Search

The cytosolic GST (glutathione transferase) superfamily has been annotated in the Drosophila melanogaster genome database. Of 36 genes, four undergo alternative splicing to yield a total of 41 GST proteins. In the present study, we have obtained the 41 transcripts encoding proteins by RT (reverse transcription)-PCR using RNA template from Drosophila S2 cells, an embryonic cell line. This observation suggests that all of the annotated DmGSTs (D. melanogaster GSTs) in the proteome are expressed in the late embryonic stages of D. melanogaster. To avoid confusion in naming these numerous DmGSTs, we have designated them following the universal GST nomenclature as well as previous designations that fit within this classification. Furthermore, in the cell line, we identified an apparent processed pseudogene, gste8, in addition to two isoforms from the Delta class that have been published previously. Only approximately one-third of the expressed DmGSTs could be purified by conventional GSH affinity chromatography. The diverse kinetic properties as well as physiological substrate specificity of the DmGSTs are such that each individual enzyme displayed a unique character even compared with members from the same class.

Cytosolic GSTs (glutathione transferases) are a multifunctional group of enzymes widely distributed in Nature and involved in cellular detoxification processes. The three-dimensional structure of GmGSTU4-4 (Glycine max GST Tau 4-4) complexed with GSH was determined by the molecular replacement method at 2.7 A (1 A=0.1 nm) resolution. The bound GSH is located in a region formed by the beginning of alpha-helices H1, H2 and H3 in the N-terminal domain of the enzyme. Significant differences in the G-site (GSH-binding site) as compared with the structure determined in complex with Nb-GSH [S-(p-nitrobenzyl)-glutathione] were found. These differences were identified in the hydrogen-bonding and electrostatic interaction pattern and, consequently, GSH was found bound in two different conformations. In one subunit, the enzyme forms a complex with the ionized form of GSH, whereas in the other subunit it can form a complex with the non-ionized form. However, only the ionized form of GSH may form a productive and catalytically competent complex. Furthermore, a comparison of the GSH-bound structure with the Nb-GSH-bound structure shows a significant movement of the upper part of alpha-helix H4 and the C-terminal. This indicates an intrasubunit modulation between the G-site and the H-site (electrophile-binding site), suggesting that the enzyme recognizes the xenobiotic substrates by an induced-fit mechanism. The reorganization of Arg111 and Tyr107 upon xenobiotic substrate binding appears to govern the intrasubunit structural communication between the G- and H-site and the binding of GSH. The structural observations were further verified by steady-state kinetic analysis and site-directed mutagenesis studies.

GSTs (glutathione transferases) are a multifunctional group of enzymes, widely distributed and involved in cellular detoxification processes. In the xenobiotic-degrading bacterium Ochrobactrum anthropi, GST is overexpressed in the presence of toxic concentrations of aromatic compounds such as 4-chlorophenol and atrazine. We have determined the crystal structure of the GST from O. anthropi (OaGST) in complex with GSH. Like other bacterial GSTs, OaGST belongs to the Beta class and shows a similar binding pocket for GSH. However, in contrast with the structure of Proteus mirabilis GST, GSH is not covalently bound to Cys10, but is present in the thiolate form. In our investigation of the structural basis for GSH stabilization, we have identified a conserved network of hydrogen-bond interactions, mediated by the presence of a structural water molecule that links Ser11 to Glu198. Partial disruption of this network, by mutagenesis of Ser11 to alanine, increases the K(m) for GSH 15-fold and decreases the catalytic efficiency 4-fold, even though Ser11 is not involved in GSH binding. Thermal- and chemical-induced unfolding studies point to a global effect of the mutation on the stability of the protein and to a central role of these residues in zippering the terminal helix of the C-terminal domain to the starting helix of the N-terminal domain.

The eye colour mutant sepia (se1) is defective in PDA {6-acetyl-2-amino-3,7,8,9-tetrahydro-4H-pyrimido[4,5-b]-[1,4]diazepin-4-one or pyrimidodiazepine} synthase involved in the conversion of 6-PTP (2-amino-4-oxo-6-pyruvoyl-5,6,7,8-tetrahydropteridine; also known as 6-pyruvoyltetrahydropterin) into PDA, a key intermediate in drosopterin biosynthesis. However, the identity of the gene encoding this enzyme, as well as its molecular properties, have not yet been established. Here, we identify and characterize the gene encoding PDA synthase and show that it is the structural gene for sepia. Based on previously reported information [Wiederrecht, Paton and Brown (1984) J. Biol. Chem. 259, 2195-2200; Wiederrecht and Brown (1984) J. Biol. Chem. 259, 14121-14127; Andres (1945) Drosoph. Inf. Serv. 19, 45; Ingham, Pinchin, Howard and Ish-Horowicz (1985) Genetics 111, 463-486; Howard, Ingham and Rushlow (1988) Genes Dev. 2, 1037-1046], we isolated five candidate genes predicted to encode GSTs (glutathione S-transferases) from the presumed sepia locus (region 66D5 on chromosome 3L). All cloned and expressed candidates exhibited relatively high thiol transferase and dehydroascorbate reductase activities and low activity towards 1-chloro-2,4-dinitrobenzene, characteristic of Omega class GSTs, whereas only CG6781 catalysed the synthesis of PDA in vitro. The molecular mass of recombinant CG6781 was estimated to be 28 kDa by SDS/PAGE and 56 kDa by gel filtration, indicating that it is a homodimer under native conditions. Sequencing of the genomic region spanning CG6781 revealed that the se1 allele has a frameshift mutation from 'AAGAA' to 'GTG' at nt 190-194, and that this generates a premature stop codon. Expression of the CG6781 open reading frame in an se1 background rescued the eye colour defect as well as PDA synthase activity and drosopterins content. The extent of rescue was dependent on the dosage of transgenic CG6781. In conclusion, we have discovered a new catalytic activity for an Omega class GST and that CG6781 is the structural gene for sepia which encodes PDA synthase.

