Structures of class pi glutathione S-transferase from human placenta in complex with substrate, transition-state analogue and inhibitor.

BACKGROUND: Glutathione S-transferases (GSTs) are detoxification enzymes, found in all aerobic organisms, which catalyse the conjugation of glutathione with a wide range of hydrophobic electrophilic substrates, thereby protecting the cell from serious damage caused by electrophilic compounds. GSTs are classified into five distinct classes (alpha, mu, pi, sigma and theta) by their substrate specificity and primary structure. Human GSTs are of interest because tumour cells show increased levels of expression of single classes of GSTs, which leads to drug resistance. Structural differences between classes of GST can therefore be utilised to develop new anti-cancer drugs. Many mutational and structural studies have been carried out on the mu and alpha classes of GST to elucidate the reaction mechanism, whereas knowledge about the pi class is still limited. RESULTS: We have solved the structures of the pi class GST hP1-1 in complex with its substrate, glutathione, a transition-state complex, the Meisenheimer complex, and an inhibitor, S-(rho-bromobenzyl)-glutathione, and refined them to resolutions of 1.8 A, 2.0 A and 1.9 A, respectively. All ligand molecules are well-defined in the electron density. In all three structures, an additionally bound N-morpholino-ethansulfonic acid molecule from the buffer solution was found. CONCLUSIONS: In the structure of the GST-glutathione complex, two conserved water molecules are observed, one of which hydrogen bonds directly to the sulphur atom of glutathione and the other forms hydrogen bonds with residues around the glutathione-binding site. These water molecules are absent from the structure of the Meisenheimer complex bound to GST, implicating that deprotonation of the cysteine occurs during formation of the ternary complex which involves expulsion of the inner bound water molecule. The comparison of our structures with known mu class GST structures show differences in the location of the electrophile-binding site (H-site), explaining the different substrate specificities of the two classes. Fluorescence measurements are in agreement with the position of the N-morpholino-ethansulfonic acid, close to Trp28, identifying a possible ligandin-substrate binding site.

Determination of a binding site for a non-substrate ligand in mammalian cytosolic glutathione S-transferases by means of fluorescence-resonance energy transfer.

To determine the location of the non-substrate-ligand-binding region in mammalian glutathione S-transferases, fluorescence-resonance energy transfer was used to calculate distances between tryptophan residues and protein-bound 8-anilinonaphthalene 1-sulphonate (an anionic ligand) in the human class-alpha glutathione S-transferase, and in a human Trp28-->Phe mutant class-pi glutathione S-transferase. Distance values of 2.21 nm and 1.82 nm were calculated for the class-alpha and class-pi enzymes, respectively. Since glutathione S-transferases bind one non-substrate ligand/protein dimer, the ligand-binding region, according to the calculated distances, is found to be located in the dimer interface near the twofold axis. This region is the same as that in which the parasitic helminth Schistosoma japonicum glutathione S-transferase binds praziquantel, a non-substrate drug used to treat schistosomiasis [McTigue, M. A., Williams, D. R. & Tainer, J. A. (1995) J. Mol. Biol. 246, 21-27]. Since the overall folding topology is conserved and certain features at the dimer interface are similar throughout the superfamily, it is reasonable to expect that all cytosolic glutathione S-transferases bind non-substrate ligands in the amphipathic groove at the dimer interface.

Cell-specific expression of a recombinant rat glutathione S-transferase Ya gene in transgenic mice.

Transgenic mice have been generated which carry a cDNA encoding the rat Ya isozyme of glutathione S-transferase (GST) under the transcriptional control of the SV40 early region promoter-enhancer. Expression of the GST transgene was highly tissue-specific, with the highest expression detected in the convoluted tubular epithelium of the mouse kidney cortex. GST Ya mRNA abundance in these cells was greater than that found for GST Ya mRNA in normal rat liver. GST Ya protein was observed in the convoluted tubule cells of the founder mouse as well as an F1 offspring. The transmission of the foreign gene was followed for two generations, and an erratic pattern of inheritance was observed. These animals provide a model for the in vivo study of GST modulation of carcinogenesis and drug toxicity.

Structural studies on human glutathione S-transferase pi. Substitution mutations to determine amino acids necessary for binding glutathione.

