NADPH- and linoleic acid hydroperoxide-induced lipid peroxidation and destruction of cytochrome P-450 in hepatic microsomes

NADPH- and iron-dependent lipid peroxidation inhibit aromatase activity in human placental microsomes

During pregnancy placenta is the most significant source of lipid hydroperoxides and other reactive oxygen species (ROS). The increased production of lipid peroxides and other ROS is often linked to pre-eclampsia. It is already proved that placental endoplasmic reticulum may be an important place of lipid peroxides and superoxide radical production. In the present study we revealed that NADPH- and iron-dependent lipid peroxidation in human placental microsomes (HPM) inhibit placental aromatase--a key enzyme of estrogen biosynthesis in human placenta. We showed that significant inhibition of this enzyme is caused by small lipid peroxidation (TBARS (thiobarbituric acid-reactive substances)<4nmol/mg microsomal protein (m.p.)). More intensive lipid peroxidation (TBARS>9nmol/mg microsomal protein) diminishes aromatase activity to value being less than 5% of initial value. NADPH- and iron-dependent lipid peroxidation also causes disappearance of cytochrome P450 parallel to observed aromatase activity inhibition. EDTA, alpha-tocopherol, MgCl(2) and superoxide dismutase (SOD) prevent aromatase activity inhibition and cytochrome P450(AROM) degradation. Mannitol and catalase have not effect on TBARS synthesis, aromatase activity and cytochrome P450 degradation. In view of the above we postulate that the inhibition of aromatase activity observed is mainly a consequence of cytochrome P450(AROM) degradation induced by lipid radicals. The role of hydroxyl radical in cytochrome P450 degradation is negligible in our experimental conditions. The results presented here also suggest that the inhibition of aromatase activity can also take place in placenta at in vivo conditions.

Pulmonary Cyp1A1 and CYP1A2 levels and activities in adult male and female offspring of rats exposed during gestation and lactation to 2,3,7, 8-tetrachlorodibenzo-p-dioxin

The levels and activities of pulmonary microsomal CYP1A1 and CYP1A2 in 40-day-old male and female, and 120-day-old male offspring of pregnant rats treated with five weekly 0.1 microg/kg doses of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) during gestation and lactation were compared with those in age-matched offspring of untreated dams. The CYP1A1-preferential activity, ethoxyresorufin O-deethylase (EROD), was comparably induced 5.3- and 6.4-fold in 40-day-old male and female offspring, respectively, but was not induced in 120-day-old male offspring, of TCDD-treated dams. Similarly, CYP1A1 protein was induced in 40-day-old female or male offspring of untreated dams but was undetectable in 120-day-old offspring of untreated or treated dams. CYP1A2 activity, as measured by the bioactivation of 2-amino-3,4-dimethylimidazo[4,5-f]quinoline (MeIQ) to mutagens in the Ames assay, was elevated 11.1- and 5.5-fold in 40-day-old female and male offspring, respectively, of TCDD-treated dams, but was unaffected by TCDD exposure in 120-day-old offspring. CYP1A2 protein was undetectable in 40-day-old male or female offspring of untreated dams or in 120-day-old male offspring of treated or untreated dams; it was detected in 40-day-old offspring of treated dams, at a level that was higher in females than in males. The results show that gestational and lactational exposure to TCDD causes long-lasting and gender-preferential induction of CYP1A1 as well as CYPIA2 in the lungs of rat offspring.

Effect of gestational and lactational 2,3,7, 8-tetrachlorodibenzo-p-dioxin exposure on the level and catalytic activities of hepatic microsomal CYP1A in prepubertal and adult rats

We determined the inducibility, as well as the persistence of the induction, of hepatic microsomal CYP1A1 and CYP1A2 (by western blot analysis), and their catalytic activities (as measured by resorufin ether O-dealkylation) in prepubertal (25-day-old) and adult (120-day-old) offspring of timed-pregnant Sprague-Dawley rats treated with 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD). TCDD treatment was subcutaneous, at a low dose of 0.1 microg/kg, on gestational days 7, 14, and 20, and on lactational days 7 and 14. CYP1A1 protein was induced significantly (23-fold) in prepubertal but not in adult offspring of TCDD-exposed dams, whereas ethoxyresorufin O-deethylase (EROD) activity, which is CYP1A1-preferential, was induced less extensively (5-fold) and slightly (1.7-fold) in the prepubertal and adult offspring, respectively. Benzyloxyresorufin O-debenzylase (BROD) activity, which is CYP2B-preferential but has been reported to be catalyzed by CYP1A1, was also induced 5- and 6-fold in prepubertal and adult offspring, respectively, of TCDD-exposed dams. However, the induced BROD activity was neither inhibited by antibody against CYP1A1 nor accompanied by an elevated level of microsomal CYP2B. CYP1A2 was induced slightly only in prepubertal offspring of TCDD-treated dams. There was suggestive evidence of enhanced lipid peroxidation in hepatic microsomes from prepubertal but not adult offspring of TCDD-treated dams. These data showed that in utero plus lactational TCDD exposure effected transient induction of hepatic microsomal CYP1A1 but sustained induction of BROD activity, which may be catalyzed by enzymes other than CYP1A or CYP2B.