The Saccharomyces cerevisiae genome encodes three proteins that display similarities with human GSTOs (Omega class glutathione S-transferases) hGSTO1-1 and hGSTO2-2. The three yeast proteins have been named Gto1, Gto2 and Gto3, and their purified recombinant forms are active as thiol transferases (glutaredoxins) against HED (beta-hydroxyethyl disulphide), as dehydroascorbate reductases and as dimethylarsinic acid reductases, while they are not active against the standard GST substrate CDNB (1-chloro-2,4-dinitrobenzene). Their glutaredoxin activity is also detectable in yeast cell extracts. The enzyme activity characteristics of the Gto proteins contrast with those of another yeast GST, Gtt1. The latter is active against CDNB and also displays glutathione peroxidase activity against organic hydroperoxides such as cumene hydroperoxide, but is not active as a thiol transferase. Analysis of point mutants derived from wild-type Gto2 indicates that, among the three cysteine residues of the molecule, only the residue at position 46 is required for the glutaredoxin activity. This indicates that the thiol transferase acts through a monothiol mechanism. Replacing the active site of the yeast monothiol glutaredoxin Grx5 with the proposed Gto2 active site containing Cys46 allows Grx5 to retain some activity against HED. Therefore the residues adjacent to the respective active cysteine residues in Gto2 and Grx5 are important determinants for the thiol transferase activity against small disulphide-containing molecules.

Structural investigations of a GST (glutathione transferase), adGSTD4-4, from the malaria vector Anopheles dirus show a novel lock-and-key 'Clasp' motif in the dimer interface of the Delta class enzyme. This motif also appears to be highly conserved across several insect GST classes, but differs from a previously reported mammalian lock-and-key motif. The aromatic 'key' residue not only inserts into a hydrophobic pocket, the 'lock', of the neighbouring subunit, but also acts as part of the 'lock' for the other subunit 'key'. The 'key' residues from both subunits show aromatic ring stacking with each other in a pi-pi interaction, generating a 'Clasp' in the middle of the subunit interface. Enzyme catalytic and structural characterizations revealed that single amino acid replacements in this 'Clasp' motif impacted on catalytic efficiencies, substrate selectivity and stability. Substitutions to the 'key' residue create strong positive co-operativity for glutathione binding, with a Hill coefficient approaching 2. The lock-and-key motif in general and especially the 'Clasp' motif with the pi-pi interaction appear to play a pivotal role in subunit communication between active sites, as well as in stabilizing the quaternary structure. Evidence of allosteric effects suggests an important role for this particular intersubunit architecture in regulating catalytic activity through conformational transitions of subunits. The observation of co-operativity in the mutants also implies that glutathione ligand binding and dimerization are linked. Quaternary structural changes of all mutants suggest that subunit assembly or dimerization basically manipulates subunit communication.

The insect GST (glutathione transferase) supergene family encodes a varied group of proteins belonging to at least six individual classes. Interest in insect GSTs has focused on their role in conferring insecticide resistance. Previously from the mosquito malaria vector Anopheles dirus, two genes encoding five Delta class GSTs have been characterized for structural as well as enzyme activities. We have obtained a new Delta class GST gene and isoenzyme from A. dirus, which we name adGSTD5-5. The adGSTD5-5 isoenzyme was identified and was only detectably expressed in A. dirus adult females. A putative promoter analysis suggests that this GST has an involvement in oogenesis. The enzyme displayed little activity for classical GST substrates, although it possessed the greatest activity for DDT [1,1,1-trichloro-2,2-bis-(p-chlorophenyl)ethane] observed for Delta GSTs. However, GST activity was inhibited or enhanced in the presence of various fatty acids, suggesting that the enzyme may be modulated by fatty acids. We obtained a crystal structure for adGSTD5-5 and compared it with other Delta GSTs, which showed that adGSTD5-5 possesses an elongated and more polar active-site topology.