In order to identify amino acids involved in binding the co-substrate glutathione to the human glutathione S-transferase (GST) pi enzyme, we assembled three criteria to implicate amino acids whose role in binding and catalysis could be tested. Presence of a residue in the highly conserved exon 4 of the GST gene, positional conservation of a residue in 12 glutathione S-transferase amino acid sequences, and results from published chemical modification studies were used to implicate 14 residues. A bacterial expression vector (pUC120 pi), which enabled abundant production (2-26% of soluble Escherichia coli protein) of wild-type or mutant GST pi, was constructed, and, following nonconservative substitution mutation of the 14 implicated residues, five mutants (R13S, D57K, Q64R, I68Y, L72F) showed a greater than 95% decrease in specific activity. A quantitative assay was developed which rapidly measured the ability of wild-type or mutant glutathione S-transferase to bind to glutathione-agarose. Using this assay, each of the five loss of function mutants showed a greater than 20-fold decrease in binding glutathione, an observation consistent with a recent crystal structure analysis showing that several of these residues help to form the glutathione-binding cleft.

Mutational substitution of residues implicated by crystal structure in binding the substrate glutathione to human glutathione S-transferase pi.

Site-directed substitution mutations were introduced into a cDNA expression vector (pUC120 pi) that encoded a human glutathione S-transferase pi isozyme to non-conservatively replace four residues (Tyr7, Arg13, Gln62 and Asp96). Our earlier X-ray crystallographic analysis implicated these residues in binding and/or chemically activating the substrate glutathione. Each substitution mutation decreased the specific activity of the enzyme to less than 2% of the wild-type. Glutathione-binding was also reduced; however, the Tyr7----Phe mutant still retained 27% of the wild-type capacity to bind glutathione, underlining the primary role that this residue is likely to play in chemically activating the glutathione molecule during catalysis.

Expression of tandem glutathione S-transferase recombinant genes in COS cells for analysis of efficiency of protein expression and associated drug resistance.

Expression vectors were designed and constructed to achieve optimum production of two different isozymes of rat glutathione S-transferase (GST) (EC 2.5.1.18) in COS cells, for studies of drug resistance. Promoter-enhancer elements from the simian virus 40 (SV40) early-region or the mouse alpha 2(I)-collagen gene, GST cDNAs encoding the rat Ya or Yb1 isozymes, and an SV40 replicative origin (ori) were positioned in the vector to express two GSTs at high levels in the same cell. The optimized construct yielded levels of both GST proteins (1% of postmitochondrial protein fraction) that were up to 1.3-fold greater than the sum of those produced individually by two single-unit expression constructs. The best production of the tandem recombinant gene products was observed when the genes were placed in a head to head orientation in close proximity (1 kilobase). With the recombinant genes configured in this way, the plasmid DNA was also amplified in COS cells to higher levels (30% increase over single-unit expression constructs), as ori elements were placed on both DNA strands. Cells expressing the recombinant GSTs were viably sorted by flow cytometry on the basis of a GST-catalyzed conjugation of glutathione to monochlorobimane. Sorted COS cells that expressed both GST Ya and Yb1 from recombinant genes in a tandem, head to head configuration were 25 or 70% more resistant to the alkylating agent chlorambucil than cells that expressed GST Ya or Yb1 alone.

Recombinant glutathione S-transferase (GST) expressing cells purified by flow cytometry on the basis of a GST-catalyzed intracellular conjugation of glutathione to monochlorobimane.

COS cells transiently expressing glutathione S-transferase (GST) pi, Ya, or Yb1 (human Pi, rat Alpha or Mu, cytosolic classes) were purified by flow cytometry and used in colony-forming assays to show that GST confers cellular resistance to the carcinogen benzo[a]pyrene (+/-)-anti-diol epoxide (anti-BPDE). We developed a sorting technique to viably separate recombinant GST+ cells (20%) from the nonexpressing electroporated population (80%) on the basis of a GST-catalyzed intracellular conjugation of glutathione to the fluorescent labeling reagent monochlorobimane (mClB). The concentration of mClB, length of time cells are exposed to mClB, and activity of the expressed GST isozyme determined the degree to which recombinant GST+ cells fluoresced more intensely than controls. On-line reagent addition ensured that all cells were exposed to 25 microM mClB for 30-35 s during transit before being analyzed for fluorescence intensity and sorted. The apparent Km for mClB of the endogenous COS cell GST-catalyzed intracellular reaction was 88 microM. Stained GST Ya+ or Yb1+ cells catalyzed the conjugation 2 or 5 times more effectively than GST pi+ cells. Enzyme activity in cytosolic fractions prepared from sorted recombinant GST+ cells was 1.8 +/- 0.3-fold greater than that of the control (80 +/- 4 nmol/min/mg protein). Upon a 5-fold purification of GST pi+ cells in the electroporated population, resistance to anti-BPDE in colony-forming assays increased 5 times, from 1.1-fold (unsorted) to 1.5-fold (sorted) (P less than 0.001).

