Molecular cloning and characterization of θ-class glutathione S-transferase (GST-T) from the hermaphroditic fish Rivulus marmoratus and biochemical comparisons with α-class glutathione S-transferase (GST-A)

BBRC Biochemical and Biophysical Research Communications 346 (2006) 1053–1061 www.elsevier.com/locate/ybbrc

Molecular cloning and characterization of h-class glutathione S-transferase (GST-T) from the hermaphroditic fish Rivulus marmoratus and biochemical comparisons with a-class glutathione S-transferase (GST-A) Young-Mi Lee a, Jung Soo Seo a, Sang-Oun Jung a, Il-Chan Kim b, Jae-Seong Lee a

a,*

Department of Molecular and Environmental Bioscience, and the National Research Laboratory of Marine Molecular and Environmental Bioscience, Graduate School, Hanyang University, Seoul 133-791, South Korea b Polar BioCenter, Korea Polar Research Institute, Korea Ocean Research and Development Institute, Incheon 406-840, South Korea Received 27 May 2006 Available online 12 June 2006

Abstract We cloned and sequenced full-length cDNA of a h-class-like glutathione S-transferase (GST-T) from liver tissue of the self-fertilizing fish Rivulus marmoratus. The full-length cDNA of rm-GST-T was 907 bp in length containing an open reading frame of 666 bp that encoded a 221-amino acid putative protein. Its derived amino acid sequence was clustered with other vertebrate h-class GSTs in a phylogenetic tree. The deduced amino acid sequence of h-like rm-GST (rm-GST-T) was compared with both classes (a and h) of GST and a-class rm-GST (rm-GST-A). Tissue-specific expression of two rm-GST mRNAs was investigated using real-time RT-PCR. To further characterize the catalytic properties of this enzyme along with rm-GST-A, we constructed the recombinant h-like rm-GST plasmid with a 6·His-Tag at the N-terminal of rm-GST-T cDNA. Recombinant rm-GST-T was highly expressed in transformed Escherichia coli, and its soluble fraction was purified by His-Tag affinity column chromatography. The kinetic properties and effects of pH and temperature on rm-GST-T were further studied, along with enzyme activity and inhibition effects, and compared with recombinant rm-GST-A. These results suggest that recombinant rm-GSTs such as rm-GST-A and rm-GST-T play a conserved functional role in R. marmoratus.  2006 Elsevier Inc. All rights reserved. Keywords: Glutathione S-transferase; His-tag affinity column; Kinetic properties; CDNB; GSH; Fish; Rivulus marmoratus; Recombinant protein

All living organisms have refined defense mechanisms, including detoxification and antioxidant defense systems involving glutathione S-transferase (GST), glutathione reductase, and glutathione peroxidase against xenobiotic compounds and oxidative stress, respectively [1]. Of natural defense systems, GSTs play a major role in detoxification of deleterious electrophilic xenobiotics such as anti-cancer drugs, herbicides, pesticides, chemical carcinogens, and environmental pollutants [2]. These enzymes catalyze the conjugation of reduced glutathione (GSH) *

Corresponding author. Fax: +82 2 2299 9450. E-mail address: [email protected] (J.-S. Lee).

0006-291X/$ - see front matter  2006 Elsevier Inc. All rights reserved. doi:10.1016/j.bbrc.2006.06.014

with exogenous and endogenous toxic compounds, rendering them more water soluble, less toxic, and easier to excrete. In addition, they are responsible for various resistance mechanisms including chemotherapeutic drug resistance in tumor cells [3], bacterial antibiotic resistance [4], herbicide resistance in plants [5], and insecticide resistance in insects [6]. Most mammalian GSTs are cytosolic enzymes, existing as homodimers or heterodimers of subunits whose molecular weight ranges from 23 to 28 kDa. The cytosolic GSTs have been classified into at least twelve different classes (a, b, d, j, l, x, /, p, r, s, h, and f) on the basis of N-terminal amino acid sequence, substrate specificity, antibody

1054

Y.-M. Lee et al. / Biochemical and Biophysical Research Communications 346 (2006) 1053–1061

cross-reactivity, and sensitivity to inhibitors [7,8]. To date, many novel classes of GST sequences have been identified and classified from non-mammalian organisms. GST cDNA sequences have been isolated and characterized in several fish species such as plaice [9,10], largemouth bass [11], red seabream [12,13], and Rivulus [14]. Fish GST isoenzymes from the liver of brown bullheads and largemouth bass [15], sea bass [16], and rainbow trout [17] have been purified and partially characterized. The hermaphroditic fish Rivulus marmoratus (Cyprinodontiformes, Rivulidae) is the only known vertebrate reproducing by internal self-fertilization [18], resulting in genetic homogeneity among individuals within each selffertilizing population. Therefore, R. marmoratus has been regarded as a useful model organism for developmental biology, cancer research, and toxicology studies with aquatic pollutants [14,19–21]. We have been interested in their tolerance to extreme environmental conditions such as a broad range of salinity (0–114 &) and temperature (7–45 C), hypoxia (<1 mg/L O2), high levels of hydrogen sulfide (H2S), and prolonged air-exposure (>1 month) [22–25], along with revealing the mechanisms of defense against oxidative damages. In this species, the study of biochemical characteristics of GST isoenzymes could provide a better understanding of the resistance mechanisms to various environmental stresses, particularly because GST isoenzymes have been regarded as useful biomarkers in fishes for monitoring exposure to environmental stress [26,27]. In this study, we isolated the R. marmoratus GST gene, and sequenced full-length h-class rm-GST (rm-GST-T) using a 5 0 - and 3 0 -RACE strategy, and compared the deduced amino acid sequence with different GSTs from other vertebrates and a previously isolated R. marmoratus a-class GST (rm-GST-A) [14]. To analyze the different catalytic properties of both rm-GST-T and rm-GST-A, we first constructed an expression plasmid construct containing full-length cDNA sequences of rm-GST-T and rm-GST-A using 6·His-Tagged pCRT7 TOPO TA expression vector. The recombinant rm-GST isoenzymes were highly expressed in Escherichia coli, and then the soluble fractions of these isozymes were further purified using His-Tag affinity column chromatography. After enzyme purification, we determined the enzyme kinetic parameters and the effects of pH and temperature on recombinant rm-GSTs, along with inhibition effects. In this paper, we describe the potential biochemical characteristics of both rm-GST isoenzymes. Materials and methods Reagents. Isopropyl-b-D-thiogalactopyranoside (IPTG) and a glutathione S-transferase assay kit containing 1-chloro-2,4-dinitrobenzene (CDNB), reduced glutathione, and phosphate-buffered saline (PBS) were purchased from Sigma chemicals, Inc. (St. Louis, MO). Ni+–NTA columns and prestained broad range protein molecular weight markers were obtained from Qiagen (Hilden, Germany) and from Bio-Rad (CA, USA), respectively. All other chemicals (molecular biology grade) were from Sigma unless otherwise described.

