Chaperone activity of recombinant maize chloroplast protein synthesis elongation factor, EF-Tu Damodara Rao 1 , Ivana Momcilovic 1 , Satoru Kobayashi 1 , Eduardo Callegari 2 and Zoran Ristic 1 1 Department of Biology, University of South Dakota and 2 Department of Basic Biomedical Sciences, University of South Dakota School of Medicine, Vermillion, SD, USA The p rotein synthesis e longation factor, EF-Tu, is a protein that carries aminoacyl-tRNA to the A -site o f the ribosome during t he elongation phase of protein synthesis. In maize (Zea mays L) th is protein has been implicated in heat tol- erance, andit has been hypothesized that EF-Tu confers heat tolerance by a cting as a molec ular chaperone and protecting heat-labile proteins from thermal aggregation and inactiva- tion. In this study w e i nvestigated t he effect of the recom- binant precursor of maize EF-Tu (pre-EF-Tu) o n t hermal aggregation and inactivation of t he heat-labile proteins, c it- rate synthase and malate dehydrogenase. T he recombinant pre-EF-Tu was purified from Escherichia coli expressing this protein, and m ass s pectrometry confirmed that the isolated protein was indeed maiz e EF-Tu. The purified protein was capable of binding GDP (indicative of protein activity) and was stable at 45 °C, the highest temperature used in this study to test this protein f or possible c haperone activity. Importantly, t he recombinant m aize pre-EF-Tu displayed chaperone activity. It protected citrate synthase and malate dehydrogenase from th ermal aggregation and inactivation. To our knowledge, this is the first observation of chaperone activity by a plant/eukaryotic pre-EF-Tu protein. The results of this study support the hypothesis that maize EF-Tu plays a role in heat tolerance b y a cting a s a molecular chaperone and protecting chloroplast proteins from thermal aggregation a nd inactiv ation. Keywords: chloroplast protein synthesis elongation factor (EF-Tu); chaperones ; heat stress; heat tolerance; Zea mays. Chloroplast protein s ynthesis elongation factor, EF-Tu, i s a protein (45–46 kDa) that plays a key role i n t he elongation phase of protein synthesis [1–3]. This p rotein catalyzes the GTP-dependent binding of aminoacyl-tRNA to the A-site of the ribosome [ 3]. In land plants, EF-Tu is encoded b y the nuclear genome and synthesized in the cytosol [4]. Chloro- plast EF-Tu is highly conserved, and it shows a high sequence similarity to prokaryotic EF-Tu [3,5]. Studies from our laboratory have s uggested that in maize (Zea mays L) chloroplast EF-Tu may play a role in the development of heat tolerance. The evidence for this conclusion includes: (a) positive correlation between the heat-induced accumulation of EF-Tu and plant ability to tolerate heat stress in several genotypes of maize [5–7], (b) association between the heat-induced synthesis of EF-Tu and the maize heat t olerance phenotype [8], (c) increased tolerance to heat st ress in Escherichia coli expressing maize EF-Tu [9], (d) decreased t olerance to heat stress in a m aize mutant with reduced capacity to accu mulate EF-Tu [10], and (e) increased thermal stability o f chloroplast p roteins in maize with higher levels o f EF-Tu [10,11]. (It should be noted that in the previous studies [6–8] maize EF-Tu was referred to as a 45–46 kDa heat shock p rotein because the identity of this pro tein was not known until the r eport of Bhadula et al . [5].) A h ypothesis h as been developed that m aize EF-Tu may confer heat to lerance by protecting o ther proteins from heat-induced aggregation and inactivation (thermal dam- age), thus acting as a molecular chaperone [10,11]. I n t his study we investigated the effect of th e recombinant p recur- sor of m aize EF-Tu (pre-EF-Tu, w hich has a 58 amino acid long chloroplast targeting sequence a t the N-terminal end [5]) on thermal aggregation and inactivation of the heat- labile proteins, citrate synthase (CS) and malate dehydro- genase (MDH). H ere we r eport, for the firs t time, that t he recombinant maize pre-EF-Tu displays ch aperone proper- ties, as it protected heat-labile proteins from thermal damage. Materials and methods Expression of maize pre-EF-Tu in Escherichia coli E. coli expressing maize pre-EF-Tu was previously trans- formed [9] using a cDNA for maize (Z. mays L) EF-Tu, designated as Zme ftu1 [5]. Zme ftu1 was s ubcloned into the expression vector pTrcHis2A, which adds C-terminal c-myc and polyhistidine tags to the protein, and the pTrcHis2A vector carrying Zmeftu1 w as used t o transform competent E. coli cells of the s train DH5 a [9]. In the c urrent study, the induction of expression of maize pre-EF-Tu in E. coli was carried o ut according to Moriarty et al . [9]. F ollowing induction, the r ecombinant protein was isolated and purified from the E. coli culture. Correspondence to Z. Ristic, Department of Biology, University of South Dakota, Vermillion, SD 57069, U S A. Fax: +1 605 677 6557, E-mail: zristic@usd.edu Abbreviations: CS, citrate synthase; MDH, malate dehydrogenase; HSPs, heat shock proteins; sHSPs, small heat shock proteins. (Received 5 May 2004, revised 1 4 July 2 004, accepted 27 July 200 4) Eur. J. Biochem. 