclR inactivated; therefore, the glyoxylatebypass operon is not constitutively expressed. StrainHL276k can produce succinate only from the oxidativebranch of the TCA cycle. Another mutant strainselected for study, described earlier, is strainHL51276k (iclR, icd, sdhAB, (ackA-pta), and poxB).The TCA cycle in this strain has been reconstructed tobecome just the glyoxylate cycle. Succinate produced bythis strain should thus originate only from the ICLpathway of the glyoxylate cycle. Strain HL51276kprovides an interesting comparison with strainHL2765k, which has both the glyoxylate cycle and theoxidative branch of the TCA cycle active for succinateproduction. The last mutant strain selected for char-acterization is strain HL2761k (sdhAB, (ackA-pta),poxB, icd). This strain neither has the glyoxylate cyclenor the oxidative branch of the TCA cycle active forsuccinate production. An icd mutation in the strainHL2761k inactivates the TCA cycle, and without amutation in the iclR, there is no bypass for the carbonflux entering the TCA cycle. Strain HL2761k providesan interesting comparison with strain HL51276k, sincethey differ in that strain HL51276k has an activeglyoxylate cycle and strain HL2761k does not. StrainHL276k also serves as the control for strains HL2761kand HL51276k. Comparison of strain HL2761k withstrain HL276k provides understanding of the effect oficd inactivation on the carbon flux toward the branchedTCA cycle. The five mutant strains selected for study inchemostat culture, and their genotypes are summarizedin Table 1.

2. Materials and methods

2.1. Strains

Mutations were created in the laboratory wildtypeGJT001, a spontaneous cadR mutant of MC4100(Tolentino et al., 1992). A library of mutant strainswas created during the process of constructing thefinal mutant strains for aerobic succinate production

ARTICLE IN PRESS

Table 2

Library of mutant strains constructed and studied

Strains Genotype Reference

GJT001 Spontaneous cadR mutant of MC4100(ATCC35695) Tolentino et al. (1992)

Dlac(arg-lac)U169rpsL150relA1ptsF SmR

HL2k GJT001(sdhAB::KmR) Lin et al. (2005a)

HL27k GJT001(sdhAB, ackA-pta::KmR) Lin et al. (2005a)

HL276k GJT001(sdhAB, (ackA-pta), poxB::KmR) Lin et al. (2005a)

HL2761k GJT001(sdhAB, (ackA-pta), poxB, icd::KmR) Lin et al. (2005a)

HL2765k GJT001(sdhAB, (ackA-pta), poxB, iclR::KmR) Lin et al. (2005b)

HL27659k GJT001(sdhAB, (ackA-pta), poxB, iclR, ptsG::KmR) Lin et al. (2005b)

HL5k GJT001(iclR::KmR) Lin et al. (2005a)

HL51k GJT001(iclR, icd::KmR) Lin et al. (2005a)

HL512k GJT001(iclR, icd, sdhAB::KmR) Lin et al. (2005a)

HL5127k GJT001(iclR, icd, sdhAB, (ackA-pta)::KmR) Lin et al. (2005a)

HL51276k GJT001(iclR, icd, sdhAB, (ackA-pta), poxB::KmR) Lin et al. (2005a)


(Table 2). Knockouts were created in SDH (sdhAB),pyruvate oxidase (poxB), acetate kinase-phosphotransa-cetylase (ackA-pta), isocitrate dehydrogenase (icd),aceBAK operon repressor (iclR), and the glucosephosphotransferase system (ptsG). Each one of thesemutations is designated by a number (1-icd, 2-sdhAB,5-iclR, 6-poxB, 7-(ackA-pta), 9-ptsG) used in thenaming of the mutant strains (Fig. 1). The kanamycincassette was left in the final mutant strains to provideselective pressure during the fermentation experiments.The mutant strains selected for study in chemostatwere HL27659k, HL2765k, HL276k, HL2761k, andHL51276k. Strain HL276k is the background controlstrain for all the other strains.

2.2. Chemostat culture medium and conditions

The medium used was Luria-Bertani broth (LB), 10 g/Ltryptone, 5 g/L yeast extract, and 10 g/L NaCl (Sam-brook et al., 1989), added with 2 g/L NaHCO3 and20 g/L of glucose. The medium used for inoculumpreparation was also the same medium, except glucosewas not added. NaHCO3 was added to the culturemedium because it promoted cell growth and succinateproduction due to its pH-buffering capacity and itsability to supply CO2. CO2 is important for succinateproduction because it is required for carboxylation byPEPC for conversion of PEP to OAA. Kanamycin wasadded to the medium at a concentration of 50mg/L.

Chemostat experiments were performed under aero-bic conditions at a dilution rate of 0.1/h. This dilutionrate was selected based on specific growth rates of thefive mutant strains, obtained from log phase growthdata of previous batch culture studies. The workingvolume was maintained at 600ml in a 1.0-L NewBrunswick Scientific Bioflo 110 fermenter. The pH wasmeasured using a glass electrode and controlled at 7.0using 1.5N HNO3 and 2N Na2CO3. The temperaturewas maintained at 37 1C, and the agitation speed

was constant at 500 rpm. The inlet airflow used was0.6 L/min. The dissolved oxygen was monitored using apolarographic oxygen electrode (New Brunswick Scien-tific) and was maintained above 50% saturationthroughout the experiment. A 1% (v/v) inoculum froman overnight culture grown from a single colony for 12 hwas used to inoculate the bioreactor. The culture wasallowed to grow in batch mode for 12–14 h before thefeed pump and waste pump were turned on to start thechemostat. The continuous culture reached steady stateafter five residence times. Optical density (OD) andmetabolites were measured from samples at five and sixresidence times and then compared to ensure that steadystate had been established. Actual samples used forobtaining the results presented were taken after sixresidence times.

2.3. Analytical techniques

OD was measured at 600 nm with a spectrophot-ometer (Bausch & Lomb Spectronic 1001); the culturewas diluted to the linear range with 0.15M NaCl and thedilution factor was accounted in the calculation of OD.The measured OD was correlated with dry cell weightusing established proportional constants. The propor-tional constants were obtained by washing and resus-pending samples of cell with 0.15M NaCl. Thesuspension was then dried until constant weight at55 1C and corrected for the weight of NaCl in the wash.The correlation between OD and dry weight was thendetermined. For analyzing the extracellular metabolites,1ml of culture was centrifuged at 8000g for 5min andthe supernatant was then filtered through a 0.2 mmPVDF membrane syringe filter and stored frozen at�20 1C until analysis.

