THE JOURNAL OF BIOLOGICAL CHEMISTRY Val. 257, No. 17, Issue of September 10, pp. 10446-10457. 1982 Printed in U.S.A.

Phenobarbital-induced Rat Liver Cytochrome P-450 PURIFICATION AND CHARACTERIZATION OF TWO CLOSELY RELATED ISOZYMIC FORMS*

(Received for publication, February 16, 1982)

David J. Waxman$ and Christopher Walsh From the Departments of Chemistry and Biology, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139

The “major” phenobarbital (PB)-induced cytochrome P-450 species present in livers of male Sprague-Dawley rats was resolved into two catalytically active heme- protein fractions on diethylaminoethyl cellulose. The two species, P-450 PB-4 (Mr = 49,000) and P-450 PB-5 (M, = 51,000), were purified to homogeneity, and their chromatographic, spectral, catalytic, and structural properties were compared. P-450 PB-5 eluted earlier on hydroxylapatite and exhibited a more significant cholate-induced Type I spectral shift than P-450 PB-4. Very similar substrate specificity profiles were evident when the two isozymes were reconstituted with lipid, cytochrome P-450 reductase, and cytochrome bs for oxidative metabolism of several xenobiotics, although P-450 PB-4 exhibited a higher specific catalytic activity (25-fold) with all substrates tested. Marked differences were also observed in the sensitivities of both isozymes to several P-450 inhibitors. In addition, P-450 PB-4 was 210-fold more susceptible than P-450 PB-5 to suicide inactivation by two allyl-containing compounds, al- lylisopropylacetamide and secobarbital, providing a possible explanation of the previously observed partial inactivation by such compounds of phenobarbital-in- duced P-450 activity in liver microsomes.

One-dimensional peptide maps of the two isozymes were highly similar. Antibody raised against purified Long Evans rat liver P-450b (Thomas, P. E., Korzen- iowski, D., Ryan, D., and Levin, W. (1979) Arch. Bio- chem. Biophys. 192, 524-532) cross-reacted with P-450 PB-4 and P-450 PB-5. NHz-terminal sequence analysis demonstrated that the first 31 residues of both PB-4 and PB-5 were identical. These sequences indicated that a highly hydrophobic terminal segment, observed previously for other P-450s as well, is followed by a cluster of basic residues, suggesting that the NHZ-ter- minal portion of these P-450s might be involved in membrane anchoring. Although it is unclear whether P-450 PB-4 and P-450 PB-5 are separate gene products or are related by post-translational modifications, this present demonstration of closely related isozymic forms suggests the possible added complexity of micro- heterogeneity for this family of microsomal monooxy- genases.

Mammalian liver contains a family of membrane-bound hemeprotein monooxygenases, known as cytochromes P-450,

* This research was supported in part by Grants GM21643 and GM20011 from the National Institutes of Health and a grant from the Monsanto Company. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. + Recipient of a Damon Runyon-Walter Winchell Cancer Fund Fellowship, 1980-1981 (DRG-439F) and a National Institutes of Health National Research Service Award, 1981-1982.

which catalyze the oxidative metabolism of a wide variety of compounds, including steroids, fatty acids, and xenobiotics (reviewed by Wislocki et al., 1980; Lu and West, 1978). P-450’ can be induced to high levels in liver and other tissues by various foreign compounds including drugs, pesticides, and carcinogens. Molecules such as PB on the one hand and polycyclic aromatic hydrocarbons (e.g. benzo[ alpyrene, 3- methylcholanthrene, P-naphthoflavone) on the other have been used as inducers of two distinct classes of P-450 isozymes. Several groups have reported the purification to apparent homogeneity of P-450 species induced by both classes of compounds (reviewed by Guengerich, 1979; Lu and West, 1980) in efforts directed toward the eventual goal of under- standing the role of each isozyme in the oxidative metabolism of both xenobiotics and endogenous substrates. These studies have demonstrated that both endogenous and induced P-450 isozymes exhibit broad and overlapping substrate specificities.

A question of physiologic and pharmacologic relevance is how many P-450 species (and how many genes encoding them) are present in mammalian tissue. More than one P-450 iso- zyme is inducible by phenobarbital in several animal systems, and it is possible that even the “major” PB-induced P-450 could consist of more than one isozymic form in a given animal. Immunological and biochemical studies (e.g. Guen- gerich et al., 1981; Dent et al., 1980) also raise the possibility of distinguishable P-450 isozymes in different strains of the same animal. Some of these questions are addressed in the present study, where purification of the major PB-induced isozyme from Sprague-Dawley rats is shown to reproducibly resolve the enzyme into two catalytically active P-450 heme- protein fractions, each of which has been purified to homo- geneity. We report several chromatographic and catalytic properties which allow distinction of these two P-450 isozymes (by our nomenclature P-450 PB-4 and P-450 PB-5) but also demonstrate structural properties and substrate specificity profiles for the two isozymes that are much more similar than those of other P-450 isozymes described previously. The re- sults presented also explain the previous observation that allyl-containing sedatives and other olefinic compounds cause only partial loss of PB-induced microsomal monooxygenase activity (e.g. Ortiz de Montellano and Mico, 1980; Loosemore et al., 1980) in that one of the PB-induced P-450 isozymes purified is much more sensitive to suicide inactivation by these compounds than is the other.

P-450; PB, phenobarbital, P-450 PB-4 or P-450 PB-5, P-450 isozyme I The abbreviations and trivial names used are: P-450, cytochrome

4 or 5 purified from PB-induced rat liver microsomes; bs, cytochrome b,; P-450 reductase, NADPH-cytochrome P-450 reductase; EDTA, disodium ethylenediaminetetraacetate; NADPH, reduced nicotin- amide adenine dinucleotide phosphate; SDS, sodium dodecyl sulfate; AIA or allylisopropylacetamide, 2-isopropyl-4-pentenamide; metpa- pone, 2-methyl-1,2-di-3-pyidyl-l-propanone; SKF-525A, diethylami- noethyldiphenylvalerate; secobarbital, 5-aUyl-5-[l-methylbutyl]barr- biturate.

10446

Phenobarbital-induced Rat Liver P-450 Isozymes 10447

MATERIALS AND METHODS

Analytical Procedures-All spectral determinations were per- formed a t 18 "C using a Perkin-Elmer model 554 recording spectro- photometer. Cytochrome P-450 concentrations were determined from the CO-reduced difference spectra using E = 91,OOO M-' cm" according to Omura and Sato (1967). Cytochrome bB was determined from its reduced versus oxidized difference spectrum using t4D9.426 ,,,,, = 185,000 M" cm-' (Estabrook and Werringloer, 1978). Protein concentrations were determined by the Lowry assay using bovine serum albumin as a standard. Discontinuous SDS-gel electrophoresis used the buffer system of Laemmli (1970) unless indicated otherwise. Apparent mo- lecular weights were determined from plots of log molecular weight versus mobility, constructed with data obtained from the following protein standards: phosphorylase a (94,000), bovine serum albumin (S$,OOO), aldolase (40,000), carbonic anhydrase (29,000). lysozyme (14,500), and cytochrome c (12,000). Immunochemical comparisons were kindly performed by Dr. P. E. Thomas, Hoffman-LaRoche, using antibodies and procedures described previously (Thomas et al., 1981).

P-450 Purifzcation-Phosphate buffers (KP,) were pH 7.4, emulgen 911 (Kao Atlas, Japan) was at a concentration of 0.2% (w/v), and glycerol was a t 20% (v/v) throughout, unless noted otherwise. Male Sprague-Dawley rats (40 days old, -140 g each; Charles River Breed- ing Laboratories, Inc.) were induced for 6 to 8 days with 0.1% sodium PB (drinking water), and a liver microsomal fraction was prepared by standard methods (Van der Hoeven and Coon, 1974). Microsomes were solubilized on ice with 2.3% cholate (Sigma cholic acid, recrys- tallized from 50% ethanol) to give 8 to 12 mg of protein per ml of 0.1 M KP,, 30% glycerol, 1 mM EDTA, and a 10 to 16% polyethylene glycol 60oO (Baker) cut was taken, as described by West et at. (1979). This fraction, containing 45 to 50% of the microsomal P-450, was homogenized in buffer A (10 mM KP,, 0.1 mM EDTA, glycerol, 0.5% cholate, emulgen 911). giving A<I~"" , = 6 to 8, and then applied at room temperature (Warner et al., 1978) to a column of Whatman DE52 (2.2 X 40 cm for P-450 derived from 2 to 2.5 g of microsomal protein; resin prewashed and equilibrated in buffer A as described below). After washing with one-half column volume of buffer A, the column was further washed with four column volumes of buffer A + 20 mM KCI.' At this point two hemeprotein fractions remained bound to the column, a dark red band, equivalent to P-450 fraction C of West et at. (1979), at -40 to 60% down the column, and a somewhat less intense red band containing cytochrome b5 at tbe top. The resin was then gently extruded from the column, and the two major hemeprotein fractions were collected. Cytochrome b:, was further purified as described below. The major P-450 fraction was transferred onto a fresh DE52 column (2.2 X 40 cm), and a linear gradient of 20 mM KC1 to 110 mM KC1 in buffer A (900 ml total) was applied. Two partially resolved hemeprotein peaks, termed P-450 PB-4 and P-450 PB-5, were obtained by this method.

