Search

Expression of the three major cytosolic classes of glutathione S-transferases (GST; Pi, Alpha and Mu) was examined by 2D gel analysis and Western blotting of biopsies from 26 patients diagnosed with ovarian carcinoma. In contrast to other tissues, at least one 'constitutive' subunit from each of the three major cytosolic GST classes was expressed. In most cases, pi appeared to be the major form present, although levels of alpha and mu subunit expression were approximately equal to pi in some patients. There was no detectable effect of prior chemotherapy on enzyme activity. Mean transferase activity for primary carcinoma was 79.9 +/- 11.9 (mean +/- SEM; nmol min-1 mg-1), with three pair-matched normal tissues showing minor decreases in transferase activity. One sample, in which a 32% increase in tumour enzyme activity was noted, was from a patient with primary disease and was associated with marked overexpression of a relatively basic form of alpha which was absent from the matching normal tissue, but present in 20% of all tumours examined. RFLP analysis of genomic tumour DNA using a human mu class cDNA probe indicated that at least two of the three mu forms (the 'constitutive' form and one other) observed in ovarian tissue were allelic variants, as a one-to-one correlation was observed between the presence of two Hind III fragments at 13.1 and 2.2 kb and expression of a second, more basic, variable form. This latter form was positively identified as the mu class subunit mu based on Southern analysis and was seen to be present in 40% of the samples examined. However, in the absence of mu expression, at least one other mu class subunit probably corresponding to GST psi, was seen to be present. Thus, at least in ovarian tissues, absence of the mu subunit does not necessarily imply a lack of ability to metabolize mu substrates, as psi has similar catalytic activity. A third mu subunit, probably corresponding to GST phi based on its relatively acidic pI, was also noted in 72% of samples examined, but has unknown substrate specificity. Increased expression of both alpha and mu forms may be of relevance to disease diagnosis and drug response.

A series of glutathione S-transferase pi (GST-pi) transfectant cell lines have been constructed in activated c-H-ras-transformed NIH-3T3 cells (pT22-3) by using a pKOneo plasmid and an expression vector containing cDNA for GST-pi with a beta-actin gene promoter. From the wild type pT22-3 cells, two clones were selected and designated RGN1 and RGN2. The degree of overexpression of GST-pi was estimated by Northern and Southern blot analysis to be incrementally higher in RGN2 compared with RGN1. Translation of mRNA was estimated by Western blot analysis using isozyme-specific polyclonal antibodies and confirmed the relative GST-pi levels. Each cell line, including the wild type, expressed alpha and mu class isozymes to the same degree and had similar but negligible expression of the mdr 1 gene. Sensitivity to various anticancer drugs and radiation was estimated by a series of cytotoxicity assays. The data confirmed that GST-pi provided a degree of protection against the toxicity of ethacrynic acid and adriamycin, but sensitivity to alkylating agents such as chlorambucil, melphalan, and cis-platinum was not influenced by GST-pi. Similarly, the response to ionizing radiation was similar for each line. Since the levels of intracellular GSH were also not significantly different, the availability of co-substrate was not a factor in determining response. In creating the GST-pi transfectants, these data establish that while increased isozyme levels can play a role in determining sensitivity to some agents, the protective effect is selective.

The aim of this study was to find factors that could explain the accumulation difference of mitoxantrone in the BCRP1-negative GLC4-MITO cell line compared to GLC4. Comparative genomic hybridisation (CGH) was applied to determine chromosomal differences between GLC4 and GLC4-MITO. Comparative genomic hybridisation analysis revealed gain of 2q, 6p, 9q, 13q, 14q, 15q, 19q and Xp and loss of 1p, 2q, 3p, 3q, 4q, 6q, 8q, 11p, 16p, 17q, 18p, 20p and Xq. In the over-represented chromosomal areas, seven transporter genes were identified: ABCB6, ABCB2 (TAP1), ABCB3 (TAP2), ABCF1 (ABC50), ABCC10 (MRP7), ABCA2 (ABC2) and ABCC4 (MRP4). No RNA or protein upregulation was observed for ABCB6, ABCF1, ABCC10, ABCC4, ABCB2 and ABCB3, but an increased expression was detected for ABCA2 mRNA in GLC4-MITO. ABCA2 is known to be involved in resistance to estramustine. In the MTT assay, GLC4-MITO was two-fold resistant to estramustine compared to GLC4. Coincubation with estramustine and mitoxantrone increased mitoxantrone accumulation in GLC4-MITO, while this was not affected in GLC4. This suggests that estramustine is able to block mitoxantrone efflux in GLC4-MITO cells. These data reveal that cellular reduction of mitoxantrone in a mitoxantrone-resistant cell line is associated with overexpression of ABCA2.

Glutathione S-transferase P1-1 (GSTpi) is an abundant and ubiquitously expressed protein in normal and malignant mammalian tissues and possesses catalytic and ligand binding properties. Our present data suggest that the protein contributes to the regulation of cell proliferation. Mouse embryo fibroblasts (MEFs) isolated from mice with a GSTP1-1 [glutathione S-transferase P1-1 (isozyme in nonhepatic tissue)] null genotype (GSTpi(-/-)) doubled their population in 26.2 h versus 33.6 h for the wild type (GSTpi(+/+)). Retroviral transfection of GSTP1-1 into GSTpi(-/-) MEF cells slowed the doubling time to 30.4 h. Both early passage and immortalized MEF cells from GSTpi(-/-) animals expressed significantly elevated activity of extracellular signal-regulated kinases ERK1/ERK2, kinases linked to cell proliferation pathways. In vivo, GSTpi(-/-) mice had higher basal levels of circulating white blood cells compared with GSTpi(+/+). Administration of a peptidomimetic inhibitor of GSTP1-1, TLK199, (gamma-glutamyl-S-(benzyl)cysteinyl-R-phenyl glycine diethyl ester), stimulated both lymphocyte production and bone marrow progenitor (colony-forming unit-granulocyte macrophage) proliferation, but only in GSTpi(+/+) and not in GSTpi(-/-) animals. Selection of a resistant clone of an HL60 tumor cell line through chronic exposure to TLK199 resulted in cells with elevated activities of c-Jun NH2 terminal kinase (JNK1) and ERK1/ERK2, and allowed the cells to proliferate under stress conditions that induced high levels of apoptosis in the wild type cells. The in vitro and in vivo data are consistent with the principle that GSTP1-1 influences cell proliferation.

The aminothiol WR-1065 (the active form of amifostine) protects normal tissues from the toxic effects of certain cancer drugs, while leaving their antitumor effects unchanged. The present data address the mechanism of action of this dichotomous effect. (35)S-Labeled WR-1065 bound directly to the transcription factors nuclear factor-kappaB, activator protein-1, and p53, resulting in enhanced binding of these proteins to target regulatory DNA sequences and subsequent transactivation of a number of downstream genes. Since other small molecular thiols could mimic WR-1065, the redox potential of the sulfhydryl is an important determinant of its activity. In nontransformed cells, WR-1065 protected cells from the cytotoxic effects of paclitaxel in a p53-dependent manner. However, in a transformed human tumor cell line, there was no cytoprotectivity by WR-1065, consistent with the premise that p53-dependent growth arrest is the basis for the protective effect of this compound, and that this pathway is abrogated in human tumors. The combined data support the principle that the cellular effects of the aminothiol WR-1065 are mediated through an impact on transcriptional regulation and are not only a consequence of radical scavenging.