For analyzing intracellular metabolites includingacetyl-CoA, 40 OD units (OD660nm� vol (ml) ¼ 40) ofcell culture were taken into a precooled centrifuge tube,immediately chilled on ice, and centrifuged at 5000g at


4 1C for 10min. The method used for preparingintracellular metabolite supernatant is a modifiedmethod of Nakano et al. (1998). The pellet was washedonce with 4 1C 0.1M Tris-HCl (pH 8.0) buffer andresuspended in 1ml of 4 1C 6% perchloric acid to lysethe cells and then placed on an ice bath for 10min. Then0.3ml of 4 1C 3M potassium carbonate was added whilevortexing to neutralize the acid. The solution was thencentrifuged again at 4 1C for 10min and the supernatantwas filtered through a 0.2 mm PVDF membrane syringefilter. The filtered supernatant was then stored at �80 1Cuntil analysis. Approximate intracellular concentrationsrelative to the cell volume were calculated based on thecell volume constant of 2.7 ml/mg dry weight (Winklerand Wilson, 1966).

The HPLC system (Shimadzu-10A Systems, Shimad-zu, Columbia, MD) used was equipped with a cation-exchange column (HPX-87H, BioRad Labs, Hercules,CA), a UV detector (Shimadzu SPD-10A) and adifferential refractive index (RI) detector (Waters 2410,Waters, Milford, MA). A 0.6ml/min mobile phase using2.5mM H2SO4 solution was applied to the column. Thecolumn was operated at 55 1C. Standards were preparedfor glucose, succinate, acetate, pyruvate, and glyoxylatefor both the RI detector and UV detector, andcalibration curves were created. Glucose, succinate,acetate, and glyoxylate were measured by the RIdetector and pyruvate was measured by the UV detectorat 210 nm. Citrate, isocitrate, and malate were measuredby enzyme assay kits obtained from R-Biopharm. Thedetection limit by the enzyme assay kit for citrate is0.5mg/L, for isocitrate is 1mg/L, and for malate is0.5mg/L. All samples were measured three times foreach metabolite by either HPLC or enzyme assay.

The quantification of acetyl-CoA was based on amodified protocol of Boynton (Boynton et al., 1994) andused previously by others (Vadali et al., 2004; Lin et al.,2004). Acetyl-CoA was analyzed by HPLC (Thermo-finnigan, San Jose, CA) using a UV detector set at254 nm. The column used was a 5 mm octyldecyl silanecolumn (Cell Technologies, Inc., Houston, TX) pre-ceded by an Allsphere ODS-2 (C18) guard column(Alltech, Deerfield, IL). The column was operated atroom temperature. Two mobile phases of buffer wereused at a flow rate of 1ml/min. One buffer was 0.2Msodium phosphate (pH 5.0) and the other buffer was800ml of 0.25M sodium phosphate (pH 5.0) mixed with200ml of 100% acetonitrile). The run profile wasadopted from Boynton (Boynton et al., 1994). Allsamples were measured three times for acetyl-CoA.

2.4. Enzyme assays

Crude extracts for all enzyme assays were prepared bytaking 20 OD units of culture (OD660 nm� vol (ml) ¼ 20)and centrifuging the appropriate volume at 5000g and

4 1C for 20min. The cell pellet was then washed once in15ml of the appropriate buffer for each type of enzymeassay. The pellet was then centrifuged again andresuspended in 10ml of that buffer. The cells were thensubjected to sonication for 10min in an ice bath. Thesonicated cells were centrifuged at 5000g and 4 1C for60min to remove cell debris. The supernatant was thenused for the enzyme assay. Total protein concentrationof the crude extract was measured by Lowry’s method(Sigma Lowry Reagent, Modified) using bovine serumalbumin as standard. All enzyme activities wereexpressed in units of U/mg, which is the micromole ofsubstrate converted to product per minute time and mgprotein. All enzyme assays for each sample wereperformed in triplicate.

Citrate synthase (CS) activity was measured by amodified method of Aoshima et al. (2003). CS producescitrate and CoA from the condensation of acetyl-CoAand OAA. DTNB, 5,5-dithiobis(2-nitrobenzoic acid),was added to react with the sulfhydryl group of theCoA, which could be measured at 412 nm. The molarextinction coefficient used was 13.6/mM/cm. The rate ofincrease in absorbance was used to calculate CS activity.

Malate dehydrogenase (MDH) was measured by amodified method of Zeikus et al. (1977). MDH reducesOAA to malate through the oxidation of NADH toNAD+. This reaction was measured at 340 nm and themolar extinction coefficient used was 6.22/mM/cm. Therate of decrease in absorbance was used to calculateMDH activity.

Phosphoenolpyruvate carboxylase was measured by amodified method of Terada et al. (1991). This assay is acoupled enzyme assay with MDH. Phosphoenolpyru-vate carboxylase converts PEP to OAA through acarboxylation reaction. MDH then reduces OAA tomalate with the oxidation of NADH. The rate ofdecrease in absorbance was used to calculate phosphoe-nolpyruvate carboxylase activity.

Isocitrate dehydrogenase was measured by a modifiedmethod of Soundar et al. (1996). Isocitrate dehydrogen-ase oxidizes isocitrate to 2-ketoglutarate with a con-comitant release of CO2 and reduction of NADP+ toNADPH. The reaction was measured at 340 nm and themolar extinction coefficient used was 6.22/mM/cm. Therate of increase in absorbance was used to calculateisocitrate dehydrogenase activity.

MS was measured by the method of de Jong-Gubbelset al. (1995). The assay is a coupled enzyme assay inwhich the malate formed by MS is oxidized to OAA byMDH with the concomitant reduction of NAD+ toNADH. The reaction was measured at 340 nm and themolar extinction coefficient used was 6.22/mM/cm. Therate of increase in absorbance was used to calculate MSactivity.

ICL was measured by a modified method of Dixonand Kornberg (1959). ICL catalyzes the hydrolysis of

ARTICLE IN PRESS

Table 3

List of genes studied and primers designed for quantitative PCR

Genes Primers

AceA Forward: 50-ATCTGATCACCTCCGATTGC-30

Reverse: 50-CACCAGACCAGGTCAGCATA-30

AceB Forward: 50-TTCTGACTGAGCTGGTGACG-30

Reverse: 50-GCGAATTTTCCAATCAGCAT-30

AceE Forward: 50-ACGTACCGGCTGACGACTAC-30

Reverse: 50-CTTATCGATTTCGCCACGTT-30

AceF Forward: 50-AAGTGACCGAAATCCTGGTG-30

Reverse: 50-CACTTTGTCACCCACGTTCA-30

gltA Forward: 50-ATGATTCTTTCCGCCTGATG-30

Reverse: 50-TTCCAGCTCCATAGCCACTT-30

icd Forward: 50-TCCGTGAAAACTCGGAAGAC-30

Reverse: 50-TCCGAACACGGCTTAATACC-30

mdh Forward: 50-GGCGTTAGTTTTACCGAGCA-30

Reverse: 50-GTGCACGAACCAGAGACAGA-30

ppc Forward: 50-GGTCCGTTTTACTTCGTGGA-30

Reverse: 50-CGACAGTTCAGAAACCAGCA-30

ptsG Forward: 50-AGGTCGGTAAATCGCTGATG-30

Reverse: 50-ATGTTTGCAAAGACGGAACC-30

pykA Forward: 50-ACTGACGCTGTGATGCTGTC-30

Reverse: 50-CCACATTGTCGAACTGAACG-30

pykF Forward: 50-TATCCGTGCACGTAAAGTCG-30

Reverse: 50-TTACCTTTTGCGGATTCACC-30

rrsA Forward: 50-CTGTCGTCAGCTCGTGTTGT-30

Reverse: 50-AAGGGCCATGATGACTTGAC-30


isocitrate into glyoxylate and succinate. The glyoxylateformed in the presence of phenylhydrazine was mea-sured as glyoxylic acid phenylhydrazone at 324 nm. Themolar extinction coefficient used was 17/mM/cm. Therate of increase in absorbance in the linear range wasused to calculate ICL activity.

