Molecular basis for complex formation between methylamine dehydrogenase and amicyanin revealed by inverse mutagenesis of an interprotein salt bridge.

Methylamine dehydrogenase (MADH) and amicyanin form a physiologic complex which is required for interprotein electron transfer. The crystal structure of this protein complex is known, and the importance of certain residues on amicyanin in its interaction with MADH has been demonstrated by site-directed mutagenesis. In this study, site-directed mutagenesis of MADH, kinetic data, and thermodynamic analysis are used to probe the molecular basis for stabilization of the protein complex by an interprotein salt bridge between Arg99 of amicyanin and Asp180 of the alpha subunit of MADH. This paper reports the first site-directed mutagenesis of MADH, as well as the construction, heterologous expression, and characterization of a six-His-tagged MADH. alpha Asp180 of MADH was converted to arginine to examine the effect on complex formation with native and mutant amicyanins. This mutation had no effect on the parameters for methylamine oxidation by MADH, but significantly affected its interaction with amicyanin. Of the native and mutant proteins that were studied, their observed order of affinity for each other was as follows: native MADH and native amicyanin > native MADH and R99D amicyanin > alpha D180R MADH and native amicyanin > alpha D180R MADH and R99D amicyanin, and alpha D180R MADH and R99L amicyanin. The alpha D180R mutation also eliminated the ionic strength dependence of the reaction of MADH with amicyanin that is observed with wild-type MADH. Interestingly, the inverse mutation pair of alpha D180R MADH and R99D amicyanin did not restore the favorable salt bridge, but instead disrupted complex formation much more severely than did either individual mutation. These results are explained using molecular modeling and thermodynamic analysis of the kinetic data to correlate the energy contributions of specific stabilizing and destabilizing interactions that are present in the wild-type and mutant complexes. A model is also proposed to describe the sequence of events that leads to stable complex formation between MADH and amicyanin.

Tyr(30) of amicyanin is not critical for electron transfer to cytochrome c-551i: implications for predicting electron transfer pathways.

A Pathways analysis of the methylamine dehydrogenase-amicyanin-cytochrome c-551i protein electron transfer (ET) complex predicts two sets of ET pathways of comparable efficiency from the type I copper of amicyanin to the heme of cytochrome c-551i. In one pathway, the electron exits copper via the Cys(92) copper ligand, and in the other, it exits via the Met(98) copper ligand. If the Pathways algorithm is modified to include contributions from the anisotropy of metal-ligand coupling, independent of differences in copper-ligand bond length, then the pathways via Cys(92) are predicted to be at least 100-fold more strongly coupled than the pathways via any of the other copper ligands. All of the favored pathways via Cys(92) include a through-space jump from Cys(92) to the side chain of Tyr(30). To determine whether or not the pathways via Cys(92) are preferentially used for ET, Tyr(30) was changed to other amino acid residues by site-directed mutagenesis. Some mutant proteins were very unstable suggesting a role for Tyr(30) in stabilizing the protein structure. Y30F and Y30I mutant amicyanins could be isolated and analyzed. For the Y30I mutant, the modified Pathways analysis which favors ET via Cys(92) predicts a decrease in ET rate of at least two orders of magnitude, whereas the standard Pathways analysis predicts no change in ET rate since ET via Met(98) is not affected. Experimentally, the ET rates of the Y30I and Y30F mutants were indistinguishable from that of wild-type amicyanin. Likely explanations for these observations are discussed as are their implications for predicting pathways for ET reactions of metalloproteins.

Heterologous expression of correctly assembled methylamine dehydrogenase in Rhodobacter sphaeroides.