To elucidate drug deposition and metabolism in cultured marine fishes, in a previous study we isolated and purified the GSTs (glutathione S-transferases) from the hepatopancreas of the red sea bream Pagrus major that contained 25 and 28 kDa GST subunits. The 25 kDa GST subunits encoded by two genes (GSTA1 and GSTA2) have been identified as Alpha-class GSTs. In the present study, we performed the molecular cloning and characterization of the GSTR1 gene encoding the 28 kDa GST subunit from the Pa. major hepatopancreas. The nucleotide sequence of GSTR1 was composed of an ORF (open reading frame) of 675 bp encoding a protein of 225 residues with a predicted molecular mass of 25.925 Da. A search of the BLAST protein database revealed that the deduced amino acid sequence of GSTR1 was structurally similar to that of GSTs derived from other fishes such as largemouth bass (Micropterus salmoides) and plaice (Pleuronectes platessa). The genomic DNA containing the GSTR1 gene was found to consist of six exons and five introns quite distinct from mammalian Theta-class GSTs. We have purified and characterized the recombinant GSTR1 enzyme (pmGSTR1-1) which showed activity only towards 1-chloro-2,4-dinitrobenzene, although it had no detectable activity towards cumene hydroperoxide, 1,2-dichloro-4-nitrobenzene, ethacrynic acid, 4-hydroxynonenal and p-nitrobenzyl chloride. Moreover, pmGSTR1-1 revealed remarkable heat instability (melting temperature Tm=30.3+/-0.11 degrees C). Collectively, our results indicated that the characteristic GST genes including GSTR1 have been conserved and functional in fishes. Therefore we designate them 'Rho-class', a new class of GSTs.

The Kappa class of GSTs (glutathione transferases) comprises soluble enzymes originally isolated from the mitochondrial matrix of rats. We have characterized a Kappa class cDNA from human breast. The cDNA is derived from a single gene comprising eight exons and seven introns located on chromosome 7q34-35. Recombinant hGSTK1-1 was expressed in Escherichia coli as a homodimer (subunit molecular mass approximately 25.5 kDa). Significant glutathione-conjugating activity was found only with the model substrate CDNB (1-chloro-2,4-ditnitrobenzene). Hyperbolic kinetics were obtained for GSH (parameters: K(m)app, 3.3+/-0.95 mM; V(max)app, 21.4+/-1.8 micromol/min per mg of enzyme), while sigmoidal kinetics were obtained for CDNB (parameters: S0.5app, 1.5+/-1.0 mM; V(max)app, 40.3+/-0.3 micromol/min per mg of enzyme; Hill coefficient, 1.3), reflecting low affinities for both substrates. Sequence analyses, homology modelling and secondary structure predictions show that hGSTK1 has (a) most similarity to bacterial HCCA (2-hydroxychromene-2-carboxylate) isomerases and (b) a predicted C-terminal domain structure that is almost identical to that of bacterial disulphide-bond-forming DsbA oxidoreductase (root mean square deviation 0.5-0.6 A). The structures of hGSTK1 and HCCA isomerase are predicted to possess a thioredoxin fold with a polyhelical domain (alpha(x)) embedded between the beta-strands (betaalphabetaalpha(x)betabetaalpha, where the underlined elements represent the N and C motifs of the thioredoxin fold), as occurs in the bacterial disulphide-bond-forming oxidoreductases. This is in contrast with the cytosolic GSTs, where the helical domain occurs exclusively at the C-terminus (betaalphabetaalphabetabetaalphaalpha(x)). Although hGSTK1-1 catalyses some typical GST reactions, we propose that it is structurally distinct from other classes of cytosolic GSTs. The present study suggests that the Kappa class may have arisen in prokaryotes well before the divergence of the cytosolic GSTs.

The class Kappa family of glutathione S-transferases (GSTs) currently comprises a single rat subunit (rGSTK1), originally isolated from the matrix of liver mitochondria [Harris, Meyer, Coles and Ketterer (1991) Biochem. J. 278, 137-141; Pemble, Wardle and Taylor (1996) Biochem. J. 319, 749-754]. In the present study, an expressed sequence tag (EST) clone has been identified which encodes a mouse class Kappa GST (designated mGSTK1). The EST clone contains an open reading frame of 678 bp, encoding a protein composed of 226 amino acid residues with 86% sequence identity with the rGSTK1 polypeptide. The mGSTK1 and rGSTK1 proteins have been heterologously expressed in Escherichia coli and purified by affinity chromatography. Both mouse and rat transferases were found to exhibit GSH-conjugating and GSH-peroxidase activities towards model substrates. Analysis of expression levels in a range of mouse and rat tissues revealed that the mRNA encoding these enzymes is expressed predominantly in heart, kidney, liver and skeletal muscle. Although other soluble GST isoenzymes are believed to reside primarily within the cytosol, subcellular fractionation of mouse liver demonstrates that this novel murine class Kappa GST is associated with mitochondrial fractions. Through the use of bioinformatics, the genes encoding the mouse and rat class Kappa GSTs have been identified. Both genes comprise eight exons, the protein coding region of which spans approx. 4.3 kb and 4.1 kb of DNA for mGSTK1 and rGSTK1 respectively. This conservation in primary structure, catalytic properties, tissue-specific expression, subcellular localization and gene structure between mouse and rat class Kappa GSTs indicates that they perform similar physiological functions. Furthermore, the association of these enzymes with mitochondrial fractions is consistent with them performing a specific conserved biological role within this organelle.

