Different roles of functional residues in the hydrophobic binding site of two sweet orange tau

Different roles of functional residues in the hydrophobic binding site of two sweet orange tau glutathione S-transferases Angela R. Lo Piero, Valeria Mercurio, Ivana Puglisi and Goffredo Petrone

Dipartimento di Scienze Agronomiche, Agrochimiche e delle Produzioni Animali (DACPA), Universita` di Catania, Italy

Keywords glutathione S-transferase; H-site; site-directed mutagenesis; sweet orange; tau class glutathione S-transferase

Correspondence A. R. Lo Piero, Dipartimento di Scienze Agronomiche, Agrochimiche e delle Produzioni Animali (DACPA), Universita` di Catania, Via S. Sofia 98, 95123, Catania, Italy Fax: +39 95 7141581 Tel: +39 95 7580238 E-mail: rlopiero@unict.it

(Received 27 July 2009, revised 7 October 2009, accepted 5 November 2009)

doi:10.1111/j.1742-4658.2009.07481.x

Glutathione S-transferases (GSTs) catalyze the conjugation of glutathione to hydrophobic compounds, contributing to the metabolism of toxic chemicals. In this study, we show that two naturally occurring tau GSTs (GSTUs) exhibit distinctive kinetic parameters towards 1-chloro-2,4-dini- trobenzene (CDNB), although they differ only in three amino acids (Arg89, Glu117 and Ile172 in GSTU1 are replaced by Pro89, Lys117 and Val172 in GSTU2). In order to understand the effects of the single mis- matched residues, several mutant GSTs were generated through site-direc- ted mutagenesis. The analysis of the kinetic parameters of the mutants led to the conclusion that Glu117 provides a critical contribution to the maintenance of a high-affinity CDNB-binding site. However, the substitu- tion E117K gives rise to mutants showing increased kcat values for CDNB, suggesting that Lys117 might positively influence the formation of the transition state during catalysis. No changes in the Km values towards glutathione were found between the naturally occurring GSTs and mutants, except for the mutant caused by the substitution R89P in GSTU1, which showed a sharp increase in Km. Moreover, the analysis of enzyme reactivation after denaturation showed that this R89P substitution leads to a two-fold enhancement of the refolded enzyme yield, suggesting that the insertion of proline might induce critical structural modifications. In contrast, the substitution P89R in GSTU2 does not modify the reacti- vation yield and does not impair the affinity of the mutant for glutathi- one, suggesting that all three residues investigated in this work are fundamental in the creation of enzymes characterized by unique biochem- ical properties.

Introduction

The glutathione S-transferases (GSTs; EC 2.5.1.18) are members of a multifunctional superfamily of enzymes catalyzing the conjugation of glutathione (GSH) to the electrophilic groups of hydrophobic and usually cytotoxic molecules of either endogenous or exogenous

origin [1–4]. GSTs are widely distributed in nature from humans to bacteria [5–7]. In plants, the GSH addition reaction is coupled to the vacuolar compart- mentation of the GSH conjugates because of the lack of an effective excretion pathway, which is active in

FEBS Journal 277 (2010) 255–262 ª 2009 The Authors Journal compilation ª 2009 FEBS

255

Abbreviations 4-NPB, 4-nitrophenethyl bromide; CDNB, 1-chloro-2,4-dinitrobenzene; ECA, ethacrynic acid; GmGSTU-4-4, Glycine max tau glutathione S-transferase-4-4; GSH, glutathione; G-site, glutathione-binding site; GST, glutathione S-transferase; GSTU, tau glutathione S-transferase; H-site, hydrophobic binding site; NBD-Cl, 7-chloro-4-nitrobenzo-2-oxa-1,3-diazole.

A. R. Lo Piero et al. Functional role of the GST H-site residues

negatively and positively charged amino acids stabi- lized by an array of hydrogen bonds, and appears to be a functionally conserved motif in all GST classes [27]. In the Glycine max GSTU-4-4 (GmGSTU4-4)– GSH complex, the strictly conserved residues Arg18, Glu66, Ser67 and Asp103 appear to form the proposed electron-sharing network [26]. Studies of the catalytic properties of Pinus tabulaeformis GSTU1 by site-direc- ted mutagenesis, which demonstrate the crucial role of both Ser67 and Glu66 in GSH binding, corroborate these findings [28]. In contrast, the H-sites of GSTs exhibit a low degree of sequence identity, and hence unique structures that reflect different functions in vivo [18]. In a previous study, we isolated from sweet orange leaves two distinct GSTU genes sharing 98.6% homology at the nucleotide levels and containing a 651 bp ORF. The encoded proteins differ in only three amino acids: the triad Arg89, Glu117 and Ile172 found in the isoform GSTU1 is replaced by the triad Pro89, Lys117 and Val172 in the GSTU2 isoform, all of the mismatches being located in the H-sites of the enzymes [11]. As long as the enzymes were expressed in vitro and purified, they exhibited different specific activity with 1-chloro-2,4-dinitrobenzene (CDNB) as substrate, GSTU1 showing a value three-fold lower than that observed for GSTU2 [11]. In the present work, site- directed mutagenesis was used to evaluate the effects of the single mismatched residues on the kinetic prop- erties of isoforms GSTU1 and GSTU2, and also to address questions regarding the functional roles of these H-site residues. The results have both academic relevance and practical importance, as they will form the basis for the design of new engineered GSTs show- ing altered substrate specificity and enhanced activity towards xenobiotics.