The biosynthesis of methylamine dehydrogenase (MADH) from Paracoccus denitrificans requires four genes in addition to those that encode the two structural protein subunits, mauB and mauA. The accessory gene products appear to be required for proper export of the protein to the periplasm, synthesis of the tryptophan tryptophylquinone (TTQ) prosthetic group, and formation of several structural disulfide bonds. To accomplish the heterologous expression of correctly assembled MADH, eight genes from the methylamine utilization gene cluster of P. denitrificans, mauFBEDACJG, were placed under the regulatory control of the coxII promoter of Rhodobacter sphaeroides and introduced into R. sphaeroides by using a broad-host-range vector. The heterologous expression of MADH was constitutive with respect to carbon source, whereas the native mau promoter allows induction only when cells are grown in the presence of methylamine as a sole carbon source and is repressed by other carbon sources. The recombinant MADH was localized exclusively in the periplasm, and its physical, spectroscopic, kinetic and redox properties were indistinguishable from those of the enzyme isolated from P. denitrificans. These results indicate that mauM and mauN are not required for MADH or TTQ biosynthesis and that mauFBEDACJG are sufficient for TTQ biosynthesis, since R. sphaeroides cannot synthesize TTQ. A similar construct introduced into Escherichia coli did not produce detectable MADH activity or accumulation of the mauB and mauA gene products but did lead to synthesizes of amicyanin, the mauC gene product. This finding suggests that active recombinant MADH is not expressed in E. coli because one of the accessory gene products is not functionally expressed. This study illustrates the potential utility of R. sphaeroides and the coxII promoter for heterologous expression of complex enzymes such as MADH which cannot be expressed in E. coli. These results also provide the foundation for future studies on the molecular mechanisms of MADH and TTQ biosynthesis, as well as a system for performing site-directed mutagenesis of the MADH gene and other mau genes.

Factors which stabilize the methylamine dehydrogenase-amicyanin electron transfer protein complex revealed by site-directed mutagenesis.

Methylamine dehydrogenase (MADH) and amicyanin form a physiologic complex within which electrons are transferred from the tryptophan tryptophylquinone (TTQ) cofactor of MADH to the type 1 copper of amicyanin. Interactions responsible for complex formation may be inferred from the crystal structures of complexes of these proteins. Site-directed mutagenesis has been performed to probe the roles of specific amino acid residues of amicyanin in stabilizing the MADH-amicyanin complex and determining the observed ionic strength dependence of complex formation. Conversion of Phe97 to Glu severely disrupted binding, establishing the importance of hydrophobic interactions involving this residue. Conversion of Arg99 to either Asp or to Leu increased the Kd for complex formation by 2 orders of magnitude at low ionic strength, establishing the importance of ionic interactions which were inferred from the crystal structure involving Arg99. Conversion of Lys68 to Ala did not disrupt binding at low ionic strength, but it did greatly diminish the observed ionic strength dependence of complex formation that is seen with wild-type amicyanin. These results demonstrate that the physiologic interaction between MADH and amicyanin is stabilized by a combination of ionic and van der Waals interactions and that individual amino acid residues on the protein surface are able to dictate specific interactions between these soluble redox proteins. These results also indicate that the orientation of MADH and amicyanin when they react with each other in solution is the same as the orientation of the proteins which is seen in the structure of the crystallized protein complex.

Mechanism of reaction of allylamine with the quinoprotein methylamine dehydrogenase.

Allylamine did not serve as an efficient substrate for methylamine dehydrogenase (EC 1.4.99.3) in a steady-state assay of activity and appeared to act as a competitive inhibitor of methylamine oxidation by methylamine dehydrogenase. Transient kinetic studies, however, revealed that allylamine rapidly reduced the tryptophan tryptophylquinone (TTQ) cofactor of methylamine dehydrogenase. The rate of TTQ reduction by allylamine was 322 s-1, slightly faster than the rate of reduction by methylamine. These data were explained by a kinetic mechanism in which allylamine and methylamine are alternative substrates for methylamine dehydrogenase. The apparent competitive inhibition by allylamine is due to a very slow rate of release of the aldehyde product, 0.28 s-1, relative to a rate of 18.6 s-1 for the release of the aldehyde product of methylamine oxidation. A reaction mechanism is proposed for the oxidative deamination of allylamine by methylamine dehydrogenase. This mechanism is discussed in relation to the reaction mechanisms of topa-bearing quinoprotein amine oxidases, the flavoprotein monoamine oxidase and the mammalian semicarbazide-sensitive amine oxidase.

Binding constants for a physiologic electron-transfer protein complex between methylamine dehydrogenase and amicyanin. Effects of ionic strength and bound copper on binding.

Two soluble proteins, methylamine dehydrogenase and amicyanin, form a physiologically relevant complex in which intermolecular electron transfer occurs. To characterize and quantitate the binding of these two weakly-associating proteins, an ultrafiltration binding assay has been developed which involves brief centrifugation of mixtures of proteins in centrifuge concentrators followed by quantitation of proteins on each side of the filtration membrane by HPLC. Under low ionic strength conditions which are optimal for the redox reaction between these proteins, a Kd of 4.5 microM was measured for the methylamine dehydrogenase-amicyanin complex. The Kd increased by 8-fold in the presence of added salt. Apoamicyanin, which is known from crystallographic analysis to be structurally very similar to amicyanin, exhibited a much higher Kd and much less specific binding than did the holoprotein. Apoamicyanin also exhibited apparent self-association at low ionic strength which was not observed with amicyanin. These observations are correlated with the known crystal structures of these proteins, free and in complex, and with the available biochemical information on the interactions of these two proteins.