We have isolated the full-length cDNA for human ATP-binding cassette, sub-family A, member 2 transporter (ABCA2). The ORF of this cDNA encodes a protein consisting of 2436 amino acids with apparent molecular weight of M(r) 270,000. Accordingly, ABCA2 is the largest known mammalian ABC transporter described thus far. Analysis of mRNA expression levels indicated that ABCA2 is highest in human brain and has a broad expression pattern in a panel of tumor cell lines. Using specific antibodies to ABCA2 and various organelle marker proteins, ABCA2 was found to colocalize with the lysosomal/endosomal marker LAMP1, forming discrete, punctate intracellular vesicles. In ABCA2-transfected cells, the transporter also colocalized with a fluorescently labeled steroid analogue, estramustine. The sequestration of the steroid into the lysosomal/endosomal compartment indicates a potential substrate specificity for ABCA2. Furthermore, the presence of a lipocalin signature motif in the ABCA2 sequence suggests a possible broad role for this protein in the transport of steroids, lipids, and related molecules.

In higher eukaryotes, reactive oxygen species (ROS) are generated during respiration in mitochondria in the course of reduction of molecular oxygen as well as by distinct enzyme systems. ROS have been implicated in the regulation of diverse cellular functions including defense against pathogens, intracellular signaling, transcriptional activation, proliferation, and apoptosis. The reduction-oxidation (redox) state of the cell is primarily a consequence of the precise balance between the levels of ROS and endogenous thiol buffers present in the cell, such as glutathione and thioredoxin, which protect cells from oxidative damage. Dramatic elevation of ROS, exceeding compensatory changes in the level of the endogenous thiol buffers, may result in the sustained activation of signaling pathways and expression of genes that induce apoptosis in affected cells. Many cytotoxic drugs function selectively to kill cancer cells by the abrogation of proliferative signals, leading to cell death, and numerous reports have demonstrated that ROS are generated following treatment with these drugs. In this review, we will summarize recent contributions to our understanding of the importance of cytotoxic drug-induced modulation of cellular redox status for signaling and transcription leading to activation of apoptotic effector mechanisms.

Glutathione (GSH), glutathione S-transferases (GSTs), and the multidrug resistance-associated protein 1 (MRP1) have been independently studied for their contributions to drug resistance. Single cDNA transfection experiments have provided inconsistent and disparate conclusions with respect to the importance of GSH and GST in conferring a resistant phenotype. Because these three proteins can act as a concerted coordinated pathway, we reasoned that equivalent increases may be required for enhanced resistance to be expressed. We have assembled these proteins together, or in various combinations, to determine whether they show cooperativity in determining drug response. Increased expression through single cDNA transfection of GSTpi, gamma-glutamylcysteine synthetase (gamma-GCS) (regulatory plus catalytic subunits), or MRP1 enhanced resistance to a number of anticancer drugs. Cotransfection of GSTpi and GCS, gave higher resistance to doxorubicin, etoposide, and vincristine than with either alone. Resistance toward chlorambucil and ethacrynic acid was similar in cells overexpressing either component or overexpressing GST alone. Coexpression of GSTpi with MRP1 conferred significant resistance above that seen for MRP1 alone to chlorambucil, etoposide, ethacrynic acid, and vincristine. The combination of GCS and MRP1 did not afford additional resistance above MRP1 alone. When all three were transfected, significantly higher levels of resistance were found for doxorubicin and etoposide. These results support the concept that coordinate enhancement of focal thiol elements of detoxification pathways provides a more efficient protective phenotype than do single components alone.

The peptidomimetic drug gamma-glutamyl-S-(benzyl)cysteinyl-R-(-)-phenyl glycine diethyl ester (TER199) is an analog of glutathione designed to be an isozyme-specific inhibitor of GSTP1-1 protein1-1. This compound (and the de-esterified moiety) is shown to be an effective inhibitor of multidrug resistance-associated protein1 (MRP1)-mediated drug resistance. Kinetic analyses revealed that gamma-glutamyl-S-(benzyl)cysteinyl-R-(-)-phenyl glycine reversibly inhibits the transport of 2,4-dinitrophenyl-S-glutathione with a K(i) of 752 microM. TER199 reversed the accumulation deficit of daunorubicin in MRP1-transfected NIH3T3 fibroblasts and maintained intracellular levels for2 h after daunorubicin removal. Cytotoxicity assays revealed that TER199 significantly reversed the resistance of MRP1-transfected NIH3T3 cells for vincristine, doxorubicin, etoposide, and mitoxantrone. HL-60 cells made resistant to TER199 by chronic, long-term selection had increased mRNA and protein levels of multidrug resistance-associated protein, MRP1, and gamma-glutamyl cysteine synthetase heavy and light subunits (the rate-limiting enzyme in GSH synthesis). In spite of increased gamma-glutamyl cysteine synthetase, their glutathione content was reduced approximately 35% from that of parental HL-60 cells. These cells also exhibited a drug resistance profile commensurate with the previously described MRP1 overexpressing phenotype, with resistance to Vinca alkaloids, epipodophyllotoxins, and anthracyclines; additional cross-resistance to paclitaxel (Taxol), mitoxantrone, and 5-fluorouracil was observed.

A photoactivatable glutathione-drug conjugate (35)S-labeled-azidophenacyl-glutathione (APA-SG) was synthesized and used to identify protein(s) involved in recognition and/or transport of glutathione conjugates of electrophilic drug species. A approximately 460-kDa protein was found to be highly labeled by (35)S-labeled APA-SG in an Adriamycin-resistant HL-60 (HL-60/ADR) cell line and identified as the catalytic subunit of DNA-dependent protein kinase (DNA-PKcs) by amino acid sequence analysis, Western blot, and immunoprecipitation with specific antibodies. Binding specificity was confirmed by competition isotope dilution assays with purified proteins. A 15- to 20-fold increase in DNA-PKcs expression in the HL-60/ADR cell line was accompanied by an equivalent increase in (35)S-labeled APA-SG binding. APA-SG, along with other glutathione conjugates and analogs inhibited the DNA-PK-mediated phosphorylation of an in vitro peptide substrate in a concentration-dependent manner. Using different antibodies to immunoprecipitate the individual components of the DNA-PK complex (DNA-PKcs, Ku70, and Ku80), it was shown that APA-SG caused a destabilization of the trimeric holoenzyme complex by dissociating the catalytic subunit from the Ku heterodimer. These data suggest that the kinase-mediated signaling is inhibited when glutathione conjugates bind to DNA-PKcs and may also indicate a possible strategy for design of novel DNA-PK inhibitors.

Since several in vitro and animal studies of an extract from Crassostrea gigas (JCOE) have demonstrated its antioxidant properties and other interesting effects, a preliminary human trial was carried out. Seven healthy male volunteers aged 23-37 received orally 3 x 2 capsules of JCOE per day for 8 days. On days 0, 1, 4, 8 and 15 (7 days after completion of the schedule) blood samples were drawn and the antioxidant potential of serum was tested. A statistically significant increase in the buffering effects of serum against hemoglobin (Hb) and lactate dehydrogenase (LDH) release from red blood cells treated with the free radical generator azobis amidino propane (AAPH) was found following JCOE treatment. At 8 days, the oxidative effects were reduced by90% of the pretreatment values. In these same individuals, serum levels of reduced glutathione were increased by an average of 1.5-fold over the time course of treatment. It is concluded that in normal human volunteers, JCOE capsules provide an orally available formula for enhancing the antioxidant capacity of blood serum. While the extract is known to contain some direct acting antioxidant components, at least a portion of the protective effect is facilitated by enhancement of GSH biosynthesis.