2.5. Gene expression analysis

2.5.1. RNA preparation

Total RNA was isolated with the Promega SV TotalRNA Isolation System (Promega Corporation, Madi-son, WI) according to the manufacturer’s protocol. Theisolated RNA was then treated again with DNase(Promega) and RNase Inhibitor (Promega). The reac-tion was incubated at 37 1C for 30min. The RNA wasthen extracted with phenol once, phenol/chloroform(50:50) once, and then chloroform twice. The RNA wasthen precipitated with ethanol and resuspended inRNase-free H2O (treated with diethyl pyrocarbonate(DEPC)). The RNase-free H2O (DEPC treated) wasprepared by adding DEPC (1 ml/ml) to distilled water,then storing it overnight at 37 1C, and then autoclavingit. The concentration of RNA was quantified bymeasuring the absorbance at 260 nm and applying theformula, concentration (mg/ml) ¼ A260� 40� dilutionfactor. The purity of the RNA was determined byreading the absorbance at 260 and 280 nm. RNAsamples used were ensured to have the A260/A280 ratiobetween 1.8 and 2.1 for clean RNA.

2.5.2. cDNA synthesis and quantitative real-time PCR

Quantitative real-time RT-PCR was performed in atwo-step process. The first step is the cDNA synthesis,and the second step is the quantitative real-time PCRusing the cDNA synthesized. cDNA was synthesized byreverse transcription using the RNA prepared astemplate. This was performed by using the PromegaReverse Transcription System (Promega, Madison, WI)and the reaction was carried out in a RoboCyclerGradient 96 (Stratagene, La Jolla, CA). The cDNA wassynthesized in a total reaction mixture volume of 60 mlcontaining 1 mg of RNA template. The reaction mixturewas incubated for 10min at room temperature forprimer extension, 30min at 50 1C for reverse transcrip-tion, and then 5min at 95 1C and 10min at 6 1C forinactivation of the reverse transcriptase. Control sam-ples of the cDNA were also prepared following the sameprotocol, except reverse transcriptase was not added tothe reaction mixture. The single-stranded cDNA samplewas then diluted into the working range with nucleasefree H2O and stored at �20 1C until further use.

Quantitative real-time PCR was performed in a ABIPrism 7000 Sequence Detection System (Applied Bio-systems, Foster City, CA) using the SYBR Green PCRMaster Mix (Applied Biosystems, Foster City, CA).

Direct detection of PCR product is monitored bymeasuring the increase in fluorescence caused by thebinding of SYBR Green dye to double-stranded DNA(SYBR Green PCR Master Mix and RT-PCR Protocol,Applied Biosystems). Forward and reverse primers weredesigned for each gene studied. The list of genes studiedand primers designed is shown in Table 3. Reactionswere carried out in a 96-well plate using the cDNAprepared as template. Each well contains a reactionmixture consisting of cDNA, forward and reverseprimers, SYBR Green PCR Master Mix, and nucleasefree water. PCR reaction was performed three timesunder identical reaction conditions for every gene usinga particular strain’s cDNA in order to calculate astandard deviation; reaction using the control cDNAsample of that strain was also performed concurrently toensure that there was no contamination.

Following the manufacturer’s protocol, the compara-tive CT method for relative quantification of geneexpression was used (ABI Prism 7700 Sequence Detec-tion System User Bulletin #2, Applied Biosystems). Thethreshold cycle, CT, was the data obtained. The rrsA

gene encoding rrnA 16S ribosomal RNA was used as anendogenous control in order to standardize the amount


of sample DNA added to a reaction. This gene is notsubjected to variable expression because its expression isabundant and relatively constant in most cells. Inaddition, since ribosome level varies with cell growthrate, the chemostat culture of m ¼ 0:1=h should maintaina steady level of this expression between strains. Thedifference between the CT of the studied gene and therrsA gene was calculated (DCT) and a sensitivity test toconfirm that the DCT was independent of the RNAconcentration was performed for each gene. The DDCT

was then calculated by taking the difference between theDCT of a gene in one strain and the DCT of the samegene in the control strain. The relative expression of aparticular gene between a strain and its control strain isgiven by the formula 2�DDCT . Using this method, it isimportant to emphasize that the expression levels ofdifferent genes cannot be compared. Expression levelscan only be compared between different strains for thesame gene, since the primers in those reactions would bethe same.

3. Results and discussion

3.1. Comparison of metabolite profiles

The production of succinate, pyruvate, and acetateunder aerobic conditions was compared between the

0

20

40

60

80

100

120

140

160

180

0

20

40

60

80

100

120

(a)

(c)

Ext

race

llula

r p

yru

vate

(m

M)

HL27659k HL2765k HL276k HL2761k HL51276k


Glu

cose

co

nsu

med

(m

M)

Fig. 4. Glucose consumed and extracellular metabolite production by s

five mutant strains HL27659k, HL2765k, HL276k,HL2761k, and HL51276k in chemostat cultures at0.1/h dilution rate. Succinate production is the focusof this study, since it is the valuable product that E. coli

was genetically engineered to produce under aerobicconditions. The biomass concentrations of the fivemutant strains were HL27659k (3.4 g/L), HL2765k(3.6 g/L), HL276k (3.4 g/L), HL2761k (2.4 g/L) andHL51276k (2.6 g/L). Substantial levels of succinate wereobtained for strains HL27659k, HL2765k, and HL276k(Fig. 4b). Strain HL27659k produced 57mM ofsuccinate, strain HL2765k produced 61mM, andstrain HL276k produced 58mM. Succinate productionwas significantly lower for strains HL2761k andHL51276k than the other three strains (Fig. 4b). StrainHL2761k only produced 2mM of succinate andstrain HL51276k produced only 3mM of succinate.Strain HL27659k achieved the highest succinate yieldamong all the strains. It obtained 0.91 mole succinateper mole glucose (Fig. 5a). This is 91% of the maximumtheoretical succinate yield, which is 1.0mol/mol glucoseunder aerobic conditions. Strains HL2765k, HL276k,HL2761k, and HL51276k reached succinate yields of0.74, 0.73, 0.02 and 0.03, respectively (Fig. 5a).

