Enhancement of over expression and chaperone assisted yield of folded recombinant aconitase in Escherichia coli in bioreactor cultures

Journal of Bioscience and BioengineeringVOL. 107 No. 2, 102–107, 2009

www.elsevier.com/locate/jbiosc

Enhancement of over expression and chaperone assisted yield of folded recombinantaconitase in Escherichia coli in bioreactor cultures

Parul Gupta, Anand Ghosalkar, Saroj Mishra, and Tapan Kumar Chaudhuri⁎

Abbreviatioammonium pe⁎ Correspond

E-mail add

1389-1723/$doi:10.1016/j.

Department of Biochemical Engineering and Biotechnology, Indian Institute of Technology Delhi, Hauz Khas, New Delhi-110016, India

Received 7 August 2008; accepted 17 October 2008

A major portion of the over expressed yeast mitochondrial aconitase, a large 82 kDa monomeric TCA cycle enzyme, inEscherichia coli led to the formation of inclusion bodies. Bacterial chaperonin GroEL mediated the correct folding of aconitasewith the assistance of its co-chaperonin GroES in an ATP dependent manner. Till date the chaperonin assisted folding ofaconitase was limited to the shake flask studies with relatively low yields of folded aconitase. No attempt had yet been madeto enhance the yield of chaperone mediated folding of aconitase using a bioreactor. The current report deals with the effect ofco-expression of GroEL/GroES in the production of soluble, biologically active recombinant aconitase in E. coli by cultivation ina bioreactor at different temperatures under optimized conditions. It revealed that the yield of functional aconitase wasenhanced, either in presence of co-expressed GroEL/ES or at low temperature cultivation. However, the outcome from thechaperone assisted folding of aconitase was more pronounced at lower temperature. A 3-fold enhancement in the yield offunctional aconitase from the bioreactor based chaperone assisted folding was obtained as compared to the shake flask study.Hence, the present study provides optimized conditions for increasing the yield of functional aconitase by batch cultivation ina bioreactor.

© 2008, The Society for Biotechnology, Japan. All rights reserved.

[Key words: Aconitase; Bioreactor; GroEL and GroES; in vivo folding; Low temperature cultivation]

Escherichia coli expression system, with its ability to grow rapidly,well characterized genetic background, high expression level ofrecombinant proteins, and cheapness, is the most attractive andideal system for heterologous protein expression. However, thegreatest disadvantage of this expression system is the problem thathigh level expression of heterologous recombinant proteins oftenleads to the incorrect folding of the newly synthesized polypeptidesand thus forms insoluble inclusion bodies. As a result, the yield offolded recombinant proteins is usually low (1). Since inclusion bodiesconsist of the protein of interest and are easily isolated bycentrifugation, their formation has often been exploited to simplifypurification schemes. The recovery of biologically active productsfrom the aggregated state is typically accomplished by unfolding withchaotropic agents, followed by dilution into optimized refoldingbuffers. However, many polypeptides (e.g. structurally complexoligomeric proteins and those containing multiple disulfide bonds)do not easily adopt an active conformation following chemicaldenaturation. In such cases, maximizing the yields of recombinantproteins in a soluble and active form, in vivo, becomes an attractivealternative to in vitro refolding (2).

Several strategies have been exploited over last two decades inorder to circumvent the ‘protein folding’ problem and enable the

ns: AMPR, ampicillin resistance; LB, Luria broth; TB, Terrific broth; APS,rsulfate; IPTG, isopropyl β-D-thiogalactoside.ing author. Tel.: +91 11 2659 1012; fax: +91 11 2658 2282.ress: [email protected] (T.K. Chaudhuri).

- see front matter © 2008, The Society for Biotechnology, Japan. Alljbiosc.2008.10.020

efficient and successful use of E. coli as a host for heterologous proteinexpression. Use of fusion partners like maltose binding protein (MBP),thioredoxin and glutathione S-transferase greatly improve thesolubility of passenger proteins by rapidly reaching a nativeconformation as it emerges from the ribosome and promote theacquisition of correct structure in downstream folding units (3, 4). Histag in combination with MBP (5) or alone (6) has greatly enhancedproduction levels of soluble protein in E. coli. Coexpression of variousfolding accessory proteins like peptidyl–prolyl cis–trans isomerase(PPIase) (7), trigger factor (8), DsbC and DsbG (9), GroEL-GroES (10–12) is another strategy successfully implemented for heterologousprotein expression in E. coli. The most extensively studied chaperonesare the chaperonin – GroEL and GroES – from E. coli (13). Two mostcommon strategies to maximize folded protein production in E. coliare the use of GroEL-GroES in combination with other foldingaccessory proteins like Dna K/Dna J (14); and the use of GroEL-GroES at low culture temperature (15, 16).