Male Sprague Dawley rats received various amounts of extract of Crassostrea gigas by gavage every day for 2 weeks or one month. At these times, groups of animals were sacrificed and samples of major organs analyzed for levels of glutathione (GSH) and glutathione S-transferase (GST) activities. Following the two week protocol, GSH levels were significantly increased in the mucosa of the large intestine; at one month the small intestine and spleen were elevated. GST activity increased in liver under both schedules and at one month, activity was also elevated in kidney and small intestine. Since the Crassostrea gigas extract contains high levels of a variety of important amino acids, it is concluded that biologically available peptides are taken up in target organs and stimulate GSH metabolism. Enhanced levels of GSH and associated enzymes may contribute to a more effective detoxification phenotype, thus providing enhanced chemoprotective capacity.

Maintenance of cellular homeostasis is a critical survival trait when cells are exposed to electrophilic chemicals. Because conjugation and elimination of these toxins is dependent upon sequential and coordinated metabolic pathways, acquired resistance through a gradual adaptive response would rarely be expected to be the consequence of changes in one gene product. Human HT29 colon cancer cells chronically exposed to EA have acquired resistance to the drug. Commensurate with resistance, EA is more effectively conjugated to GSH and effluxed from the resistant cells. Using directed and random (differential display) approaches, a number of detoxification and/or protective gene products have been shown to be expressed at elevated levels. These include gamma-GCS (approximately 3-fold), GST-pi (approximately 3-fold), MRP (approximately 3-fold), NQO1 (approximately 3-fold), DDH (20-fold), and SSP 3521, a transcriptional regulator (approximately 3-fold). Multiple mechanisms contribute to these increases, including enhanced transcriptional rate and prolonged mRNA and protein half lives. Further indications for the involvement of transcriptional regulators is found in HL60 adriamycin-resistant cells which overexpress MRP, GST-pi and gamma-GCS and also have 15-20-fold more DNA-dependent protein kinase. It is possible that this enzyme serves as an early stress response gene which may activate downstream transcription factors. Intriguingly, the catalytic subunit of DNA-dependent protein kinase has a high avidity for [35S]azidophenacyl-GSH. High levels of GSH conjugates indicate cell stress and it would seem reasonable to speculate that DNA-dependent protein kinase may serve as a receiver and transmitter of signals which contribute to drug resistance and maintain cell viability.

TER286 is a latent drug activated by human glutathione S-transferase (GST) isoforms P1-1 and A1-1 to produce a nitrogen mustard alkylating agent. M7609 human colon carcinoma, selected for resistance to doxorubicin, and MCF-7 human breast carcinoma, selected for resistance to cyclophosphamide, both showed increased sensitivity to TER286 over their parental lines in parallel with increased expression of GST P1-1. In primary human tumor clonogenic assays, the spectrum of cytotoxic activity observed for TER286 was both broad and unusual when compared to a variety of current drugs. In murine xenografts of M7609 engineered to have high, medium, or low GST P1-1, responses to TER286 were positively correlated with the level of P1-1. Cytotoxicity was also observed in several other cell culture and xenograft models. In xenografts of the MX-1 human breast carcinoma, tumor growth inhibition or regression was observed in nearly all of the animals treated with an aggressive regimen of five daily doses. This schedule resulted in a 24-h posttreatment decline in bone marrow progenitors to 60% of control and was no worse than for a single dose of TER286. These studies have motivated election of TER286 as a clinical candidate.

An estramustine-resistant human ovarian carcinoma cell line, SKEM, was generated to explore resistance mechanisms associated with this agent. Cytogenetic analysis revealed that SKEM cells have a homogeneously staining region (hsr) at chromosome 9q34. Microdissection of the hsr, followed by fluorescence in situ hybridization to SKEM and normal metaphase spreads, confirmed that the amplified region was derived from sequences from 9q34. In situ hybridization with a probe specific for ABC2, a gene located at 9q34 that encodes an ATP-binding cassette 2 (ABC2) transporter, indicated that this gene is amplified approximately 6-fold in the estramustine-resistant cells. Southern analysis confirmed that ABC2 was amplified in SKEM, and Northern analysis indicated that the ABC2 transcript was overexpressed approximately 5-fold. The ABC1 gene located at 9q22-31 was not amplified in the resistant cells, and mRNA levels of several other ABC transporter genes were unaltered. Consistent with the concept that increased ABC2 expression contributes to the resistant phenotype, we observed that the rate of efflux of dansylated estramustine was increased in SKEM compared with control cells. In addition, antisense treatment directed toward ABC2 mRNA sensitized the resistant cells to estramustine. Together, these results suggest that amplification and overexpression of ABC2 contributes to estramustine resistance and provides the first indication of a potential cellular function for this product.

Maintenance of cellular homeostasis is a critical survival trait in tumors when exposed to anticancer drugs. Because conjugation and elimination of drugs and their metabolites is dependent upon sequential and coordinated pathways, acquired drug resistance through a gradual adaptive response would rarely be expected to be the consequence of changes in the expression of one gene product. We have used a number of drug-resistant human cell lines to characterize those genes that are implicated in maintaining a resistant phenotype. Human HT29 colon cancer cells chronically exposed to ethacrynic acid (EA) [a glutathione (GSH) and glutathione S-transferase (GST) modulator] have acquired resistance to the drug. Commensurate with resistance, EA is more effectively conjugated to GSH and effluxed from the resistant cells. Using directed and random (differential display) approaches, a number of detoxification and/or protective gene products have been shown to be expressed at elevated levels. These include: gamma-glutamyl cysteine synthetase (gamma-GCS, the rate-limiting enzyme in GSH biosynthesis); GST pi (the enzyme catalyzing the conjugation reaction); multidrug resistance associated protein (MRP) (the membrane pump responsible for effluxing the conjugate from the cell interior). In addition, other gene products not directly linked with EA metabolism were induced, including dihydrodiol dehydrogenase (an alpha-ketoreductase) (30-fold), DT-diaphorase (threefold), and a transcriptional regulator SSP 3521 (threefold). HL60 cells resistant to a GSH paralog Ter199 also show increased expression of some of these gene products. Furthermore, an adriamycin-resistant human HL60 cell line also shows overexpression of GST pi, gamma-GCS, and MRP, but in addition has approximately 20-fold more DNA-dependent protein kinase catalytic subunit (DNA-PKcs). This enzyme is an early stress response gene that can phosphorylate and activate downstream transcription factors. Such overexpression could impact on the transcriptional control of the other detoxification gene products. Both adriamycin and a typical drug-GSH conjugate (APA-SG) are inhibitors of DNA-PK. Because cellular levels of these conjugates would presumably be a good indicator of stress, it would seem reasonable to speculate that DNA-PK may act as a receiver and transmitter of signals that are crucial to the drug-resistant phenotype. Additionally, this enzyme may prove to be a potentially important target for drug design based upon the inhibitory activity of GSH conjugates.