The structural gene for glutathione S-transferase (CeGST1-1) in the fungus Cunninghamella elegans was cloned by screening a cDNA library using a degenerate oligonucleotide probe based on the N-terminal sequence of the purified protein. Open reading frame analysis indicated that the cegst1 gene encodes a protein of 210 amino acid residues. The deduced amino acid sequence showed 25% sequence identity with the sequence of the Pi-class GST from Danio rerio (zebrafish). Similarity was also shown with the Alpha-class GST from Fasciola hepatica (liver fluke; 23% identity), the Mu class from Mus musculus (22%) and the Sigma class from Ommastrephes sloani (squid; 21%). Further screening of a cDNA library with the cegst1 gene probe revealed the presence of another GST isoenzyme (CeGST2-2) in this fungus, which shows 84% sequence identity with CeGST1-1 at the amino acid level. Reverse transcription PCR revealed that cegst2 was also expressed at the mRNA level in the fungus C. elegans. Both cegst genes were overexpressed in Escherichia coli using the expression vector pQE51, displaying specific activities with 1-chloro-2,4-dinitrobenzene of 2.04 and 0.75 micromol/min per mg of protein respectively. Both enzymes exhibited a similar substrate specificity and inhibition profile, indicating that CeGST1-1 and CeGST2-2 belong to the same GST class. Mutagenesis analysis revealed that Tyr(10) in the N-terminal region is essential for catalysis of CeGST1-1. We propose from these results that the CeGSTs are novel Gamma-class GSTs and designated as GSTG1-1 and GSTG2-2 respectively.

GSH-dependent prostaglandin D(2) synthase (PGDS) enzymes represent the only vertebrate members of class Sigma glutathione S-transferases (GSTs) identified to date. Complementary DNA clones encoding the orthologous human and rat GSH-dependent PGDS (hPGDS and rPGDS, respectively) have been expressed in Escherichia coli, and the recombinant proteins isolated by affinity chromatography. The purified enzymes were both shown to catalyse specifically the isomerization of prostaglandin (PG) H(2) to PGD(2). Each transferase also exhibited GSH-conjugating and GSH-peroxidase activities. The ability of hPGDS to catalyse the conjugation of aryl halides and isothiocyanates with GSH was found to be less than that of the rat enzyme. Whilst there is no difference between the enzymes with respect to their K(m) values for 1-chloro-2,4-dinitrobenzene, marked differences were found to exist with respect to their K(m) for GSH (8 mM versus 0.3 mM for hPGDS and rPGDS, respectively). Using molecular modelling techniques, amino acid substitutions have been identified in the N-terminal domain of these enzymes that lie outside the proposed GSH-binding site, which may explain these catalytic differences. The tissue-specific expression of PGDS also varies significantly between human and rat; amongst the tissues examined, variation in expression between the two species was most apparent in spleen and bone marrow. Differences in catalytic properties and tissue-specific expression of hPGDS and rPGDS appears to reflect distinct physiological roles for class Sigma GST between species. The evolution of divergent functions for the hPGDS and rPGDS is discussed in the context of the orthologous enzyme from chicken.

The sequence and cytological location of five Anopheles gambiae glutathione S-transferase (GST) genes are described. Three of these genes, aggst1-8, aggst1-9 and aggst1-10, belong to the insect class I family and are located on chromosome 2R, in close proximity to previously described members of this gene family. The remaining two genes, aggst3-1 and aggst3-2, have a low sequence similarity to either of the two previously recognized classes of insect GSTs and this prompted a re-evaluation of the classification of insect GST enzymes. We provide evidence for seven possible classes of insect protein with GST-like subunits. Four of these contain sequences with significant similarities to mammalian GSTs. The largest novel insect GST class, class III, contains functional GST enzymes including two of the A. gambiae GSTs described in this report and GSTs from Drosophila melanogaster, Musca domestica, Manduca sexta and Plutella xylostella. The genes encoding the class III GST of A. gambiae map to a region of the genome on chromosome 3R that contains a major DDT [1,1,1-trichloro-2,2-bis-(p-chlorophenyl)ethane] resistance gene, suggesting that this gene family is involved in GST-based resistance in this important malaria vector. In further support of their role in resistance, we show that the mRNA levels of aggst3-2 are approx. 5-fold higher in a DDT resistant strain than in the susceptible strain and demonstrate that recombinant AgGST3-2 has very high DDT dehydrochlorinase activity.

Potentiometric, spectroscopic and stopped-flow experiments have been performed to dissect the binding mechanism of GSH to selected glutathione S-transferases (GSTs), A1-1, M2-2 and Lucilia cuprina GST, belonging to Alpha, Mu and Delta classes respectively. Both Alpha and Mu isoenzymes quantitatively release the thiol proton of the substrate when the binary complex is formed. Proton extrusion, quenching of intrinsic fluorescence and thiolate formation, diagnostic of different steps along the binding pathway, have been monitored by stopped-flow analysis. Kinetic data are consistent with a multi-step binding mechanism: the substrate is initially bound to form an un-ionized pre-complex [k(1)>/=(2-5)x10(6) M(-1).s(-1)], which is slowly converted into the final Michaelis complex (k(2)=1100-1200 s(-1)). Ionization of GSH, fluorescence quenching and proton extrusion are fast events that occur either synchronously or rapidly after the final complex formation. The Delta isoenzyme shows an interesting difference: proton extrusion is almost stoichiometric with thiolate formed at the active site only up to pH 7.0. Above this pH, at least one protein residue acts as internal base to neutralize the thiol proton. These results suggest that the Alpha and Mu enzymes retain not only a similar catalytic outcome and overall three-dimensional structure but also share a similar kinetic mechanism for GSH binding. The Delta GST, which is closely related to the mammalian Theta class enzymes and is distantly related to Alpha and Mu GSTs in the evolutionary pathway, might display a different activation mechanism for GSH.