The succinate yields of strains HL27659k andHL51276k in chemostat cultures are different from thatof batch reactor cultures. In chemostat culture, strainHL27659k achieved significantly higher succinate yield

0

10

20

30

40

50

60

70

0

2

4

6

8

10

12

14

16

18

(b)

(d)

HL27659k

14

16

Ext

race

llula

r su

ccin

ate

(mM

)E

xtra

cellu

lar

acet

ate

(mM

)

HL2765k HL276k HL2761k HL51276k


trains HL27659k, HL2765k, HL276k, HL2761k, and HL51276k.

ARTICLE IN PRESS

00.10.20.30.40.50.6

0.91

0

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

0.80.7

Su

ccin

ate

yiel

d (

mo

l/mo

l)A

ceta

te y

ield

(m

ol/m

ol)

Pyr

uva

te y

ield

(m

ol/m

ol)

(a) HL27659k HL2765k HL276k HL2761k HL51276k

(b) HL27659k HL2765k HL276k HL2761k HL51276k

(c) HL27659k HL2765k HL276k HL2761k HL51276k

0.16

0.14

0.12

0.1

0.08

0.06

0.04

0.02

Fig. 5. Succinate, acetate, and pyruvate yields of strains HL27659k,

HL2765k, HL276k, HL2761k, and HL51276k. Yield is mole of

product produced per mole of glucose consumed.


than in batch culture (0.91mol/mol yield compared to0.78mol/mol yield (Lin et al., 2005b)). Strain HL51276kachieved significantly lower succinate yield in chemostatculture than in batch culture (0.03mol/mol yieldcompared to 0.65mol/mol yield (Lin et al., 2005b)). Apossible explanation why strain HL51276k did notobtain similar succinate yield in batch culture andchemostat culture is that under chemostat conditions,strain HL51276k is not capable of complete glucosemetabolism resulting in excretion of pyruvate. In thisstrain, the glucose uptake rate is likely faster than therates of the glyoxylate cycle pathways resulting inimbalanced glucose metabolism. Instead, a majority ofthe metabolized glucose accumulated as pyruvate forstrain HL51276k. Strain HL51276k produced 151mMof pyruvate with a yield of 1.38mol/mol glucose(Figs. 4c and 5c). The acetate produced was 14mMwith a yield of 0.13mol/mol glucose; this is higher thanany of the other four strains (Figs. 4d and 5b). StrainHL2761k also exhibited the same metabolite productioncharacteristics as strain HL51276k. It produced 146mMof pyruvate, yield of 1.35mol/mol glucose, and 12mM

of acetate, yield of 0.11mol/mol glucose (Figs. 4c, d, 5b, c).Both strains HL51276k and HL2761k produce pyruvateas their major product. The similarity can be explainedby the icd inactivation in both strains, rendering theTCA cycle inactive. Inactivation of icd severely ham-pered succinate production and caused pyruvate tobecome the major product in chemostat cultures.

Strains HL2765k and HL276k produced 67 and74mM of pyruvate, respectively, at yields of 0.80mol/mol glucose and 0.93mol/mol glucose, respectively(Figs. 4c and 5c). Strains HL2765k and HL276k,precursors to strain HL27659k, accumulated significantamounts of pyruvate, which decreased their succinateyields. Strain HL27659k did not produce any pyruvate(Fig. 4c) and its acetate production was the lowestamong all the strains (Fig. 4d). Succinate was the onlymajor product of strain HL27659k. This demonstratesthe efficiency of the pathway design in strain HL27659kin allowing a majority of the carbon flux to be channeledtoward the succinate pathways.

The inactivation of ptsG was pivotal in improvingcarbon-throughput and succinate yield in strainHL27659k in chemostat cultures. Once ptsG wasdisrupted in strain HL27659k, pyruvate accumulationwas eliminated. Precursor strain HL2765k produced67mM of pyruvate and strain HL27659k produced nopyruvate (Fig. 4c). Succinate yield also increased from0.73 to 0.91mol/mol glucose, allowing strain HL27659kto obtain the highest succinate yield among the fivemutant strains (Fig. 5a). There was also minimal acetateproduction, and no pyruvate accumulation. The inacti-vation of ptsG did decrease the amount of glucoseconsumed as shown by Fig. 4a. Strain HL27659kconsumed 64mM of glucose where its precursor strainHL2765k consumed 85mM of glucose. Presumably, thelower glucose consumption allowed more balancedglucose metabolism and more efficient carbon-through-put. It is interesting to find that strains HL2761k andHL51276k consumed the most amount of glucose (110and 111mM, respectively). Since these two strains haveinactivation in icd that inhibits carbon flux through theoxidative pathways of the TCA cycle, no substantialenergy can be generated via the TCA cycle. It is possiblethat strains HL2761k and HL51276k are consumingmore glucose to use glycolysis for generating ATP.

Citrate, isocitrate, glyoxylate and malate were alsomeasured in the cultures of the five mutant strains, but,interestingly, none of these metabolites were detected.Strains HL51276k and HL2761k were previouslyobserved to accumulate citrate and isocitrate whengrown in the batch reactor mode (Lin et al., 2005b).Apparently, growth of these two strains in chemostat at0.1/h dilution rate caused pyruvate to be accumulated,instead, as the major product. At this dilution rate, theTCA cycle of strains HL51276k and HL2761k does notcontribute to glucose metabolism.

ARTICLE IN PRESS

0

0

0.5

1

1.5

2

2.5

3

3.5

4

4.5

5

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8

2

0

1

2

3

4

5

6

7 0.7

0.6

0.5

0.4

0.3

0.2

0.1

(b)(a)

(d)(c)

HL27659k HL2765k HL276k HL2761k HL51276kHL27659k HL2765k HL276k HL2761k HL51276k

HL27659k HL2765k HL276k HL2761k HL51276k HL27659k HL2765k HL276k HL2761k HL51276k

Sp

ecif

ic g

luco

se c

on

sum

pti

on

(m

mo

l/g/h

r)

Sp

ecif

ic s

ucc

inat

e p

rod

uct

ivit

y(m

mo

l/g/h

r)S

pec

ific

ace

tate

pro

du

ctiv

ity

(mm

ol/g

/hr)

Sp

ecif

ic p

yru

vate

pro

du

ctiv

ity

(mm

ol/g

/hr)

Fig. 6. Specific glucose consumption rate and specific metabolite productivities of strains HL27659k, HL2765k, HL276k, HL2761k, and HL51276k.

Specific rate is mmol of metabolite per gram of biomass per hour time.


The results of specific glucose consumption rate andspecific productivities of succinate, pyruvate, andacetate correspond well with the results of the metabo-lite profiles and yields. Strain HL27659k has the lowestspecific glucose consumption rate among all five mutantstrains, since it is inactivated in ptsG (Fig. 6a). Thespecific succinate productivity of strain HL27659k wasequivalent to the specific succinate productivities ofstrains HL2765k and HL276k (Fig. 6b), even thoughstrain HL27659k obtained a higher succinate yield thanstrains HL2765k and HL276k. Specific pyruvate andacetate productivities follow the same trend as pyruvateand acetate production for all five mutant strains(Figs. 6c and d).