Aconitase was brought in to focus when it was discovered that onimporting aconitase into yeast mitochondria genetically deficient foreither Hsp60 or Hsp10, the homologues of GroEL and GroES, theprotein was found in insoluble aggregates (17). Aconitase being largerthan the cis cavity of GroEL became an important link for studyingGroEL/ESmediated folding of large substrate by trans mechanism (18).Refolding yields of aconitase, both in vivo and in vitro, underspontaneous conditions was found to be extremely small. Even inpresence of GroEL/ES, a maximum of 50% aconitase was found to beactive at shake flask level, in vivo (19). Thus, in the present study,

rights reserved.

GROEL/ES ASSISTED INCREASE IN PRODUCTION OF ACTIVE ACONITASE 103VOL. 107, 2009

GroEL/ES co-expression with yeast mitochondrial aconitase wasconducted in a simple batch cultivation using a bioreactor in orderto increase the biomass yield and the protein expressed. Cultivationtemperature, biomass concentration before induction and duration ofinduction were optimized to achieve maximized folding of theexpressed aconitase, which often gets limited due to reduced timeavailable when protein synthesis becomes faster. The effect of co-expression of the GroEL/ES on the production of soluble and activeaconitase was evaluated. The results revealed that low temperature orpresence of co-over expressed GroEL/ES alone could enhance theproduction of active aconitase; however, combinational employmentof lower temperature cultivation (25 °C) together with co-expressionof GroEL/ES was more effective, additively, to prevent aggregation ofaconitase and thereby producing higher extent of folded protein.

MATERIALS AND METHODS

Chemicals and reagents Luria broth (LB), yeast extract and tryptone as mediacomponents for E. coli growth and antibiotics kanamycin, ampicillin and tetracyclinewere obtained from HiMedia (India). HEPES, 1, 4-dithiothreitol (DTT), acrylamide, bis-acrylamide, standard molecular mass markers, ammonium persulfate (APS) andisopropyl β-D-thiogalactoside (IPTG) were obtained from Bangalore Genei, (India).Isocitrate dehydrogenase (porcine) was obtained from Sigma Chemicals (USA). Otherreagents and chemicals used were from Merck (Germany) and Sigma.

Strains and plasmids The gene for yeast mitochondrial aconitase, cloned in thepQE60 vector from Qiagen (AMPR selectable marker) with ColE1 origin of replicationand Lac-promoter, was obtained from Dr. Sabine Rospert, Germany. Over expression ofaconitase required induction by IPTG. The constructs, pACYCEL over expressing GroELand pACYCELS over expressing GroEL and GroES (with tetracycline resistance), weregenerous gifts from Dr. Arthur L. Horwich, USA. GroEL/GroES were constitutivelyexpressed from pACYC184 vector, due to presence of σ32 promoter upstream of groELand groES. M15 E. coli strain, a K12 derivative was used for the over expression ofvarious plasmids. M15 strain, containing multiple copies of pREP4 plasmid, wasmaintained in presence of kanamycin. pREP4 plasmid carries the lacI gene whichencodes the lac repressor and regulates the expression of aconitase from pQE60Aco.M15 cells were transformed with various plasmids, pQE60Aco, pACYCEL and pACYCELSto over express aconitase, GroEL and GroEL-GroES, respectively. The antibioticconcentrations used for the optimum growth of the cells were 25 μg ml−1, 80 μg ml−1

and 12.5 μg ml−1 for kanamycin, ampicillin and tetracycline, respectively.Optimization of complex medium for maximum aconitase expression LB

medium containing 1% casein enzymic hydrolysate, 0.5% yeast extract and 0.5% NaCl;Terrific Broth (TB) containing 1.2% yeast extract, 1.2% tryptone, 0.4% glycerol, 0.231%KH2PO4 and 1.25% K2HPO4; and M9 minimal medium containing 0.05% NaCl, 0.1%NH4Cl, 0.3% K2HPO4, 0.1 mM CaCl2, 2 mMMgSO4 and 0.2% glucose (20) were comparedfor expression of aconitase.

Optimization of biomass level before IPTG induction Transformed M15 E.coli strain expressing aconitase/GroEL/GroES was cultured in 100ml of TB. 10 ml culturebroth at different OD600 values was aseptically withdrawn and induced with 100 μMIPTG (21). Over expression of aconitase, GroEL and GroES were observed on 15% SDS-PAGE (22, 23). About 200 μl of culture was centrifuged. (The volume of the culture takenfor analysis was such that, it was inversely proportional to the OD values of the culturein each case. This was done in order to ensure similar number of cells analyzed in eachcase.) The pellet obtained was separated from the supernatant. The pellet wasresuspended in SDS-PAGE loading dye, boiled at 100 °C for 5 min in order to fractionatethe cells and denature the proteins; and loaded onto the SDS-PAGE. The protein bandscorresponding to the whole cell proteins were obtained on the gel and analyzed by therelative intensity measurements using the gel documentation unit (24).

Expression of aconitase in shake flask Primary inoculum of varioustransformed M15 E. coli strain over expressing aconitase was grown by inoculating10 ml TB, containing antibiotics. The cultures were grown overnight in a shakingincubator at 37 °C and 25 °C at 200 rpm. These cultures were grown till OD600 of 0.8 wasachieved and induced with 100 μM IPTG. Recombinant protein production, followinginduction, was carried out to study the effect temperature in minimizing aggregationand enhancing the yield of folded aconitase.