The structure and organization of the human Theta-class glutathione S-transferase (GST) genes have been determined. GSTT1 and GSTT2 are separated by approx. 50 kb. They have a similar structure, being composed of five exons with identical exon/intron boundaries. GSTT1 is 8.1 kb in length, while GSTT2 is only 3.7 kb. The GSTT2 gene lies head-to-head with a gene encoding d-dopachrome tautomerase (DDCT), which extends over 8.5 kb and contains four exons. The sequence between GSTT2 and DDCT may contain a bidirectional promoter. The GSTT2 and DDCT genes have been duplicated in an inverted repeat. Sequence analysis of the duplicated GSTT2 gene has identified an exon 2/intron 2 splice site abnormality and a premature translation stop signal at codon 196. These changes suggest that the duplicate gene is a pseudogene, and it has been named GSTT2P.

Human glutathione transferases (GSTs) from Alpha (A), Mu (M) and Theta (T) classes exhibited glutathione peroxidase activity towards phospholipid hydroperoxide. The specific activities are in the order: GST A1-1>GST T1-1>GST M1-1>GST A2-2>GST A4-4. Using a specific and sensitive HPLC method, specific activities towards the phospholipid hydroperoxide,1-palmitoyl-2-(13-hydroper oxy-cis-9, trans-11 -octadecadienoyl)-l-3-phosphatidylcholine (PLPC-OOH) were determined to be in the range of 0.8-20 nmol/min per mg of protein. Two human class Pi (P) enzymes (GST P1-1 with Ile or Val at position 105) displayed no activity towards the phospholipid hydroperoxide. Michaelis-Menten kinetics were followed only for glutathione, whereas there was a linear dependence of rate with PLPC-OOH concentration. Unlike the selenium-dependent phospholipid hydroperoxide glutathione peroxidase (Se-PHGPx), the presence of detergent inhibited the activity of GST A1-1 on PLPC-OOH. Also, in contrast with Se-PHGPx, only glutathione could act as the reducing agent for GST A1-1. A GST A1-1 mutant (Arg15Lys), which retains the positive charge between the GSH- and hydrophobic binding sites, exhibited a decreased kcat for PLPC-OOH but not for CDNB, suggesting that the correct topography of the GSH site is more critical for the phospholipid substrate. A Met208Ala mutation, which gives a modified hydrophobic site, decreased the kcat for CDNB and PLPC-OOH by comparable amounts. These results indicate that Alpha, Mu and Theta class human GSTs provide protection against accumulation of cellular phospholipid hydroperoxides.

Dichloroacetic acid (DCA), a common drinking-water contaminant, is hepatocarcinogenic in rats and mice, and is a therapeutic agent used clinically in the management of lactic acidosis. DCA is biotransformed to glyoxylic acid by glutathione-dependent cytosolic enzymes in vitro and is metabolized to glyoxylic acid in vivo. The enzymes that catalyse the oxygenation of DCA to glyoxylic acid have not, however, been identified or characterized. In the present investigation, an enzyme that catalyses the glutathione-dependent oxygenation of DCA was purified to homogeneity (587-fold) from rat liver cytosol. SDS/PAGE and HPLC gel-filtration chromatography showed that the purified enzyme had a molecular mass of 27-28 kDa. Sequence analysis showed that the N-terminus of the purified protein was blocked. An internal sequence of 30 amino acid residues was obtained that matched the recently discovered human glutathione transferase Zeta well [Board, Baker, Chelvanayagam and Jermiin (1997) Biochem. J. 328, 929-935]. Western-blot analysis showed that the purified rat-liver enzyme cross-reacted with rabbit antiserum raised against recombinant human glutathione transferase Zeta. The apparent Km and Vmax values of the purified enzyme with DCA as the variable substrate were 71.4 microM and 1334 nmol/min per mg of protein, respectively; the Km for glutathione was 59 microM. Both the purified rat-liver enzyme and the recombinant human enzyme showed high activity with DCA as the substrate. These results demonstrate that the glutathione-dependent oxygenation of DCA to glyoxylic acid is catalysed by a Zeta-class glutathione transferase.

Recent studies have revealed binding of mitochondrial enoyl-CoA isomerase (ECI) to S-hexylglutathione-Sepharose, an affinity matrix used for purification of glutathione transferases (GSTs), and the enzyme has been suggested to be identical with the Alpha class form of GST with a subunit molecular mass of about 30 kDa. In the present study, S-hexylglutathione-binding proteins of human hepatocellular carcinomas were characterized to examine their identity. Supernatant fractions of carcinoma and surrounding tissues were applied to an affinity column, and bound fractions were resolved into three proteins with subunit molecular masses/pI values of 33 kDa/7.0, 30 kDa/5.8 and 29 kDa/5.8 in addition to the well-characterized four GST subunits, A1, A2, M1 and P1, by two-dimensional gel electrophoresis. The proteins were further purified by chromatofocusing at pH 7.4-4.0. The 30 and 29 kDa proteins were eluted at pH 4.9 and by 1 M NaCl respectively, and could be clearly separated from each other. The 29 kDa protein exhibited a low but significant activity towards 1-chloro-2,4-dinitrobenzene (4.25 micromol/min per mg of protein) and reacted with anti-(GST A1-2) antibody, suggesting that it is a member of the GST Alpha class. The 30 kDa protein did not react with anti-GST antibodies and was identified as ECI by immunoblotting and N-terminal-amino-acid-sequencing analyses. The results thus indicated that the Alpha class GST form composed of the 29 kDa subunits and ECI are two different proteins. The 33 kDa protein was eluted from the chromatofocusing column at pH 7.0 and did not react with either anti-GST antibodies or antibodies against mitochondrial enzymes involved in the beta-oxidation of fatty acids. However, it exhibited a carbonyl reductase activity with menadione and ubiquinone, and amino acid sequences of its peptides cleaved by Staphylococcus aureus V8 proteinase were consistent with those reported for the enzyme. Thus this protein binding to S-hexylglutathione-Sepharose was identified as carbonyl reductase.