Reactions of benzylamines with methylamine dehydrogenase. Evidence for a carbanionic reaction intermediate and reaction mechanism similar to eukaryotic quinoproteins.

It had been previously reported that aromatic amines were not substrates for the bacterial quinoprotein methylamine dehydrogenase. In this study, benzylamine-dependent activity was also not observed in the steady-state assay of this enzyme with the artificial electron acceptor phenazine ethosulfate (PES). Benzylamines did, however, stoichiometrically reduce the protein-bound tryptophan tryptophylquinone (TTQ) prosthetic group and acted as reversible competitive inhibitors of methylamine oxidation when the enzyme was assayed with PES. When methylamine dehydrogenase activity was monitored using a steady-state assay which employed its physiological electron acceptor amicyanin instead of PES, very low but detectable benzylamine-dependent activity was observed. The reactions of a series of para-substituted benzylamines with methylamine dehydrogenase were examined. A Hammett plot of the log of Ki values for the competitive inhibition by these amines against sigma p exhibited a negative slope. Rapid kinetic measurements allowed the determination of values of k3 and Ks for the reduction of TTQ by each of these amines. A Hammett plot of log k3 versus sigma p exhibited a positive slope, which suggests that the oxidation of these amines by methylamine dehydrogenase proceeds through a carbanionic reaction intermediate. A negative slope was observed for the correlation between log Ks and sigma p. Plots of log k3 and log Ks against substituent constants which reflected either resonance or field/inductive parameters for each para substituent indicated that the magnitude of k3 was primarily influenced by field/inductive effects while Ks was primarily influenced by resonance effects. No correlation was observed between either k3 or Ks and the relative hydrophobicity of the para-substituted benzylamines or steric parameters.(ABSTRACT TRUNCATED AT 250 WORDS)

Comparison of antibody response by use of synthetic adjuvant system and Freund complete adjuvant in rabbits.

Two commercially available synthetic adjuvant systems, trehalose dimycolate (TDM) and TDM + monophosphoryl lipid A (MPL), were compared with Freund complete adjuvant (FCA) for the ability to stimulate antibody production in New Zealand White rabbits (Oryctolagus cuniculus). In addition, each animal was evaluated for adverse reactions. The antigen, rat liver microsomal epoxide hydrolase, was administered SC emulsified with FCA, TDM, or TDM + MPL. Serum antibody titers were stimulated with all 3 adjuvant-antigen combinations. The highest titer was produced by use of FCA; TDM + MPL produced an intermediate response, and TDM produced the lowest titer. All of the rabbits immunized with FCA developed sterile subcutaneous abscesses. Rabbits immunized with TDM or TDM + MPL developed no abscesses, and only slight reactions at the injection site. The synthetic adjuvant system TDM + MPL is recommended for use in rabbits, considering its adequate stimulation of antibody production with minimal adverse reactions.

Strain difference in rat renal microsomal epoxide hydrolase elevation after mercuric chloride treatment.

Administration of the nephrotoxicant mercuric chloride (HgCl2) was found to increase microsomal epoxide hydrolase (EH) activity in the kidneys of Fischer 344 (F344) and Sprague-Dawley (SD) rats, but the increases observed were 2- to 4-fold greater in SD rats than in F344 rats. This study was designed to characterize HgCl2-mediated increases in renal EH activity, and to investigate possible biochemical mechanisms underlying the strain difference. In male SD rats killed 24 h after the last of three daily i.p. injections of HgCl2 (0.1-1 mg/kg), increases in renal EH activity were dose dependent, reaching a maximum of 550% of control. Renal EH activities in identically treated F344 rats were enhanced only to 200% of control values, however, the extent of nephrotoxicity was similar in both strains. Following a single HgCl2 treatment (1 mg/kg), maximal increases in renal EH activities were observed in SD rats (450% of control) at 3 days, and in F344 rats (225%) at 1-2 days. Hepatic glutathione (GSH) concentrations were unaffected by HgCl2 treatment, whereas renal GSH was slightly elevated in both strains. Hepatic metallothionein (MT) concentrations were increased at 1 day to 300% and 400% of control in F344 and SD rats, respectively. Maximal renal MT concentrations were observed at 2 and 3 days in F344 (300% of control) and SD (225%) rats, respectively. Pretreatment with Zn(OAc)2, a potent inducer of renal and hepatic MT, reduced the nephrotoxicity of HgCl2, but did not alter HgCl2-mediated renal EH increases in either strain. In addition, possible strain differences in 203HgCl2 distribution were assessed, but 203Hg distribution was similar in both strains. These studies have demonstrated that renal EH activity is induced by HgCl2, and that there is a strain difference in this response. Differences in MT, GSH and organ distribution of Hg do not account for the strain difference. The possibility remains that other, yet to be defined, protection pathways may exist. Alternatively, renal EH may be differentially regulated between the two strains.