Fractions in each peak were pooled and dialyzed overnight a t 4 "C against 210 volumes of buffer B (emulgen 911, glycerol, 20 p~ EDTA) + 10 mM KPi (for P-450 PB-4) or buffer B + 15 mM KPi (for P-450 PB-5) and applied to hydroxylapatite (Bio-Rad HTP) equilibrated in the same buffer (6 to 8 ml of bed volume (1.4 X 6 cm) per 120 nmol of cytochrome P-450) and run at room temperature. After washing with the equilibration buffer, PB-4 was washed with buffer B + 35 mM KP, and then eluted with buffer B + 90 mM KP,. Little or no red color remained on the column. Initial studies indicated that PB-5 bound less tightly to hydroxylapatite than PB-4, facilitating removal of the PB-4 contaminant present in the DE52 pool of PB-5. Thus the PB-5-containing hydroxylapatite column was first washed with equil- ibration buffer, then with buffer B + 45 mM KP, to elute P-450 PB-5 free of PB-4, and finally with buffer B + 90 mM KP, to elute the remaining hemeprotein. Fractions were analyzed by SDS-gel electro- phoresis, and those fractions free of PB-4 were then pooled. Overall recovery of hemeprotein from hydroxylapatite was 270%. Approxi- mately 25 to 35% of the applied PB-5 could be recovered in fractions free of PB-4. P-450 isozymes thus fractionated by DE52 and further purified by hydroxylapatite chromatography were judged greater than 95% pure by SDS-gel electrophoresis and NH2-terminal sequence

first DE52 column. The fwst, termed P-450 PB-1, has been purified ' Under these conditions two hemeprotein fractions elute from this

and partly characterized (Waxman et al., 1982), and the second, termed P-450 PB-2/PB-3 has been partially purified and shown to contain a t least two distinct cytochrome P-450 isozymes (Waxman and Walsh, 1982 and manuscript in preparation).

analysis (see under "Results"). Specific P-450 contents typically ranged from 11 to 16 nmol of P-450 per mg of protein when using a deoxycholate precipitation method (Bensadoun and Weinstein, 1976) to improve reproducibility of the Lowry assay by removing interfering detergents and buffers. The good yield of NHZ-terminal coupling (- 10 nmol of phenylthiohydantoin derivative per 13 to 15 nmol of spectrally determined P-450 applied to the sequenator cup; see Table V below) and presence of a unique amino acid sequence suggest that these somewhat low specific contents reflect neither heme loss nor low protein purity but rather inaccuracy of the results obtained by the Lowry assay as performed.

Detergent Removal-Nonionic detergent (emulgen 91 1) was re- moved from purified p-450 isozymes PB-4 and PB-5 by detergent exchange while the cytochromes were bound to hydroxylapatite. Samples were dialyzed against buffer B + 10 mM KP, and then applied to hydroxylapat,ite equilibrated in the same buffer. Removal of emul- gen, detected by AYXCI ,,,,,, was accomplished by first washing with 10 mM KP,, glycerol, 0.05% emulgen 911, then with 10 mM KP,, glycerol, 0.5% cholate, m d finally with 50 mM KP,, glycerol, and 0.5% cholate. P-450 was eluted as a relatively sharp band with 150 mM KP,, glycerol, and 1.5% cholate. Both isozymes were dialyzed extensively (23 days a t 4 "C with 3 changes) against buffer C (0.1 M KP,, glycerol, 20 ,UM EDTA) to remove the cholate. P-450 isozymes prepared detergent free in this manner were recovered in high yields (>80%) and were soluble and fully active. Readdition of cholate to 0.2% (final concen- tration in the assay) inhibited 7-ethoxycoumarin 0-deethylase activ- ity (see below) by only 20 to 25%. Samples were stored frozen at -70 "C. Attempts to remove detergents by other methods including detergent exchange with 1% octyl-P-D-glucopyranoside (Calbiochem- Behring) instead of cholate, batchwise calcium phosphate treatment, treatment with Bio-Beads SM-2 (Bio-Rad), or repeated precipitation with polyethylene glycol 6000 yielded cytochrome which, although fully active catalytically, was either partially or completely insoluble in buffer C, even after readdition of ionic and/or nonionic detergents. That these fractions were active catalytically is consistent with the finding that P-450 activity is not very dependent on the enzyme's state of aggregation (Guengerich and Holladay, 1979).

Cytochrome b.; Purification-Cytochrome b, was purified from the hemeprotein fraction adsorbed to the top of the first DE52 column (see P-450 purification, above) using the following methods, adapted in part from a procedure originally published for the rabbit enzyme (Strittmatter et al., 1978). The bz-containing DE52 resin (-30 to 40 ml) was first loaded into an empty column (2.6 X 11 cm) and washed with 100 to 150 ml of 20 mM Tris-acetate (pH $.I), 0.1 mM EDTA, 0.4% deoxycholate (buffer D). The resin was then transferred to the top of a fresh column of DE52 (2.2 X 40 cm) equilibrated in buffer I1 and eluted overnight a t room temperature with a linear gradient of buffer D to buffer D + 0.2 M NaSCN (1.5 liters, total volume). Cytochrome bs-containing fractions were pooled, dialyzed against buffer D to reduce the salt concentration, and passed through a 2-ml column of 2',5"ADP Sepharose to remove P-450 reductase. The cytochrome was applied to a column of hydroxylapatite (20 ml of resin per 300 nmol of cytochrome) equilibrated in 10 mM Tris-acetate (pH 8.1), 50 mM KCI, 0.1 mM EDTA, which was then washed with the same buffer until no more emulgen 911 eluted, followed by that same buffer + 20 mM KP,. Cytochrome bh was eluted by addition of deoxycholate (to 0.4%) to this last wash buffer. The peak fractions were pooled and then washed with buffer D and concentrated to -2 ml using an Amicon PM-10 membrane. Residual nonionic detergent and other minor contaminants were removed by gel fdtration through Sephadex G-75 (superfine, 1.3 X 115 cm) equilibrated in buffer D a t 4 "C, in some cases followed by gel filtration through Bio-Gel A-1.5m (1.3 X 60 cm) cquililraced in 0.1 M KCI, 20 mM Tris-acetate (pH 8.1), 0.2 mM EDTA. The final preparation was >95% pure by SDS-gel electrophoresis, had a specific content of 31.6 nmol/mg, and, when entirely free of residual emulgen 911, had an A,,., to A2%, ratio of 2.7 to 2.8. These values are compared to 35.5 nmol/mg and A41.1/A2H,, = 2.78 for the purified steer liver enzyme (Strittmatter et al., 1978). Typical yields were 12 to 179, (with microsomes = 100%) and 28 to 3570 (with the 10 to 16% polyethylene glycol fraction = I@%.).

P-450 Reductase Purification-P-450 reductase was purified from liver microsomes prepared from PB-induced Sprague-Dawley rats (retired breeders obtained from Charles River Breeding Laboratories, Inc.) by emulgen 913 (Kao Atlas, Japan) solubilization followed by DEAE-Sephadex A-25 chromatography as described by Dignam and Strobel (1977). Final purification by affinity chromatography was according to Yasukochi and Masters (1976) excepting that the final washes and NADP' (1 mM) elution were performed in the presence

10448 Phenobarbital-induced Rat Liver P-450 Isozymes

of 0.1% deoxycholate to remove emulgen 913. The protein obtained had a specific activity of >40 pmol of cytochrome c (Sigma) reduced per min per mg of protein when assayed in 0.3 M KPi (pH 7.7) at 30 “C and was >95% pure by SDS-gel electrophoresis. The yield was 0.7 to 0.9 mg of enzyme from 1 g of liver microsomal protein.

Enzyme Assays-Unless otherwise noted, all assavs were ner- formed in 0.1 M KP,, glycerol, 20 pM EDTA (buffer C) with shaking at 37 “C. Dilaurovlphosphatidvlcholine (Sigma) was nrenared as an 0.2% aqueous solutibn in 6.1 mM EDTA (stored at room temperature 57 days) and sonicated before use. Purified enzymes in buffer C were reconstituted with lipid as follows (amounts given for a l-ml assay volume): P-450 PB-4 or P-450 PB-5 (30 pmol) was added to cyto- chrome 65 (35 pmol) and P-450 reductase (45 pmol) to give a final volume of 100 to 200 h. Freshly sonicated dilauroylphosphatidylcho- line (20 h of 200 pg/ml) was added, and the samples were incubated for 3 to 15 min at room temperature, following which substrate (&inhibitor; diluted from either a stock solution in water or methanol; final concentration of methanol %2.00/o) was added in buffer C to give a final volume of 950 X. Samples were warmed to 37 “C (4 min), and the reactions were initiated by addition of NADPH (50 h) to 0.3 mru. Product analyses were as described below. Under these conditions, P- 450 reductase saturated the reconstituted P-450 isozymes half-maxi- mally (see under “Results”).

7-Ethoxycoumarin 0-deethylase activity was determined at a sub- strate concentration of 1 mM (Aldrich; substrate was added from a freshly prepared 100 mM solution in methanol) in a volume of 300 h. Reactions were stopped by addition of 1 N HCl (150 X) after 3 to 10 min, and the fluorescent 7-hydroxycoumarin was determined at room temperature by excitation at 370 nm and emission at 455 nm after chloroform extraction and back extraction into aqueous base as described previously (Waxman et al., 1982). When low activities were determined (e.g. in the inactivation experiments shown in Fig. 6), emission at 400 nm was subtracted from that at 455 nm as a back- ground correction. Hydroxylation of coumarin (1 IIIM) to yield 7- hydroxycoumarin was determined as for 7-ethoxycoumarin O-deeth- ylation.

Sulfoxidation and p-methyl group hydroxylation of p-tolylethyl- sulfide (0.5 mru) were determined with product analysis by high performance liquid chromatography (Waxman et al., 1982). [methyl- “‘C]N,N-Dimethylaniline was synthesized and used as a substrate (0.5 mM) for N-demethylase activity as described by Loosemore et al. (1980).

Toluene oxidation was determined at 5 mM substrate after extrac- tion of reaction mixtures with hexane, essentially as described for sulfoxidation of p-tolylethylsulfide (Waxman et al., 1982). Hexane extracts (0.6 ml) were analyzed by high performance liquid chroma- tography after direct injection onto a aPorasti column (Waters Asso- ciates, Inc.) run at 8 ml of 1.5% isopropanol/98.5% hexane (v/v) per min. Compounds were detected by A zl,, nm using a Micromeritics 786 variable wavelength detector and quantitated by comparison to in- tegrated areas obtained from hexane extracts of standard products. Extraction yields were 23.0% for the three isomeric cresols and 12.8% for benzyl alcohol. Elution times were: toluene, 0.68 min; o-cresol, 1.07 min; m-cresol, 1.22 min; p-cresol, 1.22 min; benzyl alcohol, 2.00 min. Decreasing the isopropanol content to 0.5% (elution time of benzyl alcohol = 9.4 min) did not effect resolution of m- and p-cresol.