We have identified an N-capping box motif (Ser/Thr-Xaa-Xaa-Asp) that is strictly conserved, at the beginning of alpha 6 helix, in all glutathione S-transferases (GSTs) and most of the related superfamily proteins. By using CD and peptide modelling we have demonstrated that the capping box residues have an important role in determining the helical conformation adopted by this fragment in the hydrophobic environment of the protein. This is an example in which a local motif, contributing to nucleation of a structural element essential to the global folding of the protein, is strictly conserved in a superfamily of homologous proteins.

The effect of recombinant interleukin 6 (rIL-6) on the transcript levels of rat glutathione S-transferase (GST) genes rGSTA2, rGSTP1, rGSTM1 and rGSTM2 was examined in primary cultures of rat hepatocytes. rIL-6 had little effect on the increase in expression of rGSTP1 that occurs in cultured hepatocytes. Dexamethasone (DEX), in contrast, prevented the expression of rGSTP1 by hepatocytes, and rIL-6 in combination with DEX had no additional effect. Neither rIL-6 nor DEX alone had a significant effect on the transcript levels of rGSTA2, rGSTM1 and rGSTM2 in cultured hepatocytes. However, when both were present (15 ng/ml rIL-6 and 10(-7) M DEX) the transcript levels of rGSTA2, rGSTM1 and rGSTM2 decreased significantly (P < 0.05) after 48 h in culture. If the rIL-6 was removed from the cultures after 24 h, the levels of transcripts recovered and were the same at 48 h as cells cultured without rIL-6 for the entire period. Dose-response relationships of rIL-6 with 10(-7) M DEX were determined for transcripts of each GST isoenzyme and the IC50 values were between 1.5 and 7.5 ng/ml. Declines in transcript levels of rGSTA2 were observed with rIL-6 plus 10(-8) or 10(-7) M DEX but not with rIL-6 plus 10(-9), 10(-6), or 10(-5) M DEX. To determine if the cytokine and glucocorticoid effects were mediated by sequences in the 5'-flanking sequence of rGSTA2, a plasmid construct containing a 1.6 kb fragment of the 5'-flanking sequence of the rGSTA2 gene and the chloramphenicol acetyltransferase (CAT) reporter gene was used to transfect rat hepatocytes in primary culture. The addition of rIL-6 and DEX to the culture medium caused a significant (P < 0.05) decrease in CAT activity after 48 h in culture. If rIL-6 was removed after 24 h in culture, CAT activity after an additional 24 h in culture was greater than the CAT activity in cells cultured for 48 h without rIL-6. Therefore cytokines and glucocorticoids may be important physiological regulators of GST expression.

A cDNA encoding the human Theta-class glutathione transferase GSTT2-2 was expressed in Escherichia coli as a ubiquitin fusion protein. The co-translational removal of the ubiquitin by a cloned ubiquitin-specific protease, Ubp1, generates enzymically active GSTT2-2 without any additional N-terminal residues. The recombinant isoenzyme was purified to apparent homogeneity by DEAE anion-exchange, gel filtration, dye ligand chromatography and high resolution anion-exchange chromatography on Mono Q FPLC. The recombinant enzyme had significant activity with a range of substrates, including cumene hydroperoxide and 1-menapthyl sulphate. The activity of GSTT2-2 with a range of secondary lipid peroxidation products such as the trans,trans-alka-2,4-dienals and trans-alk-2-enals, as well as its glutathione peroxidase activity with organic hydroperoxides, suggest that it may play a significant role in protection against the products of lipid peroxidation.

Comparison of partial primary sequences of sigma-class glutathione S-transferases (GSH) of parasitic helminths and a GSH-dependent prostaglandin (PG)-H D-isomerase of rat immune accessory cells suggested that some of the helminth enzymes may also be involved in PG biosynthesis [Meyer and Thomas (1995) Biochem. J. 311, 739-742]. A soluble GSH transferase of the parasitic nematode Ascaridia galli has now been purified which shows high activity and specificity in the GSH-dependent isomerization of PGH to PGE, comparable to that of the rat spleen enzyme in its isomerization of PGH to PGD, and similarly stimulates the activity of prostaglandin H synthase. The enzyme subunit is structurally related to the rat spleen enzyme and sigma-class GSH transferases of helminths according to the partial primary sequence. The data support the hypothesis that some sigma-class GSH transferases of helminth parasites are involved in PG biosynthesis which, in the case of PGE, is likely to be associated with the subversion or suppression of host immunity. A PG-H E-isomerase of comparable specificity and activity has not previously been isolated.