Dinitrotoluene isomer-specific hepatocarcinogenesis in F344 rats.

The hepatocarcinogenicity of 2,4-dinitrotoluene [(2,4-DNT) CAS: 121-14-2], 2,6-DNT (CAS: 606-20-2), and a representative technical-grade DNT (TDNT) containing 76% 2,4-DNT and 18% 2,6-DNT was studied in male F344 rats. Rats were fed diets containing 2,4-DNT, 2,6-DNT, or TDNT at concentrations that resulted in doses of 27 mg/kg/day for 2,4-DNT, 7 or 14 mg/kg/day for 2,6-DNT, and 35 mg/kg/day for TDNT. The carcinogenic effects were evaluated after 1 year of treatment. Administration of 2,6-DNT produced hepatocellular carcinomas in 100 and 85% of animals receiving 14 and 7 mg/kg, respectively. In contrast to the 2,6-DNT results, feeding of 2,4-DNT for 1 year caused no hepatic tumors. Treatment with both isomers (TDNT) resulted in a 47% incidence of hepatocellular tumors. The majority of tumors had a trabecular pattern, and pulmonary metastases were present in the 14- and 7-mg/kg 2,6-DNT-fed groups. These results have demonstrated that 2,6-DNT is a potent and complete hepatocarcinogen and that 2,4-DNT, under these conditions, is nonhepatocarcinogenic. In addition, these data indicate that 2,6-DNT accounts for the majority of the carcinogenic activity of TDNT.

Effects of the chrysotherapeutic agents auranofin and gold sodium thiomalate on hepatic and renal drug metabolism and heme metabolism.

These studies were designed to investigate the effects of the chrysotherapeutic agents auranofin and myochrysine (GST) on hepatic and renal drug-metabolizing enzymes and heme metabolism. Male Sprague-Dawley rats were either administered a single dose of auranofin (17, 34, or 68 mg/kg, p.o.) or administered daily doses of auranofin (0.2, 0.6, 2, 9, or 40 mg/kg/day, p.o.) or GST (1.2 or 5.8 mg/kg/day, i.p.) for 3 or 14 days. Rats were killed 24 h after the final treatment, and subcellular fractions of liver and kidney were prepared. Cytochrome P-450 (P-450) content and ethoxycoumarin-O-deethylase (ECOD), benzphetamine-N-demethylase (BPND), delta-aminolevulinic acid (ALA) synthetase, and heme oxygenase activities were determined. Twenty-four hours following single doses of auranofin, no effects on hepatic P-450, ECOD, or BPND were observed. Treatment with the positive control compounds, CoCl2 (60 mg/kg) and Co-protophorphyrin IX (33 mg/kg), produced decreases in all three variables at 24 hr. Auranofin, at 2 mg/kg, and GST treatment, at both doses, reduced hepatic P-450 and ECOD activity at 3 days. This effect was reversed with continued treatment for 14 days. BPND activity was unaffected at 3 days but was decreased at 14 days. Heme oxygenase activity was enhanced at 3 days and had returned to control activity at 14 days, while ALA synthetase was unaffected. With the exception of heme oxygenase, which was increased, renal variables were unaltered at 3 days. At 14 days, renal P-450 content was decreased in the high-dose auranofin group, heme oxygenase activity was increased in all groups, and ALA synthetase activity was elevated in high-dose auranofin animals. These data indicate that, at doses twenty times the human dose, auranofin and GST administration produced reversible decreases in hepatic and renal P-450 which may be the result of altered heme metabolism.

