Directed synthesis of 2',5'-oligoadenylates with increased stability towards phosphodiesterases

Phosphodiesterase stability of synthetic analogs of 2',5'-oligoadenylates, the mediators of antiviral and antiproliferative action of interferons was analysed. The analogs with a 3'-terminal acyclic nucleoside residue were prepared. These analogs were treated with NIH3T3 cell lysate, mice liver homogenate and snake venom phosphodiesterase. All analogs have demonstrated a high stability as compared with the natural 2',5'-oligoadenylate and its 3'-deoxyderivative. The possible biological activity of these stable analogs of 2',5'-oligoadenylates is discussed.     

A misaligned double-stranded RNA, poly(I).poly(C12,U), induces accumulation of 2',5'-oligoadenylates in mouse tissues

2',5'-Oligoadenylate synthetase was induced 3-2000-fold in spleen, liver, kidney and brain of NIH Swiss mice injected intravenously with 2-200 micrograms of the misaligned dsRNA, poly(I).poly(C12,U). Levels of 2',5'-oligoadenylates extracted from these tissues were also elevated, although the amount of 2',5'-oligoadenylates extracted did not correlate directly with the amount of enzyme present. These results suggest that double-stranded portions of the misaligned polymer survived intracellularly and activated the 2',5'-oligoadenylate synthetase, and that the level of dsRNA may contribute to the control of 2',5'-oligoadenylate metabolism.     

Purification and characterization of a 2'-phosphodiesterase from bovine spleen

Interferon-induced 2',5'-oligoadenylates are transiently produced during viral infection and are believed to play a role in the interferon-mediated inhibition of replication of at least some viruses. 2',5'-Oligoadenylates must be catabolized but are resistant to degradation by most known ribonucleases. A 2'-phosphodiesterase that degrades 2',5'-oligoadenylates was purified 1500-fold from a low speed homogenate of bovine spleen by precipitation at pH 5.2, ammonium sulfate fractionation, differential ultrafiltration, and successive chromatography on DEAE-Sephacel, hydroxylapatite, and a fast protein liquid chromatography Mono P column. No other 2-5A-degrading activity was observed during the purification procedure. The molecular mass of the enzyme estimated by gel filtration and sodium dodecyl sulfate-polyacrylamide gel electrophoresis is 65,000. The enzyme is distinct from bovine spleen phosphodiesterase II. The 2'-phosphodiesterase cleaves 2',5'- and 3',5'-linked oligonucleotides, as well as branched oligoadenylate, A(2'pA)(3'pA), but appears to be most active on 3',5'-oligoribonucleotides. The enzyme cleaves 5'-AMP from the 2' terminus of 2',5'-oligoadenylates and appears to require a free 2' terminus and a 3'-oxygen on the penultimate nucleotide. Substrate length, 5'-phosphorylation, and base composition do not appear to be critical factors in determining enzyme activity. The effects of pH, Mg2+, Mn2+, EDTA, phosphate, 2-mercaptoethanol, and N-ethylmaleimide are also described. This enzyme may be involved in the catabolism of the interferon-induced 2',5'-oligoadenylates and other 2',5'-linked RNAs in the cell.     

Sensitive radioimmuno assay for 2',5'-oligoadenylates using a novel 125I-labeled derivative of 2',5'-triadenylate 5'-triphosphate

A novel 125I-labeled derivative of 2',5'-triadenylate 5'-triphosphate, pppA2'p5'A2'p5'A, with high specific radioactivity was synthesized by coupling of periodate-oxidized pA2'p5'A2'p5'A with beta-alanyltyrosine methyl ester followed by 5'-triphosphorylation and iodination with 125I. Antisera toward 2',5'-oligoadenylate 5'-triphosphate were produced in rabbits by immunization with the conjugate of pppA2'p5'A2'p5'A2'p5'A with bovine serum albumin, and an antiserum with high specificity and high sensitivity for 2',5'-oligoadenylates was selected and tested extensively. Radioimmuno assaying of 2',5'-oligoadenylates was carried out by a competitive double antibody method in which the amount of the antibody bound to the 125I-labeled probe was measured after precipitation with goat anti-rabbit IgG. The concentration of pppA2'p5'A2'p5'A required for 50% inhibition of the binding between the antiserum and the probe was 0.6 nM. The cross reactivity of the antiserum with the 3',5'-triadenylate was more than 10,000 times weaker compared to in the case of 2',5'-oligoadenylates. Very low or no cross reaction was observed with ATP, AMP, and adenosine. The radioimmuno assay using the 125I-labeled compound and the antiserum allows the direct analysis of 2',5'-oligoadenylates in the range of 4 fmol to 1 pmol (0.04-10 nM in a 100 microliter sample). This assay was applied to the measurement of the activity of 2',5'-oligoadenylate synthetase in cells stimulated by interferon. The properties of the 125I-labeled derivative of pppA2'p5'A2'p5'A are described.     