Fish. The self-fertilizing fish R. marmoratus were maintained at an aquarium in the Department of Molecular and Environmental Bioscience, Graduate School, Hanyang University, Seoul, South Korea, with a water temperature of 25 C, a photoperiod of 12 h light/12 h darkness, and 10 & salinity. The fish were fed hatched larvae of Artemia nauplii (<24 h after hatching) once a day. Total RNA extraction, reverse transcription, rapid amplification of cDNA ends (RACE), and tissue distribution of rm-GST-T gene. Total RNAs were isolated with TRIZOL reagent (Invitrogen, Paisley, Scotland, UK) from each tissue (brain, eye, gonad, intestine, liver, muscle, and skin) of adult hermaphrodites according to the manufacturer’s instructions. Single-stranded cDNA was synthesized from 2 lg of total RNA using oligo(dT)20 primer for reverse transcription in 20 ll reactions using the SuperScript III RT kit (Invitrogen, Carlsbad, CA). Oligonucleotides used in this study are listed in Table 1. The partial cDNA clone of the rmGST-T gene was obtained from a differentially expressed gene pool using the Genefishing Kit (Seegene, Seoul, Korea) after exposure to bisphenol A (600 lg/L, 96 h). To get a full-length cDNA of the h-like rm-GST gene, the 5 0 -end was amplified using the GeneRacer Kit (Invitrogen) according to the manufacturer’s instructions. To analyze tissue distribution of the rm-GST-T gene, real-time reverse transcriptase PCR (RT-PCR) was conducted with cDNAs from brain, eye, gonad, intestine, liver, muscle, and skin of an adult hermaphroditic fish. For real-time PCR amplification, each reaction included 1 ll cDNA and 0.2 lM primer (Rm-GST-T-RT-F/R and b-actin RT-F/R; see Table 1). Other reaction conditions were as described earlier [20]. Phylogenetic analysis. To place the rm-GST-T gene in a phylogenetic tree, we aligned it with GST genes of diverse species at the level of deduced amino acid sequence by Clustal X (1.83) [28]. Gaps and missing data were completely excluded from the data analysis. The generated data matrix was converted to nexus format. This data matrix was analyzed with Mr. Bayesian’s program using the GTR model. A total of 1,000,000 generations were conducted, and the sampling frequency was assigned as every 100 generations. After analysis, the first 2000 generations were deleted as the burn-in process, and the consensus tree was constructed and then visualized with Tree View of PHYLIP [29]. Bacterial strains and plasmid vector. All bacterial strains and plasmid vector used in this study were purchased from Invitrogen (Paisley, Scotland, UK). The plasmid vector, pCRT7/NT-TOPO containing Nterminal 6·HisG (-HHHHHHG-), was used for cloning and expression of rm-GSTs. The recombinant plasmids were transformed into E. coli strain TOP10F’ to analyze positive clones. The E. coli strain BL21(DE3)pLysS was used for expression of recombinant rm-GST isoenzymes. Construction of expression plasmids and expression of recombinant rm-GST in E. coli. Expression plasmids were constructed as previously described [14]. Briefly, the open-reading frame (ORF) region of rmGST-T was amplified with primer rm-GST-T-pro-F/R (Table 1) using an iCycler (Bio-Rad, USA), eluted from a 1% agarose gel by elution kit (Qiagen, USA), and directly inserted into expression vector 6·His-tagged pCRT7/NT-TOPO. To highly express recombinant rm-GST protein, E. coli BL21(DE3)pLysS cells were transformed by rm-GST/pCR T7 NT-TOPO vector in the presence of ampicillin (100 lg/ml) and chloramphenicol (34 lg/ml). At OD600 of 0.5, the expression of recombinant protein was induced by adding IPTG at a final concentration of 1 mM. Bacterial cells were harvested after an 18-h incubation at 30 C and directly analyzed by 12% SDS–polyacrylamide gel electrophoresis (SDS–PAGE). Collected cells were resuspended in ice-cold 1· homogenizing buffer (20 mM Tris/pH 7.9, 0.5 M NaCl, and 5 mM imidazole) at 10 ml/g wet weight of pellet. Then, cells were sonicated 3 times for 10 min with a sonicator (Branson Co., USA) at a setting of 30%. The homogenate was centrifuged at 20,000g at 4 C for 20 min. The supernatant was used for protein purification and enzyme kinetic experiments. Self-ligated pCRT7 TOPO TA expression vector was used as a negative control in all experiments. Purification of soluble rm-GSTs through His-Tag affinity column chromatography and 12% SDS–PAGE. Soluble recombinant proteins were purified using His-Tag affinity columns (15 ml Ni+–NTA resin) according