Detection of free radicals produced from the reaction of cytochrome P-450 with linoleic acid hydroperoxide

The ESR spin-trapping technique was employed to investigate the reaction of rabbit cytochrome P-450 1A2 (P450) with linoleic acid hydroperoxide. This system was compared with chemical systems where FeSO4 or FeCl3 was used in place of P450. The spin trap 5, 5'-dimethyl-1-pyrroline N-oxide (DMPO) was employed to detect and identify radical species. The DMPO adducts of hydroxyl, O2-., peroxyl, methyl and acyl radicals were detected in the P450 system. The reaction did not require NADPH-cytochrome P-450 reductase or NADPH. The same DMPO-radical adducts were detected in the FeSO4 system. Only DMPO-.OH radical adduct and carbon-centred radical adducts were detected in the FeCl3 system. Peroxyl radical production was completely O2-dependent. We propose that polyunsaturated fatty acids are initially reduced to form alkoxyl radicals, which then undergo intramolecular rearrangement to form epoxyalkyl radicals. Each epoxyalkyl radical reacts with O2, forming a peroxyl radical. Subsequent unimolecular decomposition of this peroxyl radical eliminates O2-. radical.

Differential cumene hydroperoxide sensitivity of cytochrome P-450 enzymes IA1 and IIB1 determined by their way of membrane incorporation

The cytochrome P-450-dependent O-dealkylation of alkoxyresorufins was used to study the effect of cumene hydroperoxide on cytochrome P-450 IIB1 and IA1 in microsomal and reconstituted systems. In liver microsomal systems from respectively phenobarbital and 3-methylcholanthrene pretreated male Wistar rats, cytochrome P-450 IIB1-dependent pentoxyresorufin-O-dealkylation appeared to be more sensitive to cumene hydroperoxide treatment than cytochrome P-450 IA1-dependent ethoxyresorufin-O-dealkylation. This phenomenon was also observed when the cumene hydroperoxide sensitivity of P-450 IIB1 and IA1 was studied in an isosafrole pretreated rat liver microsomal system. The decrease in alkoxy-O-dealkylating activities appeared to proceed by destruction of the cytochrome P-450 component of the enzyme system. Purification and reconstitution of the enzyme system components in a system in which the isolated proteins were not incorporated into a membrane resulted in the disappearance of the difference in sensitivity between the two P-450 enzymes. However, in a reconstituted system with membrane incorporated proteins, again cytochrome P-450 IIB1 expressed a higher sensitivity towards cumene hydroperoxide than cytochrome P-450 IA1. From this it was concluded that the differential cumene hydroperoxide sensitivity of cytochrome P-450 IIB1 and IA1 is not caused by an intrinsic difference in their sensitivity but by a differential effect of membrane incorporation on their cumene hydroperoxide sensitivity.

Fluorospectroscopic analysis of the fluorescent substances in peroxidized microsomes of rat liver

The fluorescent substances formed in rat liver microsomes in the course of lipid peroxidation were investigated by fluorescence techniques. The fluorescence emitted from peroxidizing microsomes continuously increased as lipid peroxidation progressed, while the steady-state fluorescence anisotropy increased and then reached a plateau. A similar increase was observed in the steady-state fluorescence anisotropy of 1,6-diphenyl-1,3,5-hexatriene in peroxidizing microsomes. The fluorescence from peroxidized microsomes consisted of at least three species having short, middle or long fluorescence lifetimes. The lifetimes and relative amplitudes of fluorescence were unaffected by the extent of lipid peroxidation. Both fluorescence of the chromolipids extracted and the proteins isolated from peroxidized microsomes had the same characteristics in fluorescence lifetimes as the fluorescence from whole peroxidized microsomes. Thus, these lipids and proteins appear to be the major biological substances responsible for the fluorescence emanating from whole peroxidized microsomes. Furthermore, fluorescent substances formed in microsomes seem to increase in quantity rather than change in quality as lipid peroxidation progresses.