Raman spectroscopy of interferon-induced 2',5'-linked oligoadenylates

Raman spectra of model compounds and of 2',5'-oligoadenylates in D2O were utilized to assign the Raman bands of 2',5'-oligoadenylates. The Raman spectra of A2'pA2'pA, pA2'pA2'pA, and pppA2'pA2'pA contained features that were similar to those of adenosine, adenosine 5'-monophosphate (AMP), and adenosine 5'-triphosphate, respectively. When AMP and pA2'pA2'pA were titrated from pH 2 to 9, the normalized Raman intensity of their ionized (980 cm-1) and protonated (1080 cm-1) phosphate bands revealed similar pKa's for the 5'-monophosphates. The Raman spectrum of pA2'pA2'pA was altered slightly by elevations in temperature, but not in a manner supporting the postulate that 2-5A possesses intermolecular base stacking. Major differences in the Raman spectrum of 2',5'- and 3',5'-oligoadenylates were observed in the 600-1200-cm-1 portion of the spectrum that arises predominately from ribose and phosphate vibrational modes. Phosphodiester backbone modes in A3'pA3'pA and pA3'pA3'pA produced a broad band at 802 cm-1 with a shoulder at 820 cm-1, whereas all 2',5'-oligoadenylates contained a major phosphodiester band at 823 cm-1 with a shoulder at 802 cm-1. The backbone mode of pppA2'pA2'pA contained the sharpest band at 823 cm-1, suggesting that the phosphodiester backbone may be more restrained in the biologically active, 5'-triphosphorylated molecule. The Raman band assignments for 2',5'-oligoadenylates provide a foundation for using Raman spectroscopy to explore the mechanism of binding of 2',5'-oligoadenylates to proteins.     

Effects of 2-arylbenzimidazoles on rat hepatic microsomal monooxygenase system

1. The effects of eight newly synthesized 2-aryl substituted benzimidazole derivatives on control and phenobarbital (PB) treated rat liver microsomal aniline 4-hydroxylase and ethylmorphine N-demethylase activities, and their binding to control and PB-treated rat liver microsomal oxidized cytochrome P-450 are presented. 2. All compounds inhibited ethylmorphine N-demethylase activity with I50 values ranging from 8.50 x 10(-4) M to 27.83 x 10(-4) M in control and ranging from 2.80 x 10(-4) M to 15.79 x 10(-4) M in PB-treated rats. 3. Aniline 4-hydroxylase activity was inhibited by all of the compounds tested having I50 values in the range of 7.04 x 10(-4) M-31.37 x 10(-4) M in PB-treated rats, but only five of the compounds showed inhibitory activity in control rats. 4. Only a few significant regression coefficients could be found between the parameters of the chemicals studied and their inhibitory patterns. 5. No correlation has been observed between the binding of the derivatives and their inhibitory pattern.     

Inhibition of CCl4 metabolism by oxygen varies between isoenzymes of cytochrome P-450

Oxygen inhibition of CCl4 metabolism by different isoenzymes of cytochrome P-450 was assessed by studying liver microsomes isolated from control rats and rats treated with phenobarbital or isoniazid. Rates of CCl4 metabolism were similar for all microsomes under a nitrogen atmosphere. An air atmosphere inhibited metabolism by microsomes from control rats to 12% of the value under nitrogen and metabolism by microsomes from rats treated with phenobarbital to 5%. It inhibited metabolism by microsomes from rats treated with isoniazid only to 32%. Rats treated with phenobarbital, which increases hepatic cytochrome P-450 content, or isoniazid, which does not increase hepatic cytochrome P-450 content, both metabolized more CCl4 than control rats as indicated by exhalation of greater quantities of CCl4 metabolites and by an increase in CCl4 toxicity. These results indicate that some isoenzymes of cytochrome P-450 are more effective than others in metabolizing CCl4 when oxygen is present.     