Y.-M. Lee et al. / Biochemical and Biophysical Research Communications 346 (2006) 1053–1061

1055

Table 1 Primers used in this study Gene

Oligo name

Sequences (5 0 - > 3 0 )

Nucleotide position

Remarks

rm-GST-T

RT-F RT-R 5GSP1 5GSP2 5GSP3 pro-F pro-R RT-F RT-R

GGTGCAGCGATGGGAGAACTAC ATGCTGGGCCTGTCCTTCAG TCACGGCACAAAAACA GAGGGACTTCAGGGTGTCCATTC GGGGGGCCAGGTGCTTTTGATGC ATGGCCAAGGACATGACTCTGTACTG TCAGAGGGACTTCAGGGTGTCCATTCC CTTGCGGAATCCACGAGACC CCAGGGCTGTGATCTCCTTCTG

474–495 628–647 836–851 686–708 644–666 46–71 685–711 812–831 940–960

rm-GST-T cDNA amplification

b-Actin

to the LP system (Bio-Rad, USA). After the supernatant fraction was applied to the column, the column was washed with 1· wash buffer (20 mM Tris/pH 7.9, 60 mM imidazole, and 0.5 M NaCl) for 30 min with a flow rate of 1 ml/min. Subsequently, rm-GST was eluted in ten fractions with 1· elution buffer (20 mM Tris/pH 7.9, 1 M imidazole, and 0.5 M NaCl). The eluted fraction was dialyzed in 1· dialysis buffer (20 mM Tris/ pH 7.9, 0.5 M NaCl, 5 mM imidazole, and 0.5 mM EDTA, pH 8.0) overnight. The pooled fractions were analyzed by 12% SDS–PAGE and Western blotting. Protein concentration was determined with a Bio-Rad protein assay solution (Bio-Rad, USA). As described by Laemmli [30], 12% SDS–PAGE was carried out. All samples were denatured in 1· sample buffer (60 mM Tris/pH 6.8, 25% glycerol, 2% SDS, 14.4 mM b-mercaptoethanol, and 0.1% bromophenol blue), boiled for 5 min, and separated by 12% SDS–polyacrylamide gel (Bio-Rad, USA). Electrophoresed proteins were transferred to a nitrocellulose membrane (Schleicher & Schuell Co., USA) according to the method of Towbin and Gordon [31] using a Bio-Rad Mini Protean III transblotting system (Bio-Rad, USA). Following the transfer of proteins to the membrane, the membrane was blocked with 5% BSA in 0.1% TTBS (200 mM Tris/pH 7.0, 1.37 M NaCl, and 1% Tween 20) for 1 h at room temperature. The membrane was then incubated with anti-His G-HRP antibody (1:5000) (Invitrogen, USA) at room temperature for 3 h, rinsed, and washed three times with 0.1% TTBS for 60 min at room temperature. Detection was accomplished with an ECL Plus Western blotting kit (Amersham, USA). Glutathione S-transferase (GST) assay. GST activity was analyzed with a glutathione S-transferase assay kit according to the manufacturer’s instructions. The specific activity for rm-GSTs was spectrophotometrically determined with 1 mM CDNB and 2 mM GSH as substrates at 340 nm and 25 C. Kinetic study. The response of the purified rm-GSTs with various concentrations of CDNB and GSH was studied at 25 C in phosphate-buffered saline (PBS). The effect of CDNB on rm-GST activity was examined between 0.25 and 4.0 mM with a fixed GSH concentration of 2 mM. Meanwhile, the effect of GSH was studied between 0.25 and 4.0 mM with the fixed CDNB concentration of 1 mM. The data were plotted on a Lineweaver-Burk plot to determine the Km. Effect of pH on activity of recombinant rm-GSTs. The effect of pH on the activity of rm-GST isoenzymes was analyzed at 25 C after enzymes were equilibrated for 3 min in 50 mM of the following buffer: acetate (pH 5), phosphate (pH 6.0 and 7.0), Tris–HCl (pH 8.0 and 9.0), and glycine (pH 10). The remaining activity was then measured with 1 mM CDNB and 2 mM GSH as substrates. Effect of temperature on activity of recombinant rm-GSTs. The enzymes were incubated in PBS at temperatures ranging from 25 to 50 C for 5 min and then the remaining enzyme activity was measured with 1 mM CDNB and 2 mM GSH as substrates. Inhibition assay. An inhibition assay with Cibacron blue (0.0–10 lM), hematin (0.0–10 lM), and N-ethylmaleimide (0.0–1000 lM) was carried out according to the method of Tahir et al. [32] by measuring specific activity at 25 C in PBS buffer in the presence of 1 mM CDNB and 2 mM GSH.