Results and Discussion

sweet orange GST subunits

Wild-type (GSTU1 and GSTU2) and mutant GSTs were expressed and purified as described in Experimen- tal procedures. After purification, all recombinant pro- teins showed a single band in SDS ⁄ PAGE, and an identical molecular mass of about 26.0 kDa, corre- the sponding to the calculated molecular mass of recombinant (Fig. 1). CDNB is generally considered to be the classic GST substrate, because most GST isoenzymes display high activity towards it. It has been shown that the sweet orange GST isoforms exhibit quite different specific activities with CDNB as substrate, GSTU2 being three-fold more active than GSTU1 [11]. Conse- quently, steady-state kinetic characterization of both isoforms with respect to CDNB was performed prior

animals [8,9]. GSTs, in addition to transfer of GSH to toxic compounds, act as GSH-dependent peroxidase, isomerase and oxidoreductases, playing pivotal roles in plant cell protection, as reviewed by Moons [10]. Plant GSTs are abundantly expressed, and show major tran- scriptional regulation. It has been reported that GST transcript levels can markedly increase in response to a wide variety of stressful conditions, such as herbi- cides [1,11], chilling [11,12], hypoxic stress [13], dehy- dration [14], wounding [15], and pathogen attack [16,17]. Most GSTs are active as dimers, either homo- dimers or heterodimers of subunits ranging from 23 to 30 kDa in size. Subunits of all known GST structures exhibit a two-domain fold, the N-terminal domain and C-terminal domain, including the highly conserved GSH-binding site (G-site) and the more divergent cosubstrate-binding domain or hydrophobic binding site (H-site) [7,18]. The G-site includes both a-helices and b-strands as secondary structure elements. The topological arrangement of these elements is usually bababba, similar to the thioredoxin fold of other GSH-binding or cysteine-binding proteins. The H-site is entirely helical, with a variable number of a-helices, depending on the specific enzymes [19]. Mechanisti- cally, the catalysis of the nucleophilic aromatic substi- tution reactions comprises the substrate binding to the enzyme’s active site, ionization of GSH to form the highly reactive thiolate anion, and nucleophilic attack by the thiolate at the substrate electrophilic center. A hydroxyl group, provided in most plant GSTs by a serine, is considered to be required for the correct ori- entation and stabilization of the deprotonated thiolate anion in the active site of the enzyme [20]. On the basis of protein sequence similarity, active site residue and gene organization plant GSTs are grouped in four main classes (phi, tau, zeta, and theta) [18,21,22]. The majority of the plant GSTs belongs to the tau (GSTU) and phi classes, which are plant-specific. Among the plant GSTs, GSTUs are the most numerous, and members of this class overlap in their function of enhancing crop stress tolerance. Despite the important roles of the GSTUs, extensive analysis of critical resi- dues in both domain sites is needed, although the highly conserved nature of the G-site [23] and the recent crystallographic characterization of two GSTUs [24–26] have allowed researchers to formulate some general considerations about the N-terminal domains of the enzymes. Moreover, the existence of a conserved electron-sharing network that helps the glutamyl c-car- boxylate of GSH to function as a catalytic base, accepting the proton from the thiol group to form an anionic GSH, has been reported [27]. This network is characterized by an electrostatic interaction between

FEBS Journal 277 (2010) 255–262 ª 2009 The Authors Journal compilation ª 2009 FEBS

256

A. R. Lo Piero et al. Functional role of the GST H-site residues

6

7

1

2

3

4

5

sources

from different plant

36 kDa 29 kDa

Fig. 1. SDS ⁄ PAGE of the purified wild-type and mutant GSTs. Lane 1: unpurified E. coli extract. Lane 2: purified wild-type GSTU1 (REI). Lane 3: purified wild-type GSTU2 (PKV). Lane 4: purified PEI mutant. Lane 5: RKV mutant. Lane 6: purified PKI mutant. Lane 7: molecular mass marker. The mutant enzymes are referred to by their distinctive triad of amino acids located, respectively, at posi- tions 89, 117, and 172.