271, 3684–3692 (2004) Ó FEBS 2004 doi:10.1111/j.1432-1033.2004.04309.x Isolation and purification of recombinant pre-EF-Tu from E. coli E. coli cells expressing maize p re-EF-Tu were collected by centrifugation (50 mL cell c ulture; 10 000 g,30min), washed, and suspended in isolation buffer [20 m M Tris/ HCl pH 8.0 containing 20 m M NH 4 Cl, 10 m M MgCl 2 , 2m M dithiothreitol, 0.1 m M EDTA, 10% (v/v) glycerol, 1m M phenylmethanesulfonyl fluoride] according to S tanzel et al . [12]. Crude protein extract from harvested cells was then p repared by the lysozyme/EDTA method [13]. Cells were sonicated at a medium intensity setting, holding the suspension on ice. After sonication, i nsoluble debris w as removed by centrifugation at 5000 g for 15 min. The supernatant (lysate) w as then passed through a 0.8 lm syringe filter and stored at )70 °C until further use. Purification of recombinant pre-EF-Tu was c arried out according to Stanzel et al. [12]. Recombinant protein was purified by SP-Sepharose, Q-Sepharose and gel filtration chromatography. SP-Sepharose was packed in a column (25 cm · 1 cm), equilibrated w ith e ight column volumes o f 20 m M acetate buffer (pH 4.8) consi sting of 10 m M MgCl 2 , 2m M dithiothreitol, 0.1 m M EDTA, 10% (v/v) glycerol, and 1 m M phenylmethanesulfonyl fluoride. The fractions from the S P-Sepharose column were analyzed by 1D SDS/ PAGE, and fractions having prominent bands between 45 kDa and 55 kDa were pooled [typically, pooled fractions had a total of four to five bands with molecular masses ranging f rom 20 t o 100 kDa but the prominent bands (1–2) were between 45 and 55 kDa range] and d ialyzed against isolation buffer overnight. A fter dialysis , t he dialysate was applied to Q-Sepharose column (30 cm · 1 cm), f ollowed by washing the column with the same buffer. The bound recombinant pre-EF-Tu was eluted with a linear gradient of 0.0–0.5 M NaCl in the isolation buffer. Fractions (2 mL) were collected and an alyzed by 1D SDS/PAGE. The fractions containing purified pre-EF-Tu were pooled and concentrated using a centrifuge filter device Amicon )50 (Millipore, Bedford, MA). The concentrated protein was then applied to Sephacryl SS-100 (50 cm · 2.5 cm), eluted with the isolation buffer, and protein concentration was determined using the Br adford Ass ay (Bio-Rad, CA). The purity of r ecombinant pre-EF-Tu prepar ation was checked using 1 D SDS/PAGE and Western blotting [5], the identity of the purified protein was verified using mass spectrometry, and the ability of the protein to bind GDP (indicative of EF-Tu activity [14]) was assessed using the [ 3 H]GDP exchange assay [14,15]. In addition, the heat stability of purified pre-EF-Tu was also assessed as described below. One-dimensional SDS/PAGE and Western blotting One-dimensional SDS/PAGE of purified recombinant pre- EF-Tu was carried out according to Laemmli [16]. In separate trials, 1D SDS/PAGE g els were s tained using Coomassie Brilliant Blue R250 and Silver stain (Amersham). Western blot analysis w as performed as described b y Moriarty et al. [9]. The purified protein was resolved on 10% (w/v) polyacrylamide gel with SDS, and then trans- ferred to a nitrocellulose membrane (Bio-Rad, Hercules, CA). The blot w as probed for recombinant p rotein using the E CL che miluminescent development me thod with primary antibody raised against the my c epitope, which is included near the C-terminus of recombinant protein [9]. Mass spectrometry Mass spectrometry analysis was performed according to Koc et al . [17]. Purified protein was separated on a 10% (w/v) polyacrylamide gel with SDS and stained by Coomassie Brilliant Blue R250. The stained protein band was then excised from t he gel, and the protein spot was digested in-gel with Tryp sin (Promega, Madison, WI) [18]. The peptides produced were injected in to a C apillary LC (Waters Corporation, Milford, MA) to be desalted and separated using a C18 RP PepMap, 75 lm (internal diameter) column (LC P ackings, Dionex, San Francisco, CA). The standard gradient us ed was a s follows: 0 –2 min, 3% B isocratic; 2–40 min, 3–80% B linear. Mobile phase A w as water/ acetonitrile/formic acid ( 98.9 : 1 : 0.1, v /v/v), and phase B was acetonitrile/water/formic acid (99 : 0.9 : 0.1, v/v/v). The total solvent fl ow w as 8 lLÆmin )1 . Samples were analyzed under nano-ESI/MS and nano-ESI-MS/MS u sing a Q-TOF micro mass spectrometer (Micromass, Manchester, UK). The s pectrum data were acquired by MASSLYNX 4.0 s oftware (Micromass), and peptide matching and protein searches were performed auto matically using the PROTEINLYNX 1.1 Global Server ( Micromass). T he peptide m asses a nd sequence tags were searched against the NCBI nonredundant protein database. [ 3 H]GDP exchange assay The a ctivity of recombinant pre-EF-Tu was assessed b y the [ 3 H]GDP exchange assay [14]. Various amounts of purified protein were added to scintillation vials containing 40 lLof binding buffer ( 250 m M Tris/HCl, pH 7.4, 50 m M MgCl 2 , 250 m M NH 4 Cl, 25 m M dithiothreitol) and 4.5 nCi of [ 3 H]GDP (specific activity 27.7 mCi Æmg )1 or 12.3 CiÆ mmol )1 ; t otal volume of reaction mixture, 200 lL). As controls, BSA an d ovalbumin (Sigma) were used. The reaction mixtures were allowed to equilibrate for 10 m in at 37 °C then were diluted with 2 mL of wash buffer (10 m M Tris/HCl pH 7.4, 10 m M MgCl 2 ,10m M NH 4 Cl), filtered using Millipore discs (pore s ize, 0.2 lm; diameter, 5 0 mm), and washed t hree times w ith 3 mL of the same buffer . The filters were dissolved in 5 mL of scintillation fluid, and the radioactivity w as monit ored u sing a Beckman LS 6500 scintillation counter. Heat stability of recombinant pre-EF-Tu The h eat stability of r ecombinant pre-EF-Tu was assessed using two approaches. In the first approach, 1 mL samples of purified protein (0.5 l M ) were incubated at 25 °C (control) or 45 °C (heated) for 45 min. After incubation, samples were centrifuged and the supernatant of each sample (control a nd heated) was analyzed for t he presence of pre-EF-Tu using Western blotting and anti-myc Ig as described above. Equal volumes of protein s amples were loaded in each gel lane. In the second approach, 1 mL aliquot of recombinant protein (0.5 l M ) was incubated at 45 °C for 45 min i n c overed quartz cuvettes, and t he heat stability o f EF-Tu was a ssessed by monitoring light Ó FEBS 2004 Chaperone activity of recombinant maize EF-Tu (Eur. J. Biochem. 