rm-GST-T 5 0 -RACE

rm-GST-T ORF amplification b-Actin cDNA amplification

Results and discussion Cloning of a h-like rm-GST gene and sequence comparison with an a-class rm-GST gene A partial h-like rm-GST cDNA was obtained from adult R. marmoratus following bisphenol A exposure, and the full-length cDNA sequence was completed using a 5 0 RACE strategy. The complete cDNA sequence of the hlike rm-GST gene was 907 bp in length and contained an ORF of 666 bp encoding a polypeptide of 221-amino acids. Full-length cDNA sequences were submitted to the GenBank Accession No. DQ525602 and are shown in Fig. 1A. The putative protein encoded by the h-like rm-GST cDNA (rm-GST-T) was calculated to have a theoretical pKa of 8.5 and molecular weight of 25.8 kDa. The deduced amino acid sequence of rm-GST-T revealed low similarity (14%) to rm-GST-A (data not shown), even though the theoretical pKa and molecular mass of rm-GST-T were similar to those of rm-GST-A (pKa = 7.8, Mw 25.3 kDa). By searching motifs from the Pfam database (www.sanger.ac.uk/Software/Pfam), the rm-GST-T N terminal domain (3–79 aa) and GST C-terminal domain (110–199 aa) were found using an E-value of 1.0 as a cut-off score (Fig. 1A). In the case of rm-GST-A, the GST N-terminal domain was found at 5–77 aa and the C-terminal domain at 99–188 aa. Substitution of residues on the GST N-terminal domain is responsible for different GSH-binding ability [33] and the C-terminal region modulates the catalytic and non-substrate ligand-binding functions in a-class GST [34]. Tissue distribution of rm-GST isoenzymes To investigate the tissue-specific expression of rm-GSTT mRNA, real-time RT-PCR was carried out with total RNA extracted from seven tissues (brain, eyes, gonad, intestine, liver, muscle, and skin) of adult R. marmoratus and compared with that of rm-GST-A. Expression of rmGST-T was high in the liver and intestine but relatively low in the brain, eyes, gonad, muscle, and skin. rm-GSTA was ubiquitously distributed in tissues of adult Rivulus with high expression in the liver and intestine. rm-GST-A mRNA expression in the liver was approximately 8 times higher than that of rm-GST-T (Fig. 1B). In the case of

1056

Y.-M. Lee et al. / Biochemical and Biophysical Research Communications 346 (2006) 1053–1061

A

B

Fig. 1. (A) Nucleotides and deduced amino acid sequence of the rm-GSTT gene of the self-fertilizing fish Rivulus marmoratus. The stop codon is denoted by an asterisk (*). The start codon, stop codon, and polyadenylation signal are shown in bold and underlined. The boxes show the GST N-terminal (3–79 aa) and GST C-terminal domain (110–199 aa), respectively. (B) Tissue-specific expression of rm-GST-T and rm-GST-A mRNAs in seven different tissues from adult R. marmoratus by real time RT-PCR. Amplification of b-actin gene was used as an internal control. B, brain; E, eyes; G, gonad; I, intestine; L, liver; M, muscle; S, skin. All experiments were analyzed in triplicate.

rm-GST-T, the pattern of tissue expression was close to bass GST-A, implying differences in tissues susceptible to antioxidant damage or the existence of other detoxification enzymes such as aldehyde metabolizing enzymes (i.e., aldehyde dehydrogenases, aldo–keto reductase) in those tissues which are deficient in GST-A [11]. The tissue-specific expression pattern of rm-GST-A cDNA was similar to that of plaice GST-A, suggesting that it may function as a housekeeping gene, protecting the cell against endogenous oxidative stress or xenobiotics [9]. These enzymes play an essential role in organisms [11], although their tissue distribution may vary. Multiple alignment and phylogenetic tree A BLAST search using the deduced amino acid sequence of h-like rm-GST revealed the highest similarity to GST sequences of two other fishes: plaice (Pleuronectes

platessa, GenBank Accession No. Q92103) and largemouth bass (Micropterus salmoides, GenBank Accession No. AY335905). Despite low sequence similarity with other hclass GSTs, our sequence data indicate that rm-GST-T would be designated h-class, based on terminal amino acid sequence similarity [7]. The rm-GST-T protein had 64% identity with the plaice h-class GST, 63% identity with the red seabream q-class GST, and 62% identity with the gilthead seabream h-class GST. Therefore, the deduced amino acid sequence of rm-GST-T showed a high similarity to h-class and h-class related GST of some other vertebrates, as shown in Fig. 2. However, h-like rm-GST shared relatively low similarity (11–18%) with mammal and bird hclass GST and even zebrafish h-class GST (17%). Hayes et al. [35] mentioned that there are no clearly established criteria concerning the extent of sequence similarity required for placing a GST in a particular class. Consequently, it is generally accepted that GSTs share greater than 40% identity within a class, and those with less than 25% identity are assigned to separate classes. In fact, since h-class GSTs show generally lower amino acid sequence identities compared with other classes, many GSTs would be classified as h-class. Since deduced amino acid sequences of zebrafish h-class GST share approximately 50% identity with mammalian h-class GSTs, it seems that zebrafish GSTs should be classified as h-class GSTs. Also, Leaver et al. [9] suggested that fish h-class GST might belong in a primitive GST class. So far, a- and h-class related GSTs hve been isolated in several fishes such as the red seabream [12,13], largemouth bass [11], and plaice [10]. To investigate the phylogenetic relationship of rm-GSTT with different classes of GST enzymes from other vertebrates along with a-class rm-GST, a phylogenetic tree was constructed with deduced amino acid sequences using Mr. Bayesian’s method and visualized with TreeView (Fig. 3). The ln (L) value of the constructed tree was 10887.80 and the PSRF value (convergence diagnostic factor) was 1.008. When x-class GST was designated as an out-group, the resultant phylogenetic tree revealed that GST enzymes were divided into two clades (I and II) as shown in Fig. 3. In clade I, a-class and p-class GST demonstrated a sister relationship. In clade II, most of the enzymes clustered to the h-class. In particular, the h-class was divided into two subclades (A and B), where subclade A was grouped with fish species such as plaice, largemouth bass, and gilthead seabream, and subclade B was grouped with vertebrates such as mammals, birds, and zebrafish, and invertebrates such as silk worm. A part of h-class GST in subclade A was assigned to the h-class related GST q-class. The a-class rm-GST agreed with previous results in which it clustered with a-class GSTs of other fish species. As expected, rm-GST-T was located in the same clade as other fish h-class GST and q-class GST. In this study, fish h-class and q-class GSTs were distinctly clustered from mammalian h-class GSTs. Konishi et al. [12] reported that h-class GSTs of aquatic organisms such as algae, amphibians, and fishes were also distinctly clustered