Batch culture in a bioreactor Transformed M15 E. coli strain, over expressingaconitase, was grown in 20 ml TB mediumwith antibiotics for 10 h (primary inoculum).2ml of the primary inoculum of OD600∼1was used to inoculate 100ml TBmediumwithantibiotics and allowed to grow for 8 h (secondary inoculum). Both primary andsecondary inoculum was cultured at the temperature used for bioreactor cultivation.20 ml secondary inoculum of OD600 ∼2.5 was then used to inoculate the bioreactor.Three-liter bioreactor with a working volume of 2.2 l, and equipped with pH,temperature and dissolved oxygen measurement probes was used for the cultivation(Bioengineering, Switzerland). pH was maintained constant at 7.2 by acid/base additionthrough automatic feedback control. Airwas sparged at a rate of 1 vvm (volume of air perunit volume of medium per minute) and dissolved oxygenwas maintained at 20% of airsaturation. Cultivation temperatures were maintained at 37 °C and 25 °C. Inductionwas

carried out at OD600 of 2.5. Recombinant proteinproduction, after inductionwith 100 μMIPTG was carried out to study the effect of temperature in minimizing aggregation andenhancing the production of folded recombinant aconitase in a bioreactor.

Determination of growth profile E. coli strains were grown in shake flask indifferent media and 1 ml portions of cell suspension were withdrawn at various timesfor turbidity measurements at 600 nm using DU 800 spectrophotometer (BeckmanCoulter Instruments, USA). Induction was done in mid log phase (OD600=1 for LB,OD600=2.5 for TB) with 100 μM IPTG for over expression of aconitase. Expression wasobserved by running various samples on 15% SDS-PAGE (using sample preparationmethod as described earlier) and analyzed by relative intensity measurements by geldocumentation unit.

Determination of the specific growth rate for recombinant E. coli cells Thespecific growth rate for E. coli cells grown in different media was calculated from theslope of absorbance versus time plot using the following equation:

X = X0elt

where X=biomass at time ‘t’; X0=biomass at time ‘t=0’; μ=specific growth rate; t=timein hours.

Estimation of the relative intensities of the bands in the SDS-PAGE A BioRad(USA) gel documentation unit was used for estimating the relative quantities of proteinpresent in the various bands observed on the gel, following the method given by user'smanual provided by the manufacturer and described by Chaudhuri et al. (24). Relativequantity of a particular band is the quantity measured by its intensity, expressed as apercentage of the total intensity of all the bands in the lane.

In vivo folding of aconitase Amount of the folded protein in a cell can beestimated based on the principle that the proteins with folded structure are soluble inthe cytoplasm and in aqueous buffer, however, denatured or un-folded proteins occuras aggregates and are insoluble (19). Thus, to estimate the extent of correct folding ofaconitase in vivo, culture broths (∼10 ml) of different transformed strains overexpressing aconitase at different temperatures and time intervals were harvested andresuspended in 1.5 ml lysis buffer containing 50 mM HEPES (pH 7.4), 0.5 mM MgCl2,1 mM DTT (19). Normalization of the cell culture was done, such that same number ofcells was taken for the analysis of each sample. These cells were disrupted usingultrasonicator at 4 °C (15 rounds of 30 s. pulse with 1 min intervals), followed bycentrifugation at 10,000 rpm for 45 min. The supernatant and pellet obtained wereresuspended in the loading buffer and analyzed by SDS-PAGE. Aconitase activity assaywas done taking 50 μg of total protein in each case.

Aconitase assay Aconitase activity was quantitated using a coupled enzymeassay. Aconitase catalyzes conversion of citrate to isocitrate, which in turn is convertedto α-ketoglutarate in presence of isocitrate dehydrogenase along with the formation ofNADPH from NADP (25). NADP coenzyme was used instead of NAD (commonly presentin the krebs cycle) as isocitrate dehydrogenase used in the estimation is from porcineand is specific for NADP. The assay was performed by taking 20–50 μg of protein in a1 ml reaction volume (0.1 M Tris–HCl pH 8, 0.66 mM sodium citrate, 0.66 mM MnSO4,0.5 mg ml−1 β-NADP and 0.17 mg ml−1 isocitrate dehydrogenase). The formation ofNADPH was monitored at 340 nm using time/kinetics application mode on BeckmanCoulter DU 800 spectrophotometer.

Aconitase purification Pelleted E. coli cells (2 g of dry cell weight (DCW)),containing over expressed aconitase, were resuspended in 10 ml of lysis buffer anddisrupted in a French press (3 passes at 13,000 psi), DNAse I, 10 U μl−1 (1 μl ml−1 cellsuspension) and PMSF (1 mM final concentration) were added, followed by incubationat 37 °C for 30 min. Cell debris was removed by centrifugation in Beckmanultracentrifuge at 40,000 rpm (114,000 ×g) for 40 min. The clear lysate was applied toa 20 ml Q Sepharose Fast Flow column (Pharmacia) attached in tandem to a 25 ml CMSepharose Fast Flow column (Pharmacia). The columns were washed with 20 mMHEPES (pH 7.4) until no absorbance at 280 nm was detected. The Q Sepharose columnwas removed, and the CM Sepharose column was further washed with 50–75 ml of20 mM HEPES (pH 7.4). Aconitase was eluted from CM Sepharose column with 20 mMHEPES (pH 7.4), containing 0.5 mM cis-aconitate (19).

