An invasive cleavage assay for direct quantitation of specific RNAs.

RNA quantitation is becoming increasingly important in basic, pharmaceutical, and clinical research. For example, quantitation of viral RNAs can predict disease progression and therapeutic efficacy. Likewise, gene expression analysis of diseased versus normal, or untreated versus treated, tissue can identify relevant biological responses or assess the effects of pharmacological agents. As the focus of the Human Genome Project moves toward gene expression analysis, the field will require a flexible RNA analysis technology that can quantitatively monitor multiple forms of alternatively transcribed and/or processed RNAs (refs 3,4). We have applied the principles of invasive cleavage and engineered an improved 5'-nuclease to develop an isothermal, fluorescence resonance energy transfer (FRET)-based signal amplification method for detecting RNA in both total RNA and cell lysate samples. This detection format, termed the RNA invasive cleavage assay, obviates the need for target amplification or additional enzymatic signal enhancement. In this report, we describe the assay and present data demonstrating its capabilities for sensitive (<100 copies per reaction), specific (discrimination of 95% homologous sequences, 1 in > or =20,000), and quantitative (1.2-fold changes in RNA levels) detection of unamplified RNA in both single- and biplex-reaction formats.

Probable free radical effects on rat liver nuclei during early hepatocarcinogenesis with a choline-devoid low methionine diet.

Fischer-344 rats fed a choline-devoid diet show lipid peroxidation in the liver nuclei, beginning at 1 day, reaching a peak at 3 days, and subsequently declining by 35 days. Lipid peroxidation in the mitochondria was seen first at 3 days, increased to a maximum at 28 days, and decreased after 35 days to undetectable values at 49 days. Lipid peroxidation was found in both nuclear and mitochondrial fractions both before and after stripping of their outer membranes. No microsomal lipid peroxidation could be detected at any time up to 63 days. The animals fed the same diet supplemented with choline showed no lipid peroxidation in any liver fraction. Animals given CCl4 showed the expected lipid peroxidation in the microsomes but not in the nuclear fraction. The administration of the free radical trapping agent, N-tert-butyl-alpha-phenylnitrone, prevented completely or almost so, microsomal lipid peroxidation induced by CCl4 and nuclear lipid peroxidation in the animals fed the choline-devoid, low methionine diet. The genesis of free radicals in the livers of rats fed a choline-devoid diet is considered as a likely hypothesis for the observed lipid peroxidation. The lipid peroxidation in turn is considered to be closely related to the induction of liver cell death and to the production of alterations in DNA. The DNA alterations coupled with regenerative liver cell proliferation suggest an attractive hypothesis for the initiation of hepatocarcinogenesis in rats fed a choline-devoid diet.

Purification and characterization of P-52 (glutathione S-transferase-P or 7-7) from normal liver and putative preneoplastic liver nodules.

A previous study from our laboratory (L.C. Eriksson et al., Biochem. Biophys. Res. Commun., 117: 740-745, 1983) revealed that a cytosolic polypeptide of approximate Mr 21,000 (designated P-21) was markedly elevated in amount in hepatocyte nodules induced by six different regimens. The molecular weight of this polypeptide, subsequently revised to approximately 26,000, was redesignated P-26 and was identified (T.H. Rushmore et al., Biochem. Biophys. Res. Commun., 143: 98-103, 1987) as a subunit of a placental form of glutathione S-transferase (K. Sato et al., Gann 75: 199-202, 1984), also named glutathione S-transferase 7-7 (H. Jensson et al., FEBS Lett., 187: 115-120, 1985). We describe here a convenient method for purifying relatively large amounts of P-26 from hepatocyte nodules involving the sequential use of affinity chromatography on S-hexyl glutathione-Sepharose 4B, CM-Sephadex, and DEAE-Sephacel. Evidence is presented that P-26 exists as a dimer of approximate Mr 52,000 (P-52). Analyses by two-dimensional electrophoresis have indicated that the subunits of Mr 26,000 may consist of five separate charged isomers. Investigations using appropriate antisera and analysis by amino acid sequencing have provided additional confirmation that P-52 is probably identical to rat placental glutathione S-transferase. Antibodies to P-52 are proving to be useful as a marker of new cell populations that appear regularly during hepatocarcinogenesis.

Genetic association between sensitivity to warfarin and expression of CYP2C9*3.