Target organ-specific inactivation of drug metabolizing enzymes in kidney of hamsters treated with estradiol

Chronic treatment of hamsters with estradiol for several months has previously been shown to decrease the specific content of cytochrome P450 in the kidney, a target of hormonal carcinogenesis, but not in liver. The reason for this decrease in metabolic enzyme activity is unknown and has been examined in this investigation. We now report that the decrease in specific content of renal cytochrome P450 by 73% in response to estradiol was not affected by co-treatment with tamoxifen for 1 month. The subcutaneous infusion of 250 micrograms/day estradiol for 7 days lowered renal cytochrome P450 by 71% from control values and was therefore used for further mechanistic studies. This treatment decreased renal activities of estradiol 2- or 4-hydroxylase by 77 to 80%, of 7-ethoxycoumarin-O-deethylase by 66% of control values, respectively, and completely eliminated aryl hydrocarbon hydroxylase activities, whereas liver enzymes remained unaffected. After 7 days of infusion of estradiol, fluorescent products of lipid peroxidation were more than doubled in hamster kidney but remained unchanged in liver. The possibility of enzyme destruction by binding of estradiol 2,3-quinone to metabolizing enzymes was investigated in vitro. In the presence of 2-hydroxyestradiol, cumene hydroperoxide, and microsomes, conditions known to favor the oxidation of the steroid to quinone, the binding of catechol estrogen metabolite to microsomal protein increased 60 fold over control values in the absence of cofactor. Purified rat liver cytochrome P450c also oxidized 2-hydroxyestradiol to 2,3-estradiol quinone. The rate of oxidation was linear for the first 2-3 min, but thereafter decreased with time.(ABSTRACT TRUNCATED AT 250 WORDS)

Mechanism of paraquat-stimulated lipid peroxidation in mouse brain and pulmonary microsomes

Paraquat-stimulated NADPH-dependent lipid peroxidation in mouse brain and pulmonary microsomes was inhibited by superoxide dismutase and singlet oxygen quenchers, but not by catalase or hydroxyl radical scavengers. MnCl2, which might form a salt with unsaturated lipid, inhibited the lipid peroxidation in brain microsomes, but not that in pulmonary microsomes. These findings suggest that activated oxygen species, especially superoxide and singlet oxygen, may play a major role in the stimulation of microsomal lipid peroxidation by paraquat in both brain and lung, and that the nature of the lipids exposed to peroxidative attack may be different in microsomes of the two organs.

Oxidative modification by low levels of HOOH can transform myoglobin to an oxidase

It is generally thought that the oxidative modification of hemoproteins leads to their inactivation. In the current study, however, a transiently activated form of myoglobin was shown to be formed when the prosthetic heme group became covalently bound to the polypeptide during the reaction of myoglobin with low levels of HOOH. In the presence of an enzymatic metmyoglobin reducing system containing diaphorase and methylene blue with excess NADH, this HOOH-altered myoglobin catalyzed NADH oxidation and oxygen consumption; the overall stoichiometry indicated a two-electron reduction of oxygen to HOOH. This reaction was not catalyzed by iron released from heme, as desferrioxamine had no effect on the activity. Stoichiometric amounts of HOOH were sufficient to produce the activated oxidase state of myoglobin, whereas larger amounts of HOOH lead to heme destruction, iron release, and inactivation of the oxidase activity. The alteration of myoglobin to an enzyme that can form toxic oxygen metabolites may have pathological importance, especially in myocardial injury caused by ischemia and reperfusion, where myoglobin is present in large amounts and HOOH is formed. Furthermore, the oxidase form may be involved in the mechanism of destruction of the heme seen with oxidative treatment of myoglobin.