NH,- Terminal Sequence Analyses-Purified P-450 samples (13 to 15 nmol each) were prepared for NHs-terminal sequence analysis by precipitation with trichloroacetic acid (6% w/v) followed by heme extraction with acidic acetone (Strittmatter, 1960). Samples were then lyophilized, dissolved in 88% formic acid, and lyophilized again. Samples were subjected to automated Edman degradation using a spinning cup sequenator (Beckman Instruments, 890 C) maintained by Dr. R. T. Sauer, Department of Biology, Massachusetts Institute of Technology. Reagents for sequencing were obtained from Beck- man. Sequence analysis was performed using a 0.1 M Quadrol single cleavage program (Brauer et al., 1975) operating at a 93.5% repetitive yield (myoglobin). Phenylthiohydantoin derivatives were identified and quantitated by a combination of gas chromatography, high per- formance liquid chromatography, and back hydrolysis using methods described previously (Waxman and Strominger, 1981).

Other Materials-Carbon monoxide, 99.5% minimum purity, was obtained from Matheson. Diethylaminoethyl cellulose was obtained from Whatman (DE%) and prepared for chromatography as follows. The resin was soaked overnight in distilled water, washed on a large funnel with distilled water (-5 bed volumes), followed by 0.5 M ammonium sulfate (-5 bed volumes) two or three times. After a final exhaustive wash with water, the pH was adjusted to 7.4, and the resin

was then equilibrated (while in the funnel) with buffer A until the supernatant liquid had the same pH and conductivity. Columns were then poured and washed with 1 bed volume of buffer A before use at room temperature. DE52 was not reused to improve the reproduci- bility of chromatography. Hydroxylapatite (Bio-Rad HTP) was washed according to the manufacturer’s specifications before use. Cholate and deoxycholate were recrystallized from 50% ethanol and added to buffers as the sodium salt.

Metyrapone, p-hydroxyphenyl imidazole, 7-ethoxycoumarin, 7-hy- droxvcoumarin, coumarin, N,N-dimethylaniline, toluene oxidation prod&s, and a-naphthoflavone’were obtained from Aldrich, and PB, n-octvlamine. antinvrine. NADPH, and cholic acid were from Sigma. AIA was from Ho&an-La Roche, Na secobarbital was from Lilly, benzphetamine was from The Upjohn Co., and SKF-525A was from Smith Kline and French. Sulfides and sulfoxides were those described previously (Light et al., 1982).

Enzyme Nomenclature-The nomenclature P-450 PB-1, P-450 PB- 2, P-450 PB-3, etc. is used to identify distinct isozymes isolated from PB-induced animals, with the isozymes numbered in the order of their elution from DE52. P-450 PB-4 corresuonds most closelv to isozymes P-4501, (Ryan et al., 1979), fraction-c (West et al., 1979), and P-450 PB-B (Guengerich 1977; 1978a) with respect to substrate specificity and chromatographic properties. Isozvmes corresponding to P-450 PB-5 have apparently not been reported previously.- As it is unclear whether P-450 PB-4 and P-450 PB-5 are derived from sena- rate genes or are related by post-translational modification (see under “Results”), the term isozyme is used in this paper to include both genetically independent proteins and secondary isozymes in accord- ance with the suggestions of the IUPAC-IUB Commission on Bio- chemical Nomenclature (1976).

RESULTS

Purification of P-450 Isozymes PB-4 and PB-5-Liver microsomes prepared from PB-induced male Sprague-Dawley rats were cholate-solubilized and fractionated with polyethyl- ene glycol as described under “Materials and Methods.” Cy- tochrome P-450 isozymes were then resolved using two col- umns of diethylaminoethyl cellulose (DE52) essentially ac- cording to the methods of West et al. (1979). In contrast to the single major PB isozyme (“fraction C”) obtained from the second DE52 column by these authors, two distinct P-450- containing peaks were obtained in the present study (Fig. 1A). The major protein bands of the two peaks were readily distin- guished by SDS-gel electrophoresis (Fig. 1B) and are termed P-450 PB-4 (M, - 49,000) and P-450 PB-5 (Mr - 51,000), respectively. Approximately 40 to 45% of the total P-450 recovered from the two DE52 columns was typically present within these two peaks, with the ratio of PB-4 to PB-5 = 2.5 + 0.3. Inclusion of 0.5 mM phenylmethanesulfonyl fluoride (an inhibitor of chymotrypsin-like proteases) during preparation of the liver microsomes did not affect this PB-4 to PB-5 ratio.

PB-4 and PB-5 were both purified to apparent protein homogeneity (Fig. 2) by hydroxylapatite chromatography un- der conditions which permitted removal of the residual PB-4 which contaminated the DE52 pool of PB-5 (lane 4 versus lane 5; see under “Materials and Methods” for details). Non- ionic detergent (emulgen 911) was removed from the purified hemeprotein bound to hydroxylapatite by cholate exchange (see under “Materials and Methods”), a procedure which gave good recovery of soluble active cytochrome P-450. Typically, A407 to AzRO ratios of 0.7 to 0.9 were obtained, suggesting -50 to 75 nmol of residual emulgen bound per nmol of P-450.”

Spectral Properties of Purified Isozymes-P-450 PB-4 and P-450 PB-5 exhibited very similar absolute spectra in the oxidized, reduced, and CO-reduced forms (Fig. 3). Maxima at -417, 533, and 568 nm were seen with the oxidized cyto-

” Calculated assuming •2~0 rmhrn = 1300 M-] cm-l,e;;F = gem M-1 r cm- /( and &A “pr,pr”mn = -30,000 M-’ cm-‘, with the last value based

on published amino acid compositions of the major rat liver PB isozyme purified in two different laboratories (Guengerich, 1978a; Bothelho et al., 1979).

Phenobarbital-induced Rat Liver P-450 Isozymes 10449

PE-4 PB-5 I

' t E 0.6 8 a

0.2

I\ 1 : 0.6

w \ "(0.2

0 2 0 4 0 60 F R A C T I O N

e

-94.000

.I -68,000

~29.000

4 -DYE

40 4 2 45 47 4 2 52 54 57

FIG. 1. Resolution of isozymes PB-4 and PB-5 on DE52. The major hemeprotein-containing band excised from a first DE52 column (not shown) was further purified by gradient elution on a fresh DE52 column (see under "Materials and Methods"). Protein and hemepro- tein are detected by absorbance at 295 nm and 417-650 nm, respec- tively (A ) . Aliquots (10 PI) from selected fractions (as indicated) were subjected to SDS-gel electrophoresis ( B ) , confirming the presence of distinct electrophoretic species in each of the hemeprotein peaks shown in A. Fractions were pooled as indicated in A to minimize contamination of the PB-5 pool by PB-4. Molecular weights of stan- dard proteins are shown on the right. Apparent molecular weights for P-450 isozymes PB-4 and PB-5 were 49,000 and 51,000, respectively.

chromes, with significant reduction in intensity of the Soret band and a small shift toward shorter wavelength seen upon reduction with dithionite. Reduction in the presence of CO yielded the 450-nm chromophore indicative of cytochromes P-450. In addition, a much smaller shoulder at -420 nm was present in some preparations (e.g. Fig. 3), suggesting denatur- ation of 10 to 15%4 of the cytochrome to the P-420 form.

As isolated, both isozymes PB-4 and PB-5 contained pre- dominantly low spin heme, as suggested by the A,,, at -417 nm for the oxidized cytochromes. The precise Amax was, how- ever, somewhat variable and, especially with PB-5, was shifted to lower wavelength by -2 to 3 nm at various stages of the purification. The experiments of Fig. 4 suggest that this vari- ability might reflect a detergent-induced partial interconver- sion of low and high spin forms. Addition of cholate to either PB-4 or PB-5, each in 0.2% emulgen 911 and essentially cholate free, yielded a type I difference spectrum (Fig. 4, insets) with a decrease in A4,*,,, coupled to an increase in ARS2 ,,,,,. Soret maxima at 418 nm and 395 nm probably repre- sent low spin and high spin states, respectively (Jefcoate and

Calculated from the direct spectra using extinction coefficients of 110,000 M~ ' cm" for P-450 and 213,000 M-' cm-I for P-420. Similar P- 420 levels were determined from the CO-reduced difference spectra using the method of Omura and Sato (1967).

Wa

Gaylor, 1969). The isosbestic point at -407-408 nm which characterized these cholate-induced spectral changes has also been observed for the high to low spin state change induced by addition of Renex 690 to rabbit liver isozyme LM4 (Haugen and Coon, 1976).

CO-reduced difference spectra for P-450 PB-4 and P-450 PB-5 are shown in Fig. 5. Using an extinction coefficient of 91,000 M" cm" to quantitate P-450 (Omura and Sato, 1967), 6 4 1 7 n m = -120,000 M" cm" was calculated for the Soret maximum of PB-4, with a variable €417 nm = 100,000 to 120,000 M-' cm" for PB-5. Again, the variability with PB-5 probably reflects shifts between low and high spin states, dependent on the precise buffer and detergent composition. In contrast to this sensitivity of the oxidized spectra, the CO-reduced spectra were unaffected by the presence or absence of detergents, excepting that dithionite-induced denaturation of P-450 to P- 420 was more significant in the absence of cholate.

It has recently been shown that the metyrapone-reduced uersus reduced cytochrome P-450 difference spectrum ("P- 446," Hildebrant et al., 1969) is specific for a major phenobar- bital-induced rat liver isozyme (P-4501,) and is not formed with two other isozymes, P-450,, a testosterone 70-hydroxylase and P-450,., a major 3-methylcholanthrene-induced isozyme (Luu- The et al., 1980). As with isozyme P-4501,, both P-450 PB-4 and P-450 PB-5 yielded a metyrapone-reduced difference spectrum characterized by a Amax at 445.5 nm (Fig. 5). The extinction coefficients measured for these P-446 complexes,

.J- 51,000 9- 49.000

-DYE

1 2

C29.000

*DYE

3 4 5 FIG. 2. SDS-gel electrophoresis of purified P-450 isozymes.