Y.-M. Lee et al. / Biochemical and Biophysical Research Communications 346 (2006) 1053–1061

1057

Rivulus marmoratus Theta-like Pagrus major Rho-class Pagrus major Rho-class Pleuronectes platessa Theta-like Sparus aurata theta-class Danio rerio theta 3 Mus musculus theta-class Homo sapiens theta 1 Gallus gallus theta 1

Rivulus marmoratus Theta-like Pagrus major Rho-class Pagrus major Rho-class Pleuronectes platessa Theta-like Sparus aurata theta-class Danio rerio theta 3 Mus musculus theta-class Homo sapiens theta 1 Gallus gallus theta 1

Rivulus marmoratus Theta-like Pagrus major Rho-class Pagrus major Rho-class Pleuronectes platessa Theta-like Sparus aurata theta-class Danio rerio theta 3 Mus musculus theta-class Homo sapiens theta 1 Gallus gallus theta 1

Rivulus marmoratus Theta-like Pagrus major Rho-class Pagrus major Rho-class Pleuronectes platessa Theta-like Sparus aurata theta-class Danio rerio theta 3 Mus musculus theta-class Homo sapiens theta 1 Gallus gallus theta 1

Fig. 2. Multiple alignment of the deduced amino acid sequence of rm-GST-T with those of h-class along with h-class related GST from other species. The shaded boxes show conserved residues.

with those of mammals and birds in a dendrogram tree constructed by neighbor-joining. These results suggest that GST genes of aquatic organisms could play a different role in xenobiotic detoxification. Characterization of the recombinant rm-GSTs To examine their catalytic properties, recombinant rmGST-A and rm-GST-T were purified in an active form using His-tag affinity columns. The molecular weight and purity of rm-GST-A and rm-GST-T were analyzed by 12% SDS–PAGE (Fig. 4A). Each protein formed a single band with a molecular mass of approximately 29 kDa for rm-GST-A and 30 kDa for rm-GST-T. The molecular mass of both rm-GST subunits differed from that calculated from cDNA sequences of each rm-GST. This size difference results from the 3.5-kDa structure of the constructed expression vector with co-expressed parts, including 6· His. Taking this into account, the molecular masses of the recombinant rm-GST-T and rm-GST-A are in good agreement with the calculated values of 25.3 and 25.8 kDa, respectively. Of purified GST derived from the liver of other fishes, two kinds of GST subunits have been found to be common. Two kinds of GST subunits (25.5 and 23 kDa) were purified in the liver cytosol of a rainbow trout [17]. Leaver

et al. [9] reported that two kinds of GSTs (i.e., 26-kDa GST subunit for GST-A and 24-kDa GST subunit for GST-B) were found in plaice (P. platessa) liver cytosol. In the case of grey mullet (M. cephalus), two bands (25.2-kDa GST subunit and 23.4-kDa GST subunit) were immunochemically reacted with antibodies from plaice GST A and GST B, respectively [36]. Moreover, two kinds of GST subunits (26.5 and 23.5 kDa) purified from the liver of a sea bass (D. labrax) were characterized as h-class related GST and a-class GST, respectively [16]. Our data support the previous finding that both a-class GST and h-class related GST are constitutively expressed in liver from several species [13], even though molecular masses of the recombinant rm-GST-A and rm-GST-T differ slightly from the previous data. The recombinant rm-GST-A and rm-GST-T preserved their activity in the range of 25–50 C (Fig. 4B). Konishi et al. [12] reported that q-class GSTR1, a h-class related GST purified from subtropical inhabitant P. major, was inactivated at temperatures higher than 30 C. In case of R. marmoratus, which is a tropical fish, activity of rmGST-A and rm-GST-T remained above 80% at 50 C. Therefore, our results support the hypothesis that GST enzymes might have adapted to temperature through evolution. The effect of pH on rm-GST-A and rm-GST-T was observed in the range of pH 5–10. The residual activity

1058

Y.-M. Lee et al. / Biochemical and Biophysical Research Communications 346 (2006) 1053–1061

0.98 0.99

Mouse GST A Domestic guinea pig GST A1

0.87

Chicken GST A1 1.00 Red seabream GSTA2

0.99 1.00 0.99

Red seabream GSTA1 Zebrafish alpha-like protein

I

Zebrafish GST pi

0.99 0.67

Zebrafish GST pi 2 European eel GST Pi-class

0.99

GST-pi

Sockeye salmon GST pi

0.99

GST-alpha

Rivulus GST-A

0.36

0.48

Human GST A2

African clawed frog GST pi

0.85

Golden hamster GST class-mu 1.00

African clawed frog GST Mu-class

African clawed frog GST sigma class 0.99 Red seabream GST 0.99 Red seabream GST 0.92 Gilthead seabream GST theta-class

GST-mu GST-sigma GST-rho

Largemouth bass GST 0.92 0.99 Plaice GST 0.99 0.93

A

Plaice GST A Plaice GST A1

II

Rivulus GST-T * 0.99

0.98 0.43 0.99 0.47

Chicken GST theta 1

GST-theta

Zebrafish GST theta 3

B

Human GST theta 1 Mouse GST theta Domestic silkworm GST theta

Fugu GST omega-class

GST-omega

0.1

Fig. 3. Phylogenetic tree of Rivulus marmoratus h-class GST genes with GST genes of other classes and other species along with a-class rm-GST. Tree was constructed by Bayesian method. Fugu x-class GST was assigned as an outgroup.