RESULTS

Aconitase gene present in pQE60Aco is under an IPTG inducible lacpromoter. As the over expression of GroEL and GroES is constitutiveand expression of aconitase requires induction by IPTG, the overexpression of aconitase commences after an initial accumulation ofchaperonin in the E. coli cells. Hua et al. have already reported that theearly expression of molecular chaperones followed by the expressionof the target proteins could modestly improve the production ofsoluble recombinant protein (1).

Over expression of aconitase in shake flask cultures Recom-binant M15 E. coli strain, over expressing aconitase, was grown in LB,TB and M9 medium, in order to study the media specific expressionlevels of aconitase. Cultures in all three media were induced at OD600

of ∼1with 100 μM IPTG and aconitase expressionwas analyzed on 15%

TABLE 1. Comparison of growth rates and the final biomass attained in the threedifferent media at 37 °C at the shake flask level

Medium Specific growth rate (h−1) Final biomass attained (DCW g l−1)

Luria broth 0.40 0.57Terrific broth 0.45 1.41Minimal medium 0.18 0.30

104 GUPTA ET AL. J. BIOSCI. BIOENG.,

SDS-PAGE. Expression of aconitase in TB was found to be 1.9 fold (7.2%of total cellular protein) more than in LB (3.9% of total cellular protein).The growth inM9minimalmediumwas found to be severely inhibitedand almost no aconitase expression was observed (Fig. 1). The finalOD600 achieved after 12 h of induction was 3.7 in case of TB ascompared to 1.5 and 0.8 in LB and M9 media, respectively. Thus, thevolumetric biomass yield in TBmediumwas ∼2.5 and ∼4.6 fold higherthan that obtained in LB and M9 medium, respectively (Table 1).

Recombinant M15 E. coli strain was grown in TB and inductionstrategy was optimized for maximum aconitase expression. Whilelarger biomass content before induction is necessary for enhancedexpression of the recombinant protein, the cells overproducingrecombinant protein should also be sufficiently active at the time ofinduction. Thus, optimizing the value of OD600 for induction plays acrucial role for maximizing expression of recombinant proteins. Theliterature states that a post induction period of 11 h at 37 °C and 14 h at25 °C was required to maximize aconitase expression at the shakeflask level (24). Post induction samples were collected at differentOD600 values and analyzed by 15% SDS-PAGE for checking theexpression levels of aconitase (Fig. 1). Maximum expression level inTerrific broth was 7.4% of the total cellular protein (Fig. 1) wheninduced at OD600 of 2.5 (∼1.6 fold higher than aconitase expression inLBwhen induced at same OD600, data not shown). At higher OD600, theexpression of aconitase reduced significantly with increase in biomasslevels (Fig. 1). A basal level expression of aconitase was observed (Fig.1, lane 5 and Figs. 2A and B, lane 2), even in absence of IPTG. Aconitasehas been cloned in pQE60 under the T5 promoter (a derivative of T7promoter) which is recognized by the bacterial RNA polymerase,however, it is a leaky promoter which results in the basal levelexpression of the over expressed protein even in absence of aninducer.

Over expression of aconitase in a bioreactor Co-expression offolding accessory proteins (26) and trigger factor (8) has been shownto enhance production of active proteins in various cultures. Hence,we examined the effects of co-expressing bacterial chaperonin GroELand GroES on the solubility of heterologous yeast mitochondrialprotein over expressed in E. coli. Recombinant E. coli strain was grownupto OD600 of 2.5 and induced with 100 μM IPTG. Post inductionsamples were withdrawn at different intervals and analyzed forsoluble aconitase by SDS-PAGE after fractionation and aconitaseactivity. In the absence of over expressed chaperonin at 37 °C, only 25%of aconitase was found to be soluble and the expressed aconitase wasabout 16% of the total cellular protein. When GroEL/GroES were overexpressed along with aconitase, the solubility of aconitase wasimproved to 45% and the expressed aconitase was about 13% of thetotal cellular protein (Fig. 2B, lane 4; Fig. 3B, lane 6 and Fig. 3C).