Cytochrome P4502C9 (CYP2C9) is largely responsible for terminating the anticoagulant effect of racemic warfarin via hydroxylation of the pharmacologically more potent S-enantiomer to inactive metabolites. Mutations in the CYP2C9 gene result in the expression of three allelic variants, CYP2C9*1, CYP2C9*2 and CYP2C9*3. Both CYP2C9*2 and CYP2C9*3 exhibit altered catalytic properties in vitro relative to the wild-type enzyme. In the present study, a patient was genotyped who had proven unusually sensitive to warfarin therapy and could tolerate no more than 0.5 mg of the racemic drug/day. PCR-amplification of exons 3 and 7 of the CYP2C9 gene, followed by restriction digest or sequence analysis, showed that this individual was homozygous for CYP2C9*3. In addition, patient plasma warfarin enantiomer ratios and urinary 7-hydroxywarfarin enantiomer ratios were determined by chiral-phase high performance liquid chromotography in order to investigate whether either parameter might be of diagnostic value in place of a genotypic test. Control patients receiving 4-8 mg warfarin/day exhibited plasma S:R ratios of 0.50 +/- 0.25:1, whereas the patient on very low-dose warfarin exhibited an S:R ratio of 3.9:1. In contrast, the urinary 7-hydroxywarfarin S:R ratio of 4:1 showed the same stereoselectivity as that reported for control patients. Therefore, expression of CYP2C9*3 is associated with diminished clearance of S-warfarin and a dangerously exacerbated therapeutic response to normal doses of the racemic drug. Analysis of the plasma S:R warfarin ratio may serve as a useful alternative test to genotyping for this genetic defect.

In-vitro metabolism of celecoxib, a cyclooxygenase-2 inhibitor, by allelic variant forms of human liver microsomal cytochrome P450 2C9: correlation with CYP2C9 genotype and in-vivo pharmacokinetics.

In-vitro studies were conducted to assess the impact of CYP2C9 genotype on the metabolism (methyl hydroxylation) and pharmacokinetics of celecoxib, a novel cyclooxygenase-2 inhibitor and CYP2C9 substrate. When compared to cDNA-expressed wild-type CYP2C9 (CYP2C9*1), the Vmax/Km ratio for celecoxib methyl hydroxylation was reduced by 34% and 90% in the presence of recombinant CYP2C9*2 and CYP2C9*3, respectively. These data indicated that the amino acid substitution at position 359 (Ile to Leu) elicited a more pronounced effect on the metabolism of celecoxib than did a substitution at position 144 (Arg to Cys). The Vmax/Km ratio was also decreased in microsomes of livers genotyped CYP2C9*1/*2 (47% decrease, mean of two livers), or CYP2C9*1/*3 (59% decrease, one liver). In all cases, these changes were largely reflective of a decrease in Vmax, with a minimal change in Km. Based on simulations of the in-vitro data obtained with the recombinant CYP2C9 proteins, it was anticipated that the pharmacokinetics of celecoxib (as a much as a five-fold increase in plasma AUC) would be altered (versus CYP2C9*1/*1 subjects) in subjects genotyped heterozygous or homozygous for the CYP2C9*2 (Cys144) or CYP2C9*3 (Leu359) allele. In a subsequent clinical study, the AUC of celecoxib was increased (versus CYP2C9*1/*1 subjects) approximately 2.2-fold (range, 1.6-3-fold) in two CYP2C9*1/*3 subjects and one CYP2C9*3/*3 subject receiving a single oral dose (200 mg) of the drug. In contrast, there was no significant change in celecoxib AUC in two subjects genotyped CYP2C9*1/*2.

Enzyme kinetics of cytochrome P450-mediated reactions.

The most common drug-drug interactions may be understood in terms of alterations of metabolism, associated primarily with changes in the activity of cytochrome P450 (CYP) enzymes. Kinetic parameters such as Km, Vmax, Ki and Ka, which describe metabolism-based drug interactions, are usually determined by appropriate kinetic models and may be used to predict the pharmacokinetic consequences of exposure to one or multiple drugs. According to classic Michaelis-Menten (M-M) kinetics, one binding site models can be employed to simply interpret inhibition (pure competitive, non-competitive and uncompetitive) or activation of the enzyme. However, some cytochromes P450, in particular CYP3A4, exhibit unusual kinetic characteristics. In this instance, the changes in apparent kinetic constants in the presence of inhibitor or activator or second substrate do not obey the rules of M-M kinetics, and the resulting kinetics are not straightforward and hamper mechanistic interpretation of the interaction in question. These unusual kinetics include substrate activation (autoactivation), substrate inhibition, partial inhibition, activation, differential kinetics and others. To address this problem, several kinetic models can be proposed, based upon the assumption that multiple substrate binding sites exist at the active site of a particular P450, and the resulting kinetic constants are, therefore, solved to adequately describe the observed interaction between multiple drugs. The following is an overview of some cytochrome P450-mediated classic and atypical enzyme kinetics, and the associated kinetic models. Applications of these kinetic models can provide some new insights into the mechanism of P450-mediated drug-drug interactions.

