Characterization of a Major Form of Rat Hepatic Microsomal Cytochrome P-450 Induced by Isoniazid*

Cytochrome P-450j has been purified to electropho- retic homogeneity from isoniazid-treated adult male rats; and this enzyme appears to be a major protein induced in hepatic microsomes after administration of isoniazid, as judged by sodium dodecyl sulfate-polyac- rylamide gel electrophoresis. The hemoprotein has a minimum molecular weight of approximately 5 1,500, and the ferrous-carbonyl complex of cytochrome P- 450j has a Soret maximum at 451-452 nm. The oxidized heme iron appears to be predominately in the high spin state as deduced from the Soret maximum at 395 nm. Ethylisocyanide binds to ferrous cytochrome P-450j to yield spectral maxima at approximately 458 and 430 nm with a resultant 4581430 ratio of 0.7 at 7.4. Cytochrome P-450j has no measurable catalytic activity for the metabolism of benzo[a]pyrene (3- and 9-hydroxylation), hexobarbital, testosterone, and Low, but detectable,

Isoniazid (isonicotinic hydrazide) is a primary drug used in the treatment of tuberculosis; the major pathway for metabolism of the compound in man is acetylation to acetylisoniazid followed by hydrolysis to isonicotinic acid and acetylhydrazine (1-3). Enhanced metabolism of the ether anesthetic enflurane, as reflected in high plasma levels of inorganic fluoride, has been observed in some surgical patients who had been treated chronically with isoniazid (4). Hepatic microsomes from adult male rats administered isoniazid exhibit markedly increased rates of defluorination of enflurane, methoxyflurane, sevoflurane, and isoflurane compared to metabolism by microsomes from control rats (5, 6). An induction of microsomal anesthetic defluorination has also been reported for male rats treated with acetylhydrazine or hydrazine sulfate but not isonicotinic acid (7). Rice and Talcott (5) have compared the catalytic activities of hepatic microsomes from rats treated with isoniazid, phenobarbital, or P-naphthoflavone toward a variety of substrates. The pattern of metabolism associated with isoniazid treatment did not resemble that of either phenobarbital or @naphthoflavone induction. Specifically, the administration of isoniazid to male rats resulted in increased rates of metabolism of aniline, p-nitroanisole, ethoxyresorufin, and four ether anesthetics (enflurane, methoxyflurane, isoflurane, sevoflurane) and a depression of aminopyrine demethylation relative to control animals. Although no overall increase in microsomal cytochrome P-450 content was observed, isoniazid treatment resulted in an upward shift in the Soret maximum of the CO-reduced difference spectrum to 451 nm. Based on these findings, these investigators (5) proposed that isoniazid is a "new" class of inducer unlike phenobarbital or

P-
naphthoflavone. This report describes the purification and characterization of cytochrome P-450j from isoniazid-treated adult male rats. Based on results of SDS-polyacrylamide gel electrophoresis, this hemoprotein is apparently a major protein induced in hepatic microsomes after treatment of rats with isoniazid. Several properties of cytochrome P-450j clearly distinguish this enzyme from nine other rat cytochrome P-450 isozymes (P450a-P-450i) previously purified (8,9) in this laboratory. The results of this study indicate that isoniazid may be a "unique" inducer of cytochrome P-450 in the rat.

EXPERIMENTAL PROCEDURES
Purification of Cytochrome P-450j"Forty adult male Long Evans rats (Blue Spruce Farms, Altamont, NY) a t 8 weeks of age were The abbreviations used are: SDS, sodium dodecyl sulfate; DTT, dithiothreitol.