of both rm-GSTs was not detected at pH 5 and was fairly low at pH 10 (Fig. 4C). The highest activity was shown at pH 9 for rm-GST-A and pH 8 for rm-GST-T. Purification of recombinant rm-GST-A and rm-GST-T yielded 16.64% and 29.48%, respectively, of the total activity collected after Ni+–NTA column (Table 2). Rm-GST-A and rm-GST-T had specific activities toward CDNB of 4.84 ± 0.15 and 9.94 ± 0.17 lmol/mg/min, respectively, suggesting that the binding ability of rm-GST-A to GSHcomplexes was lower than that of rm-GST-T [13]. To compare catalytic properties between rm-GST-A and rm-GST-T, Michelis–Menten kinetic parameters (Vmax and Km) were calculated for both recombinant rm-GSTs from initial velocity studies. All data were plotted as a double reciprocal, Lineweaver–Burk plot (Table 3). Vmax of rmGST-T was approximately 4.9 times higher than rmGST-A. This means that rm-GST-T would be a better catalyst of CDNB conjugation than rm-GST-A where CDNB is known as a GST class-nonspecific substrate. Generally, a-class GST shows more catalytic activity toward cumene hydroperoxide (CuooH), D5-androstene-3,17-dione (ADI), ethacrynic acid (ECA), and 4-hydroxynonenal (4HNE), while h-class GST appears to have more substrate-conjugating activity toward nitrobenzyl chloride (NBC) [11,13]. However, there are many findings that

structurally different GST enzymes may have a similar function in evolutionarily divergent species [11]. Thus, this would be a good example to compare the specificity of GST activity to CDNB. For example, although largemouth bass GST-A was structurally similar to plaice GST-A, which was classified into h-class [9], it was highly active in 4HNE conjugation, which more closely resembles a-class GST [37]. h-Class GSTs (GST-T) from mammals and plants, which were very close to rm-GST-T, have been found to be efficient catalysts of fatty acid hydroperoxides and 4-hydroalkenals, which are predominantly generated as an outcome of polyunsaturated fatty acid peroxidation [38,39]. Pham et al. [37] reported that largemouth bass GST-A had a remarkable ability to conjugate 4-HNE produced during the peroxidation of lipids. Accordingly, it would be interesting to carry out further studies to determine if rm-GST-T has similar catalytic activity to other h-class GSTs toward fatty acid metabolites such as fatty acid hydroperoxide or 4-hydroalkenal. To analyze inhibition of the CDNB-conjugating activity of rm-GST-A and rm-GST-T, several inhibitors (e.g., hematin, Cibacron blue, and N-ethylmaleimide) were assayed as shown in Fig. 5. The sensitivity of rm-GST-T to Cibacron blue and hematin was different from rmGST-A, while the inhibition effect of N-ethylmaleimide

Y.-M. Lee et al. / Biochemical and Biophysical Research Communications 346 (2006) 1053–1061

A

M

1

2

3

4

5

6

7

8

Table 3 The effect of various concentrations of GSH (0.25–4 mM) and CDNB (0.25–4 mM) in the presence of 1 mM CDNB and 2 mM GSH, respectively

9

kDa

36 26

B 120

Constant

Rm-GST-A

Rm-GST-T

K GSH (mM) M V GSH max (lmol/mg/min) K CDNB (mM) M V CDNB (lmol/mg/min) max

0.29 ± 0.01 2.79 ± 0.05 1.02 ± 0.04 5.41 ± 0.15

0.44 ± 0.01 13.66 ± 0.29 0.71 ± 0.03 15.17 ± 0.19

on both of rm-GSTs was similar. Mammalian GST a isoenzymes had a low IC50 value (0.05–2 lM) for hematin and a high value (0.6–20 lM) for Cibacron blue, respectively [40]. However, the IC50 value of rm-GST-A for both Cibacron blue and hematin was estimated to be between 1 and 10 lM, which was slightly higher than that of mammalian GST a for hematin. There is little information available regarding the sensitivity of h-class GSTs to inhibitors used in this study. Hematin and Cibacron blue inhibited rm-GST-T more easily than rm-GST-A with an IC50 between 0.01 and 0.1 lM. Sheehan et al. [7] reported that the range of kinetic properties such as substrate spec-

Rm-GST-T Rm-GST-A

100

Residual activity (%)

1059

80 60 40 20 0 37

50

Temperature (˚C)

A

C 120 GST-A GST-T

80 60

Hematin Cibacron blue (CB) N-ethylmaleimide (NEM)

120 100 80 60 40

rm-GST-A

20 0

40

0.001

0.01

0.1

1

10

100

1000

Inhibitor concentration ( M)

20 0 5

6

7

8

9

10

pH Fig. 4. (A) Twelve percentage SDS–PAGE of purified recombinant rmGST-T (staining, Coomassie solution). M, prestained protein size marker; 1, vector control; 2, total crude extract; 3, total soluble fraction; 4, soluble fraction #9; 5, soluble fraction #10; 6, soluble fraction #11; 7, soluble fraction #12; 8, soluble fraction #13; and 9, soluble fraction #14. (B) Effect of temperature on CDNB-conjugating activities of rm-GST-T and rm-GST-A. Experiments are replicated three times. (C) Effect of pH on activities of rm-GST-T and rm-GST-A. The enzyme was incubated in 50 mM of the following buffers: acetate (pH 5), phosphate (pH 6 and 7), Tris–HCl (pH 8 and 9), and glycine (pH 10) for 5 min. The residual activities were assayed with 1 mM CDNB and 2 mM GSH as substrates at 25 C. Experiments were replicated three times.

B

120

Residual activity (%)

Residual activity (%)

100

Residual activity (%)

25

100

Cibacron blue (CB) N-ethylmaleimide (NEM) Hematin

80 60 40 20

rm-GST-T

0 0.01

0.1

1

10

100

1000

10000

Inhibitor concentration ( M)

Fig. 5. Effect of inhibitors such as Cibacron blue (0.0–10 lM), hematin (0.0–10 lM), and N-etylmaleimide (0.0–1000 lM) on the activities of (A) rm-GST-A and (B) rm-GST-T. The remaining activities were recorded as percentage relative to control. Experiments were replicated three times.