Elevation of hepatic microsomal epoxide hydrolase activity by 2-acetylaminofluorene: role of metabolism.

Brief exposure of rats to hepatocarcinogenic agents causes a rapid elevation in hepatic microsomal epoxide hydrolase (EH) activity. Previous studies have demonstrated that animals which are resistant to the hepatocarcinogenic effects of 2-acetylaminofluorene (AAF) are also resistant to EH induction by this compound. The studies described here were designed to examine the role of several individual metabolic pathways on the induction of EH by AAF. EH was increased 4-fold in male Fischer 344 (F-344) or Sprague-Dawley (SD) rats treated with AAF; in female rats, deficient in N-hydroxylase and sulfotransferase activities, the activity was increased only 2-fold. Pretreatment of male F-344 rats with inducers of cytochrome P-448 activity caused a reduction in the EH response to AAF, probably due to a greater increase in ring-hydroxylation than N-hydroxylation of the AAF. Although AAF elicited only a small EH elevation in female F-344 rats, N-hydroxy-2-acetylaminofluorene (N-OH-AAF) caused a large increase in these animals. The N-OH-AAF-induced increase was partially blocked by pretreatment with pentachlorophenol, an inhibitor of sulfotransferase activity, in both male and female F-344s. In female SD rats, possessing minimal sulfotransferase activity, N-OH-AAF treatment caused only a slight elevation of EH activity. Pretreatment of male F-344 rats with inhibitors of deacetylase activity had no effect on N-OH-AAF-dependent EH induction. These observations are consistent with the suggestion that formation of the sulfate conjugate of N-OH-AAF is necessary for elevation of EH by this compound.

Effects of methapyrilene on rat hepatic xenobiotic metabolizing enzymes and liver morphology.

Short-term treatment of rats with hepatocarcinogens elicits a consistent pattern of phenotypic changes in hepatic drug metabolizing enzymes, the most striking of which is a marked increase in microsomal epoxide hydrolase (EH) activity. The antihistaminic drug methapyrilene induces a high incidence of hepatocellular carcinoma in F-344 rats. The studies reported here were designed to assess the effects of methapyrilene on hepatic EH activity, cytochrome P-450-dependent mixed-function oxidase activities, liver morphology, and liver-derived serum enzymes. Male F-344 rats were treated with three daily oral doses of methapyrilene-HCl, up to 300 mg/kg/day, and were sacrificed 48 hr after the last dose. Hepatic microsomal EH and cytosolic DT-diaphorase activities were increased in a dose-related fashion, to 420 and 230% of control, respectively. Cytochrome P-450 content and benzphetamine-N-demethylase and ethoxycoumarin-O-deethylase activities were concomitantly decreased to 35-50% of control. Serum gamma-glutamyl transpeptidase and alanine aminotransferase activities were elevated 22- to 27-fold, and serum bile acids to 36-fold by treatment with methapyrilene. Periportal lesions, characterized by inflammation, nuclear and nucleolar enlargement, bile duct hyperplasia, and hepatocellular necrosis, were observed following methapyrilene administration. The severity of the periportal lesion correlated with elevations in the serum chemistry parameters. The increases noted in microsomal EH activity supports the suggestion that this enzyme may be a useful biochemical marker for exposure to hepatocarcinogens.

Elevation of hepatic microsomal epoxide hydrolase activity by 2-acetylaminofluorene: strain and species differences.

Hepatocarcinogens have been shown to cause marked elevation of hepatic microsomal epoxide hydrolyase activity in the rat at short intervals after administration. The present studies were designed to characterize 2-acetylaminofluorene (AAF) mediated epoxide hydrolase elevation and to investigate the relationship between epoxide hydrolase increases, AAF metabolism, and hepatocarcinogenicity. Oral or i.p. administration of AAF to F-344 rats produced log-linear dose-response curves for epoxide hydrolase elevation, measured with either benzo[a]pyrene-4,5-oxide or styrene oxide substrate. Following a single dose of AAF (35 mg/kg), epoxide hydrolase activity was maximally increased (560% of control) within 48 h, and the activity declined slowly, with a half-life of 17.5 days. Co-treatment with actinomycin D effectively blocked the AAF dependent increase in epoxide hydrolase, suggesting that de novo protein synthesis is associated with the increase in enzyme activity. Dose-response curves for epoxide hydrolase induction by AAF, N-hydroxy-2-acetylaminofluorene (N-OH-AAF), and 2-aminofluorene were compared, and the potencies for increasing epoxide hydrolase activity reflected the relative hepatocarcinogenic potentials of these agents. In mice, which are resistant to the hepatocarcinogenic action of AAF and deficient in AAF-N-hydroxylase activity, AAF caused no significant increase in hepatic microsomal epoxide hydrolase activity. Similarly, in Cotton rats and guinea pigs, which are lacking in ability to form the sulfate conjugate of N-OH-AAF, neither i.p. nor dietary administration of AAF elicited increases in epoxide hydrolase activity at doses which were maximally effective in F-344 rats. These results support the hypothesis that the ability of compounds to increase epoxide hydrolase activity is related to their carcinogenic potency. Furthermore, the results suggest that increases in epoxide hydrolase activity are associated with metabolism of AAF to the putative proximate carcinogen N-OH-AAF, and the subsequent conversion of this compound to the N-O-sulfate conjugate.