Triacylglycerol metabolism in the phenobarbital-treated rat

1. Various aspects of triacylglycerol metabolism were compared in rats given phenobarbital at a dose of 100mg/kg body wt. per day by intraperitoneal injection; controls were injected with an equal volume of 0.15m-NaCl by the same route. Animals were killed after 5 days of treatment. 2. Rats injected with phenobarbital demonstrated increased liver weight, and increased microsomal protein per g of liver. Other evidence of microsomal enzyme induction was provided by increased activity of aminopyrine N-demethylase and cytochrome P-450 content. Increased hepatic activity of γ-glutamyltransferase (EC 2.3.2.2) occurred in male rats, but not in females, and was not accompanied by any detectable change in the activity of this enzyme in serum. 3. Phenobarbital treatment increased the hepatic content of triacylglycerol after 5 days in starved male and female rats, as well as in non-starved male rats; non-starved females were not tested in this regard. At 5 days after withdrawal of the drug, there was no difference in hepatic triacylglycerol content or in hepatic functions of microsomal enzyme induction between the treated and control rats. 4. After 5 days, phenobarbital increased the synthesis in vitro of glycerolipids in cell-free liver fractions fortified with optimal concentrations of substrates and co-substrates when results were expressed per whole liver. The drug caused a significant increment in the activity of hepatic diacylglycerol acyltransferase (EC 2.3.1.20), but did not affect the activity per liver of phosphatidate phosphohydrolase (EC 3.1.3.4) in cytosolic or washed microsomal fractions. A remarkable sex-dependent difference was observed for this latter enzyme. In female rats, the activity of the microsomal enzyme per liver was 10-fold greater than that of the cytosolic enzyme, whereas in males, the activities of phosphohydrolases per liver from both subcellular fractions were similar. 5. The phenobarbital-mediated increase in hepatic triacylglycerol content could not be explained by a decrease in the hepatic triacylglycerol secretion rate as measured by the Triton WR1339 technique. Since the hepatic triacylglycerol showed significant correlation with microsomal enzyme induction functions, with hepatic glycerolipid synthesis in vitro and with diacylglycerol acyltransferase activity, it is likely to be due to enhanced triacylglycerol synthesis consequent on hepatic microsomal enzyme induction. 6. In contrast with rabbits and guinea pigs, rats injected with phenobarbital showed a decrease in serum triacylglycerol concentration in the starved state; this decrease persisted for up to 5 days after drug administration stopped, and did not occur in non-starved animals. It seems to be independent of the microsomal enzyme-inducing properties of the drug, and may be due to the action of phenobarbital at an extrahepatic site.     

Analysis of 2',5'-oligoadenylates in cells and tissues

Complex mixtures of 2',5'-oligoadenylates are formed in cells and tissues under several different circumstances, and methods for analyzing such mixtures are reviewed. Separation is achieved by high-performance liquid chromatography and quantitation by competition-binding assays, using three different types of antibodies or a specific binding protein, or by functional assay, using preparations of an endonuclease specifically activated by some of the 2',5'-oligoadenylates. Representative results from three different biological systems are presented. The function of 2',5'-oligoadenylates as activators of intracellular RNA degradation is discussed, along with the possibility that these compounds may serve as signals for other intracellular regulatory processes.     

Characterization of the effects of known hepatic monooxygenase inducers on male and female rat and mouse kidneys

Few studies have been designed to quantify the response of the mammalian kidney to agents known to induce monooxygenase activity of renal monooxygenase response to three agents representing different classes of inducers: 2,4,2',4'-tetrachlorobiphenyl (2,4,2',4',-TCB), representative of the barbiturate class, beta-naphthoflavone (BNF), representative of the polycyclic aromatic hydrocarbon class and isosafrole (ISO) as a novel class of inducing agent. Studies were carried out using adult rats and mice of both sexes. Treatment with BNF and ISO stimulated ethoxycoumarin and ethoxyresorufin deethylase activities in renal microsomes from male and female rats and mice, whereas treatment with 2,4,2',4',-TCB had no effect on either enzyme in rats of either sex. NADPH-cytochrome-c-reductase activity was unaffected by any treatment. In rat renal microsomes, cytochromes P-450 and b5 were increased by treatment with BNF and ISO but were not altered by 2,4,2',4'-TCB. Sodium dodecyl sulphate-polyacrylamide gel treated with BNF showed the appearance of a protein band in the 50-60000 dalton range which is similar to that observed in liver microsomes following BNF treatment. These studies confirm and extend previous observations that rat kidney is refractory to induction by inducers of the phenobarbital class, but responds to ISO and the polycyclic aromatic class of inducers. In addition, the studies have demonstrated the presence of a protein in renal microsomes after pretreatment of rats with BNF that was not apparent in microsomes from control rat kidneys.     