Localization of a transcription start point within the human Hinf element.

The Hinf family is a repetitive nucleotide sequence of the human genome. Certain structural features of Hinf DNA resemble the eukaryotic RNA polymerase II promoters. Therefore, we studied the ability of the Hinf element to function as a transcriptional promoter in mammalian cells. We placed the Hinf element upstream from the thymidine kinase-encoding (tk) sequence in a plasmid construct, pAC401 and introduced it into Ltk- mouse cells. The Hinf-tk plasmid was able to transform Ltk- cells to Tk+ phenotype. In another plasmid construct, pAC Hinf-neo, the Hinf element was inserted upstream from the sequence encoding neomycin (Nm) resistance (neo), and this plasmid was able to confer Nm resistance to HeLa cells. The nature of transcription initiation of the tk gene in four of the Tk+ clones transformed by pAC401 was examined by S1 nuclease analysis, and the transcription start point (tsp) for the tk gene in these clones was mapped within the Hinf element. The same tsp in the Hinf element was found in HeLa cells. Our studies show that the Hinf element functions as a weak promoter.

Characterization of mutagen-activated cellular oncogenes that confer anchorage independence to human fibroblasts and tumorigenicity to NIH 3T3 cells: sequence analysis of an enzymatically amplified mutant HRAS allele.

Treatment of diploid human fibroblasts with an alkylating mutagen has been shown to induce stable, anchorage-independent cell populations at frequencies (11 X 10(-4) consistent with an activating mutation. After treatment of human foreskin fibroblasts with the mutagen benzo[a]pyrene (+/-)anti- 7,8-dihydrodiol 9,10-epoxide and selection in soft agar, 17 anchorage-independent clones were isolated and expanded, and their cellular DNA was used to cotransfect NIH 3T3 cells along with pSV2neo. DNA from 11 of the 17 clones induced multiple NIH 3T3 cell tumors in recipient nude mice. Southern blot analyses showed the presence of human Alu repetitive sequences in all of the NIH 3T3 tumor cell DNAs. Intact, human HRAS sequences were observed in 2 of the 11 tumor groups, whereas no hybridization was detected when human KRAS or NRAS probes were used. Slow-migrating ras p21 proteins, consistent with codon 12 mutations, were observed i in the same two NIH 3T3 tumor cell groups that contained the human HRAS bands. Genomic DNA from one of these two human anchorage-independent cell populations (clone 21A) was used to enzymatically amplify a portion of exon 1 of the HRAS gene. Direct sequence analysis of the amplified DNA indicated equal presence of a wild-type (GGC) and mutant (GTC) allele of the HRAS gene. The results demonstrate that exposure of normal human cells to a common environmental mutagen yields HRAS GC----TA codon 12 transversions that have been commonly observed in human tumors. This oncogene as well as yet to be identified oncogene are also shown to stably confer anchorage-independence to human cells.

Partially transformed, anchorage-independent human diploid fibroblasts result from overexpression of the c-sis oncogene: mitogenic activity of an apparent monomeric platelet-derived growth factor 2 species.