Fig. 7 shows the results of intracellular meta-bolite molar concentrations in the five mutant strainsHL27659k, HL2765k, HL276k, HL2761k, andHL51276k. All intracellular metabolite concentrationsare higher than their respective extracellular metaboliteconcentrations, associated with concentration gradientsrequired for transport. Strains HL27659k, HL2765k,and HL276k retained higher intracellular succinateconcentrations than strains HL2761k and HL51276k(Fig. 7a). This is because substantially more succinatewas produced and excreted by strains HL27659k,HL2765k, and HL276k than strains HL2761k and

HL51276k (Fig. 4b). Perhaps, engineering an activesuccinate transport system in these mutant strains couldincrease the extracellular level of succinate and reducethe amount accumulated intracellularly. Intracellularpyruvate concentrations were the highest in strainsHL2761k and HL51276k, in accordance with theirlarge accumulation of extracellular pyruvate (Figs. 7band 4c). Strain HL27659k not only did not excrete anypyruvate, but also there was no detected intracellularpyruvate accumulation (Figs. 4c and 7b). Intracellularacetate concentrations follow the same trend as extra-cellular acetate concentrations for all mutant strains,except strain HL27659k (Figs. 7c and 4d). StrainHL27659k did not accumulate any intracellular acetate.The intracellular acetyl-CoA concentration results didnot present any significant comparisons between thefive mutant strains (Fig. 7d). The levels of intra-cellular acetyl-CoA concentration were similar for allfive strains.

3.2. Comparison of enzyme activities

Crucial enzyme activities of TCA cycle pathwaysinvolved in the design of the aerobic succinate produc-tion system were measured and examined for the mutantstrains HL27659k, HL2765k, HL276k, HL2761k, and

ARTICLE IN PRESS

0

IIntr

acel

lula

r su

ccin

ate

(mM

)In

trac

ellu

lar

acet

ate

(mM

)

0

0

20

40

60

80

100

120

140

160

0

10

20

30

40

50

60

0

50

100

150

200

250

300

350

400

0

100

200

300

400

500

600

Intr

acel

lula

r p

yru

vate

(m

M)

Intr

acel

lula

r ac

etyl

-Co

A (

µM

)

(b)(a)

(d)(c)





Fig. 7. Intracellular metabolite concentrations of strains HL27659k, HL2765k, HL276k, HL2761k, and HL51276k. Intracellular metabolite

concentrations are relative to cell volume.


HL51276k. Examining the enzyme activities can providea better understanding of how each mutant strain’smetabolite profile is correlated with its metabolicpathways. The following pathways relevant to succinatesynthesis through the TCA cycle were examined: CS,isocitrate dehydrogenase (ICDH), ICL, MS, MDH, andphosphoenolpyruvate carboxylase (PPC).

Strain HL27659k, the most efficient and highest yieldsuccinate producing mutant strain, has significantlymore CS activity than the other four mutant strains(Fig. 8a). Strain HL27659k has approximately five-foldhigher CS activity than strains HL2765k and HL276k.This result shows that inactivation of ptsG may besignificantly related to the increase in CS activity instrain HL27659k. The high CS activity can explain whysuccinate is so efficiently produced by strain HL27659kand there is no pyruvate accumulation in strainHL27659k. CS drives carbon flux toward the succinatesynthesis pathways of the TCA cycle thus reduces theamount of carbon accumulated at the pyruvate node(Fig. 2). Strains HL2765k and HL276k, although theyaccumulate pyruvate, can produce substantially moresuccinate than strains HL2761k and HL51276k. Thiscan also be attributed to the CS activities of strainsHL2765k and HL276k being approximately twice thatof strains HL2761k and HL51276k. Inactivation of icd

seems to result in a decrease in CS activity as shown by

the lower CS activity of strain HL2761k compared tostrain HL276k.

No ICDH activities were detected in strains HL2761kand HL51276k, as expected, since the icd was disruptedin both strains (Fig. 8b). Strains HL2765k and HL276kboth showed similar ICDH activities, which were two-fold higher than that of strain HL27659k. This impliesthat the oxidative branch of the TCA cycle is moreactive in strains HL2765k and HL276k than in strainHL27659k. The inactivation of ptsG may be related tothe decrease in ICDH activity in strain HL27659k.

Strains HL27659k, HL2765k, and HL51276k all haveICL and MS activities because iclR was inactivated inthese strains (Figs. 8c and 8d). This shows that thedisruption of iclR induces expression of the aceA andaceB for ICL and MS, respectively, when the culture isgrown on glucose. This was also shown by Gui et al.(1996). Strains HL276k and HL2761k, which do nothave the iclR mutation, show no ICL and MS activities.

Strain HL27659k has the highest MDH activityamong all five mutant strains (Fig. 8e). This may inferthat strain HL27659k has a more efficient glyoxylatecycle for succinate production than the other twomutant strains HL2765k and HL51276k with glyoxylatecycle activity. This is because MDH is required tooxidize the malate produced from MS back to OAA(Fig. 1). The OAA can then be utilized by CS to

ARTICLE IN PRESS

00

0.000

2.000

4.000

6.000

8.000

10.000

12.000

14.000

16.000

18.000

0

0.002

0.004

0.006

0.008

0.01

0.012

0.014

0.016

0.018

0.02

0

0.01

0.02

0.03

0.04

0.05

0.06

0.09

0.08

0.07

0.06

0.05

0.04

0.03

0.02

0.010.1

0.2

0.3

0.4

0.5

0.6

0

0.1

0.2

0.3

0.4

0.5

0.8

0.6

0.7

0.9

0.000

2.000

4.000

6.000

8.000

10.000

12.000

14.000

16.000

18.000

CS

(U

/mg

)

ICD

H (

U/m

g)

MS

(U

/mg

)

ICL

(U

/mg

)

PP

C (

U/m

g)

(b)(a)

(d)(c)





(f )(e) HL27659k HL2765k HL276k HL2761k HL51276k HL27659k HL2765k HL276k HL2761k HL51276k

MD

H (

U/m

g)

Fig. 8. Specific enzyme activites of strains HL27659k, HL2765k, HL276k, HL2761k, and HL51276k. Specific enzyme activities were expressed in U/

mg, which is mmol of substrate converted to product per min per mg protein.


continuously drive carbon flux through the glyoxylatecycle. Considering only strains with glyoxylate cycleactivity, strain HL27659k has approximately twice theMDH activity of strain HL2765k and approximately 11-fold higher MDH activity than strain HL51276k.Inactivation of icd seems significantly related to decreasein MDH activity since strain HL2761k has much lowerMDH activity than strain HL276k.

The PPC activities of strains HL27659k, HL2765k,and HL276k are all significantly higher than strainsHL2761k and HL51276k (Fig. 8f). Strains HL27659k,

HL2765k, and HL276k also produced substantiallymore succinate than strains HL2761k and HL51276k(Fig. 4b). PPC is essential for increasing the OAA pool,therefore increasing succinate production (Millard et al.,1996). The inactivation of icd was observed to cause adecrease in PPC activity, since strain HL2761k haslower PPC activity than strain HL276k. This negativeeffect on PPC activity by the inactivation of icd was alsoobserved for CS and MDH.