The stimulating effect of cyclic AMP, glucagon and insulin on guanidine acetate-N-methyltransferase activity in rat liver and pancreas]

N6,O2'-dibutyrylcyclo-3',5'-AMP injected to intact rats alone or in combination with theophylline increases the activity of guanidine acetate methyltransferase (GAMT) in liver and pancreas. Cyclic 3',5'-AMP and its dibutyryl analog administered immediately or two hours after the suturing of common bile duct (SCBD) stimulate the increase of pancreatic GAMT activity 2-3 fold. Glucagon, injected intraabdominally simultaneously with SCBD and administration of theophylline, dramatically increases the theophylline effect on the GAMT activity. The freezing of rat pancreas pretreated witn secretin, a hormone structurally similar to glucagon, results in a 1.5-2-fold increase of creatine synthesis from S-adenosylmethionine and guanidinacetic acid. An hour after glucagon administration to intact rats the GAMT activity of liver increases 9 times. The effect of glucagon is enhanced by insulin. Cycloheximide inhibits the increase of GAMT activity, induced by glucagon or a combination of glucagon and insulin. Experiments on tissue homogenates demonstrate that 3',5'-AMP in concentrations of 10(-8) --10(-2) M does not affect the GAMT activity or to some extent inhibits the enzyme. The homogenate incubation in a medium containing 10(-5) M epinephrine or 10(-7) M caffeine and 5 mM Mg2+ leads to an increase in the GAMT activity. Oligomycin removes the stimulating effects of caffeine and Mg2+ on the enzyme activation. This is probably due to the presence of 3',5'-AMP-dependent protein kinase in the mechanism of GAMT activation by cyclic AMP.     

Accelerated hepatic haem catabolism in the selenium-deficient rat.

1. Hepatic microsomal cytochrome P-450 concentrations are lower in selenium-deficient rats treated with phenobarbital for 4 days than in similarly treated control rats. 2. No defect in haem synthesis was found on the basis of measurements of delta-aminolaevulinate synthase (EC 2.3.1.37), delta-aminolaevulinate dehydratase (EC 4.2.1.24) and ferrochelatase (EC 4.99.1.1) activities, and urinary excretion of delta-aminolaevulinate, porphobilinogen, uroporphyrin and coproporphyrin. 3. No defect in apo-(cytochrome P-450) separated by sodium dodecyl sulphate/polyacrylamide-gel electrophoresis. 4. An increase in haem catabolism was found. An 8-fold increase in hepatic microsomal haem oxygenase (EC 1.14.99.3) activity occurred in selenium-deficient rats after phenobarbital treatment, compared with a less than 2-fold increase in control rats. Also excretion of 14CO in the breath after administration of delta-amino[5-14C]laevulinate was greater by phenobarbital-treated selenium-deficient rats than by similarly treated controls. 5. These studies demonstrate that the defective induction of cytochrome P-450 by phenobarbital in selenium-deficient rats is accompanied by increased haem catabolism. This could be due to increased breakdown of cytochrome P-450 or to catabolism of haem before it attaches to the apo-cytochrome. The role of selenium in stabilizing cytochrome P-450 and/or in protecting haem from breakdown remains to be determined.     

NMR-based metabonomic evaluation of livers from rats chronically treated with tamoxifen,mestranol, and phenobarbital

In this study, we look at the metabolic effects of long-term dosing with tamoxifen, mestranol or phenobarbital on the liver. Tamoxifen, mestranol and phenobarbital have all been reported to act as promoters of hepatic tumors. While tamoxifen and mestranol are known to have estrogenic activity, in the liver phenobarbital is a non-estrogenic compound. Aqueous and lipophilic liver extracts from control and chronically treated Fisher 344 rats were evaluated by nuclear magnetic resonance spectroscopy (NMR). In both the aqueous and lipophilic sample sets, the estrogenic action of mestranol appears to be responsible for the clustering of these samples with those animals treated with tamoxifen. Phenobarbital does not have estrogenic activity and, therefore, clusters away from the estrogenic and control groups. In the lipophilic samples, the fatty acid peak (CH₂)_n was higher in tamoxifen-treated rats than in control, phenobarbital- or mestranol-treated rats. In the aqueous samples, serine and choline levels were higher in phenobarbital-treated rats than controls, which may be an indication that the folate–homocysteine metabolic pathways were altered.     