Glutathione S-transferase CL1-2 heterodimers purified from 1-day-old chick livers were digested with Achromobacter proteinase I. The resulting fragments were separated for amino acid sequence analysis. Oligonucleotide probes were constructed based on sequence similarity to class-Theta glutathione S-transferases for PCR using a chicken liver cDNA library as template. A full-length clone (1725 bp) encoding a polypeptide comprising 261 amino acids was isolated. Including conservative substitutions, this protein has 70-73% sequence similarity with other mammalian class-Theta glutathione S-transferases. Based on known X-ray crystal structures of class-Alpha, -Mu and -Pi glutathione S-transferases, a model is constructed for the N-terminal 232 residues of CL1.

The sex-dependent expression and growth hormone (GH) regulation of rat liver glutathione S-transferase (GST) was examined using oligonucleotide probes that distinguish between closely related class Alpha (Ya1, Ya2, Yc) and class Mu (Yb1, Yb2, Yb3) GST mRNAs [Waxman, Sundseth, Srivastava and Lapenson (1992) Cancer Res. 52, 5797-5802]. Northern-blot analysis revealed that the steady-state levels of GST Ya1, Yb1 and Yb2 mRNAs are 2.5-3-fold higher in male as compared with female rat liver. In contrast, GST Yc and Ya2 mRNAs were expressed at a 2-3-fold higher level in female rat liver. Microsomal GST mRNA did not exhibit significant sex-dependent differences in rat liver. Treatment of male rats with GH by continuous infusion suppressed expression of the male-dominant GST Ya1, Yb1 and Yb2 mRNAs to levels at or below those found in female rat liver. This suppressive effect of GH was liver-specific, insofar as GH treatment did not alter kidney GST Ya1 mRNA levels. Hypophysectomy increased expression of the male-dominant GSTs, particularly in female rats (e.g. 8-fold elevation of GST Ya1 mRNA). GST Yc mRNA was increased approx. 2-fold in hypophysectomized males, indicating that this mRNA is subject to negative regulation by one or more pituitary-dependent factors. Continuous GH treatment of the hypophysectomized rats suppressed the expression of mRNA of GSTs Ya1, Yb1 and Yb2 when given as a continuous infusion, but not when given by an intermittent (twice daily) GH-injection schedule. Combination of continuous exposure to GH with thyroxine treatment resulted in a more complete suppression of GSTs Ya1, Yb1 and Yb2. In contrast, thyroxine increased the expression of GST Yc in hypophysectomized rats. These studies establish that several Alpha and Mu class GSTs are expressed in a sex-dependent fashion in adult rat liver, where they are regulated by multiple pituitary-dependent hormones through pretranslational mechanisms.

We report the cDNA sequence for rat glutathione transferase (GST) subunit 5, which is one of at least three class Theta subunits in this species. This sequence, when compared with that of subunit 12 recently published by Ogura, Nishiyama, Okada, Kajita, Narihata, Watabe, Hiratsuka & Watabe [(1991) Biochem. Biophys. Res. Commun. 181, 1294-1300] proves that Theta is a separate multigene class of GST with little amino acid sequence identity with Mu-, Alpha- or Pi-class enzymes. The amino acid sequence identity of class-Theta subunits is highly conserved in rat, the fruitfly Drosophila, maize (Zea mays) and Methylobacterium, which suggests that this family is representative of the ancient progenitor GST gene and originates from the endosymbioses of a purple bacterium leading to the mitochondrion. The high conservation of class Theta brings into prominence that Alpha-, Mu- and Pi-class enzymes, which are not present in plants, derive from a Theta-class gene duplication before the divergence of fungi and animals and, given the binding properties of the Alpha-, Mu- and Pi-classes, suggests a role for these in the evolution of fungi and animals.

The cytosolic glutathione S-transferases (GST) from human skeletal muscle were purified by a combination of affinity chromatography and anion-exchange chromatography followed by either chromatofocusing or hydroxyapatite chromatography. Pi-class and Mu-class GST, but not Alpha-class GST, were isolated from muscle. In addition to a Pi-class GST subunit, which exists as a homodimer, this tissue also contains a total of three distinct neutral-type Mu-class GST subunits, which hybridize to form homodimers or heterodimers. The neutral-type subunits are referred to as N1-N3 and are defined by the decreasing isoelectric points of the homodimers; GST N1N1, N2N2 and N3N3 have estimated pI values of 6.1, 5.3 and less than 5.0 respectively. SDS/PAGE showed that N1, N2 and N3 have Mr values of 26,700, 26,000 and 26,300 respectively. The N1, N2 and N3 subunits are catalytically distinct, with N1 possessing a high activity for trans-4-phenylbut-3-en-2-one and N2 having high activity with 1,2-dichloro-4-nitrobenzene. In skeletal muscle the expression of the N1 subunit, but not of N2 and N3 subunits, was found to differ from specimen to specimen. The N1 subunit was absent from about 50% of samples examined, and the purification results from two different specimens are presented to illustrate this inter-individual variation. Skeletal muscle from one individual (M1), which did not express N1, contained only GST N2N2, N2N3 and pi, whereas the second sample examined (M2) contained GST N1N2, N2N2 and N2N3 as well as GST pi. N-Terminal amino acid sequence analysis supported the electrophoretic evidence that the N2 subunit in GST N1N2, N2N2 and N2N3 represents the same polypeptide. The peptides obtained from CNBr digests of N2 were subjected separately to automated amino acid sequencing, and the results indicate that N2 is distinct but closely related to the protein encoded by the human Mu-class cDNA clone GTH4 [DeJong, Chang, Whang-Peng, Knutsen & Tu (1988) Nucleic Acids Res. 16, 8541-8554]. GST N2N2 is probably identical with GST 4 [Board, Suzuki & Shaw (1988) Biochim. Biophys. Acta 953, 214-217], as over the 24 N-terminal residues of GST 4 there is complete identity between the two enzymes. Our data suggest that the GST 1 and GST 4 loci are part of the same multi-gene family.