Effect of hepatocarcinogens on epoxide hydrolase and other xenobiotic metabolizing enzymes.

Hepatocarcinogens cause marked biochemical changes in the liver at short intervals after administration. The studies described were designed to investigate the effects of hepatocarcinogens and hepatotoxicants on the microsomal mixed function oxidase system. DT-diaphorase and epoxide hydrolase. Following 5 day p.o. treatment of male F-344 rats with aflatoxin B1 (AFB), 2-acetylaminofluorene (AAF), technical grade dinitrotoluene (DNT), or 2,4-diaminotoluene, microsomal cytochrome P450 dependent enzyme activities were depressed while epoxide hydrolase activity was markedly elevated (3-8 times control). Diethylnitrosamine (DEN) given at 5 mg/kg/day and DL-ethionine at 1000 mg/kg/day failed to increase epoxide hydrolase. 3-Methylcholanthrene, methylnitrosourea, carbon tetrachloride, bromobenzene and vinyl chloride all failed to increase epoxide hydrolase activity. Using 3 daily i.p. injections, dose-response relationships for increases in epoxide hydrolase were generated for the hepatocarcinogens. With the exception of p-dimethylaminoazobenzene (DAB) and DEN, the carcinogens studied produced log-linear dose response curves for increase in epoxide hydrolase. Both DEN and DAB caused increases in epoxide hydrolase but classical sigmoidal dose-response curves were not obtained. The order of potency for increasing epoxide hydrolase was AFB greater than AAF greater than 2,6-dinitrotoluene greater than 3'-methyl-N,N-dimethyl-4-aminoazobenzene greater than DNT greater than 2, 4-dinitrotoluene. The slopes of the linear portions of the log dose-response curves were not statistically different from the slope of the dose-response curve obtained with AAF suggesting that structurally diverse carcinogens elicit increases in epoxide hydrolase by a common mechanism.

Comparison of hepatic carcinogen initiation-promotion systems.

A number of model systems have been developed to study the initiating and promoting phases of neoplastic development in rats liver. Four of these protocols use diethylnitrosamine (DEN) initiation, but employ different methods of promotion. The present studies were designed to evaluate these systems under standardized laboratory conditions to determine their relative ability to induce histochemically identifiable gamma-glutamyl transpeptidase positive (GGT+) foci. Studies were also performed to examine the effects of the four promoting regimens on liver-derived serum enzymes and hepatic drug metabolism. Under standardized laboratory conditions, including the use of a single rat strain, all four systems induced GGT+ foci following DEN initiation. Within the maximum time period evaluated (8 weeks) promotion with 2-acetylaminofluorene and partial hepatectomy resulted in the highest number of GGT+ foci/cm2. In addition, the hepatic mixed-function oxidase system was markedly affected by the promoting regimens. Cytochrome P-450 content was decreased (50% of control) by three of four systems. All four promotion regimens reduced benzphetamine-N-demethylase activity (20-50% of control). Ethoxycoumarin-O-deethylase activity (P-448 related) was not changed by the promotion regimens. Three of four regimens increased epoxide hydrolase activity (150-600% of control) and DT-diaphorase activity (150-200% of control). Combining DEN initiation and each of the four promotion protocols had little additional effect on hepatic drug metabolizing enzymes. It is concluded that the four systems evaluated are reproducible under standard conditions and that the promotion regimens employed cause striking alterations in hepatic microsomal drug metabolism that are largely independent of the presence or absence of focal GGT+ lesions.