to any other investigation concerning their catalytic mechanism. The values of kinetic parameters obtained by non-linear regression analysis are listed in Table 1. The results showed that GSTU1 had a lower apparent Km for CDNB (0.75 mm) than that exhibited by GSTU2 (1 mm), this finding being consistent with a higher affinity of GSTU1 for this substrate. The Km values of GSTU1 and GSTU2 for CDNB are in the range of those of plant GSTs, whose values vary from 0.12 to 4.43 mm [24,29,30]. As expected on the basis of the complete identity of the G-site sequences, the same Km for GSH was registered for both isoforms (0.5 mm), which is in general agreement with other

published Km values for GSH [31]. However, in the case of both substrates, GSTU2 showed higher cata- lytic efficiency (kcat ⁄ Km) as well as higher kcat values (Table 1). The alignment of the H-site sequences of 20 showed that GSTUs Pro89 and the Lys117 found in GSTU2 are strictly conserved residues, whereas, at the 172 position, the isoleucine is well conserved but valine or threonine can also be found (Fig. S1). Therefore, the naturally occur- ring GSTU1 (Arg89, Glu117, and Ile172) probably has unique features, and might perform a different func- tional role in vivo than GSTU2. This assumption is also supported by the analysis of gene expression, which showed that the GSTU1 gene is induced by cad- mium sulfate, CDNB, cyhalothrin, and cold stress, whereas GSTU2 is constitutively expressed [11]. Given that these naturally occurring GSTUs provided a good opportunity to understand how specific amino acids might contribute to the enzyme’s biochemical behav- ior, several mutant GSTs, hereafter referred to by their distinctive amino acid triad, were generated by site- directed mutagenesis. In particular, the RKI and RKV mutants were obtained by individually replacing the Glu117 of GSTU1 (REI) with Lys117, and the Pro89 of GSTU2 (PKV) with Arg89, respectively. Moreover, the PEI and PKI mutants were obtained by replacing Arg89 of GSTU1 (REI) with Pro89, and Glu117 of the PEI mutant with Lys117, respectively. Steady-state kinetic analysis of purified RKI and RKV mutants showed sharp increases of approximately six-fold and five-fold in the Km value for CDNB as compared with GSTU1 (REI) (Table 1). Consequently, the results sug- gest that Lys117-containing enzymes, either wild type or mutant, do not easily accommodate CDNB in their active site, and also highlight the critical contribution of Glu117 to CDNB recognition. Accordingly, the PEI mutant, in which the Glu117 is restored as compared with the PKI mutant, shows a Km value for CDNB similar to that of GSTU1 (REI) (Table 1). However, the kcat values for both GSTU2 (PKV) and the E117K

Table 1. Steady-state kinetic constants of wild-type and mutant GSTs. The kinetic parameters were calculated by using nonlinear regression analysis with the HYPER32 program. Each value represents the mean ± standard deviation of three replicates.

CDNB GSH

Km (mM) Km (mM) GST Vmax (lMÆmin)1) kcat ⁄ Km )1Æs)1) (M Vmax (lMÆmin)1) kcat ⁄ Km )1Æs)1) (M kcat (s)1 · 103) kcat (s)1 · 103)

31.7 ± 1.45

FEBS Journal 277 (2010) 255–262 ª 2009 The Authors Journal compilation ª 2009 FEBS

257

13.8 ± 0.24 76.9 ± 0.4 147.0 ± 0.9 129.4 ± 0.5 23.8 ± 1 108.1 ± 1 257.1 ± 2 119.4 ± 1 160.7 ± 1.6 GSTU1 (REI) GSTU2 (PKV) RKI RKV PKI PEI 0.75 ± 0.03 1.0 ± 0.07 5.0 ± 0.2 4.0 ± 0.4 4.0 ± 0.4 0.85 ± 0.05 2.0 ± 0.10 4.0 ± 0.06 9.0 ± 0.5 8.0 ± 0.25 4.5 ± 0.15 3.0 ± 0.07 28.8 ± 0.07 108.1 ± 5.4 51.4 ± 2.0 29.8 ± 0.6 40.0 ± 1.5 33.9 ± 2.4 0.5 ± 0.02 0.5 ± 0.03 0.6 ± 0.03 0.35 ± 0.02 0.7 ± 0.04 4.0 ± 0.02 1.0 ± 0.1 4.0 ± 0.15 5.0 ± 0.20 3.3 ± 0.10 4.0 ± 0.15 7.0 ± 0.3 64.0 ± 0.63 70.7 ± 0.7 27.7 ± 1.15 153.8 ± 6.94 245.0 ± 4.26 369.7 ± 21.12 91.4 ± 2.16 17.6 ± 0.24