Protein production, when carried out at lower temperature, has apronounced effect on the solubility of aconitase in E. coli. The extentof aggregation is greater at higher temperature due to the strongtemperature dependence of the hydrophobic interactions, whichdominates protein aggregation (27). When expressed at 25 °C, in the

FIG.1. Aconitase expression inM9, TB and LB and induction at different levels of biomassin TB. SDS-PAGE showing standard molecular mass markers in lane 1, lane 2 showsaconitase expression in M9medium; lane 3, aconitase expression in TB medium; lane 4,aconitase expression in LB medium; lane 5, aconitase expression in uninduced culture;lanes 6 to 10, shows aconitase expression on induction at OD 600 of 0.6 (5.5 h), 1 (6.6 h),2.5 (8.5 h), 3.0 (9 h) and 3.5 (9.5 h) in Terrific Broth. Aconitase expression as percent oftotal protein expressed is given at the bottom of each lane.

absence of chaperonin, 40% of aconitase produced existed in thesoluble form and only 12% of total cellular protein was aconitase,implying that 4.8% of the total expressed protein was solubleaconitase. The combined effect of low temperature cultivation andGroEL/GroES over expression exhibited much more effective preven-tion of aggregation (Table 2). When GroEL, GroES and aconitase wereover expressed at 25 °C, the production of soluble aconitaseincreased, such that 10% of the total cellular protein was aconitaseand nearly 75% of it was in the soluble form, implying that 7.5% oftotal expressed protein was soluble aconitase (Fig. 2A, lane 7; Fig. 3A,lane 9 and Fig. 3C).

Purification of aconitase Aconitase was purified from thesoluble fraction of the cell lysate using a simple purification method.The clear lysate obtained after cell disruption, followed by DNAsetreatment and ultracentrifugation, was loaded onto Sepharose Q HR20 ml anion exchange column in tandem with CM Sepharose 20 mlcation exchange column on AKTA FPLC system (Pharmacia). At pH 7.4,GroEL, GroES and several other cellular proteins are anionic, whereas,aconitase is cationic. Thus, the major bulk of the cellular protein bindsto Sepharose Q column, significantly lowering the protein loadentering CM Sepharose column. Aconitase bound to the CM Sepharosecolumn was eluted by carrying out elution with cis-aconitate. About60 mg of soluble aconitase was produced from 1 l culture (DCW=3 g)from the bioreactor at 25 °C, from which at least 25 mg of purifiedaconitase, with a purity of SDS-PAGE homogeneity of 99%, was finallyisolated (Fig. 4).

DISCUSSION

Current industrial production processes in biotechnology arecapable of producing large amounts of recombinant proteins in verylimited time. However, to appropriate the use of biotechnologycapacities, all aspects of protein manufacturing processes must beestablished, including protein scaffolds for the gene/host systems andhigh yield cultivation systems from which robust recoveries can bemade. Various cultivation schemes are employed in the process to

FIG. 2. Expression levels of over expressed aconitase in M15 strain expressing aconitase,GroEL and GroES at various post induction time intervals. (A) Aconitase expression fromcells grown at 25 °C. SDS-PAGE showing lane 1, standard protein molecular massmarkers; lane 2, aconitase expression before induction; lane 3, 1.5 h after induction;lane 4, 3 h after induction; lane 5, 6 h after induction; lane 6, 9 h after induction and lane7, 12 h after induction. (B) Aconitase expression from cells grown at 37 °C. SDS-PAGEshowing lane 1, standard protein molecular mass markers; lane 2, aconitase expressionbefore induction; lane 3, 1.5 h after induction; lane 4, 3 h after induction; lane 5, 4.5 hafter induction; lane 6, 7.5 h after induction and lane 7, 9 h after induction. Aconitaseexpression as percent of total protein expressed is given at the bottom of each lane.

TABLE 2. Expression level of aconitase and the percentage of soluble aconitase obtained

Percentage 37 °C 25 °C

WithoutGroEL/ES

WithGroEL/ES

WithoutGroEL/ES

WithGroEL/ES

Expressed aconitasea 16% 13% 12% 10%Soluble aconitaseb 25% 45% 40% 75%

a % of the total aconitase present with respect to total expressed protein at the time ofmaximum aconitase folding.

b % of soluble aconitase with respect to total expressed aconitase.

GROEL/ES ASSISTED INCREASE IN PRODUCTION OF ACTIVE ACONITASE 105VOL. 107, 2009

enhance protein expression levels in the cell (28). Maintenance ofoptimum culture conditions of pH, temperature, agitation andaeration throughout the cultivation helps in achieving high biomassand protein productivity. It is already known that by using abioreactor, the expression of recombinant protein can be enhancedby several folds. In some cases, switching from shake flask expressionto bioreactor cultivation can give dramatic increases in yield, withreports of expression levels in bioreactor being 10-fold higher than inshake flasks (29, 30). However, the yield of correctly folded protein hasnot been considered in most studies. In many cases, the overexpressed recombinant protein lodges itself as insoluble aggregate,making the effort futile. Hence, proper folding of over expressedrecombinant proteins should be accomplished in order to enhance theproduction of usable protein.