Iron- and ascorbic acid-induced lipid peroxidation in renal microsomes isolated from rats treated with platinum compounds

Renal microsomes isolated on day 3 from cisplatin (CDDP, single i.p. injection, 4 or 6 mg/kg)-treated rats were monitored for their susceptibility to lipid peroxidation as compared with microsomes from rats treated with carboplatin (CBDCA, 30 mg/kg), transplatin (TDDP, 6 mg/kg) or CDDP hydrolysis products (4 or 6 mg/kg) or from control animals. Cephaloridine (1 g/kg daily for 4 days, i.p. injection) was used as a positive control. The effect of CDDP on renal microsomal glucose-6-phosphatase activity was investigated in vivo and in vitro. Following treatment with CDDP and CDDP hydrolysis products vs CBDCA and TDDP treatment, microsomes revealed an enhanced susceptibility to lipid peroxidation in a Fe2+ and/or ascorbic acid stimulation system. Increased lipid peroxidation, expressed as an increase in malondialdehyde (MDA) generation, paralleled the alterations in body and kidney weight and the elevations of plasma creatinine and blood urea nitrogen concentrations. Injection of the antioxidant N,N'-diphenyl-p-phenylenediamine (DPPD, 0.5 g/kg, i.p.) at 24 h prior to CDDP treatment abolished the increased vulnerability of renal microsomes to lipid peroxidation. In vivo, only CDDP hydrolysis products exhibited a significant inhibitory effect on renal glucose-6-phosphatase activity. In vitro, rat renal and hepatic microsomal glucose-6-phosphatase activity was decreased by CDDP both time- and concentration-dependently. Nephrotoxicity induced by CDDP and CDDP hydrolysis products might be attributable to iron-dependent lipid peroxidation and microsomes might represent target organelles on a subcellular level.

Inhibition of microsomal lipid peroxidation and cytochrome P-450-catalyzed reactions by nitrofuran compounds

(5-Nitro-2-furfurylidene)amino compounds bearing triazol-4-yl, benzimidazol-1-yl, pyrazol-1-yl, triazin-4-yl or related groups (a) stimulated superoxide anion radical generated by rat liver microsomes in the presence of NADPH and oxygen; (b) inhibited the NADPH-dependent, iron-catalyzed microsomal lipid peroxidation; (c) prevented the NADPH-dependent destruction of cytochrome P-450; (d) inhibited the NADPH-dependent microsomal aniline 4-hydroxylase activity; (e) failed to inhibit either the cumenyl hydroperoxide-dependent lipid peroxidation or the aniline-4-hydroxylase activity, except for the benzimidazol-1-yl and the substituted triazol-4-yl derivatives, which produced minor inhibitions. Reducing equivalents enhanced the benzimidazol-1-yl derivative inhibition of the cumenyl hydroperoxide-induced lipid peroxidation. The ESR spectrum of the benzimidazol-1-yl derivative, reduced anaerobically by NADPH-supplemented microsomes, showed characteristic spin couplings. Compounds bearing unsaturated nitrogen heterocycles were always more active than those bearing other groups, such as nifurtimox or nitrofurazone. The energy level of the lowest unoccupied molecular orbital was in fair agreement with the capability of nitrofurans for redox-cycling and related actions. It is concluded that nitrofuran inhibition of microsomal lipid peroxidation and cytochrome P-450-catalyzed reactions was mostly due to diversion of reducing equivalents from NADPH to dioxygen. Trapping of free radicals involved in propagating lipid peroxidation might contribute to the overall effect of the benzimidazol-1-yl and substituted triazol-4-yl derivatives.

Effects of mansonones on lipid peroxidation, P450 monooxygenase activity, and superoxide anion generation by rat liver microsomes

Several structurally related ortho-naphthoquinones isolated from Mansonia altissima Chev (mansonones C, E and F) (a) inhibited NADPH-dependent, iron-catalyzed microsomal lipid peroxidation; (b) prevented NADPH-dependent cytochrome P450 destruction; (c) inhibited NADPH-supported aniline 4-hydroxylase activity; (d) inhibited Fe(III)ADP reduction by NADPH-supplemented microsomes; (e) stimulated superoxide anion generation by NADPH-supplemented microsomes; and (f) stimulated ascorbate oxidation. ESR investigation of ascorbate-reduced mansonone F demonstrated semiquinone formation. Mansonone C had a greater effect than mansonones E and F on NADPH-dependent lipid peroxidation, O2- production and ascorbate oxidation, whereas mansonone E was more effective than mansonones C and F on aniline 4-hydroxylase activity. Mansonones E and F did not inhibit hydroperoxide-dependent lipid peroxidation, cytochrome P450 destruction or microsomal aniline 4-hydroxylase activity. Mansonone C inhibited to a limited degree tert-butyl hydroperoxide-dependent lipid peroxidation, this inhibition being increased by NADPH. Mansonone A, a tetrahydro orthonapthoquinone derivative, was in all respects relatively less effective than mansonones C, E and F. It is postulated that mansonones C, E and F inhibited microsomal lipid peroxidation and cytochrome P450 catalyzed reactions by diverting reducing equivalents from NADPH to dioxygen, but mansonone C (including its reduced form) may also exert direct antioxidant activity.