A. R. Lo Piero et al. Functional role of the GST H-site residues

indicate that

mutants (RKI, RKV, and PKI) were increased as com- pared with those of the Glu117-containing enzymes (REI and PEI) (Table 1). This finding suggests a nega- tive influence of Glu117 on the catalytic event. It has the addition of GSH to CDNB been shown that occurs via an addition–elimination sequence involving a short-lived r-complex intermediate (Meisenheimer complex) as transition state [32]. The structure of the indicates that the enzyme might transition state [32] provide electrophilic assistance interactions to the developing charge of the r-complex o-nitro group [33]. The comparison of the crystal structure of the rat M1- 1 GST in complex with the transition state analog and with the product reveals two completely different binding modes for the intermediate and the product, suggesting that a specific motion is associated with the intermediate into the product the collapse of [34]. More recently, Axarli et al. [26] have also shown that the catalytic reaction of GmGSTU4-4 is barely sensitive to the nature of the leaving group, as the sub- stitution of the chlorine atom with the more electro- negative fluorine in the CDNB molecule did not affect the kcat values. Altogether, these results are consistent with the idea that the rate-limiting step of the CDNB– GSH catalytic reaction is the physical event of product release, probably involving structural motions or con- formational changes of the ternary complex. In this context, we propose that the substitution in the orange GSTU1 of a nucleophilic residue by an electrophilic one (E117K) could function in strongly favoring the formation of the transition state during GSH addition to CDNB by providing the required electrophilic assis- tance to the developing transition state.

the putative structural

catalytic

Interestingly,

efficiency.

for Citrus sinensis GSTU1 (REI) and the PEI mutant, as well as for Citrus sinensis GSTU2 (PKV) and the RKV mutant, by submitting the amino acid sequences to swiss-model [35]. Then, the alignments of the 3D structural homology models were performed between the wild-type GSTs and their respective mutants at the 89 position (REI–PEI, and PKV–RKV) by the web-based program matras [36]. The analysis of the superimposed REI–PEI models revealed relevant con- formational differences between the enzyme structures, mainly involving a-helix1 (from Ser13 to Gly27) and a-helix3 (from Ser67 to Asp80), which represent the catalytic ‘core’ of the G-site [26] (numbering of helices is consistent with that of other GSTs) (Fig. 2A). In contrast, minimal structural perturbations of the afore- mentioned a-helix1 and a-helix3 were observed in the case of the superimposed PKV–RKV models, whose major nonoverlapping regions are, instead, localized in the H-site of the enzymes (Fig. 2B). These findings are in agreement with the different values of apparent Km for GSH observed between GSTU1 (REI) and the PEI the substitution mutant, and overall R89P in GSTU1 might modify the architecture of the G-site, thus negatively influencing the enzyme’s affinity for GSH. Furthermore, Table 2 shows the recovered activity following denaturation and refolding of wild- type GSTUs and the PEI and RKV mutants. The reac- tivation yield of the PEI mutant was two-fold higher than that observed in GSTU1 (REI) (Table 2), thus supporting our hypothesis that Pro89 might have a structural role in the PEI mutant. However, the recov- ered activities of both GSTU2 (PKV) and the RKV mutant were similar, suggesting that the substitution P89R in GSTU2 (PKV) does not affect the refolding process. Consequently, role assigned to the Pro89 of the PEI mutant cannot be attributed to the Pro89 of GSTU2 (PKV), as the RKV mutant, arising from the P89R substitution in GSTU2 (PKV), has both similar Km values (Table 1) reactivation yields after denaturation and similar (Table 2) as GSTU2 (PKV). Therefore, the results lead to the conclusion that all three amino acids inves- in the creation of tigated in this work take part enzymes showing unique structures and, consequently, functions.

examined,

the

As regards the analysis of kinetic parameters regard- in vivo substrate GSH, the mutants ing the natural containing Lys117 showed apparent Km values similar to those reported for the wild-type GSTs (Table 1). However, the RK and PK enzymes, both wild type and mutants, showed a strong increase in the kcat ⁄ Km values with respect to the Glu117-containing enzymes (REI and PEI) (Table 1), indicating that the substitu- tion E117K is crucial to the formation of enzymes with the PEI higher mutant showed an eight-fold increase in the Km value for GSH as compared with GSTU1 (REI) (substitution R89P) (Table 1), suggesting that such mutation of the H-site might exert its effects upon the G-site kinetic properties. Recently, Axarli et al. [25] reported the crystal structure of the GSTU4-4 from soybean, which shares 71% sequence similarity with the orange GSTUs [11]. The consistency in the fold of GSTs and structure of GmGSTU4-4 the availability of prompted us to construct molecular homology models

The substrate specificity of the sweet orange GSTs was also investigated in order to identify catalytic activities that may be related to their biological func- tion. To this end, some substrates in addition to CDNB were including 7-chloro-4-nitro- benzo-2-oxa-1,3-diazole (NBD-Cl), with which mam- malian a-class GSTs show high activity [37], the alkyl halide 4-nitrophenethyl bromide (4-NPB), related to