It has already been established that E. coli chaperonin GroEL andGroES enhances the in vivo folding of different large polypeptidesubstrates-αβ heterodimer of BKCD (31) and aconitase (19). However,all these studies have been carried out at shake flask level orlaboratory scale. Hence, we attempted to increase the yield of alarge recombinant substrate protein at a relatively higher scale usingbioreactor. Yeast mitochondrial aconitase is a monomeric, 82 kDa,Fe4S4 cluster containing enzyme of the Krebs cycle that catalyzes theisomerization of citrate to isocitrate. Rospert et al. reported earlierthat, when the precursor form of this protein is imported into

FIG. 3. Percentage of soluble aconitase in the cell lysate ofM15 strain expressing aconitase, Grostandard protein molecular mass markers; lanes 2 to 4, aconitase expression in whole cell, scell, supernatant and pellet at 9 h after induction; lanes 8 to 10, aconitase expression in whgrown at 37 °C. SDS-PAGE showing lane 1, standard protein molecular mass markers; lanes 2lanes 5 to 7, aconitase expression in whole cell, supernatant and pellet at 3 h after inductioninduction. Aconitase expression as percent of total protein expressed is given at the bottom oshown in SDS-PAGE. Activity of aconitase from cells grown at 25 °C is given by white bars a

mitochondria, deficient in either hsp60 (GroEL) or hsp10 (GroES), theimported protein was lodged as insoluble aggregates, as compared tobeing fully soluble after import into wild type mitochondria (17).Production of large amounts of this protein has always been restricteddue to its large size and requirement of chaperonin assistance to attainbiologically active form, both in vivo and in vitro.

Our studywith different bacterial growthmedia showed that a richmedium like Terrific broth supports higher specific growth rate, whichin turn improves both the expression level of aconitase and itssolubility in vivo. Extremely inhibited growth and reduced level ofexpression of aconitase in minimal medium may arise due to thedifference in the codon usage between eukaryotes and prokaryotes(3). The arginine codons AGA and AGG are the most rare codons in E.

EL and GroES. (A) Soluble aconitase from cells grown at 25 °C. SDS-PAGE showing lane 1,upernatant and pellet at 6 h after induction; lanes 5 to 7, aconitase expression in wholeole cell, supernatant and pellet at 12 h after induction. (B) Soluble aconitase from cellsto 4, aconitase expression in whole cell, supernatant and pellet at 1.5 h after induction;; lanes 8 to 10, aconitase expression in whole cell, supernatant and pellet at 7.5 h afterf each lane. (C) Graph showing the activity of the soluble aconitase in lanes 3, 6 and 9 asnd at 37 °C is given by grey bars.

FIG. 4. SDS-PAGE of purified yeast mitochondrial aconitase. Lane 1, standard proteinweight markers; lane 2, 0.04 μg of pure aconitase; lane 3, 0.07 μg of pure aconitase andlane 4, 0.14 μg of pure aconitase.

106 GUPTA ET AL. J. BIOSCI. BIOENG.,

coli genes, whereas, they are fairly common in eukaryotes. Thepresence of such codons in cloned genes affect protein accumulationlevels, mRNA and plasmid stability and, in extreme cases, inhibitsprotein synthesis and cell growth (3, 32). Aconitase has about 33arginine amino acid residues in the primary sequence. Out of these 33residues, 20 are coded by AGA and one by AGG; both are extremelyrare in E. coli codons. Thus, about 64% of arginine in aconitase is codedby rare E. coli codons. The effect of AGA codon is much pronouncedwhen the cells are grown in minimal medium (3). Thus, terrific brothwas employed to maximize both specific and volumetric productyield. Under physiological conditions, where there is no additionalfactor (like iron chelating factors or DTSB) causing iron depletionwithin the cell, the mitochondrial aconitase exists as “holo” product.The high activity of the crude fraction obtained justifies that theaconitase product was indeed in holo state. The activity of aconitaseobtained from TB (even in the absence of any external iron additive),clearly show that the expressed aconitase was in holo state. Whereas,the lack of aconitase activity from M9 was largely due to stuntedgrowth and inhibited expression of aconitase in M9 medium from thesame cells. Hence, no further studies like the effect of addition of ironin the medium were done by employing the minimal medium or TB.

The large, barrel shaped oligomers known as chaperonin play acentral and essential role of folding, monitoring, and maintainingcellular proteins in E. coli. The double ring chaperonin GroEL from E. colihas been shown to mediate ATP dependent folding of a variety ofproteins with the assistance from its co-chaperonin GroES. Pre-existence of enough endogenous chaperones in the E. coli cytosolseems to be important for their function of preventing proteinaggregation and facilitating protein folding. Induction with IPTG,carriedout at different biomass accumulation, gave significant variationin the expression levels of aconitase in recombinant E. coli cells.Induction of aconitase gene at an OD600 of 2.5 in Terrific Broth showedsignificantly higher expression of aconitase. The post inductionincubation time was also shown to have a significant effect on thesolubility of aconitase at both the temperatures. As the post inductionincubation timewas increased the solubility of aconitase also increased,whereas the fraction of aconitase in the pellet reduced. This may be adirect consequence of the unfolding action or disaggregase activity ofGroEL (19). Also, GroEL within the cell is continually recycled and itsconcentration becomes significantly high at higher time of incubation,resulting in an increase in the efficiency of folding of aconitase. Co-expression of chaperonin GroEL and GroES along with recombinantaconitase has shown ∼1.8-fold increase in soluble aconitase at both25 °C and 37 °C. It is also widely recognized that lower temperaturereduces the rate of protein synthesis and allows sufficient time for thenascent peptide chains to fold properly. The objective of the study wasto maintain high yields of aconitase expression by means of bioreactorcultivation, while ensuring optimum conditions for correct folding, inorder to obtain high yields of correctly folded aconitase in largequantities. Our study clearly shows that, irrespective of the co-expression of chaperones, lower temperature could increase the yieldof soluble aconitase by ∼1.6-fold. Furthermore, combined employment