Shown are aliquots of P-450 isozymes PB-4 (lanes I and 3) and PB- 5 (lanes 2 and 5) purified by hydroxylapatite chromatography (see under "Materials and Methods") with resolution on a 13.5% gel (lanes I and 2) or a 10% gel (lanes 3 to 5 ) . Lane 4, isozyme PB-5 as pooled off the second DE52 column, before hydroxylapatite chromatography, showing PB-4 contamination. Molecular weights of standard proteins and of the purified isozymes were as indicated.

10450

O o 8 I - w 0 06 V z a m

$ 0 0 4 LL

m a

A. P 4 5 0 P B - 4

Phenobarbital-induced Rt

- Fet3 "" F e t 2

Fe+'-CO

O o 2 V - , k- "L-

350 400 450 500 550 600 650 W A V E L E N G T H (nml

B. P 4 5 0 P B - 5

- F e f 3 "" Fe+

i / ' , L I

350 400 450 500 550 600 650 W A V E L E N G T H ( n m )

FIG. 3. Visible absorption spectra of P-450 isozymes PB-4 ( A ) and PB-5 ( B ) . Spectra were measured in buffer B + 45 KP,, a t P-450 concentrations of 0.70 nmol/ml and 0.47 nmol/ml ( A and B, respectively). Shown are spectra of the oxidized enzymes (-) and of the enzymes after ciithionite reduction (- - -) or reduction in the presence of CO (---). The small peak a t -420 nm in the CO- reduced spectrum reflects denaturation of 10 to 15Wj4 of the cyto- chromes to a P-420 form.

t445.5.1!)0 = 60,000 M" cm ' * 4,000 for P-450 PB-4 and t445.g.4!3~)

= 70,000 M" cm- I +- 5,000 for P-450 PB-5 (both a t 1 mM metyrapone) were greater than that determined for P-4501,, t = 52,000 M" cm I (Luu-The et al., 1980).

Catalytic Properties-Both P-450 isozymes PB-4 and PB- 5, when reconstituted with purified P-450 reductase and cy- tochrome 6, in a dilauroylphosphatidylcholine system (see under "Materials and Methods") catalyzed oxidative metab- olism of several model substrates (Table I). With all substrates examined, P-450s PB-4 and PB-5 showed very similar sub- strate specificities, with PB-5 exhibiting only -20% the activ- ity per nmol of P-450 as did PB-4. The turnover numbers tabulated reflect enzymatic activity under standard assay conditions, i.e. 30 nM P-450, 35 nM cytochrome bg, 45 nM P- 450 reductase, and 4 pg/ml (6.4 p ~ ) of lipid. This lipid concen- tration was optimal for 7-ethoxycoumarin 0-deethylation cat- alyzed by both PB-4 and PB-5, with significantly reduced activity both in the absence of lipid (Tabie 11, column A ) or a t higher lipid concentrations.5 Although cytochrome h, at 35 nM was apparently saturating, 45 nM P-450 reductase yielded only half-maximal saturation of catalytic activity per P-450, as determined by Lineweaver-Burk analysis. One would, therefore, expect -2-fold greater turnover numbers upon re-

" e.g. 50W, reduction in activity a t 33 kg/ml of lipid.

zt Liver P-450 Isuzymes

A . PB -4 t CHOLATE 1

340 380 420 460 500 W A V E L E N G T H (nrn)

B. P B - 5 t CHOLATE

A

0.03-

a -

I I I I I 1 t 340 38 0 4 20 4 60 500

WAVELENGTH (nrn) FIG. 4. Cholate-induced conversion of cytochromes P-450

from low to high spin forms. Cholate was added to isozymes PB- 4 and PB-5 ( A and B, respectively) each a t 0.4 nmol of P-450/ml in buffer B + 45 mM KP, to give final cholate concentrations as indicated. Appropriate volume corrections were made for control samples. Vis- ible spectra were measured with absorbances normalized a t 700 nm. The shift to high spin form (X,,,, [high spin heme I - 395 nm) occurred with an isosbestic point of 407 to 408 nm. Inset, cholate-induced difference spectra with samples in 0.067 to 1.0% cholate (as indicated) in the sample cuvette and that with 0% cholate in the reference cuvette. C, difference spectrum between 1 8 cholate + buffer and buffer alone.

constitution with saturating P-450 reductase. Inclusion of cytochrome bs in the reconstituted system

stimulated monooxygenase activity of both isozymes by 50 to 100% (Table 11, column C). This stimulation was dependent on the inclusion of lipid in the assay; in the absence of lipid cytochrome h5 inhibited catalytic activity by 60 to 70% (col- umn D ) . These results suggest that the relatively low concen- tration of dilauroylphosphatidylcholine monomers present (critical micelle concentration 2.45 p ~ ) " is sufficient to facili- tate productive interactions between P-450 reductase, bs, and

Quoted by Miwa and Lu (1981).

Phenobarbital-induced Rat Liver P-450 Isozymes 10451 41

A P 4 5 0 P B - 4 E P 4 5 0 P B - 5

+ 0 0 5 6 A

a - 0 0 1 4

~ 0028

L, I 1 ~1 L I - ~ _ L L ~ I L v 380 400 420 440 460 480

W A V E L E N G T H I n m )

FIG. 5. Metyrapone-reduced and CO-reduced difference spectra. P-450 PB-4 (0.65 nmol/ml in 35 mM KP;, glycerol, 20 PM EDTA, 0.15%) emulgen 91 1, 0.4% cholate; A ) and P-450 PB-5 (0.40 nmol/ml in buffer B + 45 mM KP, + 0.5% cholate; B ) were dithionite reduced in the presence of CO (-) or 1 mM metyrapone (MTP,

Both CO-reduced difference spectra exhibited X,,,, values a t 449.5 nm f 0.5, while the metyrapone-reduced difference spectra were blue shifted to 445.5 nm * 0.5.

”-) and the ligand/reduced uersus reduced spectra was measured.

TABLE I Catalytic activities of P-450 isozymes PB-4 and PB-5

Purified isozymes were reconstituted with P-450 reductase and cytochrome b.,, and catalytic activities were measured as described under “Materials and Methods.” Activities are expressed as turnover numbers, i.e. nanomoles of product formed per nmol of P-450 per min at the indicated substrate concentrations. Typically, P-450 PB-5 exhibited -20% the specific catalytic activity of P-450 PB-4, although with some ureuarations this value was as low as 10 to 13%.

Substrate Products determined

7-Ethoxycoumarin 7-Hydroxycoumarin Coumarin 7-Hydroxycoumarin N,N-Dimethylaniline N-Methylaniline p-Tolylethylsulfide p-Tolylethylsulfox-

p-Hydroxymethyl- idea

phenylethylsul- fide

Toluene Benzyl alcohol p/m-Cresol o-Cresol

strate Sub-

con- cen-

tion tra-

r n M

1.0 1.0 0.5 0.5

Turnover

P-450 P-450 PB-4 PB-5

min” P-450” 4.5 0.8

t0.02 t0.02 27 5.0 30 5.0

-0.1 t0.05

5.0 22 3.7 2.4 t 0 . 5 3.4 t 0 . 5

‘ I Both S- and K-sulfoxides were formed (assayed as described in Waxman et al., 1982) in a ratio of S /R -4 to 5 for both isozymes P- 450 PB-4 and P-450 PB-5.

”Isomeric cresols were not resolved (see under “Materials and Methods”). These data can be compared with product ratios reported in microsomal incubations: benzyl alcohol >> cresols; o-:m-:p-cresol = 591328 (Ullrich, 1972).

P-450 (as indicated by the stimulation in catalytic activity). This effect is most pronounced in the case of P-450 PB-1, whose characteristic 6- to %fold b5 stimulation (Waxman et al., 1982) is abolished in the absence of lipid.’ This lipid requirement is consistent with the observation that 65 reduc- tion by P-450 reductase is greatly stimulated by addition of lipid (Enoch and Strittmatter, 1979). Reconstitution in the presence of bs was also seen to magnify the effective stimula- tion of catalytic activity by dilauroylphosphatidylcholine (col- umn A uersus B ) . 6, stimulation of catalytic activity most likely reflects an increase in the net rate of transfer of the second electron from P-450 reductase to P-450.

Effects of Inhibitors and Inactiuators-Several known P- 450 inhibitors were examined to determine whether significant differential inhibition of isozymes PB-4 and PB-5 could be detected. The results obtained (Table 111) indicate that P-450 PB-4 is significantly more sensitive to inhibition by each of the compounds tested than is P-450 PB-5. Although PB-5 formed the P-446 metyrapone spectral complex as well as did PB-4 (Fig. 5), it was more resistant to inhibition by this heme ligand, suggesting that the coordination of a ligand to the enzyme’s heme is not necessarily equivalent to inhibition of catalytic activity as a result of binding to the active site. Other ligands, including a-naphthoflavone (50 p ~ ) , coumarin (1 mM), and antipyrene (1 mM) did not inhibit the catalytic activity of either PB-4 or PB-5.

TABLE I1 Influence of dilauroylphosphatidyl choline and cytochrome bi on

catalytic activity Purified isozymes were assayed for 7-ethoxycoumarin 0-deethylase

activity as described under “Materials and Methods.” Activities were determined for both isozymes in the presence of dilauroylphosphati- dylcholine (“lipid”) and cytochrome b5 and in the absence of either or both components as indicated. The observed cytochrome ha stimula- tion (column C) of both isozymes varied from - 1.4 to 2.0, with typical values as shown. Deletion of cytochrome P-450, P-450 reductase, NADPH, or 7-ethoxycoumarin reduced the observed activity to less than 5% of control values. Values in columns A to D were calculated by taking ratios of the appropriate pairs of samples for each isozyme.