Induction of rat hepatic drug metabolizing enzymes by dimethylcyclosiloxanes

Low molecular weight dimethylcyclosiloxanes (DMCS) are important precursors in the synthesis of polydimethysiloxane polymers widely used in industry, and in medical and personal care products. The objective of this study was to characterize the ability of two DMCS, octamethylcyclosiloxane (D4) and decamethylcyclopentasiloxane (D5) to induce drug metabolizing enzymes in rats. Male and female Sprague-Dawley rats were administered 1, 5, 20, or 100 mg/kg D4 or D5 in corn oil daily by gavage for 4 days. Changes in the levels of activity and/or immunoreactivity of CYP1A1/2, CYP2B1/2, CYP3A1/2 and NADPH cytochrome P450 reductase in liver microsomes were examined. Significant increases were observed in the liver to body weight ratio in female rats administered either D4 or D5 at doses > or = 20 mg/kg. Increases in the liver to body weight ratio were observed in male rats treated with > or = 100 mg/kg D5 but not with D4. Relatively large increases in CYP2B1/2 enzymatic activity and immunoreactive protein were observed with increasing concentrations of both D4 and D5. Significant increases in 7-pentoxyresorufin O-depentylase (PROD) activity were also detected in male and female rats given D4 at doses > or = 5 mg/kg. D5 increased PROD activity in male rats at doses > or = 20 mg/kg and in female rats at doses > or = 5 mg/kg. 7-Ethoxyresorufin O-deethylase (EROD) activity was increased in both male and female rats receiving > or = 20 mg/kg D4 or > or = 5 mg/kg D5; however, no changes were detected in CYP1A1/2 immunoreactive protein in rats of either sex. D4 and D5 caused significant increases in CYP3A1/2 immunoreactive protein in only male rats treated with 100 mg/kg of either compound. However, significant increases were detected in CYP3A1/2 immunoreactive protein in female rats at D4 doses > or = 20 mg/kg and D5 doses > or = 5 mg/kg. Induction of NADPH cytochrome P-450 reductase immunoreactive protein was observed with D4 in female rats and in both male and female rats with D5. Induction of CYP2B/1/2, CYP3A1/2 and NADPH cytochrome P450 reductase was observed in rats treated with 50 mg/kg phenobarbital by intraperitoneal injection. Maximal CYP2B induction detected with D4 was approximately 50% of the increase observed with phenobarbital. In summary, D4 and D5 induced CYP2B1/2 in adult rat liver in a manner similar to that observed with phenobarbital; however, differences were observed between D4 and D5 in their ability to induce CYP3A1/2 and NADPH cytochrome P450 reductase. Female rats were more sensitive to the inductive properties of low doses of both DMCS than male rats whereas male rats were more responsive to phenobarbital induction.     

ADP-ribosylation of nuclear proteins is increased by phenobarbital. Identification of the ADP-ribosylated histone fractions in rat liver nuclei

Changes in the ADP-ribosylation of total proteins and purified histones of rat liver nuclei after phenobarbital treatment (80 mg/kg, 24 h) have been studied. The [32P]NAD incorporation into total trichloroacetic acid precipitated proteins, in histone Hl and in core histones was evaluated, the specific radioactivities increasing 150, 40 and 8%, respectively. Histones Hl and H2B were the best ADP-ribose acceptors. Histone H4 did not show any 32P incorporation, as revealed by autoradiography after SDS-PAGE of the purified histones, in either the control or phenobarbital treated rats. Possible involvement of ADP-ribosylation of nuclear proteins in the adaptative response of liver to phenobarbital is discussed.     

The induction of the monooxygenase system and the incorporation of the radioactive label from 2-14C-lysine into the liver microsome fraction of rats exposed to phenobarbital against a background of deficiencies in lysine, methionine, threonine and vitamins A, C and E

The effect of diet on induction of monooxygenases and distribution of label from 2-14C-lysine in fractions of liver homogenate, muscle homogenate and blood of male rats treated with phenobarbital (80 mg/kg, three days) was studied. 2-14C-lysine was injected intraperitoneally 24 h before the first injection of phenobarbital. It was demonstrated that monoxygenase induction, increase of relative liver weight and incorporation of label from 2-14C-lysine into fractions of liver homogenate in phenobarbital-treated rats were more pronounced as compared with the similarly treated rats that were fed a balanced diet. The possibility of mobilization of deficient essential components to liver from other organs and tissues for maintenance of monoxygenase induction is discussed.     