271) 3685 scattering at 320 nm during incubation [11,19]. As a control, a heat-labile p rotein, citrate synthase, [20] was used. In addition, the a ctivity of recombinant pre-EF-Tu was also measured af ter heating (45 min at 45 °C) by the [ 3 H]GDP exchange assay [14] as outlined above. Chaperone assays The recombinant pre-EF-Tu was t ested for possible chap- erone activity using t wo approaches: (a) by monitoring thermal aggregation of heat-labile CS or MDH in the presence or absence of pre-EF-Tu, and (b) by measuring residual activity o f CS or MDH after heating in the presence or absence o f pre-EF-Tu. CS and MDH w ere chosen as model s ubstrates because they are known to be relatively heat-labile and have b een used in chaperone studies [20–23]. CS and MDH were obtained from Boehringer Mannheim. Both enzymes are homodimers, and in the t ext and figures the c oncentrations of CS and MDH refer to the 98 kDa homodimer and 70 kDa homodimer, respecti vely. The thermal aggregation of CS and MDH was tested according to Lee et al . [ 21]. In separate trials, CS (0.15 l M ) and MDH (0.3 l M ) were m ixed with various amounts of purified recombinant pre-EF-Tu (as indicated in Fig. 6 ) in 20 m M Tris/HCl buffer [7 m M MgCl 2 ,60m M NH 4 Cl, 0.2 m M EDTA, and 10% (v/v) glycerol; pH 8; total v olume 1 m L] in covered quartz cuvettes. Three controls were used: CS o r MDH a lone, CS or M DH mixed with BSA, a nd CS or MDH mixed with ovalbumin (molar concentrations are indicated in F ig. 6 ). Samples were incubated at indicated high temperatur es (CS: 41 °Cor45°C; MDH: 45 °C) for 45 min, and CS or MDH stability was estimated by monitoring light scattering at 320 nm during incubation [21]. The residual activity of CS a nd MDH w as determined using the experimental d esign of L ee et al. [21]. In separate trials, CS ( 2 l M )andMDH(0.5l M )weremixedwith4l M and 2 l M of purified recombinant pre-EF-Tu, respectively. Aliquots (1 mL) of the mixtures [CS mixture: 0.2 m M acetyl-CoA, 0.5 m M oxaloacetic acid, 0.1 m M 5,5¢-dithio- bis(2-nitrobenzoic acid) in 100 m M Tris/HCl (pH 7 .5); MDH mixture : 0 .1 m M NADH, 0.1 m M oxaloacetic acid in 50 m M Tris/HCl (pH 7.5)] were then heated at various high temperatures ( 38 °C, 41 °C, 43 °C, and 45 °C) for 30 min. As controls, C S or M DH alone and CS or MDH mixed with B SA or ovalbumin w ere used (molar concen- trations for BSA and ovalbumin are indicated in Fig. 7). After heating, aliquots were quickly cooled to room temperature and then kept on ice for up to 75 min (75 m in recovery). The residual activity of CS a nd MDH was measured at room temperature immediately after heating and at various times of r ecovery, according to S rere [24] and Banaszak & Bradshaw [25], respectively. We also investigated the possible e ffect of recombinant pre-EF-Tu on reactivation of heat-inactivated CS and MDH. In separate trials, C S (2 l M )andMDH(0.5l M ) were incubated at 43 °C for 30 min, without the presence of pre-EF-Tu. I mmediately after incubation, pre-EF-Tu was added to the heated pr otein s amples (molar concentrations for pre-EF-Tu are indicated in F ig. 7) and the r eaction mixtures were allowed t o recover for 4 5 min. The mixtures were kept on ice during recovery. The residual activity of CS and MDH was t hen m easured at various times of recovery as described above. Results One-dimensional SDS/PAGE, Western blot, and mass spectrometry analysis of purified recombinant pre-EF-Tu The r ecombinant m aize pre-EF-Tu was purified to homo- geneity f rom E. coli expressing this protein [ 9]. A previou s study has shown that the expressed maize pre-EF-Tu appears in E. coli in a highly s oluble form [ 9]. Both s ilver stained ( Fig. 1, lane 1) and Coomassie B lue s tained (Fig . 1, lane 2) 1D SDS/PAGE gels showed a single band indicating purified protein. The molecular mass of the purified protein was approximately 50–51 kDa; this molecular mass was greater than that of the native chloroplast EF-Tu (45–46 k Da) [5] because of the presence of a chloroplast targeting sequence a t the N-terminal end a nd the c-myc and polyhistidine tags at the C-terminus o f the polypeptid e [9]. Western b lot probed with anti-myc Ig, which is specific to recombinant pre-EF-Tu carrying the c- myc tag [9], also showed a single band with the molecular mass of 50–51 kDa (Fig. 1, lane 3). The identity of the purified protein was further c onfirmed by m ass spectrometry, which showed that the purified protein was indeed chloroplast pre-EF-Tu (Fig. 2). The recombinant p rotein amino acid sequence obtained by mass spectrometry (Fig. 2B,C) matched t he sequence of maize chloroplast EF-Tu [5] and chloroplast EF-Tu from Oryza s ativa L., Glycine m ax (L) Merr, Pisum sativum L., and Nicotiana silvestris Speg. (NCBI nonredundant protein database; data not shown). GDP binding activity of purified recombinant pre-EF-Tu We assessed the activity of purified pre-EF-Tu using the [ 3 H]GDP exchange assay [14]. The assay showed that the Fig. 1. One-dimensional SDS/PAGE gels and Western blot o f purified recombinant maizepre-EF-Tu.The recombinant protein was purified from E. coli expressing this protein. Lane 1, gel stained with silver stain; lane 2, gel stained with Coomassie Brilliant Blue R250; lane 3 , Western blot probed w ith a nti-myc Ig. A rrow indicates r ecombinant pre-EF-Tu ( 50–51 kDa). Note: E. c oli EF-Tu has a mass of 43 kDa [22], and the protein band of this mass is not seen in lanes 1 and 2. Hence, th e purified protein s een in lanes 1 and 2 is most p robably maize pre-EF-Tu. 3686 D. Rao et al.(Eur. J. Biochem. 