FEBS Journal 277 (2010) 255–262 ª 2009 The Authors Journal compilation ª 2009 FEBS

258

A. R. Lo Piero et al. Functional role of the GST H-site residues

A

(Table 3). Interestingly,

B

a,b-alkenals, which may be formed during oxidative stress [29] (Table 3). Overall, CDNB is the preferred substrate of both naturally occurring (REI and PKV) and mutant GSTs, with the RKV and PKV enzymes showing the highest specific activity. Relatively lower activity was detected with the mammalian GST sub- strate NBD-Cl the mutant enzyme RKV is able to conjugate 4-NPB to GSH, whereas wild types, as well as other mutant GSTs, are not, thus suggesting the crucial role of Val172, but not of Ile172, in creating the active site architecture that is suitable to accommodate the aforesaid substrate. This finding is of particular interest, as alkyl halides have interest in view of their occurrence as toxicological environmental pollutants [38]. Therefore, the RKV mutant, owing to the distinguishing ability to conju- gate 4-NPB to GSH (Table 3) and to the extremely high catalytic efficiency towards GSH (Table 1), exhib- its great potential for the development of enzymes with novel properties, e.g. the ability to reclaim strongly contaminated environments through nucleophilic sub- stitution reactions involving either aryl halide or alkyl halide xenobiotics. None of the enzymes is active with ECA, suggesting that wild-type and mutant GSTs might not be directly involved in the removal of harm- ful oxidative stress byproducts (Table 3). The data showing that recombinant orange GSTs do not exhibit in vitro glutathione peroxidase activity (not shown) support this last hypothesis.

Experimental procedures

Molecular cloning of sweet orange GSTU1 and GSTU2

Cloning of sweet orange GSTU1 and GSTU2 genes and their transfer into the expression vector pEXP1–DEST (Invitrogen, Carlsbad, CA, USA) was as described by Lo

Fig. 2. Representation of the 3D homology models of the wild-type GSTs and their respective mutants at position 89. (A) Superposition of wild-type GSTU1 (REI) and the PEI mutant (R89P). (B) Superposi- tion of the wild-type GSTU2 (PKV) and the RKV mutant (P89R). Wild-type GSTs are shown in yellow, and mutants are shown in red. The nonoverlapping regions between the superimposed 3D models appear as red areas.

Table 2. Recovered activity after denaturation and refolding of wild-type and mutant GSTs. The recovered GST activities were measured in standard conditions after dilution of the denaturating agent to an ineffective concentration. Each value represents the mean ± standard deviation of three replicates. Table 3. Specific activity of the wild-type and mutant GSTs towards different substrates. The GST assay was performed in standard conditions in the presence of 1 mM of different sub- strates. Each value represents the mean ± standard deviation of three replicates. ND, not detectable. GST Recovered activity after denaturation (%) Activity [nmol (minÆmg)1)]

GST CDNB 4-NPB NBD-Cl ECA

GSTU1 (REI) PEI GSTU2 (PKV) (RKV) 25 ± 0.5 49 ± 1 36 ± 0.8 37 ± 0.7

GSTU1 (REI) PEI RKI RKV PKI GDSTU2 (PKV) 45.7 ± 1.2 56.3 ± 1.4 66.8 ± 1.8 81.1 ± 1.7 75 ± 1.5 79.2 ± 1.6 ND ND ND 148 ± 1.5 ND ND 26.1 ± 0.8 23.1 ± 0.2 53.6 ± 0.9 17.7 ± 0.2 38.3 ± 0.6 27.2 ± 0.3 ND ND ND ND ND ND

the role of GSTs in detoxification processes, and ethac- rynic acid (ECA), a phenylacetic derivative that con- tains an electrophilic group, similar to the cytotoxic

FEBS Journal 277 (2010) 255–262 ª 2009 The Authors Journal compilation ª 2009 FEBS

259

Piero et al. [11]. The nucleotide sequences of wild-type GSTs were submitted to GenBank under the following accession numbers: EF597102 (GSTU1-coding sequence) and FJ184997 (GSTU2-coding sequence).

included in the peak core were collected and assayed for GST activity (see below). A negative control of the recombi- nant protein expression was also performed by incubating the E. coli extract with empty plasmid. SDS ⁄ PAGE was car- ried out according to the method described in Laemmli [40].