Irreversible binding of heme to microsomal protein during inactivation of cytochrome P450 by 4-alkyl analogues of 3,5-diethoxycarbonyl-1,4-dihydro-2,4,6-trimethylpyridine

The porphyrinogenicity of 4-alkyl analogues of 3,5-diethoxycarbonyl-1,4-dihydro-2,4,6-trimethylpyridine (DDC) is related to the process of mechanism-based destruction of cytochrome P450 (P450) heme, accompanied by conversion of heme to N-alkylprotoporphyrins (N-alkylPPs). Certain DDC analogues (4-isopropyl, 4-isobutyl, 4-hexyl) are weakly porphyrinogenic in comparison to the potent porphyrinogen, 4-ethyl DDC. We have examined the abilities of these DDC analogues to promote irreversible binding of radiolabeled heme to protein in rat liver microsomal preparations. The goals of this study were to determine whether DDC analogues with different porphyrinogenicities differ in the extents to which they cause heme adduct formation, and whether P450 isozymes differ in their capacities to catalyze heme covalent binding. Incubation of microsomes with NADPH alone promoted heme covalent binding, while loss of spectral P450 heme was minimal or absent. In microsomal incubations containing NADPH, the 4-ethyl, 4-isopropyl, and 4-isobutyl analogues caused heme covalent binding to extents which paralleled their P450 destructive activities. In contrast, 4-hexyl DDC caused less heme covalent binding as a function of P450 loss than the other analogues in microsomes from untreated and beta-naphthoflavone (beta NF)-treated rats. Thus, the weakly porphyrinogenic DDC analogues do not cause greater heme covalent binding than 4-ethyl DDC. Weak porphyrinogenicity, therefore, cannot be explained by diversion of the heme moiety of P450 from conversion to N-alkylPPs towards utilization for formation of heme-derived protein adducts. Treatment of rats with P450 inducing agents altered the degree to which DDC analogues caused heme covalent binding. The greatest heme adduct formation occurred in microsomes from untreated and dexamethasone (DEX)-treated rats, whereas treatment with phenobarbital and especially beta NF reduced heme covalent binding as a function of P450 loss. Thus, these microsomal studies suggest that constitutive P450 isozymes and members of the DEX-inducible P450IIIA subfamily appear to catalyze heme covalent binding, while beta NF-inducible forms such as P450IA1 (P450c) seem to be relatively inactive in this regard.

Covalent interaction of 3,3'-dichlorobenzidine with hepatic lipids. Enzymic basis and stability of the adducts

Administration of a single oral dose (20 mg/kg) of [U-14C]3,3'-dichlorobenzidine to rats resulted in the in vivo covalent binding of the compound to hepatic lipids. More than 70% of the lipid-3,3'-dichlorobenzidine adducts were accounted for in microsomes. Loss of the lipid-bound 3,3'-dichlorobenzidine residues from either total liver or endoplasmic reticulum occurred in at least two phases--an initial fast phase and a terminal slow phase. In vitro studies with hepatic microsomes in the presence of antibodies to specific P450 isozymes and chemical inhibitors to determine the enzymes that activate 3,3'-dichlorobenzidine to the lipid-binding derivative(s) implicated cytochrome P450d. The 3,3'-dichlorobenzidine-bound microsomal lipids were not mutagenic to Salmonella TA98 in the Ames test. The results suggest that adduct formation between 3,3'-dichlorobenzidine and membrane lipids may provide a measure of 3,3'-dichlorobenzidine activation. It is speculated that covalent interaction of the compound with membrane lipids may modify cellular processes, leading to either enhancement or attenuation of carcinogenesis by the chemical.