Assay condi- tions

~ . Relalive activily

I.ipid stimulation” h , stimulation*

lipid hi (A) (B) +hi -h5 +hpd -lipid

(C), (D)

0.30

I. P-450 PB-4 1. + + = 100 10.2 1.47 2. + - 67.9 2.10 3. - + 9.8 4. - 32.4

11. P-450 PB-5 5. + + I 100 2.89 1.97 6. + - 50.7 0.53 7. - + 34.6 0.36 8. -

“Activity in the presence of lipid compared to activity in the absence of lipid, with cytochrome b5 either present or absent, as indicated. The -3-fold greater lipid stimulation of P-450 PB-4 may reflect higher residual emulgen 911 in this preparation of P-450 PB-5 ( A . ~ , V / A ~ ~ ~ ~ = 0.83 for PB-4 uersus 0.72 for PB-5) effecting a partial stimulation of P-450 activity even in the absence of added lipid.

Activity in the presence of cytochrome b5 compared to activity in t,he absence of ba, with lipid either present or absent, as indicated.

-

- 95.6 .” ”. . . . ” ._ ”

TABLE 111 Effects of P-450 in,hibitors on catalytic actiuity

Isozymes PB-4 and PB-5 were reconstituted with P-450 reductase and cytochrome b5 and enzymatic activity then measured using 7- ethoxycoumarin (1 mM) as substrate in a standard assay (see under “Materials and Methods”). Listed are catalytic activities in the pres- ence of various inhibitors relative to control (uninhibited) samples.

Inhibitor Concentration PB-4 PB-5

CLM 70 control Metyrapone

Benzphetamine

n-Octylamine

SKF-525A

p-Hydroxyphenylimidazole

70 46 70 1000 9.3 35 200 24 57

lo00 8.7 25 25 21 62

200 5.5 47 25 33 66

200 8 41 12.5 30 74 50 13 61 ’ Waxman and Walsh, 1982, and manuscript in preparation.

10452

z a a a y I O * X 0 n a 5

r * I

P-

a9

Phenobarbital-induced Rat Liver P-450 Isozymes

1 1 I

A. P B - 4 + A I A

CONTROL

A

1 E. P B - 4 + SECOBARBITAL

0 CONTROL

400prn

n C. P B - 5 + A I A I SECOBARBITAL

CONTROL

‘0.4rnM SB

1 I I I I

2 0 40 T I M E (rnin)

FIG. 6. Suicide inactivation of isozymes PB-4 and PB-5 by AIA and secobarbital. P-450s PB-4 ( A and B ) and PB-5 (C) (30 nM each) were reconstituted with P-450 reductase (38 nM) and cyto-

TABLE IV Suicide inactivation of P-450 PB-4

AIA Secobarbital

I. Purified components” kinactivatton 1.6 X 10~:’s” 2.2 X 10”’s“ K, 0.52 mM 40 pM

knactivatmn 1.4 X lO-’’s” K, 1.1 mM

11. PB-induced microsomes

~

“ Determined from the data in Fig. 6 after replotting t ~ p values (time for 50% inactivation) uersus l/[inactivator], with y intercept = k~nacl~val~un/ln 2 and x intercept = -1/KI. ki...t,..t,.. is the rate constant for inactivation under conditions of saturating inactivator, and KI is the apparent K, for the inactivator.

* Data from Ortiz de Montellano and Mico (1981).

A suicide inactivation of microsomal cytochrome P-450 by several alkene-containing compounds, including AIA and al- lyl-containing barbiturates has been described (DeMatteis, 1971; Levin et al., 1972, 1973). From several studies it is apparent that AIA (as well as other alkene inactivators, e.g. Ortiz de Montellano and Mico, 1980) might be isozyme specific in that a maximum of only -70% of the dimethylaniline N- demethylase activity (Loosemore et al., 1980) or -40% of the total spectrally detected cytochrome P-450 (Ortiz de Montel- lano and Mico, 1981) can be inactivated by AIA pretreatment of phenobarbital-induced rat liver microsomes. As isozymes P-450 PB-4 and P-450 PB-5 probably account for the majority of the dimethylaniline N-demethylase activity of these micro- somes (isozymes PB-1 and PB-2/PB-32 catalyzing this reac- tion at even lower rates than PB-5),7 the sensitivity of both isozymes to AIA-induced inactivation was tested. The results obtained (Fig. 6, A and C) demonstrate that inactivation of P- 450 PB-4 is characterized by pseudo-first order kinetics, con- sistent with inactivation proceeding via a suicide mechanism (Walsh et al., 1978). Furthermore, P-450 PB-4 was signifi- cantly more sensitive to this suicide inactivation than P-450 PB-5. A similar differential sensitivity was seen in the case of the presumably analogous inactivation by secobarbital (Fig. 6, B and C). The kinetic parameters characterizing inactiva- tion of PB-4 by these two compounds are included in Table IV, where they are compared to data obtained previously in a microsomal study (Ortiz de Montellano and Mico, 1981). I t is unclear whether the much reduced sensitivity of PB-5 to these inactivators reflects an inherent resistance to the heme alkylation which likely accounts for loss of activity (Ortiz de Montellano et al., 1981a) or rather a much reduced rate of turnover which is necessary for generating the inactivating species.

Structural Studies-Peptide mapping by partial proteolysis followed by one-dimensional SDS-gel electrophoresis (Cleve- land et al., 1977) has been used to examine the structural

C ) or NADPH + secobarbital (SB; B ) , each at the indicated concen- trations for 0 to 50 min, after which 7-ethoxycoumarin was added to 1 mM. Catalytic assays were quenched by addition of 1 N HC1 after either a 3-min (PB-4) or 5 min (PB-5) reaction, with 7-hydroxycou- marin quantitated as described under “Materials and Methods.” Suicide inactivation of PB-4 by AIA and secobarbital appears to be superimposed on an inhibition of 7-ethoxycoumarin 0-deethylase activity which is most noticeable at 2 mM AIA or 0.4 mM secobarbital. The rates of inactivation of P-450 PB-5 upon incubation with 10 mM AIA or 5 mM secobarbital were essentially the same as for the concentrations shown in C, excepting that the superimposed inhibi- tory effects were larger (relative activities of 55% (10 mM AIA) and 40% (5 mM secobarbital)). Under the assay conditions utilized in the

chrome bs (35 nM) at 7.5 pg of lipid/ml, with the other conditions as present study (i.e. at lower enzyme concentrations and in the presence described under “Materials and Methods.” Samples were preincu- of b5), P-450 autoinactivation upon incubation with NADPH and 0 2

bated with NADPH alone (control) or with NADPH + AIA ( A and (Loosemore et al., 1980) was minimized (control).

Phenobarbital-induced Rat Liver P-450 Isozymes 10453

94,000-

68,000-1

14,500L I 2,000 7

FIG. 7. Peptide mapping of P-450 isozymes PB-4 and PB-5. Purified isozymes were precipitated with trichloroacetic acid (10% final concentration). The pellets were washed with acetone then dissolved in 88% formic acid containing 0.4% SDS to yield a protein concentration of -0.15 mg/ml. CNBr (-1 g/ml of acetonitrile) was added to 100 mg/ml, and digestion was allowed to proceed for 1 to 60 min at room temperature, after which 10 volumes of water were added, and the samples were lyophilized. Residues were analyzed by SDS-gel electrophoresis (-8 pg per lane). Shown is a stained 15% gel prepared using the buffer system of Maizels (1971) with molecular weights of standard proteins shown on the left. Lanes I to 3, P-450 PB-4 digested for 1, 10, and 60 min, respectively. Lanes 4 to 6, P-450 PB-5 digested for 1, 10, and 60 min, respectively.

relationship of purified cytochrome P-450 isozymes. Thus peptide mapping of rat liver isozymes P-450., P-450h, P-450,, and P-450,1 (an isosafrole-induced isozyme) (Ryan et al., 1980) and isozymes PB-A, PB-B, PB-D, and 3-methylcholanthrene- B (Guengerich, 1978b), as well as several rabbit liver isozymes (Guengerich, 1978b; Johnson, 1980; Koop et al., 1981), has clearly demonstrated that despite general similarities in mo- lecular weight, spectral properties, and substrate specificities, chromatographically distinct P-450 fractions described thus far are also distinct structurally. In contrast to these results, striking similarities in peptide maps were evident after cleav- age of P-450 PB-4 and P-450 PB-5 with cyanogen bromide (Fig. 7), Staphylococcus aureus V8 protease, or a-chymotryp- sin (data not shown) suggesting major structural similarities, i.e. similar distributions of methionine (CNBr), glutamic acid (V8 protease), and phenylalanine, tyrosine, tryptophan, or leucine residues (chymotrypsin) along the polypeptide chains of PB-4 and PB-5.

Further evidence for a close structural homology between P-450 PB-4 and P-450 PB-5 is provided by NHZ-terminal sequence analysis of each isozyme. The results obtained (Ta- ble V) clearly indicate that PB-4 and PB-5 have identical primary structures for their first 31 and probably 34 amino acid residues. Furthermore, the sequences obtained are highly hydrophobic as well as highly homologous with the 22 NH2- terminal residues of two other PB-induced isozymes, rat liver P-4501,, and rabbit liver P-450 LMn, as shown in Fig. 8. The

TABLE V NH2-terminal sequence analysis of PB-4 and PB-5

Purified isozymes (-13 to 15 nmol each) were subjected to auto- mated NH2-terminal sequence analysis with phenylthiohydantoin residues identified as described under "Materials and Methods." Yields listed refer to the number of nanomoles of residue at each cycle, with quantitation by high performance liquid chromatography, except as indicated. Contaminating residues (other than overlapping residues) at the first 3 steps were present a t 45% the level of the major residue identified. Uncertain residues are in parenthesis.

P-450 PB-4 P-450 PR-5

Residue Cycle identified Cycle identified Residue

1 2 3 4 5 6 7 8 9

10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35

Met Glu Pro Ser Ile Leu Leu Leu Leu Ala Leu Leu Val G ~ Y Phe Leu Leu Leu Leu Val k g G ~ Y His Pro LYS Ser k g GlY Asn Phe Pro

(Pro) GlY

(Pro) ?