Prevention by free radical scavenger AD5 of prooxidant effects of choline deficiency.

This study was designed to explore the possible preventive effects of a novel radicophile, N-p-methoxyphenylacetyl-dehydroalanine (AD5) and three other antioxidants, N,N'-diphenyl-p-phenylenediamine (DPPD), butylated hydroxyanisole (BHA) and a water-soluble analogue of vitamin E, trolox C, on the acute effects of the liver of feeding a choline-deficient (CD) diet. It has been suggested that some of the acute effects of a CD diet are related to free radicals, the generation or metabolism of which is disturbed in this acute dietary model. AD5 was found to be very effective in preventing nuclear lipid peroxidation, DNA damage and cell death induced by a CD diet but to have little effect on triglyceride accumulation ("fatty liver"). DPPD, BHA, and trolox C were ineffective. These results add strength to the hypothesis that oxygen free radicals might be an important component in the early events during carcinogenesis induced by feeding a CD diet.

Cloning and expression of three isoforms of the human EP3 prostanoid receptor.

Functional cDNA clones coding for three isoforms of the human prostaglandin E receptor EP3 subtype have been isolated from kidney and uterus cDNA libraries. The three isoforms, designated hEP3-I, hEP3-II and hEP3-III, have open reading frames corresponding to 390, 388 and 365 amino acids, respectively. They differ only in the length and amino acid composition of their carboxy-terminal regions, beginning at position 360. The human EP3 receptor has seven predicted transmembrane spanning domains and therefore belongs to the G-protein-coupled receptor family. The rank order of potency for prostaglandins and related analogs in competition for [3H]PGE2 specific binding to membranes prepared from transfected COS cells was comparable for all three isoforms, and as predicted for the EP3 receptor, with PGE2 = PGE1 >> PGF2 alpha = iloprost > PGD2 >> U46619. In addition, the EP3-selective agonist MB28767 was a potent competing ligand with an IC50 value of 0.3 nM, whereas the EP1-selective antagonist AH6909 gave IC50 values of 2-7 microM and the EP2-selective agonist butaprost was inactive. In summary, we have cloned three isoforms of the human EP3 receptor having comparable ligand binding properties.

Initiation of carcinogenesis by a dietary deficiency of choline in the absence of added carcinogens.

The feeding for 10 or 11 weeks of young male Fischer-344 rats, a diet devoid of choline and low in methionine, leads to the appearance of gamma-glutamyltransferase-positive foci of altered hepatocytes in the liver and to the induction of initiated resistant hepatocytes. The latter are known to contain the primary precursor cells for the ultimate development of hepatocellular carcinoma. This initiation of carcinogenesis with the choline-devoid diet is prevented by added choline. These observations indicate that a dietary deficiency may, by itself, without known contaminating or added carcinogens, initiate the carcinogenic process.

Rapid lipid peroxidation in the nuclear fraction of rat liver induced by a diet deficient in choline and methionine.

A diet deficient in choline and methionine, known to produce hepatocellular carcinoma in the absence of any added chemical carcinogen, induced lipid peroxidation in the nuclear fraction of the liver when fed to male Fischer 344 rats. This lipid peroxidation was detected within 1 day of feeding the diet by the appearance of diene conjugates and increased progressively up to 3 days. It was prevented completely by the addition of choline chloride to the diet. The close proximity of DNA may make it a possible target for attack by free radicals.

Glutathione and enzymes related to free radical metabolism in liver of rats fed a choline-devoid low-methionine diet.