The steady state expression of glutathione S-transferases (GSTs) at both the protein and mRNA level is reported for the 60 tumor cell lines that are used for the National Cancer Institute Drug Screening Program. Individual GST isozymes were separated, identified, and quantified (with reverse-phase calibration curves) through a novel high performance liquid chromatographic procedure. GSTP1 was the predominant isozyme and was found at quantifiable levels in all but two of the cell lines. This isozyme ranged from 0.03% to 2.7% of the total cytosolic protein. For the mu family, 90% of the lines had GSTM2, 68% had GSTM3, but only 28% were positive for the M1 phenotype. The M1 proportion is lower than would be expected from the standard M1 null phenotype for human populations. Isozymes of the alpha family were detected only at very low levels in 35% of the lines. Significant quantitative correlations among enzyme activity, total enzyme protein, and mRNA were shown for GSTP1. However, such relationships were not apparent for the mu or alpha families. Levels of glutathione (GSH), and the transcript levels of other enzymes involved in GSH homeostasis were determined. gamma-Glutamyl cysteine synthetase (gamma-GCS) was present in all cell lines, but did not correlate with levels of intracellular GSH. Glyoxalase-I and gamma-glutamyl transpeptidase, both involved in GSH salvage, were found in 100% and 70% of the cell lines, respectively. Using a pattern-matching computer program, COMPARE, we compared and correlated the arrays of mRNA and protein levels with the pattern of chemosensitivity or chemoresistance of the 60 cell lines with 175 agents constituting a standard agent database. This database is composed of compounds to which a putative mechanism of action has been assigned. Although Pearson correlation coefficients relating the target and drug patterns were generally modest, when the patterns for the enzyme protein and mRNA levels for GST pi were correlated to drug sensitivity patterns, the list of 30 agents most closely matching (for which P0.05) was enriched with alkylating agents. gamma-GCS also showed an enrichment of alkylating agents in the COMPARE correlations, indicating that high levels of gamma-GCS may be an important determinant of resistance. In contrast, none of the other enzymes or GSH had patterns of expression that resulted in an obvious correlation to the sensitivity or resistance of alkylating agents.

Estramustine (EM), an antimicrotubule agent, is effective against hormone-refractory prostate cancer when used in combination with vinblastine or paclitaxel. To understand the effect of EM on beta-tubulin, a cellular target for this class of drugs, human prostate carcinoma cells (DU-145) were made resistant to EM, and two cell lines were selected at 12- (EM-12) and 15-microMolar (EM-15) concentrations of the drug. These cell lines exhibited 8- to 9-fold resistance to EM and 2- to 4-fold cross-resistance to paclitaxel. Immunofluorescent staining of the cells with beta-tubulin isotype-specific antibodies showed an approximately 6-fold increase in the beta(III)-tubulin levels and moderate increase in overall beta-tubulin levels in EM-resistant cells when compared to DU-145 cells. This increase of beta(III) isotype was confirmed by Western analysis. A reverse transcriptase-PCR assay was also employed using beta-tubulin isotype-specific primers to quantify beta-tubulin isotype RNA. A 4-fold increase in beta(III) and a 3-fold increase in beta(IV alpha) transcript were seen in both EM-resistant cell lines. These results indicate that overexpression of specific beta-tubulin isotypes may play a role in the cellular defense against EM and other antimicrotubule agents.

We investigated the effects of glutathione-S-transferase (GST) inhibitor treatment on human colon HT29 cell mRNA levels of dihydrodiol dehydrogenase (DDH), glyoxalase I, and gamma-glutamylcysteine synthetase. Time- and concentration-dependent increases in both DDH and gamma-glutamylcysteine synthetase mRNAs resulted from treatment with ethacrynic acid, ethacrynic acid/glutathione conjugate, and T.199 (gamma-glutamyl-S-(benzyl)-cysteinyl-R(-)-phenyl glycine diethyl ester), a selective GST pi inhibitor. In contrast, glutathione analogue GST alpha- and GST mu-selective inhibitors did not induce expression of these genes. Treatment with ethacrynic acid or T.199 had no effect on the mRNA levels of the glutathione-dependent glyoxalase I gene. Pretreatment of cells with buthionine-DL-sulfoximine, a gamma-glutamylcysteine synthetase inhibitor and glutathione depleter, coupled with ethacrynic acid, ethacrynic acid/glutathione conjugate, or T.199 resulted in greater levels of gamma-glutamylcysteine synthetase and DDH induction compared with single treatments. Treatment with buthionine-DL-sulfoximine alone resulted in modest increases in both gamma-glutamylcysteine synthetase and DDH expression. Analyses of DDH induction by both differential Northern hybridization with specific oligonucleotides as probes and reverse transcriptase-polymerase chain reaction amplification of products, followed by diagnostic restriction digestion with endonucleases, showed that ethacrynic acid induced multiple DDH transcripts in HT29 cells and human HepG2 and SKHep1 hepatoma cells. Possible induction mechanisms include the alteration of sulfhydryl status by the electrophilic properties of EA or by elevations of endogenously generated oxidative stress via transient removal of GST pi from the cytosolic GST pool.

Estramustine previously has been shown to interact with P-glycoprotein and to restore intracellular accumulation of vinblastine and paclitaxel in cells overexpressing this drug transporter. However, the ability of estramustine to potentiate the cytotoxicities of several drugs was less than that expected. To resolve this apparent discordance, the authors examined the effects of serum on the actions of estramustine.The cytotoxicities of anticancer drugs with or without estramustine or verapamil toward MCF-7 breast carcinoma cells and a P-glycoprotein-overexpressing subline MCF-7/ADR were determined using the sulforhodamine-binding assay. The extent of intracellular accumulation of [3H]vinblastine and [3H]paclitaxel was determined for each using standard methods, and the binding of radiolabeled drugs to plasma proteins was characterized by equilibrium dialysis.Without serum, the sensitivities of MCF-7/ADR cells to several P-glycoprotein-transported drugs were increased by estramustine and verapamil. Conversely, when the cells were treated with a 10% serum, the cytotoxicities of these drugs were increased by verapamil, but not by estramustine. Without serum, intracellular accumulation of [3H]vinblastine and [3H]paclitaxel by MCF-7/ADR cells was increased markedly by verapamil and estramustine; however, serum suppressed the effects of estramustine much more strongly than those of verapamil. Equilibrium dialysis experiments demonstrated that [3H]estramustine binds to plasma proteins, predominantly albumin, whereas [3H]paclitaxel binds to albumin and alpha 1-acid-glycoprotein, and [3H]vinblastine binds predominantly to alpha 1-acid-glycoprotein.Although estramustine can bind to P-glycoprotein, its effectiveness as a reversing agent in vivo likely is limited by binding to plasma proteins.

Eukaryotic cells have evolved several mechanisms to protect cellular constituents, especially DNA, from highly reactive molecules entering from without. The greater affinity of electrophiles for thiol groups than for hydroxyl or amine groups provides a teleologic rationale that the availability of high concentrations of thiol could be protective of these other important entities. The major intracellular nonprotein thiol is the tripeptide glutathione.