Inhibition of microsomal lipid peroxidation and cytochrome P-450-catalyzed reactions by beta-lapachone and related naphthoquinones

The lipophilic o-naphthoquinones beta-lapachone, 3,4-dihydro-2-methyl-2-ethyl-2H-naphtho[1,2b]pyran-5,6-dione (CG 8-935), 3,4-dihydro-2-methyl-2-phenyl-2H-naphtho[1,2b]pyran-5,6-dione (CG 9-442), and 3,4-dihydro-2,2-dimethyl-9-chloro-2H-naphtho[1,2b]pyran-5,6-dione (CG 10-248) (a) inhibited NADPH-dependent, iron-catalyzed microsomal lipid peroxidation; (b) prevented NADPH-dependent cytochrome P-450 destruction; (c) inhibited microsomal aniline 4-hydroxylase, aminopyrine N-demethylase and 7-ethoxycoumarin deethylase; (d) did not inhibit the ascorbate- and tert-butyl hydroperoxide-dependent lipid peroxidation and the cumenyl hydroperoxide-linked aniline 4-hydroxylase reaction; and (e) stimulated NADPH oxidation, superoxide anion radical generation and Fe(III)ADP reduction by NADPH-supplemented microsomes. In the presence of ascorbate, the same o-naphthoquinones stimulated oxygen uptake and semiquinone formation, as detected by ESR measurements. The p-naphthoquinones alpha-lapachone and menadione were relatively less effective than the o-naphthoquinones. These observations support the hypothesis that, in the micromolar concentration range, o-naphthoquinones inhibit microsomal lipid peroxidation and cytochrome P-450-catalyzed reactions, by diverting reducing equivalents from NADPH to dioxygen.

Characterization of 19-nor-10-oxo-25-hydroxyvitamin D3 production by solubilized chick kidney mitochondria and bovine serum albumin

The vitamin D3 metabolite obtained from the incubation of 3-[(cholamidopropyl)dimethylammonio]-2-hydroxy-1-propanesulfonate (CHAPSO)-solubilized chick kidney mitochondria with 25-hydroxyvitamin D3 (25-OH-D3) was identified to be 5(E)-19-nor-10-oxo-25-hydroxyvitamin D3 (5(E)-19-nor). The production of 19-nor was dependent on time and on protein concentration, but was not dependent on the pH of the incubation. 19-Nor was not formed in the absence of protein or when protein had been heat-treated following detergent solubilization. 19-Nor was not further metabolized to any other product upon incubation with the CHAPSO-solubilized proteins. No 19-nor-10-oxo derivative of 1,25(OH)2D3 was formed when 1,25(OH)2D3 was used as substrate in the incubation. Kinetic analysis showed a substrate saturation with an apparent Vmax of about 4.1 pmol/min.mg and S0.5 of approximately 1.3 x 10(-6) M. The production of 19-nor was not restricted to the CHAPSO-soluble protein fraction of kidney mitochondria but was also found in both the CHAPSO-soluble and -insoluble fractions of chick liver mitochondria and CHAPSO-treated bovine serum albumin (BSA). 19-Nor production by detergent-treated BSA also showed saturation kinetics with a similar S0.5 and an apparent Vmax which was about 5-fold higher than that obtained with CHAPSO-solubilized mitochondria. The evidence suggests that the formation of 19-nor is not mediated by a traditional enzyme, but does require protein. A mechanism for the conversion of 25-OH-E3 to 19-nor is proposed, in which the naturally-occurring 5(Z)-25-OH-D3 substrate binds to protein, isomerizes to 5(E)-25-OH-D3 and is oxidized by hydrogen peroxide to 5(E)-19-nor via a dioxetane intermediate.

EDTA affects cytochrome P450-dependent biotransformation reactions during incubations for the liver microsomal assay

In order to optimize the condition of the liver microsomal assay (LMA), studies were carried out to determine the effects of EDTA on mixed-function oxidase activity and its stability under the exact incubation conditions for the LMA. Aminopyrine N-demethylase (APD) and p-nitroanisole O-demethylase (p-NAD) activities as well as lipid peroxidation development (LP) in S9 liver fractions from beta-naphthoflavone and sodium phenobarbital (beta-NF + PB)- or Aroclor 1254 (AC)-treated mice were examined during a period of preincubation with EDTA ranging from 1 to 40 mM. At 5 mM EDTA, we obtained a strong inhibition of the microsomal LP as well as the greatest value of the mean specific activity (Asp) for both APD and pNAD activities. In agreement with the biochemical data, the presence of 5 mM EDTA in the incubation mixtures for the LMA significantly increased the mitotic gene conversion, mitotic crossing-over and point-reverse mutation of the well-known premutagen cyclophosphamide (30 mM) on the diploid D7 strain of Saccharomyces cerevisiae as the outcome of a greater metabolic activity. We concluded that the systematic use of 5 mM EDTA in LMA mixtures could improve the reliability and sensitivity of such a test.