271) Ó FEBS 2004 purified pre-EF-Tu binds [ 3 H]GDP (Fig. 3) suggesting t hat this protein was probably i n a phys iologically active form. As indicated by an increase i n radioactivity (disintegration per m inute), the binding of pre-EF-Tu with GDP increased with an increase in the concentration o f recombinant protein (Fig. 3). No significant radioactivity, however, w as detected when [ 3 H]GDP was mixed with control proteins, BSA or ovalbumin (Fig. 3 ). Heat stability of recombinant maize pre-EF-Tu The h eat stability of r ecombinant pre-EF-Tu was assessed as its ability t o remain soluble and maintain activity at high temperature [20]. L ight scattering experim ents with heated aliquots of purified pre-EF-Tu showed that this protein w as heat stable (remained soluble) at 45 °C, as no increase in relative light scattering was observed when the protein was heated at this temperature ( Fig. 4A) . The control protein (CS), in c ontrast, showed no stability at 4 5 °C, indicated by an increase in relative light scattering (Fig. 4 A). The pre- EF-Tu also maintained its activity a t h igh t emperature (45 °C). As indicated by the [ 3 H]GDP exchange assay, the binding of h eated pre-EF-Tu with GDP was s imilar to the binding of nonheated (25 °C) pre-EF-Tu with GDP (Fig. 3 ). Western blot analysis of the supernatant of centrifuged heated samples of purified pre-EF-Tu corro- borated the results of light the scattering experiments. A Western b lot revealed t hat t he recombinant pre-EF-Tu was present in the soluble fraction (supernatant) at 45 °C, indicating its stability at this temperature (Fig. 4B). Recombinant maize pre-EF-Tu protected CS and MDH from thermal aggregation The recombinant maize pre-EF-Tu protected CS from thermal aggregation. When heated at either 41 °Cor45°C, CS began to form insoluble aggregates, indicated by an increase in relative light scattering (Fig. 5A,B). T he CS aggregation, however, was reduced in the presence o f various amounts of pre-EF-Tu and was almost completely suppressed a t a n pre-EF-Tu : CS molar ratio of 3.3 at 41 °C (Fig. 5A) and 6.7 a t 45 °C (Fig. 5B). Ovalbumin (0.5 l M , Fig. 5A,B) and BSA (not shown) added to C S had no protective effect on CS aggregation. Recombinant pre-EF-Tu also protected MDH from thermal aggregation. When heated at 45 °C, MDH began to form insoluble aggregates, indicated by a n increase in relative light scattering (Fig. 5C). Addition of various amounts of pre-EF-Tu, however, reduced MDH A B C Fig. 2. Mass spectrometry analysis of rec om- binant m aize pre-EF-Tu ( EF-Tu) isolated a nd purified f rom E. coli expressing this protein. (A) S core, number of matches, molecular mass, and p I (isoelectric point) of p urified protein identified by n an o-ESI-MS/MS. The score was d eterm ined by the PROTEINLYNX 1.1 Global Server (Microm ass) and is an indicato r of search result quality. (B) Matching peptides and a mino acid sequences of peptide ion spectra obtained from t he trypsin digestion of purified m aize pre-EF-Tu. (C) Matching sites of peptide p roducts in the com plete sequence of maize c hloroplast EF-Tu protein. T he peptide products [ from (B)] are s hown in red, blue, and gre en. The complete s eq uence was obtained from the database using PROTEIN- LYNX 1.1 Global S erver. Fig. 3. Binding of recombinant maize p re-EF-Tu (EF-Tu) to [ 3 H]GDP. Purified p re-EF-Tu was incubate d alone at 25 °Cor45 °Cfor45min. Following i ncubatio n, t he activity of the p rote in was assessed by t he [ 3 H]GDP exchange assay [14]. Reaction mixture (total volume 200 lL) contained binding buff er, 4.5 nmol of [ 3 H]GDP (12.3 CiÆmmol )1 ), and various a moun ts of EF-Tu. As controls, BSA and ovalbumin were used. Reaction m ixtures were allowedtoequilibrateatroomtem- perature for 10 min. Radioactivity was monitored using a B eckman LS 6500 scintillation counter. Increase in radioactivity (DPM, disin- tegration p er minute) indicates binding of pre-EF-Tu to [ 3 H]GDP [14]. Binding of pre-EF-Tu to [ 3 H]GDP suggests that this protein (pre-EF- Tu) is p robably in a physiologically active form [14]. S imilar results were obtained in a duplicate experiment. Ó FEBS 2004 Chaperone activity of recombinant maize EF-Tu (Eur. J. Biochem. 271) 3687 aggregation and almost completely suppressed it a t a pre- EF-Tu : MDH molar ratio of 10 (Fig. 5 C). O valbumin (3 l M , Fig. 5 C) and BSA (not shown), in contrast, did not protect MDH from thermal aggregation. Recombinant maize pre-EF-Tu protected CS and MDH from thermal inactivation The recombinant maize pre-EF-Tu protected CS from thermal inactivation. The enzymatic act ivity of CS was severely halted when 2 l M CS was h eated at 4 3 °C alone or in the p resence of e ither 4 l M BSA o r 4 l M ovalbumin; l ess than 20 % o f the original CS activity r emained af ter 30 min at 43 °C ( Fig. 6A). Upon temperature shift of the samples to room temperature, the CS activity did not change significantly, a s less than 20% of the original CS activity remained after 75 min of recovery (Fig. 6A). In contrast, when 2 l M CS was heated at 43 °C in the presence of 4 l M pre-EF-Tu, 46% of CS activity remained after 30 min (Fig. 6 A). During the recovery period, the activity of C S gradually increased, reaching a maximum of 68% of its original activity after 45 min (Fig. 6 A). Similar results on A B Fig. 4. Heat stability of purified recombinant maize pre-EF-Tu (EF-Tu). (A) Relative ligh t sc attering of purified pre-EF-Tu du ring in cubation at 45 °C.Aliquotofproteinsample(1 mL)wasincubatedat45 °Cfor 45 min, and light scattering was monitored at 320 nm. As a control, heat-labile CS w as used. D ata represent averages of two independe nt experiments. Bars indicate standard e rrors. N ote that during i ncu ba- tion at 45 °C there is no increase in relative light scattering indicating that the purified maize pre-EF-Tu is heat stable at this temperature. (B) Western b lot of purified pre-EF-Tu. A sample of purified prote in (1 mL ; 0.5 l M )wasincubatedat25°C (control) or 45 °C for 45 min. Following incubation, samples were centrifuged and the supernatant of each sample was analyzed for the presence of pre-EF-Tu using Westernblottingandanti-myc Ig. Equal volumes of protein samples were loaded in each lane. Note that t he pre-EF-Tu protein band is present in the sample h eated at 45 °C, in dicating pre-EF-Tu stability at this temperature. Fig. 5. Effect o f recombinant maize p re-EF-Tu (EF-Tu) o n thermal aggregation o f citrate synthase (CS; A and B) and malate dehydrogenase (MDH; C). In separa te trials, CS a nd MDH were mixed with various amounts of pre-EF-Tu . Three contro ls were used: C S or MD H alone, CS or MDH mixed with ovalbumin, and C S or MDH mixed with BSA (0.5 l M BSAwasmixedwithCSand3l M BSA was mixed with MDH; not shown). Mixtures (to tal volume o f each m ixture, 1 mL) were incubated a t in dicated tem per ature f or 45 min. During incuba- tion, samples were monitored for their absorbance a t 3 20 nm, which is indicative of l ight scattering due t o C S or MDH aggregation [ 20,21]. Data represent averages of two inde pend ent experiments. Bars indicate standard errors. Note that pre-EF-Tu p rotected CS an d MDH from thermal a ggregation. Note: BSA and ovalbumin wer e chosen as con- trol proteins because they are k nown to b e r elatively heat stable a nd have been used in chaperone (protein aggregation) studies [40]. In addition, our preliminary ligh t scattering expe rimen ts showed that BSA and ovalb umin are stable at 45 °C, as no increase in light scattering (indicative o f protein aggregation [20]) was d etected when BSA or ovalbumin were heated at this te mpera ture for 45 min (not shown). 3688 D. Rao et al.(Eur. J. Biochem. 271) Ó FEBS 2004 CS activity were obtained when this enzyme w as heated at other high temperatures, 38 °C, 41 °C, and 45 °C (not shown). The recombinant pre-EF-Tu also protected MDH from thermal inactivation. When 0.5 l M MDH was heated at 43 °C a lone or in the presence of either 2 l M BSA o r 2 l M ovalbumin, t he MDH activity was very low, le ss than 1% of its o riginal a ctivity a fter 3 0 min (Fig. 6 B). However, when MDH was heated in the p resence of 2 l M pre-EF-Tu, 5 0% of MDH activity remained immediately after heating (Fig. 6B). During the recovery period, the activity of MDH did not change significantly (Fig. 6 B). A similar effect of pre-EF-Tu on MDH activity was seen when MDH was incubated at 38 °C, 41 °C, and 45 °C (not shown). The recombinant pre-EF-Tu did not show an effect on reactivation of heat-inactivated CS and MDH (Fig. 7). When CS and MDH were heated at 43 °C, witho ut pre-EF- Tu, t he activity of CS and MDH w as severely reduc ed, a nd the addition of pre-EF-Tu after heating did not change their activity during the recovery (Fig. 7). Discussion Elevation of ambient temperature ( heat shock or heat stress) affects cell metabolism, causing changes in the rates of biochemical reactions and injuries to cellular membranes [26,27]. Moreover, increases in ambient temperature also cause denaturation and aggregation o f most proteins [ 27], but protein denaturation due to heat shock is reversible unless followed by aggregation [28]. Plants generally respond to high temperatures with the induction of heat shock proteins ( HSPs). HSPs are t hought to play a role i n heat tolerance by acting as molecular chaperones; that is, they bind and stabilize o ther proteins, protecting them from thermal aggregation a nd inactivation (thermal damage) [29–31]. Recent studies have suggested that some protein synthesis elongation factors may be involved in heat tolerance by acting as molecular chaperones. Prokaryotic elongation factors, EF-G [23] and EF-Tu [22], for example, interact with unfolded and denatured p roteins, as do molecular chaperones, and protect them from thermal aggregation. Also, E. coli EF-Tu interacts p referentially with h ydropho- bic regions of substrate proteins [32], a strategy used by molecular chaperones to p revent thermal aggregation of their substrate proteins [3 0]. Studies from our laboratory have implicated maize EF-Tu in heat tolerance [5,9–11]. Maize EF-Tu exhibits > 80% amino acid i de ntity w ith bacterial EF-Tu [5], and it has been hypothesized that, in maize, this protein may show chaperone activity similar to prokaryotic EF-Tu [5,9–11]. Fig. 6. Effect of recombinant m aize pre-EF-Tu ( EF-Tu) on the activity of citra te synthase (CS; A) and malate de hydrogenase ( MDH; B) after incubation at 43 °C. In separate trials, CS and MDH were mixed with indicated amounts of recombinant pre-EF-Tu. Reaction mixtures (total volume of each mixture , 1 m L) w ere incubated at 4 3 °Cfor 30 m in. After in cubation, th e mixtures w ere qu ickly c ooled to room temperature and then kept on ice for up to 75 min (75 min recovery). Where indicated, a mixture of CS and BSA or ovalbumin (ovalb.) was used as control. CS and MDH activity was measured at various times of rec overy. Data represent averages of three indepe ndent experiments (standard e rrors are plotted but theyareoftensmallerthanthesym- bols). Note that in the presence of recombinant pre-EF-Tu, CS and MDH showed a r elatively high activity immediately after exposure t o 43 °C ( 0 m in of recovery). Similar results were obtained at 38 °C, 41 °Cand45°C (not shown). Fig. 7. Effect of r ecombinant maize pre-EF-Tu (EF-Tu) o n the a c tivity of h eat-inactivated CS and MDH. CS and M DH were incubated at 43 °C f or 30 min. Followin g incubation, indicated amounts of pre-EF- Tu were add ed t o the he at ed prote in s amples, a nd th e r eaction m ix- tures (total volume of each mixture, 1 mL) w ere allowed to re co ver on ice for 45 min. The residual activity of CS and MDH w as measured at room temperature a t v arious tim es o f recovery. Data represent aver- ages of two independent experiments. Bars indi cate standard errors. Ó FEBS 2004 Chaperone activity of recombinant maize EF-Tu (Eur. J. Biochem. 