Fischer F-344 male rats, fed a choline-devoid diet that leads to a highly reproducible sequence of biochemical and biological changes with an ultimate development of hepatocellular carcinoma, show elevated levels of glutathione in the liver at 3, 6 and 8 days. Several enzymes related to the metabolism of free radicals, including superoxide dismutase, catalase, glutathione peroxidase, glutathione S-transferase and DT-diaphorase show neither increased nor decreased activity as measured between 12 h and 8 days on the diet. Thus, of several known cellular components related to the possible scavenger of free radicals in the liver, only glutathione responded to the feeding of the CD diet. It is tentatively concluded that a decrease in the levels of possible scavengers for free radicals is not a major basis for the nuclear and mitochondrial lipid peroxidation seen early in rats fed a choline-devoid diet.

Pathogenesis of skin and liver neoplasms in white suckers from industrially polluted areas in Lake Ontario.

Increased prevalences of epidermal and hepatobiliary neoplasms in white suckers (Catostomu commersoni) and brown bullheads (Ictalurus nebulosus) in the Western region of Lake Ontario have been associated with industrial pollution, but the identity and causative role of environmental carcinogens have not yet been established. Most epidermal tumors of lip and body skin are benign focal proliferations that occur in fish from the polluted Hamilton region, and also in fish from less polluted sites in the Great Lakes. These skin tumors in white suckers do not have consistent alterations in cellular glutathione S-transferases (GST), suggesting that growth of skin tumors is not promoted by chemicals normally detoxified by GST. However, elevated levels of glutathione peroxidase (GPO) and glutathione reductase (GR) in skin papillomas are indicative of promotional peroxidative tissue injury, either caused directly by xenobiotics or indirectly by chemical-induced inflammation. Liver tumors in white suckers from Lake Ontario include preneoplastic, benign, and malignant populations of hepatocellular and biliary cells, all of which are more prevalent in fish from polluted sites. These liver tumors are consistently associated with chronic cholangiohepatitis and segmental cholangiofibrosis, but these conditions also occur in white suckers in non-industrial locations. Thus, the natural occurrence of biliary disease, not attributable to industrial pollution, may have some influence on the development of liver tumors. Some preneoplastic lesions and the majority of neoplastic hepatocellular and biliary lesions in white suckers have low levels of total GST, indicating that these liver neoplasms are not promoted by xenobiotics normally detoxified by hepatic GSTs.(ABSTRACT TRUNCATED AT 250 WORDS)

Compendium of gene expression profiles comprising a baseline model of the human liver drug metabolism transcriptome.

Oligonucleotide microarrays were used to study the variability of pharmacokinetics and drug metabolism (PKDM)-related gene expression in 75 normal human livers. The objective was to define and use absorption, distribution, metabolism and excretion (ADME) gene expression variability to discern co-regulated genes and potential surrogate biomarkers of inducible gene expression. RNA was prepared from donor tissue and hybridized on Agilent microarrays against an RNA mass balanced pool from all donors. Clustering of PKDM gene sets revealed donors with distinct patterns of gene expression that grouped genes known to be regulated by the nuclear receptor, pregnane X-receptor (PXR). Fold range metrics and frequency distributions from the heterogeneous human population were used to define the variability of individual PKDM genes in the 75 human livers and were placed in context by comparing expression data with basal ADME gene expression variability in an inbred and diet/environment controlled population of 27 Rhesus livers. The most variable genes in the hepatic transcriptome were mainly related to drug metabolism, intermediary metabolism, inflammation and cell cycle control. Unique patterns of expression across 75 individuals of inducible ADME gene expression allowed their expression to be correlated with the expression of many other genes. Correlated genes for AhR, CAR and PXR responsive genes (CYP1A2, CYP2B6 and CYP3A4) were identified that may be co-regulated and, therefore, provide clues to the identity of surrogate gene or protein markers for CYP induction. In conclusion, microarrays were used to define the variable expression of hepatic ADME genes in a diverse human population, the expression variability of ADME genes was compared with the expression variability in an inbred population of Rhesus monkeys, and genes were defined that may be co-regulated with important inducible CYP genes.

Transcriptional regulation of the rat glutathione S-transferase Ya subunit gene. Characterization of a xenobiotic-responsive element controlling inducible expression by phenolic antioxidants.