A. R. Lo Piero et al. Functional role of the GST H-site residues

Site-directed mutagenesis

GST enzyme assay

templates

for

Site-directed mutagenesis of GSTU1 and GSTU2 was per- formed by PCR using sweet orange pEXP–GSTU1 and pEXP–GSTU2 as (Gene Tailor Site-directed mutagenesis system; Invitrogen). The PCR reaction mix- tures contained 10 lm each primer, 1 U of Accuprime Pfx DNA polymerase (Invitrogen), 0.3 mm each dNTP, 1 mm MgSO4, and 18 lg of methylated plasmid DNA, in final volume of 50 lL. PCR conditions were optimized to the following: 94 (cid:2)C for 2 min (one cycle), 94 (cid:2)C for 30 s, 55 (cid:2)C for 30 s, and 68 (cid:2)C for 5 min (20 cycles), and 68 (cid:2)C for 10 min (one cycle). All the mutagenesis primers are shown in Table 4. The PCR products were analyzed on a 1% agarose gel containing 0.5 lgÆmL)1 ethidium bromide and purified with the Qiaquick gel extraction kit (Qiagen, the mutants were confirmed by Hilden, Germany). All sequencing the plasmid DNA with the T7 promoter and T7 reverse primers.

In vitro expression and purification of sweet orange wild-type and mutant GSTs

comparison of

The GST assay was routinely performed as described in Lo Piero et al. [3]. In the substrate specificity experiment, the reaction mixture (final volume 0.5 mL), containing 1· NaCl ⁄ Pi (pH 7.4), 1 mm glutathione, 1 mm different sub- strates, and purified recombinant enzymes (10–20 lg), was incubated at 30 (cid:2)C for 15 min. GSH conjugates were detected by measuring the absorbance of samples at 340 nm. Molar extinction coefficients of 9600 m)1Æcm)1 (CDNB), 13 000 m)1Æcm)1 (NBD-Cl) and 1200 m)1Æcm)1 (4-NPB) were used. All measurements were adjusted by subtracting the absorbance values obtained for the nonen- zymatic conjugation of substrates. The apparent Km and Vmax values for GSH of both wild-type and mutant GSTs were determined in the presence of GSH in the concentra- tion range 0.1–1.0 mm and a fixed CDNB concentration of 1 mm. Alternatively, the determination of CDNB apparent Km and Vmax values, GSH was used at a fixed concentration of 1 mm and the CDNB concentration was varied in the range 0.1–1 mm. The kinetic parameters were derived using nonlinear regression analysis with the hyper32 program, available at http://homepage.ntlworld. com/john.easterby/hyper32.html. The experiments were repeated three times on independent enzyme preparations. Kinetic the His-tagged and untagged enzymes showed that the extra six histidines on the N-ter- minus did not interfere with the activity or function of the enzymes (data not shown).

Refolding studies

Refolding experiments were performed according to a slight modification of the method described by Zeng et al. [28]. All enzymes (70 lg) were incubated in a denaturation buf- fer (4 m guanidinium chloride, 0.1 m phosphate, and 1 mm EDTA, pH 6.5) at 25 (cid:2)C for 30 min. At the end of the

In vitro expression of functionally active GSTs, both wild types and mutants, was performed according to the method described in Lo Piero et al. [11]. Briefly, protein expression was achieved in a cell-free system (Expressway Plus; Invitro- gen) by incubating the plasmids (15 lg), in IVPS Plus Escherichia coli extract for 6 h at 25 (cid:2)C, to promote proper protein folding. The recombinant proteins were purified by loading the cell-free extract onto a His-graviTrap column prepacked with Ni2+–Sepharose 6 fast flow (GE Healthcare, Milwaukee, WI, USA). The unspecific bound proteins were removed by washing the column with 20 mm phosphate buffer (pH 8.0), 500 mm NaCl, and 20 mm imidazole. The His-tagged protein was eluted with 500 mm imidazole by recovering 0.2 mL fractions. Fractions were tested for pro- tein content using the Bradford method [39], and those

Table 4. Primers used in site-directed mutagenesis.

Template Mutant Primers (5¢- to 3¢)

GSTU1 RKI

GSTU2 RKV

GSTU1 PEI

FEBS Journal 277 (2010) 255–262 ª 2009 The Authors Journal compilation ª 2009 FEBS

260

PEI PKI E117K-for: AAGACATGGACCACAAAGGGAGAAGAGCAGGAG E117K-rev: TGTGGTCCATGTCTTCGTCGAAGCATC P89R-for: TGGCTTCCCTCTGATCGCTACCAGAGAGCTCAA P89R-rev: ATCAGAGGGAAGCAATGGAGCCTTGTC R89P-for: TTGCTTCCCTCTGATCCCTACCAGAGAGCTCAA R89P-rev: ATCAGAGGGAAGCAATGGAGCCTTGTC E117K-for: TTTGGAAAGTCCAGCATTGAGGCTGAGTGCCCC E117K-rev: GCTGGACTTTCCAAATGTCTCATA

transferases: implications for classification of non-mam- malian members of an ancient enzyme superfamily. Biochem J 360, 1–16.

8 Sandermann H Jr (1992) Plant metabolism of xenobiot-

incubation period, they were rapidly diluted up to an inef- fective guanidinium chloride concentration (1 : 30) in a (0.1 m phosphate, pH 6.5, 1 mm renaturation buffer EDTA), and the recovered activity towards the substrate CDNB was immediately monitored.

ics. Trends Biochem Sci 17, 82–84.