Purification and characterization of the dog hepatic cytochrome P-450 isozyme responsible for the metabolism of 2,2',4,4',5,5'-hexachlorobiphenyl

The biochemical basis for the marked difference in the rate of the hepatic metabolism of 2,2',4,4',5,5'-hexachlorobiphenyl (245-HCB) by Beagle dogs and Sprague-Dawley rats has been investigated. Control dog liver microsomes metabolize this substrate 15 times faster than control rat liver microsomes. Upon treatment with phenobarbital (PB), at least two cytochrome P-450 isozymes are induced in the dog, and the hepatic microsomal metabolism of 245-HCB is increased on both a per nanomole P-450 basis (twofold) and a per milligram protein basis (fivefold). One of the PB-induced isozymes, PBD-2, has been purified to a specific content of 17-19 nmol/mg protein and to less than 95% homogeneity, as evidenced by sodium dodecyl sulfate-polyacrylamide gel electrophoresis. In a reconstituted system containing cytochrome b5, this isozyme shows an activity toward 245-HCB which is greater than threefold that seen in intact liver microsomes from PB-induced dogs. A reconstituted system containing the major isozyme induced by PB in the rat (PB-B) metabolizes 245-HCB at 1/10 the rate observed with purified PBD-2. Antibody inhibition studies have shown that PBD-2 accounts for greater than 90% of the hepatic microsomal metabolism of 245-HCB in control and PB-induced dogs, while PB-B only accounts for about half of the metabolism of this compound by microsomes obtained from PB-treated rats. Immunoblot analysis has revealed that the level of PBD-2 in dog liver microsomes increases nearly sixfold with PB treatment, and this increase correlates well with the fivefold increase in the rate of hepatic microsomal metabolism of 245-HCB by dogs. Together these data support a primary role for isozyme PBD-2 in the hepatic metabolism of 245-HCB in control and PB-induced dogs. In addition, these results suggest that, in contrast to rats, dogs can readily metabolize 245-HCB as a result of the presence of a cytochrome P-450 isozyme with efficient 245-HCB metabolizing activity.     

Inhibition of rat liver calciferol 25-hydroxylase activity with anticonvulsant drugs.

Diphenylhydantoin or phenobarbital adminstered for 25 days to Vitamin D-deficient rats inhibited liver calciferol 25-hydroxylase activity. This inhibition was observed with either total homogenate or the microsomal fraction. Eight days following cessation of phenobarbital treatment, liver calciferol 25-hydroxylase activity had returned to control value. Addition of diphenylhydantoin or phenobarbital in vitro to liver homogenate or microsomes isolated from rachitic untreated animals also inhibits the enzymic activity. These data suggest that the impaired conversion of Vitamin D₃ to its 25-hydroxylated metabolite may be the cause of low plasma 25-hydroxycholecalciferol levels in anticonvulsant treated patients.     

Expression of four phenobarbital-inducible cytochrome P-450s in liver, kidney, and lung of rats

Specific antibodies were prepared against cytochromes P450 PB-1, PB-2, PB-4, and PB-5 purified from hepatic microsomes of male rats treated with phenobarbital. With these antibodies, the levels of these four cytochrome P450s in hepatic, renal, and pulmonary microsomes of male rats that were untreated, treated with phenobarbital, or treated with 3-methylcholanthrene were examined. P450 PB-1 and PB-2 were present in moderate amounts in hepatic microsomes of untreated male rats and were induced 2- to 3-fold with phenobarbital. Also, the expression of these forms was suppressed by 3-methylcholanthrene. These forms were not detected in the renal or pulmonary microsomes of untreated rats or rats treated with phenobarbital or 3-methylcholanthrene. P450 PB-4 and PB-5 were found in the hepatic microsomes of untreated male rats at a low level but were induced with phenobarbital more than 50-fold. P450 PB-4 and PB-5 were not detected in renal microsomes; only P450 PB-4 or a closely related form was present in the pulmonary microsomes of untreated male rats, and its level was not changed by phenobarbital treatment. The constitutive presence of P450 PB-4 in pulmonary microsomes was confirmed by the investigation of testosterone metabolism. Purified P450 PB-4 had high testosterone 16 alpha- and 16 beta-hydroxylation activity in a reconstituted system. The testosterone 16 beta-hydroxylation activity of hepatic microsomes was induced with phenobarbital, and more than 90% of the testosterone 16 beta-hydroxylation activity of hepatic microsomes from rats treated with phenobarbital was inhibited by anti-P450 PB-4 antibody.(ABSTRACT TRUNCATED AT 250 WORDS)