Normal rat liver expresses Ya (Mr 25,500), Yc (Mr 27,500) and Yk (Mr 25,000) Class Alpha glutathione S-transferase (GST) subunits. The Ya-type subunit can be resolved into two separate polypeptides, designated Ya1 and Ya2, by reverse-phase h.p.l.c. In rat livers that possess aflatoxin B1-induced pre-neoplastic nodules, a marked increase is observed in the expression of Ya1, Ya2, Yc and Yk; of these subunits, Ya2 exhibited the greatest increase in concentration. The Ya1 and Ya2 subunits isolated from nodule-bearing livers were cleaved with CNBr, and the purified peptides were subjected to automated amino-acid-sequence analysis. Differences in the primary structures of the two Ya GST subunits were found at positions 31, 34, 107 and 117. These data demonstrate that Ya1 and Ya2 are distinct polypeptides and are the products of separate genes. The amino acid sequences obtained from Ya1 and Ya2 were compared with the cloned cDNAs pGTB 38 [Pickett, Telakowski-Hopkins, Ding, Argenbright & Lu (1984) J. Biol. Chem. 259, 4112-4115] and pGTR 261 [Lai, Li, Weiss, Reddy & Tu (1984) J. Biol. Chem. 259, 5182-5188], which encode rat Ya-type subunits. From these comparisons it appears probable that Ya1 represents the GST subunit encoded by pGTR 261, whereas Ya2 represents the subunit encoded by pGTB 38. It is likely that the over-expression of Ya1 and Ya2 in nodule-bearing livers is of major significance in the acquired resistance of nodules to aflatoxin B1, since previous work [Coles, Meyer, Ketterer, Stanton & Garner (1985) Carcinogenesis 6, 693-697] has shown that the Ya-type GST subunit has high activity towards aflatoxin B1 8,9-epoxide.

Cytosolic glutathione transferases (GSTs) were purified from the rat spleen by S-hexyl-GSH-Sepharose chromatography, and two major forms were identified as GSTs 2-2 and 7-7 (GST P). Besides these forms an acidic form (pI 5.8) was purified by chromatofocusing at pH 7-4 and it accounted for about 1% of the total GST activity bound to S-hexyl-GSH-Sepharose. Two-dimensional gel electrophoresis revealed that it is a homodimer (subunit Mr 26,000 with pI 5.8). Immunoblot analysis demonstrated that it was immunologically related to GSTs 2-2 and 1-1, and its N-terminal amino acid was apparently blocked, similarly to other forms of the class Alpha. This form had a low activity towards cumene hydroperoxide or 4-hydroxynon-2-enal, indicating that this form differed from GSTs 10-10 and 8-8 as well as from GSTs 1-1 and 2-2. These results suggest that it is a new form of GST belonging to the class Alpha.

A labile GSH transferase homodimer termed 11-11 was purified from rat testis by GSH-agarose affinity chromatography followed by anion-exchange f.p.l.c. The enzyme is unstable in the absence of thiol(s) and has relatively low affinity for both 1-chloro-2,4-dinitrobenzene (Km 4.4 mM) and GSH (Km(app.) 4.4mM). Its mobility on SDS/polyacrylamide-gel electrophoresis is slightly less than that of subunits 3 and 4 and its pI is 5.2. Subunit 11 has a blocked N-terminal amino acid residue, but after CNBr cleavage fragments accounting for 113 amino acid residues were sequenced and showed 65% homology with corresponding sequences in subunit 4, indicating that it is a member of the Mu family. GSH transferase 11 is a major isoenzyme in testis, epididymis, prostate and brain and present at lower concentrations in other tissues.

Six GSH transferases with neutral/acidic isoelectric points were purified from the cytosol fraction of rat liver. Four transferases are class Mu enzymes related to the previously characterized GSH transferases 3-3, 4-4 and 6-6, as judged by structural and enzymic properties. Two additional GSH transferases are distinguished by high specific activities with 4-hydroxyalk-2-enals, toxic products of lipid peroxidation. The most abundant of these two enzymes, GSH transferase 8-8, a class Alpha enzyme, has earlier been identified in rat lung and kidney. The amino acid sequence of subunit 8 was determined and showed a typical class Alpha GSH transferase structure including an N-acetylated N-terminal methionine residue.