Strain HL27659k has higher CS and MDH activitiesthan the other four strains, HL2765k, HL276k, HL2761k,

ARTICLE IN PRESSH. Lin et al. / Metabolic Engineering 7 (2005) 337–352 349

and HL51276k. It also has an active glyoxylate cyclecomplemented by lower ICD activity. In strainHL27659k, the glyoxylate cycle is more efficient forsuccinate production than the TCA cycle oxidativepathways. The enzyme activity results showed that thesuccinate synthesis network is more efficient in strainHL27659k than in the other four strains.

3.3. Comparison of gene expression profiles

The expression of genes involved in the design ofthe aerobic succinate production system was profiledby quantitative real-time RT-PCR for the mutantstrains HL27659k, HL2765k, HL276k, HL2761k, andHL51276k. Gene expression profiles combined withenzyme activities and metabolite production can providea more holistic understanding of the connectionsbetween genes, proteins, and metabolites. For thepurpose of studying the aerobic succinate productionsystem designed, this methodology was narrowed downto a select few genes for expression profiling. The genesexamined were some TCA cycle genes, mdh, gltA, icd,aceA and aceB, and some glycolysis genes, ptsG, ppc,aceE, aceF, pykA, and pykF. The genes mdh, gltA, icd,aceA, aceB, ptsG, and ppc refer to MDH, CS, isocitratedehydrogenase, ICL, MS, glucose phosphotransferasesystem, and phosphoenolpyruvate carboxylase, respec-tively. The genes aceE and aceF are core components ofthe pyruvate dehydrogenase complex. Genes pykA andpykF express isoenzymes of pyruvate kinase. Expressionof a particular gene was compared between the fivemutant strains. The expression of a gene in the strainsHL27659k, HL2765k, HL2761k and HL51276k isrelative to the expression of that gene in the controlstrain HL276k. The results do not allow cross compar-ison of different genes. Also, because the background ofall these mutant strains has mutations in sdhAB, ackA-pta, and poxB, this may influence the gene expressionpatterns observed.

The gltA expression profile shows that strainHL2761k has the highest level of gltA expression amongthe five strains (Fig. 9a). The gene expression result didnot correspond with the enzyme activity profile of CS.Strain HL27659k was shown to exhibit the highest CSactivity whereas strains HL2761k and HL51276kexhibited the lowest CS activities (Fig. 8a). This,possibly, could be attributed to mRNA degradation orprotein degradation or instability. In vitro enzymeassays are also not very reflective of the in vivoenvironment. In the intracellular environment, enzymescould be regulated by different effectors such ascofactors, metal ions and metabolites. This environmentcannot be perfectly mimicked by in vitro assays. ThegltA expression in strain HL27659k is higher than thatof its precursor strain HL2765k and the control strainHL276k. This is the same trend observed for the CS

activity of strain HL27659k compared to strainsHL2765k and HL276k. Fig. 9e shows that strainHL2761k has the highest mdh expression, even thoughit has the lowest MDH activity (Fig. 8e). StrainHL27659k has the highest MDH activity, but not thehighest mdh expression.

The icd expression profile corresponded well with theICDH activities of the five mutant strains. StrainsHL2761k and HL51276k did not have any icd expres-sion because of the icd mutation (Fig. 9b). StrainHL27659k has lower icd expression than strainsHL2765k and HL276k and this result correlates withthe lower ICDH activity of strain HL27659k than thatof strains HL2765k and HL276k. Figs. 9c and d confirmthat there is no expression of the glyoxylate bypass genesaceA and aceB in strains HL276k and HL2761k, sincethere is functional IclR in these strains. StrainsHL27659k, HL2765k, and HL51276k, which have iclR

mutation, are the only strains with aceA and aceB

expressed, and thus possess ICL and MS activities.Strains HL27659k, HL2765k and HL276k showed

higher PPC activities than strains HL2761k andHL51276k, even though their ppc gene expression valuesdid not follow this trend (Figs. 8f and 10b). The ptsG

expression profile shows that strain HL27659k did nothave ptsG expression, and this corresponded well withits ptsG knockout (Fig. 10a). The components ofpyruvate dehydrogenase complex, aceE and aceF, wereexpressed together in the same proportion in the fivemutant strains since both are required to form pyruvatedehydrogenase (Figs. 10c and d). Strain HL27659k hasrelatively lower expression of pyruvate dehydrogenasethan the other strains. The expression profiles of pykA

and pykF, both isoenzymes of pyruvate kinase, did notfollow the same trend as each other for the five mutantstrains (Figs. 10e and f). The dissimilarity in the trend ofthe expression profiles of pykA and pykF may beexplained by the different pyruvate kinase enzymes theyencode. The pykF isoenzyme has been shown to play agreater role than the pykA isoenzyme in the activity ofpyruvate kinase when E. coli is grown on glucose; thus,the pykF is regulated differently than pykA (Ponce et al.,1995). Strain HL27659k has relatively lower expressionof pyruvate kinase than the other strains. Overall, lowerexpression of pyruvate kinase and pyruvate dehydro-genase in strain HL27659k indicates that the glycolyticpathways may be less upregulated in this strain; this maybe possibly due to its more efficient carbon throughputwith the active glyoxylate cycle and TCA oxidativepathways.

4. Conclusion

Mutant strains involved in the developmentof the aerobic succinate production systems were

ARTICLE IN PRESS

0

0.2

0.4

0.6

0.8

1

1.2

1.4

0

0.5

11.5

22.5

33.5

4

4.55

0

0.5

11.5

22.5

33.5

4

4.55

0

1

0.2

0.4

0.6

0.8

1.2

1.4

0

100

200

300

400

500

600

0

200

400

600

800

1000

0

0.5

11.5

22.5

33.5

4

4.55

(a)

(c)

(b)

(d)

(e)





Am

ou

nt

of

glt

A e

xpre

ssio

n r

elat

ive

to t

he

con

tro

l str

ain

HL

276k

Am

ou

nt

of

icd

exp

ress

ion

rel

ativ

e to

th

e co

ntr

ol s

trai

n H

L27

6kA

mo

un

t o

f ac

eB e

xpre

ssio

n r

elat

ive

to t

he

con

tro

l str

ain

HL

276k

Am

ou

nt

of

aceA

exp

ress

ion

rel

ativ

e to

th

e co

ntr

ol s

trai

n H

L27

6k

Am

ou

nt

of

md

h e

xpre

ssio

n r

elat

ive

to t

he

con

tro

l str

ain

HL

276k

Fig. 9. Gene expression profiles (gltA, icd, aceA, aceB, and mdh) of strains HL27659k, HL2765k, HL276k, HL2761k, and HL51276k. Expression

levels of all strains are relative to that of the control strain HL276k. Expression levels cannot be compared between different genes.