Table 2 Purification of recombinant rm-GST-A and rm-GST-T Step

Volume (ml)

Total protein (mg)

Specific activity (lmol/mg/min)

Total activity (lmol)

Yield

Purification fold

Purified rm-GST-A Purified rm-GST-T

6 8

0.06 7.52

4.84 ± 0.15 9.94 ± 0.17

0.29 74.78

16.64 29.48

1067.42 11.08

1060

Y.-M. Lee et al. / Biochemical and Biophysical Research Communications 346 (2006) 1053–1061

ificities and inhibitor sensitivities is sometimes broad and overlap.

[16]

Acknowledgments This work was supported by a Korea Research Foundation ABRL Grant (2003; Grant No. H00007) funded to Jae-Seong Lee.

[17]

[18]

References [19] [1] R.H. Burdon, Genes and the environment, T.J. International Ltd, Chapter 5 (1999) 69–74. [2] R. Masella, R.D. Benedetto, R. Vari, C. Filesi, C. Giovannini, Novel mechanisms of natural antioxidant compounds inbiological systems: involvement of glutathione and glutathione-related enzymes, J. Nutr. Biochem. 16 (2005) 577–586. [3] J.D. Hayes, D.J. Pulford, The glutathione S-transferase supergene family: regulation of GST and the contribution of the isoenzymes to cancer chemoprotection and drug resistance, Crit. Rev. Biochem. Mol. Biol. 30 (1995) 445–600. [4] P. Arca, C. Hardisson, J. Suarez, Purification of a glutathione Stransferase that mediates fosfomycin resistance in bacteria, Antimicrob. Agents Chemother. 34 (1997) 844–848. [5] P.J. Hatton, I. Cummins, D.J. Cole, R. Edwards, Glutathione transferase involved in herbicide detoxification in the leaves of Setaria faberi (giant toxtail), Physiol. Plant 105 (1999) 9–16. [6] H. Ranson, L. Rossiter, F. Ortelli, B. Jensen, X. Wang, C.W. Roth, F.H. Collins, J. Hemingway, Identification of a novel class of insect glutathione S-transferase involved in resistance to DDT in the malaria vector Anopheles gambiae, Biochem. J. 359 (2001) 295–304. [7] D. Sheehan, G. Meade, V.M. Foley, C.A. Dowd, Structure, function and evolution of glutathione transferases: implications for classification of non-mammalian members of an ancient enzyme superfamily, Biochem. J. 360 (2001) 1–16. [8] S.E. Pemble, A.F. Wardle, J.B. Taylor, Glutathione S-transferase class j: classification by cloning of rat mitochondrial GST and identification of a human homologue, Biochem. J. 319 (1996) 749– 754. [9] M.J. Leaver, K. Scott, S.G. George, Cloning and characterization of the major hepatic glutathione S-transferase from a marine teleost flatfish, the plaice (Pleuronectes platessa), with structural similarities to plant, insect and mammalian h class isoenzymes, Biochem. J. 292 (1993) 189–195. [10] E. Martinez-Lara, M. Leaver, S. George, Evidence from heterologous expression of glutathione S-transferases A and A1 of the plaice (Pleuronectes platessa) that their endogenous role is in detoxification of lipid peroxidation products, Mar. Environ. Res. 54 (2002) 263–266. [11] A.M. Doi, R.T. Pham, E.M. Hughes, D.S. Barber, E.P. Gallagher, Molecular cloning the characterization of a glutathione S-transferase from largemouth bass (Micropterus salmoides) liver that is involved in the detoxification of 4-hydroxynonenal, Biochem. Pharmacol. 67 (2004) 2129–2139. [12] T. Konishi, K. Kato, T. Araki, K. Shiraki, M. Takagi, Y. Tamaru, A new class of glutathione S-transferase from the hepatopancreas of the red sea bream Parus major, Biochem. J. 288 (2005) 299–307. [13] T. Konishi, K. Kato, T. Araki, K. Shiraki, M. Takagi, Y. Tamaru, Molecular cloning and characterization of a-class glutathione Stransferase genes from the hepatopancreas of red sea bream, Pagrus major, Comp. Biochem. Physiol. 140C (2005) 309–320. [14] Y.-M. Lee, S.Y. Chang, S.-O. Jung, H.-S. Kweon, J.-S. Lee, Cloning and expression of a class glutathione S-transferase gene from the small hermaphroditic fish Rivulus marmoratus (Cyprinodontiformes, Rivulidae), Mar. Pollut. Bull. 51 (2005) 776–783. [15] E.P. Gallagher, K.M. Sheehy, M.W. Lame, H.J. Sengall, In vitro kinetics of hepatic glutathione S-transferase conjugation in large-

[20]

[21]

[22]

[23]

[24]

[25]

[26]

[27]

[28]

[29] [30] [31] [32]

[33]