6385
Rat Hepatic Cytochrome P-450 Isozymes administered isoniazid (Aldrich) in drinking water for 10 days. The drinking water was composed of 0.1% (w/v) isonicotinic acid hydrazide adjusted to pH 7.4, and the animals were allowed free access to Ralston Purina Rodent Chow 5001@. The rats were killed by decapitation, and hepatic microsomes were prepared as reported (10). Microsomes prepared from the livers of 40 rats contained 7-8 g of protein, and the specific content of the preparations was approximately 1.0 nmol of cytochrome P-450/mg of protein. Except for the CM-Sepharose and phosphocellulose columns, the following purification procedure was conducted a t 4 "C.
A portion of a microsomal preparation containing 2 g of protein (-2000 nmol of cytochrome P-450) was diluted to 500 ml (4 mg protein/ml) in a buffer mixture to yield the following final concentrations: 0.1 M potassium phosphate buffer (pH 7.25), 20% glycerol (v/v), 1.0 mM EDTA, and 1.0 mM DTT. Sodium cholate, which had been recrystallized from ethanol, was added to the microsomal suspension to a final concentration of 0.6% (w/v). Recrystallized sodium cholate was used in all buffers. The mixture was stirred for 20 min, divided in half, and each portion was applied to an n-octylamine-Sepharose 4B column (2.2 X 25 cm). 1,s-Diaminooctane had been coupled to cyanogen bromide-activated Sepharose 4B as previously reported (11)(12)(13). Before sample application, the columns were each washed with 200 ml of 0.1 M potassium pliosphate buffer (pH 7.25) and equilibrated with 160 ml of the same buffer containing 20% glycerol, 1.0 mM EDTA, 1.0 mM DTT, and 0.6% sodium cholate. The samples were applied, and each column was washed with 150 ml of 0.01 M potassium phosphate buffer (pH 7.25) containing 20% glycerol, 1.0 mM EDTA, 1.0 mM DTT, and 0.42% sodium cholate followed by 2 liters of the same buffer mixture containing 0.08% Emulgen 911 and 0.33% sodium cholate. Subsequently, the cytochrome P-450j fraction was eluted from the column with 1 liter of the same buffer mixture containing 0.2% Emulgen 911 and 0.33% sodium cholate. The flow rate of the column was maintained at 1.1 ml/min. Approximately 15% of the total cytochrome P-450 applied to the column was recovered in the final elution.
Based on SDS-polyacrylamide gel patterns, the fractions containing cytochrome P-450j were pooled and applied to a hydroxylapatite column (Hypatite C, Clarkson Chemical Co., Williamsport, PA) previously equilibrated with 0.01 M potassium phosphate buffer (pH 7.25) containing 20% glycerol and 0.2% Emulgen 911. The column (2.2 X 10 cm) was washed with 40 ml of equilibration buffer, 300 ml of 0.02 M buffer with the same components, and was eluted with 125 ml of 0.25 M potassium phosphate buffer (pH 7.25) containing 20% glycerol and 0.2% Emulgen 911. The recovery of total microsomal cytochrome P-450 at this step was approximately 13%. The fractions containing cytochrome P-450j were pooled based on SDS-polyacrylamide gel profiles.
The partially purified cytochrome P-450j preparation was concentrated %fold by ultrafiltration through an Amicon PM-30 membrane and dialyzed overnight against 12 liters of 0.005 M Tris-HC1 (pH 7.7 a t 4 "C) containing 20% glycerol, 0.1 mM DTT, and 1.0 mM EDTA. After dialysis, a final concentration of 1.0 mM DTT was added to the sample. The cytochrome P-450j fraction was applied to a DEAE-Sepharose CL-GB (Pharmacia Fine Chemicals) column (1.5 X 13 cm) that had been equilibrated with 0.005 M Tris-HC1 (pH 7.7 at 4 "C) containing 20% glycerol, 1.0 mM DTT, 1.0 mM EDTA, and 0.6% Emulgen 911. The column was washed with 125 ml of equilibration buffer and eluted with a linear gradient of 0-0.4 M NaCl in 500 ml of the same buffer mixture. Cytochrome P-450j eluted from the column in the first peak of A417 which contained 6-7% of total microsomal cytochrome P-450. The appropriate fractions were pooled based on SDS-polyacrylamide gel patterns.
The pooled fractions containing cytochrome P-450j were dialyzed overnight against 12 liters of 0.005 M potassium phosphate buffer (pH 6.5) containing 20% glycerol, 0.1 mM DTT, and 1.0 mM EDTA and then allowed to come to room temperature. The sample was applied to a CM-Sepharose CL-GB (Pharmacia Fine Chemicals) column (0.9 X 15 cm) equilibrated with 0.005 M potassium phosphate buffer (pH 6.5) containing 20% glycerol, 0.1 mM DTT, 1.0 mM EDTA and 0.2% Emulgen 911 a t room temperature. After the column was washed with 30 ml of equilibration buffer, cytochrome P-450j was eluted with a linear gradient (200 ml) of 0-0.3 M NaCl in the same buffer mixture. The fractions were analyzed by SDS-polyacrylamide gel electrophoresis, and the tubes containing the enzyme were pooled. The recovery of cytochrome P-450j from total microsomal cytochrome P-450 was approximately 3% at this step.
The pooled cytochrome P-450j fraction was dialyzed overnight against 6 liters of 0.005 M potassium phosphate buffer (pH 7.4) containing 20% glycerol, 0.1 mM DTT, and 1.0 mM EDTA and subsequently applied to a phosphocellulose column (2.2 X 5 cm) previously equilibrated with 0.01 M potassium phosphate buffer (pH 7.4) containing 20% glycerol, 0.1 mM DTT, 1.0 mM EDTA, and 0.2% Emulgen 911 at room temperature. After sample application, the column was washed with 75 ml of equilibration buffer, and the homogenous enzyme was eluted with a linear gradient of 0-0.4 M NaCl in 300 ml of the same buffer mixture. The overall recovery of purified cytochrome P-450j from total microsomal cytochrome P-450j was approximately 2%.
Excess detergent was removed from the enzyme preparation by calcium phosphate gel absorption. Cytochrome P-45Oj was dialyzed overnight a t 4 "C against 6 liters of 0.01 M potassium phosphate buffer (pH 7.4) containing 20% glycerol, 0.1 mM DTT, and 0.1 mM EDTA. After dialysis, calcium phosphate gel was added to the protein a t a ratio of 20 mg/l.O A a t 417 nm. The gel was washed with 20 ml of 0.015 M potassium phosphate buffer (pH 7.4) and was eluted with a small volume of 0.4 M potassium phosphate buffer (pH 7.4) containing 20% glycerol. The eluted protein was dialyzed overnight at 4 "C against 6 liters of 0.05 M potassium phosphate buffer, pH 7.4, containing 20% glycerol and 0.1 mM DTT and concentrated by ultrafiltration through an Amicon PM-30 membrane. The purified enzyme preparation was stable for several months when stored at -90 "C. The specific content of the cytochrome P-450j preparations was 11-13 nmol/mg of protein.
Purification of Other Rat Hepatic Microsomal Enzymes-Cytochromes P-450a-P-450i were purified to apparent homogeneity as previously reported (8,9). NADPH-cytochrome c reductase was purified to a specific activity of 35,000-40,000 units/mg protein by a combination of the methods of Dignam and Strobe1 (14) and Yasukochi and Masters (15). Enzyme activity was assayed according to the procedure of Phillips and Langdon (16). One unit of reductase is defined as that amount catalyzing the reduction of 1 nmol of cytochrome c/min at 22 "C. Electrophoretically homogeneous rat hepatic microsomal epoxide hydrolase was isolated as previously described (17).
Other Methods-Protein was determined by the method of Lowry et al. (18) with crystalline bovine serum albumin as standard. Cytochrome P-450 content was calculated according to the method of Omura and Sat0 (19) from the CO-reduced difference spectrum based on an extinction coefficient of 91 mM" cm".
The binding of ethylisocyanide (2.0 mM final concentration) to ferrous cytochrome P-450 was measured as previously reported (20). The hemoprotein concentrations and buffer components used for spectral determinations are detailed in the appropriate figure legend.
SDS-polyacrylamide gel electrophoresis was performed according to the method of Laemmli (21) in a separating gel (7.5% acrylamide) 0.75-mm thick and 10-cm long. The gels were stained with Coomassie Blue R-250 and destained as described (17). Limited proteolytic digestion of the 10 purified hemoproteins in the presence of SDS was conducted as described by Cleveland et al. (22). The cytochromes P-450 were incubated with the protease for 10 min a t 37 "C a t protein ratios described in the figure legend. The proteolytic digests of the cytochromes P-450 were subjected to SDS-polyacrylamide gel electrophoresis (21) in gels containing 12.5% acrylamide that were 1.5-mm thick and 10-cm long. The amino acid composition of cytochrome P-450j was analyzed by a previously described protocol (23). The NHzterminal sequence of the acetone-precipitated protein was determined (23) by manual Edman degradation.
Catalytic activity of cytochrome P-450j was assayed under conditions in which metabolism was proportional to hemoprotein concentration and time of incubation. In each experiment, other purified cytochrome P-450 isozymes with known activity were included for reference. Saturating amounts of NADPH-cytochrome c reductase and optimal dilauroylphosphatidylcholine were used in all experiments. Dilauroylphosphatidylcholine was prepared in water and sonicated immediately before use. The following references contain the methods for assays of the various substrates: [N-methyl-"C]benzphetamine (27)