271) 3689 In this study, we isolated and purified the recombinant precursor of m aize EF-Tu from E. coli and tested it for possible chaperone activity. Before the chaperone studies were cond ucted, the recombinant p re-EF-Tu was tested f or its purity, identity, a bility t o bind GDP (indicative of EF-Tu activity [14]), and the rmal stability. The re sults showed that the recombinant pre-EF-Tu was isolated in h ighly pure form (Fig. 1 ) capable of binding GDP (Fig. 3), and that the identity of the protein, as determined by mass spectrometry, was indeed maize EF-Tu (Fig. 2). In addition, the heat stability tests showed that the recombinant pre-EF-Tu was stable a t 45 °C ( Figs 3 and 4), t he highest temperature used in our study to test th is protein for possible chaperone activity. The thermal stability of m aize pre-EF-Tu obse rved in our study was similar t o the thermal stability of bacterial EF-Tu, which has been shown to b e s table a t t emperatures ranging from 40 °Cto45°C[22]. Importantly, the recombinant maize pre-EF-Tu dis- played chaperone activity. It protected the heat-labile proteins, CS a nd MDH, from thermal a ggregation and inactivation. The protective role of pre-EF-Tu against thermal aggregation was exhibited in a concentration dependent manner with t he most effective p rotection seen when the molar ratio o f pre-EF-Tu : substrate p rotein (CS or MDH) was 6.7 for CS and 10 for MDH. The results on the influence of m aize pre-EF-Tu on thermalaggregationofCSweresimilartothosereportedfor bacterial EF-Tu [22]. Bacterial EF-Tu w as also found t o protect CS from thermal aggregation in a concentration dependent manner [22]. When 0.8 l M CS was heated at 43 °C it formed insoluble aggregates [22]. However, the addition of 2 l M bacterial EF-Tu partially reduced, and 5 l M EF-Tu completely suppressed, the thermal aggrega- tion of CS [22]. Thus, the most effective b acterial EF- Tu : CS molar ratio that suppressed CS aggregation was 6.25 [22], and this is similar to the maize pre-EF-Tu : CS molar ratio (6.7 at 45 °C) observed in our study. Bacterial EF-Tu has been found to facilitate refolding of denatured proteins [ 22,33]. Kudlicki et al.[33]haveshown that Therm us t hermophilus EF-Tu h as chaperone- like capa- city to as sist in the refolding of denatured r hodanese. A lso, Caldas et al . [22] have observed refolding of urea-denatured CS in the p resence of E. coli EF-Tu. In c ontrast to bacterial EF-Tu, however, recombinant maize pre-EF-Tu does not seem to have an effect on renaturation of denatured proteins. Rather, this protein appears t o be important in protecting proteins from thermal d amage during exposure to heat stress. As seen in our study, maize pre-EF-Tu helped CS and MDH maintain a relatively high activity during h eat stress (Fig. 6) but it had no effect on r eactivation o f t hese enzymes f ollowing their almost complete thermal inactiva- tion (Fig. 7). The ability of r ecombinant maize pre-EF-Tu to protect model substrates (CS and MDH) from thermal damage provides evidence for the possible role of native EF-Tu in heat tolerance. Native chloroplast EF-Tu is predominantly localized in the chloroplast stroma [11], and it is h ighly possible that during heat stress this protein may protect chloroplast stromal proteins from thermal damage by acting as a molecular c haperone. T his possibility is suppor- ted by Momcilovic & Ristic [11] and Ristic et al .[10] who found that chloroplast stromal proteins from maize genotypes with hi gher levels of EF-Tu display greater heat stability (lower t hermal aggregation) than chloroplast stromal proteins from genotypes with lower levels of EF-Tu. The above hypothesis is also corroborated by studies which showed that wh ole chloroplasts from a high- level EF-Tu maize line are more heat stable than whole chloroplasts from a low-level EF-Tu line [34,35] (in these previous studies, maize EF-Tu was referred to as a 45–46 kDa HSP because the identity of this protein was not known until the report of Bhadula et a l.[5]). One could argue that the chaperone activity observed in our study may be a n attribute of p re-EF-Tu, because o f the presence of chloroplast targeting sequence, and that the native EF-Tu may n ot have chaperone properties. We do not completely rule out this possibility, however, the evidence supports the hypothesis that the native maize EF-Tu displays chaperone activity. Like n ative chloroplast EF-Tu [12], the recombinant pre-EF-Tu shows the ability to bind GDP (Fig. 3), an indication that the targeting sequence does not significantly affect the activity of this protein. Furthermore, as stated earlier, the amino acid s equence of native maize EF-Tu is highly similar to t hat of bacterial EF-Tu [5], which i s known to display chaperone activity [22]. In addition, a comparison of the predicted two- dimensional ( SCRATCH servers; http://www.igb.uci.edu/ tools/scratch/) and three-dimensional [36] s tructure reveals a striking s imilarity between the native m aize EF-Tu and i ts precursor ( pre-EF-Tu) i mplying that the functional proper- ties of the native EF-Tu and pre-EF-Tu are similar. The hypothesis t hat the native maize EF-Tu acts as a c haperone and p rotects c hloroplast proteins f rom thermal aggregation is consistent with the lower thermal