[34]

mouth bass and brown bullheads, Environ. Toxicol. Chem. 19 (2000) 319–326. S. Angelucci, P. Sacchetta, P. Moio, S. Melino, R. Petruzzelli, P. Gervasi, C. Di Ilio, Purification and characterization of glutathione transferase from the sea bass (Dicentrarchus labrax) liver, Arch. Biochem. Biophys. 373 (2000) 435–441. M.J. Melgar-Riol, M.C. Mo´voa-Valin˜as, M.A. Garcia-Ferna´ndez, M. Pe´rez-Lo´pez, Glutathione S-transferase from rainbow trout liver and freshly isolated hepatocytes: purification and characterization, Comp. Biochem. Physiol. 128C (2001) 227–235. R.W. Harrington Jr., Oviparous hermaphroditic fish with internal self-fertilization, Science 134 (1961) 1749–1750. Y.-M. Lee, T.D. Williams, S.-O. Jung, J.-S. Lee, cDNA cloning and expression of a cytochrome P450 1A (CYP1A) gene from the hermaphroditic fish Rivulus marmoratus, Mar. Pollut. Bull. 51 (2005) 769–775. Y.-M. Lee, J.S. Seo, I.-C. Kim, Y.-D. Yoon, J.-S. Lee, Endocrine disrupting chemicals (bisphenol A, 4-nonylphenol, 4-tert-octylphenol) modulate expression of two distinct cytochrome P450 aromatase genes differently in gender types of the hermaphroditic fish Rivulus marmoratus, Biochem. Biophys. Res. Commun. 345 (2006) 894–903. J.S. Seo, Y.-M. Lee, S.-O. Jung, I.-C. Kim, Y.-D. Yoon, J.-S. Lee, Nonylphenol modulates expression of androgen receptor and estrogen receptor genes differently in gender types of the hermaphroditic fish Rivulus marmoratus, Biochem. Biophys. Res. Commun. 346 (2006) 213–223. J.C. King, D.C. Abel, D.R. DiBona, Effects of salinity on chloride cells in the euryhaline cyprinodontid fish Rivulus marmoratus, Cell Tissue Res. 257 (1989) 367–377. W.P. Davis, D.S. Taylor, B.J. Turner, Field observations of the ecology and habits of mangrove Rivulus (Rivulus marmoratus), in Belize and Florida (Teleostei: Cyprinodontiformes: Rivulidae), Ichthyol. Explor. Freshwaters 1 (1990) 123–134. W.A. Dunson, D.B. Dunson, Factors influencing growth and survival of the killifish, Rivulus marmoratus, held inside enclosures in mangrove swamps, Copeia 3 (1999) 661–668. D.C. Abel, C.C. Koenig, W.P. Davis, Emersion in the mangrove forest fish Rivulus marmoratus: a unique response to hydrogen sulfide, Environ. Biol. Fishes 18 (1987) 67–72. R. Van der Oost, J. Beyer, N.P.E. Vermeulen, Fish bioaccumulation and biomarkers in environmental risk, Environ. Toxicol. Pharmacol. 13 (2003) 57–149. L.L. Amado, C.E. da Rosa, A.M. Leite, L. Moraes, W.V. Pires, G.L. Pinho, C.M. Martins, R.B. Robaldo, L.E. Nery, J.M. Monserrat, A. Bianchini, P.E. Martinez, L.A. Geracitano, Biomarkers in croakers Micropogonias furnieri (Teleostei: Sciaenidae) from polluted and nonpolluted areas from the Patos Lagoon estuary (Souther Brazil): evidences of genotoxic and immunological effects, Mar. Pollut. Bull. 52 (2006) 199–206. J.D. Thompson, T.J. Gibson, F. Plewniak, F. Jeanmougin, D.G. Higgins, The ClustalX windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools, Nucleic Acids Res. 25 (1997) 4876–4882. R.D.M. Page, Treeview: an application to display phylogenetic trees on personal computers, Comp. Appl. Biol. Sci. 12 (1996) 357–358. U.K. Laemmli, Cleavage of structural proteins during the assembly of the head of bacteriophage T4, Nature 227 (1970) 680–685. H. Towbin, J. Gordon, Immunoblotting and dot immunoblotting— current status and outlook, J. Immunol. Meth. 72 (1984) 313–340. M.K. Tahir, C. Guttenberg, B. Mannervik, Inhibitors for distinction of three types of human glutathione transferase, FEBS Lett. 181 (1985) 249–252. C.S. Allardyce, P.H. McDonagh, L.Y. Lian, C.R. Wolf, G.C. Roverts, The role of tyrosine-9 and the C-terminal helix in the catalytic methanism of a-class glutathione S-transferases, Biochem. J. 343 (1999) 525–531. S. Mosebi, Y. Sayed, J. Burke, H.W. Dirr, Residue 219 impacts on the dynamics of the C-terminal region in glutathione transferase A1-1:

Y.-M. Lee et al. / Biochemical and Biophysical Research Communications 346 (2006) 1053–1061 implications for stability and catalytic and ligandin functions, Biochemistry 42 (2003) 15326–15332. [35] J.D. Hayes, J.U. Flangan, I.R. Jowsey, Glutathione transferases, Annu. Rev. Pharmacol. Toxicol. 45 (2004) 51–88. [36] E. Martines-Lara, S.G. George, J. Lopez-Barea, J.A. Barcena, Purification and characterization of multiple glutathione transferase isoenzymes from grey mullet liver, Cell. Mol. Life Sci. 53 (1997) 759–768. [37] R.T. Pham, D.S. Barber, E.P. Gallagher, GST A is a major glutathione S-transferase gene responsible for 4-hydroxynonenal conjugation in largemouth bass liver, Mar. Environ. Res. 58 (2004) 485–488.

1061

[38] K.L. Tan, P.G. Board, Purification and characterization of a recombinant human h-class glutathione transferase (GSTT2-2), Biochem. J. 315 (1996) 727–732. [39] M.J. Leaver, J. Wright, S.G. George, Structure and expression of a cluster of glutathione S-transferase genes from a marine fish, the plaice (Pleuronectes platessa), Biochem. J. 321 (1997) 405–412. ˚ lin, C. Guthenberg, H. Jensson, M.K. Tahir, M. [40] B. Mannervik, P. A Warholm, H. Jo¨rnvall, Identification of three classes of cytosolic glutathione transferase common to several mammalian species: correlation between structural data and enzymatic properties, Proc. Natl. Acad. Sci. USA 82 (1985) 7202–7206.