9 Rea PA (1999) MRP sub family ABC transporters from

plants and yeast. J Exp Bot 50, 895–913.

A. R. Lo Piero et al. Functional role of the GST H-site residues

Sequence alignment and structural modeling

10 Moons A (2005) Regulatory and functional interactions of plant growth regulators and plant glutathione trans- ferases (GSTs). Vitam Horm 72, 155–202.

11 Lo Piero AR, Mercurio V, Puglisi I & Petrone G (2009) Gene isolation and expression analysis of two distinct sweet orange [Citrus sinensis L. (Osbeck)] tau-type glutathione transferases. Gene 443, 143–150. 12 Lo Piero AR, Puglisi I, Rapisarda P & Petrone G (2005) Anthocyanin accumulation and related gene expression in red orange fruit induced by low tempera- ture storage. J Agric Food Chem 53, 9083–9088.

The comparison of the C-terminal GST protein sequences was performed by using the multiple sequence alignment program clustalw 1.8. To generate structural models, amino acid sequences of GSTs, both wild type and mutants, were submitted to swiss-model (http://swissmod- el.expasy.org/) [35]. Homology models were generated using the known X-ray structure of GmGSTU4-4 (Protein Data Bank code: 2vo4A) as template. The 3D homology models were compared with matras [36], and jmol version 2.7 was used to generate 3D images.

Acknowledgements

13 Moons A (2003) Osgstu3 and osgtu4, encoding tau class glutathione S-transferases, are heavy metal and hypoxic stress-induced and differentially salt stress-responsive in rice roots. FEBS Lett 553, 427–432.

14 Bianchi MW, Roux C & Vartanian N (2002) Drought

regulation of GST8, encoding the Arabidopsis homo- logue of ParC ⁄ Nt107 glutathione transferase ⁄ peroxi- dase. Physiol Plant 116, 96–105.

15 Vollenweider S, Weber H, Stolz S, Chetelat A &

Financial support was provided by a grant to Dr Angela Roberta Lo Piero by the University of Catania, Fondi del Bilancio Universitario, Progetti di Ricerca di Ateneo (PRA), 2006, Project title: ‘Isolamento e carat- terizzazione di un gene codificante per la glutatione transferasi coinvolta nei meccanismi di trasferimento nel vacuolo dei pigmenti antociani nella polpa di ara- nce rosse e bionde.’

Farmer EE (2000) Fatty acid ketodienes and fatty acid ketotrienes: Michael addition acceptors that accumulate in wounded and diseased Arabidopsis leaves. Plant J 24, 467–476.

16 Mauch F & Dudler R (1993) Differential induction of

References

distinct glutathione transferases of wheat by xenobiotics and by pathogen attack. Plant Physiol 102, 1193–1201.

17 Dean JD, Goodwin PH & Hsiang T (2005) Induction of glutathione S-transferase genes of Nicotiana benth- amiana following infection by Colletotrichum destructi- vum and C. orbiculare and involvement of one in resistance. J Exp Bot 56, 1525–1533.

1 Edwards R & Dixon DP (2000) The role of glutathione transferases in herbicide metabolism. In Herbicides and Their Mechanisms of Action (Cobb AH & Kirkwood RC eds), pp 38–71. Sheffield, Sheffield Academic Press. 2 Axarli I, Rigden DJ & Labrou NE (2004) Characteriza- tion of the ligandin site of maize glutathione transferase I. Biochem J 382, 885–893.

18 Edwards R, Dixon DP & Walbot V. (2000) Plant gluta-

3 Lo Piero AR, Puglisi I & Petrone G (2006) Gene isolation, analysis of expression and in vitro synthesis of a glutathi- one S-transferase from orange fruit [Citrus sinensis L. (Osbeck)]. J Agric Food Chem 54, 9227–9233.

thione transferases: enzymes with multiple functions in sickness and in health. Trends Plant Sci 5, 193–198. 19 Frova C (2006) Glutathione transferases in the genom- ics era: new insight and perspective. Biomol Eng 23, 149–169.

20 Labrou NE, Mello LV & Clonis YD (2001) Functional

4 Dixon DP, Lapthorn A, Madesis P, Mudd EA, Day A & Edwards R (2008) Binding and glutathione conjuga- tion of porphyrinogens by plant glutathione transfer- ases. J Biol Chem 283, 20268–20276.

and structural role of the glutathione binding residues in maize (Zea mays) glutathione S-transferase I. Biochem J 358, 101–110.

21 Dixon DP, Lapthorn A & Edwards R (2002) Plant

glutathione transferases. Genome Biol 3, 3004.1–3004.10.