To identify specific drug targets of the antimitotic drug estramustine, a photoaffinity analogue, 17-O-[[2-[3-(4-azido-3-[125I] iodophenyl)propionamido]ethyl]carbamyl]estradiol-3-N-bis(2- chloroethyl)carbamate, was synthesized and reacted in competition assays with cytoskeletal protein preparations. By attaching the photoaffinity ligand to the 17 beta-position of the steroid D-ring, the cytotoxic properties of the drug were maintained. In cytoskeletal protein preparations from human prostate carcinoma cells (DU 145) or a clonally selected, estramustine-resistant cell line (E4), the major microtubule-associated protein (MAP) present was MAP4. In both cytoskeletal fractions and reconstituted microtubules, 17-O-[[2-[3-(4-azido-3-[125I]iodophenyl)propionamido] ethyl]carbamyl]estradiol-3-N-bis(2-chloroethyl)carbamate bound to both MAP4 and tubulin. From competition assays, the apparent binding constant for MAP4 from DU 145 cells was 15 microM. Similar calculations for tubulin gave values of 13 microM (bovine brain), 19 microM (DU 145 wild-type cells), and 25 microM (E4 cells). The identification of these cytoskeletal proteins as specific drug targets provides a direct explanation for the antimicrotubule and antimitotic effects of estramustine.

Ethacrynic acid (EA) is an inhibitor of the glutathione S-transferases (GSTs), a family of detoxification enzymes the expression of which has been associated with resistance to several classes of anticancer drugs. We performed a two-way randomized crossover study to investigate the pharmacokinetics and bioavailability of EA and describe any toxicities of EA associated with i.v. administration. We administered EA (100 mg) either by the p.o. or i.v. route on days 1 and 2 for pharmacokinetic analysis. After i.v. administration, plasma EA disappearance was biphasic in seven patients and monophasic in two patients with a terminal half-life of 30 and 8 min, respectively. Mean total body clearance was high; 1405 ml/min in patients described by using a one-compartment model and 611 ml/min in those patients described by a two-compartment model. After p.o. administration, peak EA plasma concentrations were less than 10% of i.v. EA and the absolute bioavailability was less than 21% (range, 7-35%). The urinary output as a result of EA treatment was equal following either route of administration and together with the large first-pass effect suggests that a metabolite(s) may be the active diuretic agent(s). Burning at the injection site was the only toxicity unique to the i.v. route of EA administration. We concluded that the systemic availability of EA after p.o. administration is low and variable. This finding supports the potential utility of the i.v. route of administration for the treatment of drug resistance.

The importance of thiol-mediated detoxification of anticancer drugs that produce toxic electrophiles has been of considerable interest to many investigators. Glutathione and glutathione S-transferases (GST) are the focus of much attention in characterizing drug resistant cells. However, ambiguous and sometimes conflicting data have complicated the field. This article attempts to clarify some of the confusion. The following observations are well established: (a) tumors express high levels of GST, especially GST psi, although the isozyme components vary quite markedly between tissues and the isozymes are inducible; (b) nitrogen mustards are good substrates for the GST alpha family of isozymes which are frequently overexpressed in cells with acquired resistance to these drugs; (c) most drugs of the multidrug-resistant phenotype have not been shown to be GST substrates and although GST psi is frequently overexpressed in multidrug-resistant cells, most indications are that this is an accompaniment to, rather than a cause of, the resistant phenotype; (d) transfection of GST complementary DNAs has produced some lines with increased resistance to alkylating agents. Most studies of the relationships between GST and resistance have overlooked the potential importance of other enzymes involved in the maintenance of cellular glutathione homeostasis, and this has complicated data interpretation. Translational research aimed at applying our knowledge of glutathione pathways has produced preclinical and clinical testing of some glutathione and GST inhibitors, with some encouraging preliminary results. In brief, GSTs are important determinants of drug response for some, not all, anticancer drugs. Caution should be encouraged in assessing cause/effect relationships between GST overexpression and resistance mechanisms.

A human oxidoreductase (H-37) that is overexpressed in ethacrynic acid-resistant HT29 colon cells (Ciaccio, P. J., Stuart, J.E., and Tew, K.D. (1993) Mol. Pharmacol. 43, 845-853) has been identified as a dihydrodiol dehydrogenase. Translated protein from a dihydrodiol dehydrogenase cDNA isolated from a library prepared from ethacrynic acid-resistant HT29 cell poly(A+) RNA was recognized by anti-H-37 IgG and was identical in molecular weight with H-37. The isolated cDNA was identical in both nucleotide and amino acid sequences with the recently cloned liver dihydrodiol dehydrogenase (Stolz, A., Hammond, L., Lou, H., Takikawa, H., Ronk, M., and Shively, J.E. (1993) J. Biol. Chem. 268, 10448-10457). Using this cDNA as probe, we have examined its induction by Michael acceptors. The steady state dihydrodiol dehydrogenase mRNA level in the ethacrynic acid-resistant line was increased 30-fold relative to that of wild-type cells. Twenty-four hour treatment of wild-type cells with ethacrynic acid or dimethyl maleate increased mRNA 10-fold and 5-fold, respectively. These changes are accompanied by both increased protein expression and increased NADP-dependent 1-acenaphthenol oxidative activity in cell cytosol. In gel shift assays, compared to wild type controls, increased binding of NAD(P)H quinone oxidoreductase human antioxidant response element (hARE) DNA to redox labile protein complexes present in treated and resistant cell nuclear extract was observed. Ethacrynic acid induced CAT activity 2-fold in Hepa1 cells stably transfected with NAD(P)H quinone oxidoreductase hARE-tk-CAT chimeric gene construct. Thus, dihydrodiol dehydrogenase protein is inducible by de novo synthesis from mRNA by structurally related monofunctional inducer Michael acceptors. Altered in vitro binding of nuclear protein to the hARE is indirect evidence for the involvement of an element similar to hARE in the regulation of dihydrodiol dehydrogenase by these agents.

Glyoxalase-I (Gly-I) is part of the glyoxalase system which converts methylglyoxal to D-lactic acid via an S-D-lactoylglutathione intermediate. This glutathione (GSH)-binding protein was purified from human colon tumors and corresponding normal tissue. The GSH-affinity purified fraction from normal human colon tissue showed enzyme activity of 30.6 +/- 11.5 mumol/min per mg protein, with methylglyoxal as substrate. Corresponding fractions from carcinomas showed significantly elevated Gly-I activity of 54.5 +/- 15 mumol/min per mg protein. Polyclonal antibodies made against human Gly-I cross-reacted weakly with mouse liver Gly-I but not with yeast Gly-I. Isoelectric points of Gly-I from human, mouse and yeast were determined to be 4.6, 4.9 and 7.0, respectively, by horizontal IEF. Immunohistochemical analysis confirmed the increase of Gly-I in human colon carcinoma in 16 out of 21 samples when compared to corresponding normal tissue. The elevated levels of Gly-I in colon tumors may be an indicator of the enhanced proliferative status of the neoplastic condition.