9.4 1 5.0 2 4.7 3 4.3 4 6.8 5 8.1 6 6.7 7 8.8 8 7.7 9 7.0 10 8.1 11 8.8 12 6.6 13 4.4 14 3.5 15 3.9 16 5.0 17 5.2 18 5.6 19 3.2 20 1.6" 21 3.1 22 2.0" 23 1.4 24 4.5 25 2.9 26 1.5" 27 3.4 28 1.6 29 2.2 30 1.3 31 1.5 32 3.0 33 0.9 34

35

Met Glu Pro Ser Ile Leu Leu Leu Leu Ala Leu

Val Leu

GlY Phe Leu Leu Leu Leu Val k g G ~ Y His Pro LYS Ser k g GlY Asn Phe Pro

(Pro) GlY

(Pro) ?

7.1 6.9 5.7 6.5 7.6 7.0 8.2 7.4 8.0 8.5 6.1 7.3 4.7 4.3 3.6 4.4 6.0 5.9 6.7 4.6 3.1 a

6.7 0.9" 4.36 4.0 1.5 l.OR 1.5 1.8 3.0 2.0 1.4 2.9 1.1

" Quantitation by back hydrolysis with HI followed by amino acid analysis.

" Triple cleavage at this proline reduced the significant increase in overlap observed with PB-4, i.e. for PB-4, a 25% Val overlap at cycle 21 increased to a 788 Lys overlap at cycle 26 oersus for PB-5, a 33% Val overlap at cycle 21 increased to only a 48% Lys overlap at cycle 26.

absence of significant contaminating residues at each of the first few cycles of Edman degradation ( ~ 5 % the level of the major residue; Table V) provides additional evidence for the homogeneity of each isozyme.'

Comparisons to Other P-450 Isozymes-The NHz-terminal sequences and the chromatographic, spectral, and catalytic properties described in this report together suggest that P-450 isozymes PB-4 and PB-5, purified from livers of Sprague- Dawley rats, might be similar to the major rat liver PB isozyme purified previously from Long Evans rats, i.e Fraction C (West et al., 1979) or P-45Ob (Ryan et aL, 1979). Analysis by double diffusion immunoprecipitation in Ouchterlony plates, kindly performed by Dr. P. E. Thomas, Hoffman-LaRoche, using highly purified antibodies to Long Evans isozyme P-

" Microheterogeneity in other regions of the polypeptide chains of either P-450 PB-4 or P-450 PB-5 would not, of course, be detected in the NHn-terminal sequence runs.

10454 Phenobarbital-induced Rat Liver P-450 Isozymes l o O H 5 10

P450 PB-4 H~N-RET-GLU-PRO-S~R-ILE-LEU-LEU-~U-LEU-ALA-LEU-LEU-V~~-GLY-PHE-L~U-LEU-

P450 pa-5 H2N------ _ _ _ P450, H2N"THR"""

P450 L!$ H ~ N - - +HE-LE- --____ PHE-ALA-LEU-

1 5

20 P450 PB-4 L E U - L E U - ~ a L - ~ ~ - G L ~ - ~ ~ - P ~ o - L ~ - ~ ~ ~ - ~ ~ - G L ~ - ~ ~ N - P H E - P ~ o ~ ~ ~ o ~ G L ~ ~ P ~ o ~

P450 PB-5 -

P450, -

P 4 5 0 LT2

25 30

-~ PHE--

FIG. 8. Comparison of NH,-terminal sequences of phenobar- bital-induced P-450 isozymes. Shown are the NHz-terminal se- quences determined for rat liver isozymes PB-4 and PB-5 (data from Table V) and for the major PB isozyme previously isolated from rat liver, P-4& (Botelho et al., 1979). Also shown is the NHZ-terminal sequence determined for the major rabbit PB-induced isozyme, P-450 LM2 (Haugen et al., 1977). Charged and hydroxy-containing side chains are as indicated. Residues identified are those which differ from the identifications made with P-450 PB-4. The same 20 NH2- terminal residues determined in this study were reported by Bar-Nun et al. (1980) for a PB-induced isozyme purified from rat liver (also see under "Discussion"). The hydrophobicities of these sequences are analyzed in Table VI.

450h (Thomas et al., 1979), confirmed this expectation and indicated that isozymes PB-4 and PB-5 were each immuno- chemically indistinguishable from isozyme P-4501, (not shown). Further analysis demonstrated the absence of cross- reactivity (and thus the absence of contamination ( t 3 to 5%)) of PB-4 or PB-5 by rat liver isozymes P-450,, P-450,, and P- 450,,, as well as by epoxide hydrase and P-450 reductase.

DISCUSSION

Two major cytochrome P-450 isozymes have been purified from PB-induced rat liver microsomes. The two isozymes, termed P-450 PB-4 and P-450 PB-5, were shown to have highly similar spectral properties, substrate specificities, and peptide maps, yet were readily distinguished by their chro- matographic properties, specific catalytic activities, and sen- sitivities to several inhibitors and suicide inactivators. The highly related peptide maps, exact identity of the NHn-ter- minal31 amino acids, and similar immunochemical properties of P-450 PB-4 and P-450 PB-5 provide strong evidence that these two isozymes arise from closely related and possibly identical genes. Thus this pair of P-450 isozymes is distinct from others of the P-450 enzyme family purified from rat or rabbit liver, which have readily distinguishable peptide maps and distinct if overlapping substrate specificities (e.g. Guen- gerich, 1978a, 1978b; Ryan et al., 1980; Johnson, 1980; Koop et al., 1981).9

Using methods similar to those employed here in the puri- fication of P-450 PB-4 and P-450 PB-5, West et al. (1979) isolated a single major P-450 isozyme (termed fraction C) from PB-induced rat liver microsomes. Possibly the much smaller hemeprotein peak, eluting just after fraction C in their published DE52 chromatogram (Fig. 2 of West et al., 1979) corresponds to isozyme P-450 PB-5. If so, this would suggest a ratio of P-450 PB-4 to P-450 PB-5 of -8 to 10 in the Long Evans rats used by West et al. versus the 2.5 ratio observed in the current study with animals of Sprague-Dawley descent.

Although some properties of P-450 PB-4 and P-450 PB-5 are analogous to those of isozymes PB-B and PB-D isolated by Guenger- ich (1977, 1978a) from Sprague-Dawley rats, the latter two isozymes have distinct NH2 termini (which are also different from those re- ported here) and distinct peptide maps (Guengerich, 1977, 1978b) and, therefore, do not appear to correspond to the isozymes described in this study.

Ouchterlony double diffusion analysis of P-450 PB-4 and P- 450 PB-5 indicated the absence of detectable immunoreactiv- ity with highly specific antibodies that had been raised by Thomas et al. (1979, 1981) against rat liver isozymes P-450,, P-450,, and P-45Od (as well as P-450 reductase and epoxide hydrase), supporting the lack of such contaminants in these isozyme preparations. By contrast, both P-450 PB-4 and P- 450 PB-5 were cross-reactive and immunochemically indistin- guishable from the major Long Evans PB-induced isozyme, P-4501, (purified by Ryan et al. (1979) and most probably equivalent to fraction C of West et al. (1979)). This immuno- chemical homology between P-450 PB-4 and P-450 PB-5 seems analogous to the immunocrossreactivity of two PB- induced Long Evans rat liver isozymes (P-450b and P-450,) recently presented by Ryan and Levin (1981). P-450 PB-4 and P-450b are not, however, identical, as shown by comparative SDS-gel anaysis; although Sprague-Dawley P-450 PB-4 co- migrates with Holtzman rat P-450h,'O Long Evans P-450b displayed a slightly lower mobility and P-450 PB-5 a distinctly lower mobility (electrophoresis comparisons kindly performed by D. E. Ryan, Hoffman-LaRoche). Some immunochemical differences between PB-induced P-450 isozymes from Long Evans and Sprague-Dawley rats have been detected (Guen- gerich et al., 1981) suggesting that monoclonal antibodies specific for each distinct PB P-450 isozyme, e.g. P-450 PB-4, P-450 PB-5, and P-45&, might be developed to facilitate quantitation of each form present in hepatic microsomes, e.g. after treatment with each of the PB-type inducers described by Thomas et al. (1981).

Cytochrome P-450 inhibitors exhibiting isozyme selectivity have been described. These include a-naphthoflavone, which inhibits P-450 isozymes induced by 3-methylcholanthrene and other polycyclic aromatics but not PB isozymes (Wiebel et al., 1971) such as P-450 PB-4 and P-450 PB-5, as shown in this study. The dipyridyl heme ligand metyrapone binds to PB- induced isozymes such as P-45Oh (Luu-The et al., 1980) to form a spectrally distinct P-446 complex (Hildebrandt et al., 1969), but not to P-450 PB-1, another isozyme isolated from these PB-treated rats (Waxman et al., 1982),'. P-450,, a testosterone 7-a-hydroxylase, or P-450,, a 3-methylcholan- threne-induced isozyme (Luu-The et al., 1980). In the case of the classic P-450 inhibitor SKF-525A, discrimination between P450 isozymes PB-2 and PB-3 has been ob~erved .~ The avail- ability of isozyme-specific inhibitors in conjunction with other specific inhibitors such as suicide inactivators (e.g. AIA; see below) or monospecific antibodies should facilitate assessment of isozymic contributions to microsomal activities and ulti- mately to in vivo activities.

AIA and other olefins, including allyl-containing barbitu- rates, inactivate hepatic cytochrome P-450 as a consequence of alkylation of the P-450's active site heme by a reactive metabolite derived from the olefin inactivator (reviewed by Ortiz de Montellano et al., 1981a). Recent studies have dem- onstrated that PB-induced P-450 is significantly more sensi- tive to these inactivators than is the 3-methylcholanthrene- induced P-450 (Bradshaw et al., 1978; Ortiz de Montellano et al., 1981b). Although these inactivations most likely proceed via a suicide-type mechanism (Walsh et al., 1978), kinetics consistent with such a route have not been demonstrated previously using a purified P-450 isozyme. Inactivation kinet- ics had been studied with PB-induced rat liver microsomes, where only -70% of the N,N-dimethylaniline N-demethylase activity or -40% of the total spectrally detected P-450 was inactivated by AIA with a partition number (ratio of number

"'Purified by Drs. D. E. Ryan, P. E. Thomas, and W. Levin, Hoffman-LaRoche, from PB-induced Holtzman rats and shown to be immunochemically indistinguishable from Long Evans P-4501,.