of both chaperonins GroEL/GroES and low temperature cultivation(25 °C) enhanced the production of soluble aconitase by ∼3-fold. Use ofbioreactor helped to maintain a high level of expression of aconitase ofabout 10% of total cellular protein, which is vitally important for largescale production of aconitase.

We also used a simple, economic procedure for purificationof recombinant aconitase. Purified aconitase per unit of biomassobtained from shake flask culture was 2 mg g−1 of DCW cultured in TBmedium. The yield of purified aconitase per unit of biomass wasincreased to 6.5 mg g−1 of cells cultivated in a bioreactor using TB.Biological activity assay using coupled reaction of citrate to isocitrateto α-ketoglutarate showed that the specific activity of the purifiedpreparation was 0.27 nmol of isocitate per mg of protein per min andwas ∼15-fold more than the crude lysate.

Thus, by employing the co-expression of GroEL/ES and lowtemperature cultivation in a bioreactor, the yield of functionalaconitase can be improved from shake flask culture method. Thebioreactor based studies on folding of small proteins has beenstudied by several workers. However, the use of this technology andthe effect of various parameters for increasing the folding of largeproteins like aconitase, in order to increase the overall productionhas never been undertaken before. Thus, our study forms the basisfor further bioprocess optimization, like fed-batch cultivation andits effects, for large scale production of large protein like aconitasein E. coli.

ACKNOWLEDGMENTS

The authors acknowledge the generous gifts of pQE60Aco plasmidfrom Prof. Sabine Rospert (Germany) and pACYCEL and pACYCELSfrom Prof. A.L. Horwich (USA). Miss Parul Gupta is a recipient of seniorresearch fellowship from Council of Scientific and Industrial Research(CSIR), Govt. of India. The work has also been supported by Ministry ofHuman Resource and Development (MHRD), Govt. of India andIndustrial Research and Development division (IRD), IIT Delhi.

References

1. Xu, H., Zhang, G., Ji, X., Cao, L., Shu, L., and Hua, Z.: Expression of soluble,biologically active recombinant human endostatin in Escherichia coli, Protein Expr.Purif., 41, 252–258 (2005).

2. Thomas, J. G. and Baneyx, F.: Protein misfolding and inclusion body formation inrecombinant Escherichia coli cells overexpressing heat-shock proteins, J. Biol.Chem., 271, 11141–11147 (1996).

3. Baneyx, F.: Recombinant protein expression in Escherichia coli, Curr. Opin.Biotechnol., 10, 411–421 (1999).

4. Lee, C., Lee, S. G., Takahashi, S., and Kim, S. G.: The soluble expression of thehuman renin binding protein using fusion partners: a comparison of ubiquitin,thioredoxin, maltose binding protein and NusA, Biotechnol. Bioprocess Eng., 8,89–93 (2003).

5. Pryor, K. D. and Leiting, B.: High-level expression of soluble protein in Escherichiacoli using a His6-tag and maltose-binding-protein double-affinity fusion system,Protein Expr. Purif., 10, 309–319 (1997).

6. Fang, L., Jia, K., Tang, Y., Ma, D., Yu, M., and Hua, Z.: An improved strategy for high-level production of TEV protease in Escherichia coli and its purification andcharacterization, Protein Expr. Purif., 51, 102–109 (2007).

7. Lee, D. H., Kim, M. D., Lee, W. H., Kweon, D. H., and Seo, J. H.: Consortium of fold-catalyzing proteins increases soluble expression of cyclohexanone monooxygenasein recombinant Escherichia coli, Appl. Microbiol. Biotechnol., 63, 549–552 (2004).

8. Nishihara, K., Kanemori, M., Yanagi, H., and Yura, T.: Overexpression of triggerfactor prevents aggregation of recombinant proteins in Escherichia coli, Appl.Environ. Microbiol., 66, 884–889 (2000).

9. Zhang, Z., Li, Z. H., Wang, F., Fang, M., Yin, C. C., Zhou, Z. Y., Lin, Q., and Huang,H. L.: Overexpression of DsbC and DsbG markedly improves soluble and functionalexpression of single-chain Fv antibodies in Escherichia coli, Protein Expr. Purif., 26,218–228 (2002).

10. Horwich, A. L., Low, K. B., Fenton, W. A., Hirshfield, I. N., and Furtak, K.: Foldingin vivo of bacterial cytoplasmic proteins: Role of GroEL, Cell, 74, 909–917 (1993).