A human c-sis cDNA in an expression vector was introduced into human diploid fibroblasts by transfection or electroporation. Fibroblast clones showing an aberrant, densely packed colony morphology were isolated and found to overexpress a 3.6-kilobase sis mRNA species and associated immunoprecipitable platelet-derived growth factor (PDGF) 2 proteins. Parallel analyses in cell clones of sis mRNA expression and colony formation in agar indicated that, above a threshold, a linear, positive correlation existed between sis overexpression and acquired anchorage independence. The sis-overexpressing cells formed transient, regressing tumor nodules when injected into nude mice, consistent with the finite life span which they retained. Protein products generated from the transfected c-sis construct in two overexpressing clones were immunoprecipitated with anti-human PDGF antibodies. One clone contained an apparent PDGF dimer of 21 kilodaltons; the second clone contained only an apparent PDGF monomer of 12 kilodaltons, which was shown to account for all of the mitogenic activity present in the cells, essentially all of which was concentrated in the membrane fraction. The results demonstrate a clear link between sis overexpression and acquisition of a partially transformed, anchorage-independent phenotype, and when combined with previous observations of sis overexpression in human tumors, clearly implicate sis overexpression as a genetic mechanism which contributes to human cell transformation.

Promoter-glutathione S-transferase Ya cDNA hybrid genes. Expression and conferred resistance to an alkylating molecule in mammalian cells.

Promoter-glutathione S-transferase Ya cDNA hybrid genes were constructed and analyzed to determine the efficiency with which the Ya coding sequence was transcribed and also to determine the associated levels of Ya-specific enzyme activity in mammalian cells which had received the hybrid gene constructs via electroporation. Promoter-containing fragments from either the SV40 early region or the herpes simplex thymidine kinase gene were positioned 5' to the Ya cDNA present in the pGTB38 plasmid. Both promoters supported transcription in in vitro run-off incubations containing a rat cell extract. Efficient transcription was also observed in both monkey Cos cells and mouse C3H/10T1/2 cells. Constructs containing the SV40 promoter and a residual portion of the homopolymeric G tail used in the original Ya cDNA cloning consistently gave 4-50-fold higher levels of transcript than other promoter-cDNA configurations. Associated with transcription of the hybrid gene was the appearance of a glutathione S-transferase YaYa-specific enzyme activity (delta 5-androstene-3,17-dione isomerization) in cytosols of cells electroporated with the hybrid genes. 50-260-fold increases in Ya-specific enzyme activity were found in Cos or C3H/10T1/2 cells containing multiple, episomal copies of the plasmid constructs; enzyme levels dropped in cells containing fewer, integrated plasmid copies. When a mixed population of Cos cells containing YaYa overexpressing cells was treated with benzo(a)pyrene (+/-)-anti-diol epoxide, a cytotoxic alkylating molecule and known YaYa substrate, a 20-30-fold enrichment in clones of YaYa overexpressing cells was seen among those cells which survived the treatment. The results clearly indicate that glutathione S-transferase isozymes can be overexpressed in mammalian cells and that this is accompanied by significant biological resistance to a known alkylating molecule.

Integration of a mutant c-Ha-ras oncogene into C3H/10T1/2 cells and its relationship to tumorigenic transformation.

C3H/10T1/2-CL8 mouse cells were shown to take up and express a plasmid-cloned drug resistance gene (Ecogpt) after DNA transfection at a frequency (2-6 X 10(-4) which is acceptable for routine recovery of gene-transformed populations. Transfection of 10T1/2 cells with a mutant c-Ha-ras oncogene (pEJ6.6 plasmid) results in neoplastically transformed 10T1/2 cell populations as judged by colony morphology and tumorigenic growth in nude mice. The levels of mutant c-Ha-ras gene integration and expression in the tumorigenic cell populations and 10T1/2 cell controls were determined, and the highest level of mutant ras transcript was seen in the most tumorigenic cell population. A preliminary comparison of 10T1/2 and NIH/3T3 cells showed similar frequencies for pEJ 6.6-induced transformed foci and a similar lack of sensitivity to the transforming effects of a cloned B-lym oncogene. The results identify a genetic event, which has previously been shown to be carcinogen-inducible, that is permissive for neoplastic transformation of the widely used carcinogen-transformable 10T1/2 mouse cell line.

Mapping of the locus involved in the catabolic oxidation of D-amino acids in Pseudomonas aeruginosa PAO.

Mutants of Pseudomonas aeruginosa PAO deficient in their utilization of DL-valine have been found to have lost their capacity to utilize DL-alanine and L-proline. Use of conjugal and transductional mediated gene transfers have established the chromosomal location of this gene and also its pleotropic function in the induction of the D-amino acid oxidase, involved in the oxidative utilization of DL-valine, DL-alanine and L-proline. These point mutations are clustered in a single locus designated as Val D and mapped around the 19th minute on the chromosome.