aggregation of chloro- plast s tromal proteins in m aize genotypes with higher levels of EF-Tu as outlined above. Plant cells possess many structurally diverse chaperones [30,37,38], some of which, the small heat shock proteins (sHSPs), function in conjunction with other chaperones [21,31]. A model has been proposed for t he activity of sHSPs [31,39]. D uring high t emperature stress, sHSPs bind substrate p roteins in a n A TP-independent manner, p re- venting their aggregation and maintaining them in a state competent f or subsequent ATP-dependent refolding, which is facilitated by other chaperones (e.g. HSP70 system) [21,31,39]. This model is supported by L ee & Vierling [ 31] who demonstrated that the HSP70 system is required for refolding of a sHSP18.1-bound firefly luciferase. Some plant sHSPs, however, can facilitate r eactivation of heat-inactivated proteins d uring recovery from stress with- out the presence of o ther chaperones and ATP. Pea (Pisum sativum L) HSP17.7 and HSP18.1, for example, minimally protected C S activity a t 3 8 °C, but helped this enzyme regain 65–70% of its original activity after 60 min of recovery at 22 °C [20]. The reactivation activity o f HSP17.7 and HSP18.1, how ever, seemed to be limited to tempera- tures b elow 4 0 °C, as these two sHSPs had no effect on CS reactivation following CS exposure to 45 °C[20]. Recombinant maize pre-EF-Tu does not seem to com- pletely fit t he model proposed for the function of sHSPs, and it differs from pea HSP17.7 and HSP18.1 in s ome aspects o f its chaperone activity. Maize pre-EF-Tu appears to be effective in protecting heat-labile proteins from thermal damage without a r equirement for the presence of 3690 D. Rao et al.(Eur. J. Biochem. 271) Ó FEBS 2004 other chaperones and ATP. As our in v itro experiments showed, maize pre-EF-Tu not only protected CS and MDH from thermal aggregation (Fig. 5) but also helped CS and MDH maintain a relativ ely high activity immediately a fter exposure to h eat stress at t emperature above 40 °C (Fig. 6). We do not know, however, if and how maize pre-EF-Tu and/or native EF-Tu may function as molecular chaperones in vivo. Our obs ervation that recombinant maize pre-EF- Tu acts independently in vitro, without a requirement for o ther chaperones and ATP, does not rule out the possibility that in vivo this protein and/or its n ative form may function in cooperation with other chaperones. Further s tudies are needed to investigate this possibility. In conclusion, in this study we demonstrate t hat, in vitro, the recombinant maize pre-EF-Tu acts as a molecular chaperone and protects heat-labile proteins, CS a nd MDH, from thermal aggregation and inactivation. To o ur know- ledge, th is is the first observation of chaperone activity by a plant/eukaryotic precursor of the EF-Tu protein. Previous studies have shown t hat whole chloroplasts [34,35] and chloroplast s tromal proteins [10,11] from maize with higher levels of EF-Tu display greater heat stability (lower t hermal aggregation) than whole chloroplasts and chloroplast stromal proteins f rom maize with lower levels o f EF-Tu. Combined, our current and previous studies [10,11,34,35] strongly support the hypothesis that maize EF-Tu plays a role in heat tolerance by acting as a molecular chaperone and protecting chloroplast stromal proteins from thermal damage. Acknowledgements We acknowledge fi nancial support for this research from the United States Department of Agriculture grant (Agreement no. 99-35100-8559) to Z. Ristic. The authors are thankful to Drs Karen L. Koster and Gary D. Small, The University of South Dakota, Dr Thomas E. Elthon, The University of Nebraska – L incoln, and Dr David P. Horwath, the U .S. Department of A griculture Experimental Research Laboratory, Fargo, N D for c ritical reading of the manuscript. References 1. Brot, N. ( 1977) Translation, transloca tion. In Mo lecu lar Me cha- nisms of Protein Biosynthes is (Weissbach, H. & Pestka, S ., eds), pp. 375–411. Academic Press, N ew York. 2. Miller, D.L. & Weissbach, H. (1977) Factors involved in the transfer of aminoacyl-tRNA to the ribosome. In Molecul ar Mechanisms of Pr ot ein Biosynthesis (Weissbach,H.&Pestka,S., eds), p p. 323–373. Academic Press, Ne w York. 3. Riis, B., Rattan, S.I.S., Clark, B .F.C. & M errick, W.C. (1990) Eukaryotic protein e longation factors. Tr ends Biochem. Sci. 15 , 420–424. 4. Baldauf, S. L. & Palmer, J .D. ( 1990) Evolutionary transfer of t he chloroplas t tufA g ene to the nucleus. Na ture 344, 2 62–265. 5. Bhadula, S.K., Elthon, T.E., H abben, J.E., Helentjaris, T.G., Jiao, S. & Ristic, Z. (2001) Heat-stress induced synthesis of chloroplast protein synthesis elongation factor (EF-Tu ) in a heat-toleran t maize line. Planta 212 , 359–366. 6. Ristic,Z.,Gifford,D.J.&Cass,D.D.(1991)Heatshockproteins in two lines of Zea m ays L . that diffe r in dro ught and h eat resistance. Pl ant Physiol. 97, 1430–1434. 7. Ristic, Z., Williams, G., Yang, G., Martin, B. & Fullerton, S. (1996) Dehydration, damage to cellular membranes, and heat- shock proteins i n m aize hybrid s from d ifferent c limates. J. Pl ant Physiol. 149 , 424–432. 8. Ristic, Z., Yang, G., Martin, B. & Fu llerton, S. (1998) Evidence of association between specific heat-shock protein(s) and th e drought and h eat tolerance phen otype in m aize. J. Plant Physiol. 153, 497–505. 9. Moriarty,T.,West,R.,Small,G.,Rao,D.&Ristic,Z.(2002) Heterologous expression of maize chloroplast protein syn thesis elongation f actor (E F-Tu) e nhances Escherichia coli viability under heat s t ress. Plant Sci. 163, 1 075–1082. 