11. Johnson, J. L. and Craig, E. A.: Protein folding in vivo: Minireview unravelingcomplex pathways, Cell, 90, 201–204 (1997).

GROEL/ES ASSISTED INCREASE IN PRODUCTION OF ACTIVE ACONITASE 107VOL. 107, 2009

12. McLennan, N. and Masters, M.: GroE is vital for cell-wall synthesis, Nature, 392,139 (2004).

13. Fenton, W. A. and Horwich, A. L.: GroEL mediated protein folding, Protein Sci., 6,743–760 (1997).

14. Gragerov, A., Nudler, E., Komissarova, N., Gaitanaris, G. A., Gottesman, M. E., andNikiforov, V.: Cooperation of GroEL/GroES and Dna K/Dna J heat shock proteins inpreventing protein misfolding in Escherichia coli, Proc. Natl. Acad. Sci. USA, 89,1031–1034 (1992).

15. Park, S. L., Kwon, M. J., Kim, S. K., and Nam, S. W.: GroEL/ES chaperone and lowculture temperature synergistically enhanced the soluble expression of CGTase inE. coli, . J. Microbiol. Biotechnol., 14, 216–219 (2004).

16. Sun, Q., Chen, L., Cao, L., Fang, L., Chen, C., and Hua, Z.: An improved strategy forhigh-level production of human vasostatin 120–180, Biotechnol. Prog., 21,1048–1052 (2005).

17. Dubaquie, Y., Looser, R., Funfschilling, U., Jeno, P., and Rospert, S.: Identificationof in vivo substrates of yeast mitochondrial chaperonins reveals overlapping butnon-identical requirement for hsp60 and hsp10, EMBO J., 17, 5868–5876 (1998).

18. Farr, G. W., Fenton,W. A., Chaudhuri, T. K., Clare, D. K., Saibil, H. R., and Horwich,A. L.: Folding with and without encapsulation by cis- and trans-only GroEL-GroEScomplexes, EMBO J., 22, 3220–3230 (2003).

19. Chaudhuri, T. K., Farr, G. W., Fenton,W. A., Rospert, S., and Horwich, A. L.: GroEL/GroESmediated folding of a protein too large to be encapsulated, Cell,107, 235–246(2001).

20. Srivastava, P., Bhattacharya, P., Pandey, G., and Mukherjee, K. J.: Overexpressionand purification of recombinant human interferon alpha2b in Escherichia coli,Protein Expr. Purif., 41, 313–322 (2005).

21. Christodoulou, E. and Vorgias, C. E.: Understanding heterologous protein over-production under the T7 promoter, Biochem. Mol. Biol. Edu., 30, 189–191 (2002).

22. Laemmli, U. K.: Cleavage of structural proteins during the assembly of the head ofbacteriophage T4, Nature, 227, 680–685 (1970).

23. Sambrook, J. and Russell, D.:Molecular cloning: A laboratory manual, Cold SpringHarbor Laboratory Press, New York, 2001.

24. Gupta, P., Aggarwal, N., Batra, P., Mishra, S., and Chaudhuri, T. K.: Co-expressionof chaperonin GroEL/GroES enhances in vivo folding of yeast mitochondrialaconitase and alters the growth characteristics of Escherichia coli, Int. J. Biochem.Cell. Biol., 38, 1975–1985 (2006).

25. Morrison, J. F.: The activation of aconitase by ferrous ions and reducing agents,Biochem. J., 58, 685–692 (1954).

26. Kim, S., Kweon, D., Lee, D., Park, Y., and Seo, J.: Coexpression of folding accessoryproteins for production of active cyclodextrin glycosyltransferase of Bacillusmacerans in recombinant Escherichia coli, Protein Expr. Purif., 41, 426–432 (2005).

27. Miot, M. and Betton, J. M.: Protein quality control in the bacterial periplasm,Microb. Cell Fact., 3, 1–13 (2004).

28. Yee, L. and Blanch, H. W.: Recombinant protein expression in high cell density fed-batch cultures of Escherichia coli, BioTechnology, 10, 1550–1556 (1992).

29. Clare, J. J., Rayment, F. B., Ballantine, S. P., Sreekrishna, K., and Romanos, M. A.:High-level expression of tetanus toxin fragment C in Pichia pastoris strainscontaining multiple tandem integrations of the gene, BioTechnology, 9, 455–460(1991).

30. Fischer, R., Drossard, J., Emans, N., Commandeur, U., and Hellwig, S.: Towardsmolecular farming in the future: Pichia pastoris-based production of single-chainantibody fragments, Biotechnol. Appl. Biochem., 30, 117–120 (1999).

31. Huang, Y. S. and Chuang, D. T.:Mechanisms for GroEL/GroES-mediated folding of alarge 86-kDa fusion polypeptide in vitro, J. Biol. Chem., 274, 10405–10412 (1999).

32. Zahn, K.: Overexpression of an mRNA dependent on rare codon inhibits proteinsynthesis and cell growth, J. Bacteriol., 178, 2926–2933 (1996).