We have identified previously a xenobiotic-responsive element, which we termed the beta-naphthoflavone-responsive element, between nucleotide -722 and -682 in the 5'-flanking region of the rat glutathione S-transferase Ya subunit gene (Rushmore, T.H., King, R.G., Paulson, K.E., and Pickett, C.B. (1990) Proc. Natl. Acad. Sci. U.S.A. 87, 3826-3830). The beta-naphthoflavone-responsive element is responsible for part of the transcriptional activation of the Ya subunit gene by planar aromatic compounds but has a sequence distinct from the xenobiotic-responsive element found in multiple copies in the cytochrome P-450 IA1 gene and as a single copy in the Ya subunit gene. In the present study, we demonstrate that the beta-naphthoflavone-responsive element is required for the transcriptional activation of the Ya subunit gene by phenolic antioxidants such as t-butylhydroquinone through a mechanism that does not require functional Ah receptors. Furthermore, we present evidence that planar aromatic compounds must be metabolized before they transcriptionally activate the Ya subunit gene through the beta-naphthoflavone-responsive element. The transcriptional activation of the Ya subunit gene by planar aromatic compounds requires a functional Ah receptor. These data provide evidence that transcriptional activation of the glutathione S-transferase Ya subunit gene can be mediated by a novel xenobiotic-responsive element which is directly responsive to phenolic antioxidants such as t-butylhydroquinone. Hence we have named this new xenobiotic-responsive element the antioxidant-responsive element or ARE.

Influences of glutathione status on different cytocidal responses of monolayer rat hepatocytes exposed to aflatoxin B1 or acetaminophen.

In short-term primary monolayer cultures of rat hepatocytes, aflatoxin B1 (AFB1) causes a characteristic prelethal cytomorphological response in which peripheral attached cytoplasm contracts segmentally to form finger-like blebs. This response precedes lethal injury as detected by release of lactate dehydrogenase (LDH) into culture medium. We compared the influences of various modifiers of cellular glutathione (GSH) status on cytocidal responses of Fischer 344 rats hepatocytes exposed to AFB1 or acetaminophen (AAP), a hepatotoxin which does not produce segmental cytoplasmic contraction. N-Acetylcysteine (4 mM) reduced the degree of LDH release by AAP (4 to 16 mM) but was not protective against cell killing by AFB1, although it slightly reduced the percentage of hepatocytes with segmental cytoplasmic contraction at 6 hr. BCNU (1,3-bis(2-chloroethyl)-1-nitrosourea) at 40 microM markedly inhibited glutathione reductase and also strongly potentiated cell killing by AAP but did not significantly influence segmental cytoplasmic contraction or LDH release in response to AFB1. Diethylmaleate (40 to 160 microM), a depletor of hepatocellular GSH, and buthionine-D,L-sulfoximine (4 mM), an inhibitor of GSH synthesis, each did not alter hepatocyte killing by AFB1 but were strong potentiators of toxicity of AAP. AAP inhibited glutathione reductase but AFB1 did not. Total GSH concentrations at 6 and 18 hr were reduced by AAP and to a lesser extent by AFB1 in comparison with control cultures. These findings demonstrate that, in contrast to AAP toxicity, the characteristic mode of hepatocyte killing by AFB1 in monolayer cultures is substantially independent of induced alterations in GSH. These results indicate that GSH-dependent detoxification mechanisms do not play a major role in removing necrogenic metabolites of AFB1 in Fischer 344 rat hepatocytes. They further suggest that prelethal responses of AFB1-injured hepatocytes are not affected by GSH-dependent cytoprotective mechanisms.

Inhibition of RNA synthesis by benzene metabolites and their covalent binding to DNA in rabbit bone marrow mitochondria in vitro.

Mitoplasts from rabbit bone marrow cells in vitro incubated with benzene metabolites showed a concentration-dependent inhibition of mitochondrial RNA synthesis. The 50% molar inhibitory concentration for each metabolite was determined to be p-benzoquinone, 2 X 10(-6); phenol, 2.5 X 10(-5); hydroquinone, 5 X 10(-5); catechol, 2 X 10(-3); benzene, 1.6 X 10(-2). Activated metabolites of radiolabeled benzene covalently bound to mitochondrial DNA in vitro. Labeled deoxynucleosides, enzymatically produced from DNA isolated from mitoplasts prelabeled in DNA with [3H]dGTP and exposed to [14C]benzene in vitro, were chromatographed on a Sephadex LH-20 column to obtain deoxynucleoside-adducts. The elution profiles indicated that mtDNA contained seven deoxyguanosine-adducts. A similar experiment in which the mtDNA was prelabeled with [3H]dATP indicated that two deoxyadenosine-adducts of mtDNA were formed from benzene in vitro. Tentative identification of several deoxyguanosine-adducts indicates that p-benzoquinone, hydroquinone, phenol, and 1,2,4-benzenetriol produced from benzene form adducts with guanine.