characterized. Chemostat results distinctly showed thatstrain HL27659k was the most efficient aerobic succi-nate production strain. At a dilution rate of 0.1/h, thisstrain could reach 91% of the maximum theoreticalsuccinate yield. Strain HL27659k consumed less glucose,which allowed more balanced glucose metabolism andefficient carbon-throughput to the desired end productsuccinate. Strain HL27659k did not accumulate anypyruvate and it produced the least acetate among all themutant strains. Enzyme activity measurements showedthat CS and MDH were substantially more active instrain HL27659k than in any of the other four mutantstrains studied. The results also showed that strainHL27659k did have an active glyoxylate cycle with theinactivation of iclR. Strain HL27659k possessed ICL

and MS activity as a result of the iclR inactivation. Thisalso complemented the gene expression data for aceA

and aceB. Without iclR inactivation, absolutely no aceA

and aceB expression was observed. Strain HL27659kalso possessed lower ICDH activity than its precursorstrains HL2765k and HL276k. This was complementedby the gene expression data, which showed that strainHL27659k had lower icd expression. These resultssuggest that the glyoxylate cycle is more efficient thanthe TCA cycle oxidative pathways for producingsuccinate. These results furthermore imply that strainHL27659k possesses a more efficient metabolism thanthe other mutant strains; allowing it to producesuccinate as its major product with low residual acetateand no residual pyruvate.

ARTICLE IN PRESS

0

0.5

1

1.5

2

2.5

0

1

1.5

0.5

2

0

1

1.5

0.5

2

0

1

1.61.41.2

0.60.8

0.40.2

0

1

1.2

0.6

0.8

0.4

0.2

0

1

1.61.41.2

0.60.8

0.40.2

Am

ou

nt

of

pts

G

exp

ress

ion

rel

ativ

e to

th

e co

ntr

ol s

trai

n H

L27

6k

Am

ou

nt

of

pp

c ex

pre

ssio

n r

elat

ive

to

the

con

tro

l str

ain

HL

276k

Am

ou

nt

of

aceF

exp

ress

ion

rel

ativ

e to

th

e co

ntr

ol s

trai

n H

L27

6k

Am

ou

nt

of

aceE

exp

ress

ion

rel

ativ

e to

th

e co

ntr

ol s

trai

n H

L27

6k

Am

ou

nt

of

pyk

Aex

pre

ssio

n r

elat

ive

to

the

con

tro

l str

ain

HL

276k

Am

ou

nt

of

pyk

Fex

pre

ssio

n r

elat

ive

to

the

con

tro

l str

ain

HL

276k

(a) (b)

(c) (d)

(e) (f )




Fig. 10. Gene expression profiles (ptsG, ppc, aceE, aceF, pykA, and pykF) of strains HL27659k, HL2765k, HL276k, HL2761k, and HL51276k.

Expression levels of all strains are relative to that of the control strain HL276k. Expression levels cannot be compared between different genes.


Acknowledgments

This work was supported by grants from the NationalScience Foundation (BES-0222691, BES-0000303, andBES-0420840).

References

Aoshima, M., Ishii, M., Yamagishi, A., Oshima, T., Igarashi, Y., 2003.

Metabolic characteristics of an isocitrate dehydrogenase defective

derivative of Escherichia coli BL21(DE3). Biotechnol. Bioeng. 84,

732–737.

Boynton, Z.L., Bennett, G.N., Rudolph, F.B., 1994. Intracellular

concentrations of coenzyme A and its derivatives from Clostridium

acetobutylicum ATCC 824 and their roles in enzyme regulation.

Appl. Environ. Microbiol. 60, 39–44.

Chatterjee, R., Millard, C.S., Champion, K., Clark, D.P., Donnelly,

M.I., 2001. Mutation of the ptsG gene results in increased

production of succinate in fermentation of glucose by Escherichia

coli. Appl. Environ. Microbiol. 67, 148–154.

Clark, D.P., 1989. The fermentation pathways of Escherichia coli.

FEMS Microbiol. Rev. 63, 223–234.

de Jong-Gubbels, P., Vanrolleghem, P., Heijnen, S., van Dijken, J.P.,

Pronk, J.T., 1995. Regulation of carbon metabolism in chemostat

cultures of Saccharomyces cerevisiae grown on mixtures of glucose

and ethanol. Yeast 11, 407–418.

Dixon, G.H., Kornberg, H.L., 1959. Assay methods of key enzymes of

the glyoxylate cycle. Proc. Biochem. Soc. 72 3P.

Doelle, H.W., Ewings, K.N., Hollywood, N.W., 1981. Regulation of

glucose metabolism in bacterial systems. Adv. Biochem. Eng./

Biotechnol. 23, 1–36.

Gokarn, R.R., Eiteman, M.A., Altman, E., 1998. Expression of

pyruvate carboxylase enhances succinate production in Escherichia

coli without affecting glucose uptake. Biotechnol. Lett. 20,

795–798.

Gokarn, R.R., Eiteman, M.A., Altman, E., 2000. Metabolic analysis

of Escherichia coli in the presence and absence of the carboxylating

enzymes phosphoenolpyruvate carboxylase and pyruvate carbox-

ylase. Appl. Environ. Microbiol. 66, 1844–1850.

Gokarn, R.R., Evans, J.D., Walker, J.R., Martin, S.A., Eiteman,

M.A., Altman, E., 2001. The physiological effects and metabolic

alterations caused by the expression of Rhizobium etli pyruvate

carboxylase in Escherichia coli. Appl. Microbiol. Biotechnol. 56,

188–195.

Goldberg, I., Lonberg-Holm, K., Bagley, E.A., Stieglitz, B., 1983.

Improved conversion of fumarate to succinate by Escherichia coli

strains amplified for fumarate reductase. Appl. Environ. Microbiol.

45, 1838–1847.

Gui, L., Sunnarborg, A., Pan, B., LaPorte, D.C., 1996. Autoregulation

of iclR, the gene encoding the repressor of the glyoxylate bypass

operon. J. Bacteriol. 178, 321–324.

Hong, S.H., Lee, S.Y., 2001. Metabolic flux analysis for succinic acid

production by recombinant Escherichia coli with amplified malic

enzyme acitvity. Biotechnol. Bioeng. 74 (2), 89–95.

Lee, P.C., Lee, W.G., Kwon, S., Lee, S.Y., Chang, H.N., 2000. Batch

and continuous cultivation of Anaerobiospirillum succiniciproducens

for the production of succinic acid from whey. Appl. Microbiol.

Biotechnol. 54, 23–27.


Lee, P.C., Lee, W.G., Lee, S.Y., Chang, H.N., 2001. Succinic acid

production with reduced by-product formation in the fermentation

of Anaerobiospirillum succiniciproducens using glycerol as a carbon

source. Biotechnol. Bioeng. 72, 41–48.

Lee, P.C., Lee, S.Y., Hong, S.H., Chang, H.N., 2002. Isolation and

characterization of a new succinic acid-producing bacterium,

Mannheimia succiniciproducens MBEL55E, from bovine rumen.

Appl. Microbiol. Biotechnol. 58, 663–668.