22 Frova C (2003) The plant glutathione gene family:

5 Hayes JD, Flanagan JU & Jowsey IR (2005) Glutathione transferases. Annu Rev Pharmacol Toxicol 45, 51–88. 6 Allocati N, Federici L, Masulli M & Di Ilio C (2009) Glutathione transferases in bacteria. FEBS J, 276, 58–75.

7 Sheehan D, Meade G, Foley VM & Dowd CA (2001)

genomic structure, functions, expression and evolution. Physiol Plant 119, 469–479.

Structure, function and evolution of glutathione

FEBS Journal 277 (2010) 255–262 ª 2009 The Authors Journal compilation ª 2009 FEBS

261

34 Ji X, Armstrong RN & Gilliland GL (1993) Snapshots

along the reaction coordinate of an SNAr reaction cata- lyzed by glutathione transferase. Biochemistry 32, 12949–12954.

35 Schwede T, Kopp J, Guex N & Peitsch MC (2003)

23 Droog F (1997) Plant glutathione S-transferases, a tale of theta and tau. J Plant Growth Regul 16, 95–107. 24 Thom R, Cummins I, Dixon DP, Edwards R, Cole DJ & Lapthorn AJ (2002) Structure of a tau class gluta- thione S-transferase from wheat active in herbicide detoxification. Biochemistry 41, 7008–7020.

25 Axarli I, Dhavala P, Papageorgiou AC & Labrou NE

SWISS-MODEL: an automated protein homology-mod- eling server. Nucleic Acids Res 31, 3381–3385.

(2009) Crystallographic and functional characterization of the fluorodifen-inducible glutathione transferase from Glycine max reveals an active site topography suited for diphenylether herbicides and a novel L-site. J Mol Biol 385, 984–1002.

36 Kawabata T (2003) MATRAS: a program for protein 3D structure comparison. Nucleic Acids Res 31, 3367–3369. 37 Ricci C, Caccuri AM, Lo Bello M, Pastore A, Piemonte F & Federici G (1994) Colorimetric and fluorometric assays of glutathione transferase based on 7-chloro- 4-nitrobenzo-2-oxa-1,3-diazole. Anal Biochem 218, 463–465.

26 Axarli I, Dhavala P, Papageorgiou AC & Labrou NE (2009) Crystal structure of Glycine max glutathione transferase in complex with glutathione: investigation of the mechanism operating by the tau class glutathione transferases. Biochem J 422, 247–256.

38 Wheeler JB, Stourman NV, Thier R, Dommermuth A, Vuilleumier S, Rose JA, Armstrong R & Guengerich FP (2001) Conjugation of haloalkanes by bacterial and mammalian glutathione transferases: mono- and dihalo- methanes. Chem Res Toxicol 14, 1118–1127.

39 Bradford MM (1976) A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 72, 248–254.

27 Winayanuwattikun P & Ketterman AJ (2005) An elec- tron-sharing network involved in the catalytic mecha- nism is functionally conserved in different glutathione transferase classes. J Biol Chem 280, 31776–31782. 28 Zeng QY & Wang XR (2005) Catalytic properties of glutathione-binding residues in a s class glutathione transferase (PtGSTU1) from Pinus tabulaeformis. FEBS Lett 579, 2657–2662.

29 Kilili KG, Atanassova N, Vardanyan N, Clatot N,

40 Laemmli UK (1970) Cleavage of structural proteins during the assembly of head of bacteriophage T4. Nature 227, 680–685.

A. R. Lo Piero et al. Functional role of the GST H-site residues

Supporting information

Al-Sabarna K, Kanellopoulos PN, Makris AM & Kampranis SC (2004) Differential roles of tau class glutathione S-transferase in oxidative stress. J Biol Chem 279, 24540–24551.

30 Dixon DP, Cole DJ & Edwards R (1999) Dimerisation of maize glutathione transferases in recombinant bacte- ria. Plant Mol Biol 40, 997–1008.

The following supplementary material is available: Fig. S1. Alignment of the sweet orange GST H-site sequences with those of 18 representative members of tau class GSTs.

This supplementary material can be found in the

online version of this article.

31 Clark AG (1989) The comparative enzymology of the glutathione S-transferases from non-vertebrate organ- ism. Comp Biochem Physiol B 92, 419–446.

32 Armstrong RN (1997) Structure, catalytic mechanism, and evolution of glutathione transferase. Chem Res Toxicol 10, 2–18.

Please note: As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer-reviewed and may be re-organized for online delivery, but are not copy-edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

33 Johnson WW, Liu S, Ji X, Gilliland GL & Armstrong RN (1993) Tyrosine 115 participates both in chemical and physical steps of the catalytic mechanism of a glu- tathione transferase. J Biol Chem 268, 11508–11511.

FEBS Journal 277 (2010) 255–262 ª 2009 The Authors Journal compilation ª 2009 FEBS

262