We report the initial identification of a 37.5-kDa putative aldoketo reductase in human colon carcinoma cells. An aminoterminal trypsin fragment was sequenced and found to be identical to bovine prostaglandin F synthase in 19 of 21 amino acids. Levels of this cytosolic human aldo-keto reductase, assessed by immunoblots using polyclonal antibodies raised against this protein, increased 30-fold in cells resistant to the Michael reaction acceptor ethacrynic acid and increased with time and ethacrynic acid concentration after treatment of wild-type cells. Induction of the reductase appeared to be cell type and drug specific. It was induced by the Michael reaction acceptors dimethyl maleate, t-butylhydroquinone, and hydroquinone but not by the nitrogen mustard chlorambucil. Ethacrynic acid and dimethyl maleate induced the reductase in a second human colon cell line but not in human prostate cells. NADPH-dependent metabolism of aldoketo reductase substrates by cytosol from colon but not prostate cells was enhanced 2-3-fold when cells were grown in the presence of either ethacrynic acid or dimethyl maleate. The discrepancy between induced reductase activity and protein levels may be due to the multiplicity of constitutively expressed NADPH-dependent reductases that compete for substrate. Ethacrynic acid-resistant cells exhibited low levels of cross-resistance to Adriamycin, mitomycin C, and the bovine prostaglandin F synthase substrates phenylglyoxal and prostaglandin D2. Thus, significant overexpression of a human aldo-keto reductase structurally related to bovine prostaglandin F synthase may result from exposure of cells to Michael reaction acceptors and may give rise to an enhanced capacity to metabolize exogenous and endogenous substrates, thereby contributing to the drug-resistant phenotype.

Glyoxalase-I cDNA clones were isolated from a human colon cDNA library using polyclonal antibodies raised against the protein purified from human colon tissue. Positive clones were purified, subcloned, and their nucleotide sequence determined. The glyoxalase-I cDNA encodes a 184-amino acid protein with a predicted molecular weight of 20,774, corresponding to the monomeric subunit weight of the purified protein from human colon glyoxalase-I. The human enzyme showed 51% homology at the nucleotide level and 42% at the amino acid level with bacterial glyoxalase-I. Transfection of COS-1 cells with the 622-base pair cDNA containing the entire coding region cloned into a pMT2 vector produced an immunoreactive protein and an approximate 180-fold increase in glyoxalase-I enzyme activity as determined with methylglyoxal as a substrate. Transfection of a truncated cDNA lacking 94 base pairs of the 5'-coding sequence also produced an approximately 15-kDa immunoreactive protein, but with no detectable increase in enzyme activity. Northern analysis of the RNA showed an approximately 12-fold increase of the 2.2-kilobase glyoxalase-I transcript in carcinoma when compared to normal colon tissue from the same patient. Examination of colon carcinomas for the amplification of the glyoxalase-I gene by Southern blot analysis revealed no change in gene copy number. These results suggest induction of the glyoxalase-I gene expression in colon carcinomas.

Estramustine (EM) and taxol, two antimicrotubule agents with distinct and apparently opposing mechanisms of action, were found to be effective in combination in the preclinical treatment of EM-resistant and sensitive, wild-type human prostatic carcinoma cell lines. Estramustine combined with 1 nM taxol (concentration 100-fold less than that measured in plasma of patients treated with taxol) produced greater than additive effects on the inhibition of cell survival of both wild-type and EM-resistant cells. When taxol was used with another microtubule-destabilizing drug, vinblastine, no significantly increased cytotoxicity was observed. Other effects on wild-type and EM-resistant cells produced by the combination of EM and taxol included (a) an increased proportion of the cells in the S phase of the cell cycle; (b) no mitotic block; and (c) an increase in the percentage of micronucleated cells from a control value of less than 1% to greater than 20% after drug treatment. Immunofluorescent microscopic analysis of the effect of this drug combination on the mitotic spindle apparatus revealed specific examples of aberrant mitotic figures, including multiple asters, cells with two distinct spindles, and tripolar spindles able to traverse mitosis and complete cytokinesis. These data provide supportive preclinical evidence for the potential development of an EM/taxol combination clinical regimen either for prostate or other cancers.

The crystal and molecular structures of estramustine and two of its analogues have been determined by X-ray crystallographic techniques (a total of three different compounds). The compounds studied are estramustine [1,3,5(10)-estratriene-3,17 beta-diol-3-N,N-bis(2'- chloroethyl)carbamate] and its monohydrate, estromustine [17-oxo-1,3,5(10)-estratriene-3-yl-N,N-bis(2'-chloroethyl)carbamate], and 17-oxo-5-androsten-3 beta-yl-N,N-bis(2'-chloroethyl)carbamate. Three views of estramustine were obtained from the study of its two crystal forms. The main structural features found are as follows: (a) the geometries of the steroid moieties are closely similar to those of the parent steroids, (b) the bonds around the nitrogen atom of the nitrogen mustard grouping lie approximately in a plane in each structure, (c) the plane through the carbon atoms of the steroid A-ring lies approximately perpendicular to the plane through the carbamate atoms in each structure, (d) the carbonyl C-O of the carbamate points to the alpha side of the steroid moiety in each structure, and (e) one chlorine atom of the nitrogen mustard grouping makes a close contact [3.13 A], in each structure, to the nitrogen atom. Hydrogen bonding to the carbamate appears to occur from the alpha side of the steroid; there is no hydrogen bonding to the nitrogen atom of the carbamate group. These structural data provide some steric explanations for the resistance of the carbamate to enzymatic hydrolysis. The long in vivo half-life of the intact estramustine molecule is a result of this stability. This is responsible for the absence of alkylating ability and the propensity of the drug to bind microtubule-associated proteins and express an antimitotic mechanism of action.

The glutathione transferases comprise a family of isoenzymes, one or more of which are involved in the conjugation of alkylating agents to glutathione (GSH). Increased GSH transferase activity has been shown to underlie acquired resistance to several alkylating agents. Ethacrynic acid inhibits the isoenzymes of GSH transferase with 50% inhibitory concentration values ranging from 0.3 to 6.0 microM and has been shown to restore sensitivity to alkylating agents in drug-resistant animal tumor models. We entered 27 previously treated patients with advanced cancer on a study of ethacrynic acid (25 to 75 mg/m2 p.o. every 6 h for 3 doses) and thiotepa (30 to 55 mg/m2 i.v. 1 h after the second dose of ethacrynic acid). The major toxicity of ethacrynic acid was diuresis, which was observed at every dose level; in addition, severe metabolic abnormalities occurred at 75 mg/m2. At 50 mg/m2, the diuretic effects were manageable. Myelosuppression was the most important effect of the combination. Two of seven courses of ethacrynic acid, 50 mg/m2, and thiotepa, 55 mg/m2, were associated with grade 3 or 4 neutropenia and/or thrombocytopenia. Nausea/vomiting greater than or equal to grade 2 was observed in 16% of courses. GSH transferase activity was assayed spectrophotometrically in the peripheral mononuclear cells of all patients. At each dose level, activity decreased following ethacrynic acid administration, with recovery by 6 h. Administration of ethacrynic acid, 50 mg/m2, resulted in a mean nadir of transferase activity of 37% of control. The pharmacokinetics of thiotepa and its principal metabolite TEPA were studied in 23 patients. The plasma disappearance of thiotepa fit a two-compartment open model with a terminal half-life of approximately 2 h. Plasma TEPA levels peaked at a mean of 2.16 h following thiotepa administration. The harmonic mean terminal half-life of TEPA was 10.4 h, and the TEPA area under the curve (AUC) did not increase with increasing thiotepa dose. The AUC of thiotepa was approximately twice, and the clearance about one-half, of the values obtained in a previous study of single agent thiotepa. The AUC of TEPA was lower than that previously observed. The data suggest that ethacrynic acid inhibits enzymes involved in the metabolic disposition of thiotepa, including its oxidative desulfuration to TEPA. The severity of the platelet toxicity was correlated with the AUC of thiotepa, but not with that of TEPA. This combination of thiotepa and ethacrynic acid will be tested further in Phase II trials.