10. Ristic, Z., Wilson, K., Nelsen, C., M omcilovic, I., Kobayashi, S ., Meeley, R., Musz ynski, M. & H abben, J. (2004) A maize mutant with d ecreased c apacity to accumulate c hloroplast protein synth- esis elongation f actor (EF-Tu) displays reduced tolerance to heat stress. Plant Sci. 167, doi:10.1016/JPLANTSCI.2004.07.016. 11. Momcilovic, I. & R istic, Z. (2004) Localization a nd abundance of chloroplast p rotein synthes is elongation factor ( EF-Tu) and heat stability of c hloroplast str omal proteins in m aize. Plant Sci. 166, 81–88. 12. Stanzel, M., Schon, A. & Sprinzl, M. (1994) Discrimination against m isacylated tR NA by chloroplast elongation f actor Tu. Eur. J. Biochem. 219, 4 35–439. 13. Cull, M. & McHenry, C .S. (1990) Pr eparation of e xtracts from prokaryotes. Methods Enzymol. 182 , 147–153. 14. Zhang, Y.X., Shi, Y ., Zhou, M. & Petsko, G.A. (1994) Cloning, sequencing, and expression in Escherichia coli of the gene encoding a 45-kilodalton protein, elongation factor Tu, from Chlamydia trachomatis ser ovar F. J. Bacteriol. 176 , 1184–1187. 15. Miller, D .L. & Weissbach, H. (1974) Elongation factor Tu and aminoacyl-tRNA.EFTu GTP complex. Methods Enzymol. 30, 219–232. 16. Laemmli, U.K. (1970) Cleavage of structural proteins during the assembly of t he head of bacteriophage T4. Natur e 227, 680–685. 17. Koc, E.C., Burkhart, W., Blackburn, K., Moyer, M.B., Schlatzer, D.M., Moseley, A. & Spre mulli, L.L. (2001) The large subunit of the m ammalian mitochondrial ribosome: analysis of the comple- ment of ribosomal proteins p resent. J. Biol. Chem. 276, 43958– 43969. 18. Kinter, M. & Sherman, N.E. (2000) The preparation of protein digests for mass spectrometric sequencing experiments. In Protein Sequencing and Identification U sing Tan dem Mass Spectrometry (Desiderio, D.M. & N ibbering, N.M.M., eds), pp. 1 47–164. Wiley- Interscience, New York. 19. Jaenicke, R. & Rudolph, R. (1989) Folding Proteins. In Protein Structure: a Practical Approach (Crei ghton, T.E., ed.), pp. 191– 223. I RL Press, O xford. 20. Lee, G. J., Pokala, N. & Vierling, E. (1995) Structure and in vitro molecular chaperone activity of cytosolic small heat shock pro- teins from pea. J. Biol. C hem. 270, 1043 2–10438. 21. Lee, G.J., Roseman, A.M., Saibil, H .R. & V ierling, E. (1997) A small h eat shock protein s tably binds he at-denat ured model s ub- strates and ca n maintain a substrate i n a folding-competent state. EMBO J. 16 , 659–671. 22. Caldas, T.D., Ya agoubi, A.E. & Richarme, G. (1998) Chaperone properties of bacterial elongatio n factor. J. Biol. Chem. 273, 11478–11482. 23. Caldas, T ., Laalami, S. & Richarme, G. (2000) C haperone prop- erties of bacterial elon gation f actor EF-G a nd initiation factor IF2. J. Biol. Chem. 275, 855–860. 24. Srere, P. A. (1969) Citrate synthase. Methods Enzymol. 13 , 3–11. 25. Banaszak, L. & Bradshaw, R.L. (1975) Malate dehydrogenase. In The Enzymes, XI (Boyer, P ., ed.), pp. 369–397. Academic Press, New York. Ó FEBS 2004 Chaperone activity of recombinant maize EF-Tu (Eur. J. Biochem. 271) 3691 26. Berry, J.A. & B jorkman, O. (1980) Photosynthetic response and adaptation to temp erature in higher plants. Ann. Rev. Plant Phy- siol. 31, 4 91–543. 27. Levitt, J. (1980) Responses of Plants to Environmental Stress: Chilling, Freezing and High T emperature Stresses, 1 (Kozlowski, T.T., ed.). Academic Press, New York. 28. Tanford, C. (1968) Protein denaturation. In Advances in Protein Chemistry, Vol. 23 (Anfinsen, C.B., E dsall, J.T., A nson, M.L. & Richards, F.M. , eds), pp. 121–282. Academic Press, New York. 29. Vierling, E. (1991) The r oles of heat shock proteins in plants. Ann. Rev. Pla nt Physiol. Plant Mol. B iol. 42, 579 –620. 30. Hendrick,J.P.&Hartl,F.U.(1993) Molecular chaperone func- tions of h eat shock proteins. Ann. Rev. Bioch em. 62, 349–384. 31. Lee, G.J. & Vierling, E. (2000) A small heat shock protein cooperates with heat shock protein 70 sy stems to reactivate a heat- denatured protein. Plant Physiol. 122 , 189–197. 32.Malki,A.,Caldas,T.,Parmeggiani,A.,Kohiyama,M.& Richarme, G. ( 2002) S pecificity of e longation factor EF-Tu for h ydrophobic peptides. Biochem. Bio phys. R es. C ommun. 296, 749–754. 33. Kudlicki,W.,Coffman,A.,Kramer,G.&Hardesty,B.(1997) Renaturation of rhodanese by translational el ongation f actor (EF)Tu: protein refolding by EF -Tu flexing. J. Biol. Chem. 272, 32206–32210. 34. Ristic, Z. & Cass, D.D. (1992) Chloroplast structure after water and high-temperature stress in two lines of maize th at differ in endogenous levels of abscisic acid. Int. J. P lant Sci. 153, 186–196. 35. Ristic, Z . & Cass, D.D. (1993) Dehydration avoidance and damage to the plasma an d thylakoid membranes in lines of maize differing in endogenous levels of abscisic acid. J. Plan t Physiol. 142, 7 59–764. 36. Schwede, T., K opp, J., Guex, N. & Peitsch, M.C. ( 2003) SWISS- MODEL: an automated p rotein homology-modeling server. Nucleic A cids Res. 31, 3381–3385. 37. Georgopoulos, C. & Welch, W.J. (1993) Role of the m ajor heat shock proteins as mo lecular chaperones. Ann. Rev. Cell Biol. 9, 601–634. 38. Boston, R.S., Viitanen, P.V. & Vierling, E. (1996) Molecular chaperones and protein folding in plants. Plant Mol. Biol. 32 , 191– 222. 39. Sun, W ., Montagu, M .V. & Verbruggen, N. (2002) Small heat shock p roteins and stress tolerance i n plants. Biochim. Bio phys. Acta 1577, 1–9. 40. Oh, H.J., Che n, X. & Subjeck, J.R. (1 997) Hsp110 protec ts heat- denatured proteins and confers c ellular thermoresistance. J. Bi ol. Chem. 272, 31636–31640. 3692 D. Rao et al.(Eur. J. Biochem. 271) Ó FEBS 2004 . Chaperone activity of recombinant maize chloroplast protein synthesis elongation factor, EF-Tu Damodara Rao 1 , Ivana Momcilovic 1 ,. synthesis e longation factor, EF-Tu, is a protein that carries aminoacyl-tRNA to the A -site o f the ribosome during t he elongation phase of protein synthesis.