Phenobarbital-induced Rat Liver P-450 Isozymes LO455

of AIA molecules metabolized per P-450 molecule inactivated) of -200 (Loosemore et al., 1980, 1981; Ortiz de Montellano and Mico, 1981). Inactivation of less than 100% of the total P- 450 activity was also observed with other olefinic inactivators (Ortiz de Montellano and Mico, 1980), raising the question as to whether the residual activity reflects the presence of par- tially inactivated P-450 molecules ("wounded-enzyme" hy- pothesis) or, alternatively, a population of inactivator-resist- ant isozymes (Loosemore et al., 1980). Results obtained in the present study (Fig. 6, A and B ) demonstrate that both AIA and the allyl barbiturate secobarbital inactivate P-450 PB-4 with pseudo-fist order kinetics comparable to those observed in microsomes (Ortiz de Montellano and Mico, 1981) as well as consistent with a suicide-type inactivation pathway. Fur- thermore, the dramatically reduced inactivation rate of P-450 PB-5 (Fig. 6C) (as well as of P-450 PB-1 and P-450 PB-2/PB- 3') provides evidence that the partial loss observed with microsomal P-450 reflects an isozyme specificity on the part of these olefin-induced suicide inactivators. It remains to be determined whether the reduced sensitivity of these other isozymes reflects a higher partition ratio or lower turnover number for processing of the suicide substrates.

The NHAerminal sequences determined for P-450 PB-4 and P-450 PB-5 (Fig. 8) are identical with the sequence

published for Long Evans rat liver P-450b (Botelho et al., 1979), with the exception of the NHz-terminal methionine, absent in P-450b, and the serine at position 4, reported as a threonine by Botelho et al. Strain differences probably do not account for the serine/threonine discrepancy at residue 4, since Bar-Nun et al. (1980) obtained the sequence H,N-Met- Glu-Pro-Ser-Ile-Ile- . . . by standard Edman degradation of a P-450 isolated from PB-induced Long Evans rat liver. Ozols et al. (1981) have noted that the NHz-terminal methionine of rabbit liver P-450 LM3b is absent in some of their preparations and have suggested that the sequence reported by Botelho et al. may reflect a similar lability.

In a study of the biosynthesis of P-450 in PB-induced rat liver endoplasmic reticulum, Bar-Nun et al. (1980) demon- strated that an NHp-terminal segment of the major PB-in- duced P-450 is not proteolytically processed during matura- tion and insertion of the isozyme into the endoplasmic retic- ulum. As such, it has been suggested that the highly hydro- phobic NH,-terminal segments of PB-induced isozymes puri- fied from rabbit (P-450 LM,) and rat (P-450h, etc.) might correspond to uncleaved signal sequences, i.e. to the NH,- terminal hydrophobic peptides removed from nascent secre- tory and membrane proteins during their insertion into the endoplasmic reticulum membrane (Haugen et al., 1977; Bar-

TABLE VI Hydrophobicity of P-450 NH2-terminal sequen

Mean hydropathy values were determined for 19-residue segments as described by Kyte and Doolittle (1982). Calculations for the most hydrophobic segments of this length found in nine soluble proteins (IV) and for the membranous segments of coliphage MI3 coat protein and red blood cell glycophorin A are taken from Kyte and Doolittle (1982). Other calculations were done in the present study based on the sequences found in the references indicated. Included are all published mammalian microsomal P-450 segments at least 18 residues in length. Several are as hydrophobic as the most hvdroDhobic

ces: comparison to other hydrophobic sequences membranous segments presently known. Although the hydrophobic- ities of several of the signal sequences are larger than the most hydrophobic segments found within soluble proteins, they are smaller than the more hydrophobic membranous segments. Some of these membrane proteins might be membrane bound, in part by virtue of covalently bound fatty acyl groups (e.g. Schmidt and Schlesinger, 1980). Residues are numbered from the NH2 terminus except as indicated.

Protein Amino acid residues

I. Cytochrome P-450 isozymes (hydrophobic NH,- terminal region)

P-450 PB-4 and P-450 PB-5 (rat) Ser4-Gly2* P-450, (rat) P-450 LM2 (rabbit)

Gly5-LeuZ3 Glu2-PheZ0

P-450d (rat) P-450 testis (pig)

Leu8-LeuZ6

P-450, (rat) Met'-Pro'*(18)" Ile'-Val''

ments) 11. Membrane proteins (membrane-spanning seg-

Glycoprotein G (vesicular stomatitis virus) Ser48-Le~30 E l glycoprotein (semliki forest virus) L e ~ ~ ~ ~ - G l y ~ ~ ~ Red blood cell glycophorin A (human) IIe7'-Ileg1 Hemagglutinin (influenza virus) Ile36-Gly'8b E2 glycoprotein (semliki forest virus) Val367-Cysrn Immunoglobulin, p chain Thr"-Phe4' Membrane coat protein (M13 coliphage) Cytochrome be (rabbit)

~ ~ ~ ~ - 1 1 ~ ~ ~ Vd113-Ala'3'

111. Signal sequences (cleaved to yield mature pro- tein)'

(a) Membrane proteins Precoat protein (M13 coliphage) Leu"-Ala' Preglycoprotein G (vesicular stomatitis vi- Met"-Cys'( 16) a

Prelipoprotein (Escherichia coli)

Prelysozyme (hen oviduct) Met'8-Gly'(18)n Preparathyroid hormone (bovine) Met24-Leu6 Pre-/3-lactamase ( E . coli) His"-Ala' Preimmunoglobulin, K chain (mouse) Met2*-Gly4

W'us) MetZo-Ala'

(b) Secretory proteins

IV. Soluble proteins (buried hydrophobic stretches) (n = 9)

' I Segment length, as indicated in parenthesis, is less than 19 residues.

' Segments numbered from the cleavage site. Segments numbered from the protein's COOH terminus.

hydropathy Mean Sequence data reference

~~

2.60 2.76 2.58 2.24 1.79 1.31

2.76 2.66 2.65 2.65 2.21 2.02 1.92 1.83

1.49 ::2 I 2.16 1.77 I 1.11 I 1.09 f 0.22

This study Botelho et al., 1979 Haugen et al., 1977 Botelho et al., 1982 Nakajin et al., 1981 Botelho et al., 1979

Rose et al., 1980 Garoff et al., 1980 Furthmayr et al., 1978 Jou et al., 1980 Garoff et al., 1980 Rogers et al., 1980 Wickner, 1976 Takagaki et al., 1980

Davis and Tai, 1980

Davis and Tai, 1980

Kyte and Doolittle, 1982

10456 Phenobarbital-induced Rat Liver P-450 Isozymes

Nun et al., 1980). It should be noted, however, that P-450s differ from these other proteins in that they are not trans- ported to the cell surface. Although the hydrophobic NH2- terminal region of P-450 might function as a “signal” directing the nascent polypeptide to the membrane, the presence of a cluster of 4 basic residues immediately following the stretch of 18 uncharged residues ( i .e . at positions 21 to 27) suggests that the molecule’s hydrophobic terminus might play an im- portant role in anchoring this hemeprotein to the endoplasmic reticulum. Analysis of the hydropathy of this segment using a method recently developed by Kyte and Doolittle (1982) indicated that the NH2-terminal regions of several P-450s are distinctly more hydrophobic than several signal sequences and, in fact, are as hydrophobic as the most hydrophobic membrane-spanning polypeptide segments studied to date, including red blood cell glycophorin A, vesicular stomatitis virus glycoprotein G , and influenza virus hemagglutinin (Ta- ble VI). Hydropathic values calculated for all three groups, i.e. the P-450 NHn-terminal sequences, the membranous seg- ments, and the signal sequences, are readily distinguished from the buried (internal) segments of highest hydrophobicity found in soluble proteins (Kyte and Doolittle, 1982), as shown in Table VI. Further investigation is necessary to elucidate the structural features important for anchoring P-450 PB-4, P-450 PB-5, and other P-450 isozymes to the endoplasmic reticulum.

The resolution of P-450 isozymes PB-4 and PB-5 advances somewhat our understanding of the potential complexities of PB-induced cytochrome P-450 monooxygenases from rat liver. These findings further emphasize the caution which must be exercised when purifying closely related isozymes, i.e. even NH2-terminal sequencing might not indicate heterogeneity in the case of closely related isozymes like P-450 PB-4 and P-450 PB-5. As such, it is even possible that sequence microhetero- geneity exists in our preparations of P-450 PB-4 or P-450 PB- 5. That the AIA and secobarbital susceptibility of the purified isozymes mirror the partial susceptibility of microsomal P-450 to inactivation suggests that any proteolytic relationship be- tween P-450 PB-4 and P-450 PB-5 might reflect an in vivo cleavage. That P-450 PB-4 is not only significantly more active but is also more sensitive both to inhibitors and to suicide inactivators suggests that biosynthesis of P-450 PB-4 versus P-450 PB-5 (whether determined by transcriptional control or by post-translational modification) could play an important regulatory role in P-450 activity. Further study is required to determine the structural relationship between P-450 PB-4 and P-450 PB-5, the total number of PB-induced isozymes and corresponding genes, as well as the contributions of each isozymic form to an organism’s overall monooxygenase capa- bility.

After completion of this manuscript, Ryan et aE.” reported purification of an isozyme (termed P-450,) from PB-induced Long Evans rat liver that appears equivalent to P-450 PB-5. As with P-450 PB-5, when compared to P-450 PB-4, P-450, exhibited significantly lower catalytic activity than P-4501, as well as immunocrossreactivity and peptide maps highly simi- lar to P-450h, in agreement with the major findings of the present study. That isozymes such as P-450 PB-4 and P-450 PB-5 could be distinct gene products is supported by recent studies of Fujii-Kuriyama and co-workers,” who detected nucleotide sequence heterogeneity (corresponding to a t least 6 amino acid replacements) in two cDNA clones corresponding to such PB-induced P-450 isozymes from Sprague-Dawley

” Ryan, D. E., Thomas, P. E., and Levin, W. (1982) Arch. Biochem. Biophys. 216,272-288

’* Fuji-Kunyama, Y., Mazukami, Y., Kawajiri, K., Sogawa, K., and Muramatsu, M. (1982) Proc. Natl. Acad. Sci. U. S. A . 79,2793-2797

rats. Predicted amino acid sequences agree with the 34-residue NH,-terminal sequence obtained in the present study.