Following EMS mutagenesis, three estramustine (EM) resistant DU 145 human prostatic carcinoma cell lines were clonally selected by exposure to incrementally increasing concentrations of the drug. Although only low levels of resistance (approximately 3-fold) were attainable, this resistance was stable in the absence of continuous drug exposure. These EM-resistant clones (EMR 4,9,12) did not exhibit cross resistance to vinblastine, taxol, or adriamycin, and had collateral sensitivity to cytochalasin B. None of the lines had elevated expression of P-glycoprotein mRNA or glutathione S-transferase activity, suggesting a phenotype distinct from the classic multi-drug resistance phenotype. This conclusion was supported further by the observation that two MDR cell lines (FLC mouse erythroleukaemic and SKOV3 human ovarian carcinoma cells) showed sensitivity to EM. Fluorescent activated cell sorting analysis of the effects of EM on cell cycle traverse revealed that at EM concentrations up to 20 microM an increasing percentage of wild type cells were blocked in G2/M; no such effect occurred in EMR lines. Differential interference contrast microscopy was employed to study EM's effect on mitosis. EMR lines were able to form functional, albeit smaller, spindles at EM concentrations that resulted in chromosomal disorganisation and inhibition of mitotic progression in wild type cells. EMR lines were able to progress through mitosis and cytokinesis at the same rate as untreated cells. Tritiated EM was used to evaluate potential drug uptake/efflux mutations in ERM clones. EMR 4 and 9 incorporate less EM than wild type cells; however, they have significantly decreased cellular volumes. The initial efflux rate constants for EMR clones were greater than for wild type cells. Within 5 min greater than 70% of the drug was lost from resistant cells compared to a 50% loss by the wild type. Although the specific mechanisms of resistance have yet to be defined, the lack of collateral resistance to other MDR/anti-microtubule agents could serve as the basis for the clinical use of EM in combination chemotherapy. Images Figure 3 Figure 5

The chemotherapeutic drug, estramustine, has been shown to cause the disassembly of microtubules via binding to microtubule-associated proteins. In this report, estramustine is shown to be a potent inhibitor of mitotic progression in the human prostatic carcinoma cell line, DU 145. Examination of individual living cells via video-enhanced differential interference contrast (DIC) optics shows that the drug delays the onset of anaphase, reduces anaphase spindle-pole elongation (anaphase B), and delays cytokinesis. In addition, immunofluorescent studies demonstrate that estramustine causes a rapid disorganization of the mitotic apparatus at significantly lower concentrations than those reported previously. Electron microscopic studies show that microtubule bundles are present in drug-treated mitotic cells in association with kinetochores and centrioles.

Estramustine offers a mechanistically novel therapy for the treatment of prostatic cancer. The drug is composed of oestradiol coupled to nor-nitrogen mustard. The cytotoxic properties of estramustine act independently of its constituent molecules. In vivo and in vitro studies indicate that the drug binds microtubule associated proteins and inhibits microtubule regulated processes. We have lately shown that micromolar levels of estramustine inhibit selected processes in mitosis, suggesting that certain classes of microtubules and/or MAPs are differentially sensitive to the drug. New clinical trials using estramustine and vinblastine on patients with advanced hormone refractory disease suggest that the combination of two antimicrotubule agents acting by means of different target molecules may be important clinically.

Chronic lymphocytic leukemia (CLL) is a disease state which frequently responds to alkylating agent chemotherapy but ultimately becomes refractory through acquired resistance mechanisms. In the present study, we have examined the expression of glutathione S-transferases (GST) in both CLL and normal control lymphocytes, as these enzymes have been implicated in mechanisms of natural and acquired resistance. Lymphocyte GST was purified from samples by high-pressure liquid affinity chromatography, and subunits were identified by two-dimensional gel electrophoresis and immunoblotting by using polyclonal antibodies specific for individual subunits. Analysis of CLL lymphocyte GST activity using the general substrate 1-chloro-2,4-dinitrobenzene showed a statistically significant 2-fold increase in cells from chlorambucil-resistant patients over those from untreated patients and normal individuals. Furthermore, chlorambucil therapy was seen to cause a 1.3- to 1.5-fold elevation of enzyme activity in three previously drug-naive patients. Analysis of GST isozyme subunits indicated that 95% of the CLL patients examined were positive for the pi isozyme, and this appeared quantitatively to be the major isozyme present. The alpha and mu isozymes were also expressed in 63 and 53% of the patients, respectively. Examination of control lymphocytes, as well as separated B- and T-cell subpopulations, yielded similar results. The present study indicates that a high degree of interindividual variation occurs and that the pattern of CLL lymphocyte GST expression differs from that of other tumor tissues. While there were no obvious correlations between the disease state or stage and isozymes expressed, the quantitative increase in GST activity in chlorambucil-resistant CLL patients may be of relevance to the overall resistant phenotype.

A series of glutathione S-transferase pi (GST-pi) transfectant cell lines have been constructed in activated c-H-ras-transformed NIH-3T3 cells (pT22-3) by using a pKOneo plasmid and an expression vector containing cDNA for GST-pi with a beta-actin gene promoter. From the wild type pT22-3 cells, two clones were selected and designated RGN1 and RGN2. The degree of overexpression of GST-pi was estimated by Northern and Southern blot analysis to be incrementally higher in RGN2 compared with RGN1. Translation of mRNA was estimated by Western blot analysis using isozyme-specific polyclonal antibodies and confirmed the relative GST-pi levels. Each cell line, including the wild type, expressed alpha and mu class isozymes to the same degree and had similar but negligible expression of the mdr 1 gene. Sensitivity to various anticancer drugs and radiation was estimated by a series of cytotoxicity assays. The data confirmed that GST-pi provided a degree of protection against the toxicity of ethacrynic acid and adriamycin, but sensitivity to alkylating agents such as chlorambucil, melphalan, and cis-platinum was not influenced by GST-pi. Similarly, the response to ionizing radiation was similar for each line. Since the levels of intracellular GSH were also not significantly different, the availability of co-substrate was not a factor in determining response. In creating the GST-pi transfectants, these data establish that while increased isozyme levels can play a role in determining sensitivity to some agents, the protective effect is selective.

Cell killing, DNA-interstrand crosslinks, and DNA-protein crosslinks were assayed in nitrogen mustard-resistant Walker 256 carcinoma (WR) cells and the parent cell line (WS) after treatment with 5-[3-(2-chloroethyl)-1-triazenyl]imidazo-4-carboxamide (MCTIC). The WR cells, which also express collateral sensitivity to chloroethylnitrosoureas, were approximately twice as sensitive to the cytotoxic effects of MCTIC as were WS cells. Following treatment with 100 microM MCTIC, there was a rapid accumulation of both DNA-interstrand and DNA-protein crosslinks in the WR cell line, which reached a maximum at 6 and 12 hr, respectively. There was considerably less crosslinking in the WS cells and both cell lines were proficient in repairing most of the crosslinks by 24 hr. Measurement of guanine-O6-alkyl transferase activity showed the enzyme to be present in WS but not in WR cells. These data indicate that the collateral sensitivity of nitrogen mustard-resistant WR cells to chloroethylating drugs is in part due to the loss of guanine-O6-alkyl transferase activity which is present in the parent line.