Lee, P.C., Lee, S.Y., Hong, S.H., Chang, H.N., 2003a. Batch and

continuous cultures of Mannheimia succiniciproducens MBEL55E

for the production of succinic acid from whey and corn steep

liquor. Bioprocess Biosyst. Eng. 26, 63–67.

Lee, P.C., Lee, S.Y., Hong, S.H., Chang, H.N., Park, S.C., 2003b.

Biological conversion of wood hydrolysate to succinic acid

by Anaerobiospirillum succiniciproducens. Biotechnol. Lett. 25,

111–114.

Lin, H., Vadali, R.V., Bennett, G.N., San, K.-Y., 2004. Increasing the

acetyl-CoA pool in the presence of overexpressed phosphoenolpyr-

uvate carboxylase or pyruvate carboxylase enhances succinate

production in Escherichia coli. Biotechnol. Prog. 20, 1599–1604.

Lin, H., Bennett, G.N., San, K.-Y., 2005a. Genetic reconstruction of

the aerobic central metabolism in Escherichia coli for the absolute

aerobic production of succinate. Biotechnol. Bioeng. 89, 148–156.

Lin, H., Bennett, G.N., San, K.-Y., 2005b. Metabolic engineering of

aerobic succinate production systems in Escherichia coli to improve

process productivity and achieve the maximum theoretical

succinate yield. Metab. Eng. 7, 116–127.

Lin, H., Bennett, G.N., San, K.-Y., 2005c. Fed batch culture of a

metabolically engineered Escherichia coli strain designed for high

level succinate production and yield under aerobic conditions.

Biotechnol. Bioeng. 90, 775–779.

Millard, C.S., Chao, Y., Liao, J.C., Donnelly, M.I., 1996. Enhanced

production of succinic acid by overexpression of phosphoenolpyr-

uvate carboxylase in Escherichia coli. Appl. Environ. Microbiol. 62,

1808–1810.

Nakano, M.M., Zuber, P., Sonenshein, A.L., 1998. Anaerobic

regulation of Bacillus subtilis Krebs cycle genes. J. Bacteriol. 180,

3304–3311.

Ponce, E., Flores, N., Martinez, A., Valle, F., Bolivar, F., 1995.

Cloning of the two pyruvate kinase isoenzymes structural genes

from Escherichia coli: the relative roles of these enzymes in

pyruvate biosynthesis. J. Bacteriol. 177, 5719–5722.

Sambrook, J., Fritsch, E.F., Maniatis, T., 1989. Molecular cloning: a

laboratory manual, second ed. Cold Spring Harbor Laboratory

Press, Cold Spring Harbor, New York.

Samuelov, N.S., Datta, R., Jain, M.K., Zeikus, J.G., 1999. Whey

fermentation by Anaerobiospirillum succiniciproducens for produc-

tion of a succinate-based animal feed additive. Appl. Environ.

Microbiol. 65, 2260–2263.

Sanchez, A.M., Bennett, G.N., San, K.-Y., 2005a. Efficient succinate

production from glucose through overexpression of pyruvate

carboxylase in an Escherichia coli alcohol dehydrogenase and

lactate dehydrogenase mutant. Biotechnol. Prog. 21, 358–365.

Sanchez, A.M., Bennett, G.N., San, K.-Y., 2005b. Novel pathway

engineering design of the anaerobic central metabolic pathway in

Escherichia coli to increase succinate yield and productivity. Metab.

Eng. 7, 229–239.

Schilling, L.B., 1995. Chemicals from alternative feedstocks in the

United States. FEMS Microbiol. Rev. 16, 101–110.

Soundar, S., Jennings, G.T., McAlister-Henn, L., Colman, R.F.,

1996. Expression of pig heart mitochondrial NADP-dependent

isocitrate dehydrogenase in Escherichia coli. Protein Expr. Purif. 8,

305–312.

Stols, L., Donnelly, M.I., 1997. Production of succinic acid through

overexpression of NAD+-dependent malic enzyme in an Escher-

ichia coli mutant. Appl. Environ. Microbiol. 63, 2695–2701.

Terada, K., Murata, T., Izui, K., 1991. Site-directed mutagenesis of

phosphoenolpyruvate carboxylase from Escherichia coli: the role of

His579 in the catalytic and regulatory functions. J. Biochem. 109,

49–54.

Tolentino, G.J., Meng, S.-Y., Bennett, G.N., San, K.-Y., 1992. A pH-

regulated promoter for the expression of recombinant proteins in

Escherichia coli. Biotechnol. Lett. 14, 157–162.

Vadali, R.V., Bennett, G.N., San, K.-Y., 2004. Cofactor engineering of

intracellular CoA/acetyl-CoA and its effect on metabolic flux

redistribution in Escherichia coli. Metab. Eng. 6, 133–139.

Van der Werf, M.J., Guettler, M.V., Jain, M.K., Zeikus, J.G., 1997.

Environmental and physiological factors affecting the succinate

product ratio during carbohydrate fermentation by Actinobacillus

sp. 130Z. Arch. Microbiol. 167, 332–342.

Vemuri, G.N., Eiteman, M.A., Altman, E., 2002a. Effects of growth

mode and pyruvate carboxylase on succinic acid production by

metabolically engineered strains of Escherichia coli. Appl. Environ.

Microbiol. 68, 1715–1727.

Vemuri, G.N., Eiteman, M.A., Altman, E., 2002b. Succinate produc-

tion in dual-phase Escherichia coli fermentations depends on the

time of transition from aerobic to anaerobic conditions. J. Ind.

Microbiol. Biotechnol. 28, 325–332.

Wang, X., Gong, C.S., Tsao, G.T., 1998. Bioconversion of fumaric

acid to succinic acid by recombinant Escherichia coli. Appl.

Biochem. Biotechnol. 70–72, 919–928.

Winkler, H.H., Wilson, T.H., 1966. The role of energy coupling in the

transport of b-galactosides by Escherichia coli. J. Biol. Chem. 241,

2200–2211.

Yang, Y.-T., San, K.-Y., Bennett, G.N., 1999. Redistribution

of metabolic fluxes in Escherichia coli with fermentative lactate

dehydrogenase overexpression and deletion. Metab. Eng. 1,

141–152.

Zeikus, J.G., 1980. Chemical and fuel production by anaerobic

bacteria. Annu. Rev. Microbiol. 34, 423–464.

Zeikus, J.G., Fuchs, G., Kenealy, W., Thauer, R.K., 1977. Oxidor-

eductases involved in cell carbon synthesis of Methanobacterium

thermoautotrophicum. J. Bacteriol. 132, 604–613.

Zeikus, J.G., Jain, M.K., Elankovan, P., 1999. Biotechnology of

succinic acid production and markets for derived industrial

products. Appl. Microbiol. Biotechnol. 51, 545–552.

Chemostat culture characterization of Escherichia coli mutant strains metabolically engineered for...

Documents

Transcript of Chemostat culture characterization of Escherichia coli mutant strains metabolically engineered for...