The antioxidant responsive element. Activation by oxidative stress and identification of the DNA consensus sequence required for functional activity.

We have characterized further the antioxidant responsive element (ARE) identified in the 5'-flanking region of the rat glutathione S-transferase Ya subunit gene and the NAD(P)H:quinone reductase gene by mutational and deletion analyses. Our data suggest that the sequence, 5'-puGTGACNNNGC-3' 3'-pyCACTGNNNCG-5' where N is any nucleotide, represents the core sequence of the ARE required for transcriptional activation by phenolic antioxidants and metabolizable planar aromatic compounds (e.g. beta-naphthoflavone and 3-methylcholanthrene). We also have found that the ARE is responsive to hydrogen peroxide and phenolic antioxidants that undergo redox cycling. These latter data suggest that the ARE is responsive to reactive oxygen species and thus may represent part of a signal transduction pathway that allow eukaryotic cells to sense and respond to oxidative stress.

Transcriptional regulation of a rat liver glutathione S-transferase Ya subunit gene. Analysis of the antioxidant response element and its activation by the phorbol ester 12-O-tetradecanoylphorbol-13-acetate.

Using transfection and gel retardation assays, we have characterized further the antioxidant response element (ARE) found in the 5'-flanking region of the rat glutathione S-transferase Ya subunit gene. The ARE core sequence (5'-GTGACAAAGC-3') is sufficient for transcriptional activation of the Ya subunit gene by metabolizable planar aromatic compounds, phenolic antioxidants, and hydrogen peroxide. When the ARE sequence is ligated to a chloramphenicol acetyltransferase reporter gene and transfected into HepG2 cells, chloramphenicol acetyltransferase activity is modestly inducible by the phorbol ester 12-O-tetradecanoylphorbol-13-acetate (TPA). Since the ARE is responsive to TPA and shows some sequence similarity to an AP-1-binding site (Jun/Fos recognition motif), we have explored whether members of the Jun/Fos family of transcription factors might bind to the ARE. Using in vitro synthesized Jun and Fos, binding to the ARE could not be detected, whereas Jun/Fos binding to a classical AP-1-binding site, a TPA response element (TRE) from the human collagenase gene, could be demonstrated by gel retardation assays. If the 2 A nucleotides underlined in the ARE core sequence (5'-GTGACAAAGC-3') are changed to TC, the ARE sequence (ARE-TRE) becomes a high-affinity AP-1-binding site and retains xenobiotic inducibility. Removal of the -GC- dinucleotide at the 3'-end of the ARE or the ARE-TRE eliminates xenobiotic inducibility. However, the ARE-TRE construct without the -GC- dinucleotide is still a high-affinity AP-1 site and responsive to TPA. Taken together, our data suggest that the ARE is not a high-affinity binding site for the Jun/Fos heterodimer. Functionally, however, an AP-1-binding site can resemble an ARE in its response to various xenobiotics if a 3'-GC- dinucleotide is present.

Inhibition of hepatocarcinogenic responses to 1,2-dimethylhydrazine by diallyl sulfide, a component of garlic oil.

The mechanisms by which diallyl sulfide (DAS), a component of garlic oil, inhibits hepatocarcinogenicity of 1,2-dimethylhydrazine (DMH) were examined in male Fischer 344 rats. Rats were subjected to partial hepatectomy to stimulate hepatocellular proliferation required for initiation by DMH (50-200 mg/kg i.p.) given 12 h later. Initiation was assessed by the numbers of foci and nodules of hepatocytes that were positive for gamma-glutamyl transpeptidase (gamma-GT) or glutathione-S-transferase-P (GST-P) after 6 weeks promotion by orotic acid (1% in semi-purified diet). DAS at doses above 50 mg/kg (by gavage) administered 1 h before DMH (50 mg/kg) partially reduced the numbers of gamma-GT and GST-P-positive foci. By comparison, all doses of DAS (25-100 mg/kg) completely prevented liver necrosis by DMH (200 mg/kg). DAS substantially reduced macromolecular binding of [14C]DMH in cultured liver cells, but had no effect on their levels of glutathione-S-transferase, glutathione reductase or glutathione peroxidase at 18 h. These findings suggest that low dosages of DAS which reduce DMH binding appear more likely to inhibit hepatocarcinogenicity by reducing the promoting influences of post-necrotic regeneration than by preventing initiation.