Tumor cell resistance to alkylating agents was studied by examining Walker 256 rat mammary carcinoma cells differentially sensitive to nitrogen mustards. A resistant subpopulation (WR) was selected by exposure to chlorambucil. WR cells showed approximately a 15-fold resistance to the cytotoxic effects of nitrogen mustards and elevated glutathione S-transferase (GST) activity when compared to the sensitive parent cell line (WS). To extend these findings, the GSTs from WR and WS were purified by affinity chromatography on S-hexylglutathione coupled to epoxy-activated agarose. Substrate specificity experiments using purified GSTs demonstrated different profiles of enzyme activity for WR and WS and suggested differential isoenzyme expression in these two cell lines. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis and Western blot analysis revealed that the major GST present in both WR and WS was a 26,000-Da subunit that was immunologically distinct from the rat liver GSTs. This GST subunit cross-reacted with antibodies against anionic human placental GST. In addition, three GST forms common to rat liver (29,500, 28,500 and 27,500 molecular weight) were also identified. Overexpression of the 29,500-Da protein was observed in WR cells. These data suggest that differential expression of GST subunits may contribute to the nitrogen mustard-resistant phenotype.

The employment of traditional biochemical and pharmacological concepts, together with the development of original molecular approaches, has created a more comprehensive understanding of the mechanism of action of nitrosoureas. For example, it is clear that all nuclear macromolecules are susceptible to alkylation and/or carbamoylation by nitrosoureas. The relative importance of these respective processes is equivocal, but there is evidence to suggest that the alkylating potential of these drugs is directly responsible for their cytotoxic, mutagenic, or carcinogenic properties. The structural organization of nuclear chromatin, either at the nucleosomal or function level, is a critical determinant when considering specific macromolecules as potential targets for nitrosourea interaction. Perhaps a more complete understanding of the nuclear architecture of specific cell types will lead to a more precise understanding of nitrosourea selectivity, especially with respect to the nonmyelotoxic properties of chlorozotocin. Such knowledge would also assist in the rational development of novel nitrosourea analogues.

A Walker 256 rat mammary carcinoma cell line (WR) resistant to bifunctional nitrogen mustards has been shown to have an approximate twofold increase in bulk glutathione-S-transferase activity compared to the parent cell line. Substrate specificity studies suggest that higher levels of Yb subunit contribute to the increased activity. By exposing WR cells to additional chlorambucil, either as a single concentration (50 micrograms/ml) or at 5 micrograms/ml for 10 days, transferase activity was further increased by up to three times the normal WR level. By using colony-forming assays, mitotic index depression, or trypan blue exclusion, the increased transferase activity could be correlated with an increase in resistance of these cells to either subsequent chlorambucil or a different bifunctional nitrogen mustard, phosphoramide mustard.

A Walker 256 rat carcinoma cell line (WR) has been shown to be resistant to a broad spectrum of bifunctional nitrogen mustards (NM) in cell culture. The parent cell line (WS) from which the WR cells were selected retains marked sensitivity to this class of drugs. Karyotype analysis showed that the parent WS had a chromosome content which was significantly higher than the WR. A number of chromosome marker bands were also distinguishable, indicative of distinct nuclear structural differences. A lack of collateral resistance to haloethylnitrosoureas was demonstrated for the WR cell line. In some cases the sensitivity of the WR to nitrosoureas was greater than that of WS. In addition, the WR could be sensitized to NM by a concomitant addition of a water-soluble carbamoylating agent, N,N'-bis(trans-4-hydroxycyclohexyl)-N'-nitrosourea, which possessed no intrinsic alkylating activity. Since NM and nitrosoureas differ pharmacologically, mainly by the latter's potential to carbamoylate, this reaction would appear to be critical to the cytotoxic properties of nitrosoureas against WR cells. Heretofore, carbamoylation has been considered of little importance to the antitumor properties of nitrosoureas. Moreover, 1-(4-amino-2-methylpyrimidine-5-yl)-methyl-3-(2-chloroethyl)-3-nitrosourea hydrochloride (ACNU) has been reported as possessing only minimal carbamoylating potential, as measured in vitro by a lysine assay. While remaining an accurate assay for the modification of the psi-amino groups of lysine, it is possible that the quantitative lysine assay may not predict the physiological carbamoylating potential of ACNU in the Walker cells, since marked cytotoxicity was achieved in both WR and WS by ACNU.

The stimulation of poly(adenosine diphosphate ribose) [poly(ADP-ribose)] polymerase activity at the nuclear level after damage of HeLa cells by 1-methyl-1-nitrosourea has been previously reported. We have observed a similar activation of the enzyme after treatment of cells with MNU at the nucleosomal level of chromatin (greater than 1N). This stimulation of enzyme activity did not occur through an inhibition of the glycohydrolase enzyme which cleaves poly(ADP-ribose), or elongation of poly(ADP-ribose) chains, or an increased biosynthesis of enzyme protein. The increased activity appears to be a consequence of the generation of more acceptor sites on nuclear proteins for initiation of poly(ADP-ribose) synthesis. The data indicate that MNU increased the accessibility of nucleosome core histones for modification by poly(ADP) ribosylation.

Using polyclonal antibodies raised against a rat liver nuclear envelope protein, lamin protein A, the nuclear matrix proteins of a Walker 256 rat mammary carcinoma wild-type (WS) and a selected cell line with acquired resistance to nitrogen mustards (WR) were found to possess antigenic determinants which were recognized by the antibodies. In one-dimensional immunoblotting analysis, the nuclear matrix protein fractions of both cell lines revealed a common band at Mr 75,000; however only the WS nuclear matrix protein fraction contained a broad band at approximately Mr 70,000. Two-dimensional gel blotting studies of these proteins showed that this Mr 70,000 WS protein had a pI of approximately 7.5. Immunoprecipitation analysis revealed that the altered mobility of this protein could be a function of phosphorylation. The nuclear matrix proteins from both WS and WR cells were shown to bind 3':5'-cyclic adenylic acid (cAMP), as judged by photoaffinity labeling and gel electrophoresis studies. The WS nuclear matrix proteins showed a quantitatively greater level of cAMP binding compared to WR, with predominant binding to proteins with molecular weights of 45,000, 55,000, and 70,000. In WR cells, there was no cAMP binding in the Mr 70,000 region. These data indicate that the Mr 70,000 nuclear matrix lamin proteins are antigenically similar in WS and WR but differ in that the WR protein is hypophosphorylated and does not bind cAMP.

Ethacrynic acid and piriprost (6,9-deepoxy-6,9-(phenylimino)-delta 6,8-prostaglandin I1) have been shown to potentiate the cytotoxic activity of chlorambucil in rat and human tumor cell lines. Walker 256 rat breast carcinoma cells (WS), with acquired resistance to nitrogen mustards (WR), and two human colon carcinoma cell lines, HT 29 and BE, were sensitized to chlorambucil when either ethacrynic acid or piriprost was administered at the same time as the alkylating agent. Both as single agents and in combination with chlorambucil, there was inhibition of glutathione S-transferase activity as measured with 1-chloro-2,4-dinitrobenzene as a substrate. A depletion in intracellular glutathione was also evident following ethacrynic acid alone or in combination with chlorambucil. Thus, diuretic plant phenols or prostaglandin analogues may have potential therapeutic utility in combination with alkylating agents.