Acknowledgments-We wish to thank Joan Monks for expert technical assistance in protein purification, Dr. R. T. Sauer for use of the protein sequencer, Drs. P. E. Thomas, D. E. Ryan, and W. Levin for immunochemical and electrophoretic comparisons of our purified P-450 isozymes with those purified in their laboratory, and Drs. J. Kyte and R. F. Doolittle for providing us with a copy of their manuscript prior to publication.

REFERENCES

Bar-Nun, S., Kreibich, G., Adesnik, M., Alterman, L., Negishi, M., and Sabatini, D. D. (1980) Proc. Natl. Acad. Sci. U. S. A. 77, 965- 969

Bensadoun, A., and Weinstein, D. (1976) Anal. Biochem. 70,241-250 Botelho, L. H., Ryan, D. E., and Levin, W. (1979) J. Biol. Chem. 254,

Botelho, L. H., Ryan, D. E., Yuan, P.-M., Kutny, R., Shively, J . E.,

Bradshaw, J. J., Ziman, M. R., and Ivanetich, K. M. (1978) Biochem.

Brauer, A. W., Margolies, M. N., and Haber, E. (1975) Biochemistry

Cleveland, D. W., Fischer, S. G., Kirschner, M. W., and Laemmli, U.

Davis, B. D., and Tai, P.-C. (1980) Nature 283, 433-438 DeMatteis, F. (1971) Biochem. J. 124, 767-777 Dent, J . G., Graichen, M. E., Schnell, S., and Lasker, J. (1980) Toxicol.

Dignam, J. D., and Strobel, H. W. (1977) Biochemistry 16,1116-1123 Enoch, H. G., and Strittmatter, P. (1979) J. Biol. Chem. 254, 8976-

Estabrook, R. W., and Werringloer, J. (1978) Methods Enzymol. 52,

Furthmayr, H., Galardy, R. E., Tomita, M., and Marchesi, V. T.

Garoff, H., Frischauf, A.-M., Simons, K., Lehrach, H., and Delius, H.

Guengerich, F. P. (1977) J. Biol. Chem. 252, 3970-3979 Guengerich, F. P. (1978a) J. B i d . Chem. 253,7931-7939 Guengerich, F. P. (197813) Biochem. Biophys. Res. Commun. 82,820-

Guengerich, F. P. (1979) Pharmacol. Ther. 6, 99-121 Guengerich, F. P., and Holladay, L. A. (1979) Biochemistry 18, 5442-

Guengerich, F. P., Wang, P., Mason, P. S., and Mitchell, M. B. (1981)

Haugen, D. A., and Coon, M. J . (1976) J. Biol. Chem. 251,7929-7939 Haugen, D. A., Armes, L. G., Yasunobu, K. T., and Coon, M. J. (1977)

Hildebrandt, A. G., Leibman, K. C., and Estabrook, R. W. (1969)

IUPAC-IUB Commission on Biochemical Nomenclature (1976) J.

Jefcoate, C. R. E., and Gaylor, J . L. (1969) Biochemistry 8,3464-3474 Johnson, E. F. (1980) J. Biol. Chem. 255,304-309 Jou, W. M., Verhoeyen, M., Devos, R., Saman, E., Fang, R., Huyle-

Emtage, S. (1980) Cell 19,683-696 broeck, D., Fiers, W., Threlfall, G., Barber, C., Carey, N., and

Koop, D. R., Person, A. V., and Coon, M. J . (1981) J. Biol. Chem. 256,10704-10711

Kyte, J., and Doolittle, R. F. (1982) J. Mol. Biol. 157, 105-132 Laemmli, U. K. (1970) Nature 227,680-685 Levin, W . , Jacobson, M., and Kuntzman, R. (1972) Arch. Biochem.

Levin, W., Jacobson, M., Sernatinger, E., and Kuntzman, R. (1973)

Lieht. D. R.. Waxman. D. J., and Walsh. C. (1982) Biochemistry 21,

5635-5640

and Levin, W. (1982) Biochemistry 21, 1152-1155

Biophys. Res. Commun. 85,859-866

14,3029-3035

K. (1977) J. Biol. Chem. 252, 1102-1106

Appl. Pharmacol. 52,45-53

8981

212-220

(1978) Arch. Biochern. Biophys. 185,21-29

(1980) Nature 288,236-241

827

5449

Biochemistry 20,2370-2378

Biochem. Biophys. Res. Commun. 77,967-973

Biochem. Biophys. Res. Commun. 37,477-485

Biol. Chem. 252, 5939-5941

Biophys. 148,262-269

Drug Metab. Dispos. 1,275-285

?490-2498 Loosemore, M., Light, D. R., and Walsh, C. (1980) J. Biol. Chem.

255,9017-9020

256,8705-8712

Chemother. Toxicol. Metab. Inhibitors 2, 337-358

Loosemore, M. J., Wogan, G. N., and Walsh, C. (1981) J. Biol. Chem.

Lu, A. Y. H., and West, S. B. (1978) Pharrnacol. Ther. Part A

Lu, A. Y. H., and West, S. B. (1980) Pharrnacol. Rev. 31,277-295

Phenobarbital-induced Rat Liver P-450 Isozymes 10457

Luu-The, V., Cumps, J., and Dumont, P. (1980) Biochem. Biophys.

Maizels, J . (1971) Methods Virol. 5, 180-247 Miwa, G. T., and Lu, A. Y. H. (1981) Arch. Biochem. Biophys. 211,

Nakajin, S., Shively, J . E., Yuan, P.-M., and Hall, P. F. (1981)

Omura, T., and Sato, R. (1967) Methods Enzymol. 10,556-559 Ortiz de Montellano, P. R., and Mico, B. A. (1980) Mol. Pharmacol.

18, 128-135 Ortiz de Montellano, P. R., and Mico, B. A. (1981) Arch. Biochem.

Biophys. 206,43-50 Ortiz de Montellano, P. R., Mico, B. A., Beilan, H. S., and Kunze, K.

L. (1981a) in Molecular Basis of Drug Action (Singer, T . P., and Ondarza, R., eds) pp. 151-166, Elsevier/North-Holland, New York

Ortiz de Montellano, P. R., Mico, B. A., Mathews, J. M., Kunze, K. L., Miwa, G. T., and Lu, A. Y. H. (1981b) Arch. Biochem. Biophys. 210, 717-728

Ozols, J., Heinemann, F. S., and Johnson, E. F. (1981) J. Biol. Chem. 256, 11405-11408

Rogers, J., Early, P., Carter, C., Calame, K., Bond, M., Hood, L., and Wall, R. (1980) Cell 20, 303-312

Rose, J . K., Welch, W. J., Sefton, B. M., Esch, F. S., and Ling, N. C. (1980) Proc. Natl. Acad. Sci. U. S. A . 77,3884-3888

Ryan, D. E., Thomas, P. E., Koneniowski, D., and Levin, W. (1979) J. Biol. Chem. 254, 1365-1374

Ryan, D. E., Thomas, P. E., and Levin, W. (1980) J. Biol. Chem. 255,

Ryan, D. E., and Levin, W. (1981) Fed. Proc. 40, 1640 Schmidt, M. F. G., and Schlesinger, M. J. (1980) J. Biol. Chem. 255,

Strittmatter, P. (1960) J. Biol. Chem. 235, 2492-2497

Res. Commun. 93, 776-781

454-458

Biochemistry 20,4037-4042

7941-7955

3334-3339

Strittmatter, P., Fleming, P., Connors, M., and Corcoran, D. (1978)

Takagaki, Y., Gerber, G. E., Nihei, K., and Khorana, H. G. (1980) J.

Thomas, P. E., Korzeniowski, D., Ryan, D., and Levin, W. (1979)

Thomas, P. E., Reik, L. M., Ryan, D. E., and Levin, W. (1981) J. Biol.

Ullrich, V. (1972) Angew. Chem. Int. Ed. Engl. 11, 701-712 van der Hoeven, T. A,, and Coon, M. J. (1974) J. Biol. Chem. 249,

6302-6310 Walsh, C., Cromartie, T., Marcotte, P., and Spencer, R. (1978) Meth-

ods Enzymol. 53,437-448 Warner, M., La Marca, M. V., and Neims, A. H. (1978) Drug. Metab.

Dispos. 6, 353-362 Waxman, D. J., and Strominger, J. L. (1981) J. Biol. Chem. 256,

2067-2077 Waxman, D. J., and Walsh, C. (1982) in Cytochrome P-450, Biochem-

istry, Biophysics and Environmental Implications (Hietanen, E., ed) Elsevier/North-Holland, New York, in press

Waxman, D. J., Light, D. R., and Walsh, C. (1982) Biochemistry 21, 2499-2507

West, S. B., Huang, M.-T., Miwa, G . T., and Lu, A. Y. H. (1979) Arch. Biochem. Biophys. 193,42-50

Wickner, W. (1976) Proc. Natl. Acad. Sci. U . S. A. 73, 1159-1163 Wiebel, F. J., Leutz, J . C., Diamond, L., and Gelboin, H. V. (1971)

Arch. Biochem. Biophys. 144, 78-86 Wislocki, P. G., Miwa, G. T., and Lu, A. Y. H. (1980) in Enzymatic

Basis of Detoxification (Jakoby, W. B., ed) Vol. 1, pp. 135-182, Academic Press, New York

Yasukochi, Y., and Masters, B. S. S . (1976) J. Biol. Chem. 251,5337- 5344

Methods Enzymol. 52, 97-101

Biol. Chem. 255, 1536-1541

Arch. Biochem. Biophys. 192, 524-532

Chem. 256, 1044-1052