RESULTS
Purification of Cytochrome P-4.50;-Cytochrome P-450j was purified from hepatic microsomes of isoniazid-treated adult male rats hy chromatography on n-octylamine-Sepharose 4R, hydroxylapatite, DEAE-Sepharose, CM-Sepharose, and phosphocellulose, as detailed under "Experimental Procedures," with a yield of approximately 2% of total microsomal cytochrome P-450. The initial n-octylamine-Sepharose 4R column was based on the original approach of Imai and Sato (41) for t,he purification of rabbit liver c-ytochrome P-450 as modified by Guengerich et al. (13,42) for rat liver enzymes but utilized buffer mixtures that were most effective for the purification of cytochrome P-450j. A dramatic purification of cytochrome P-450j is achieved by chromatography on n-octylamine-Sepharose 4R; based on SDS-gel patterns, only 4-5 major contaminating proteins co-elute with this enzyme. The subsequent hydroxylapatite column effects a removal of two major contaminants from the partially purified cytochrome P-45Oj preparation. The t.hird step in the procedure, the DEAE-Sepharose column, was the most difficult to develop since cytochrome P-45O.i does not bind well to most anionic exchange resins under several conditions. Optimal recovery and purification of the hemoprotein are achieved in the presence of a high concentration of Emulgen 911 (0.6%). If this step is eliminated from the procedure, the CM-Sepharose column is ineffective at removing certain remaining contaminants. As judged by SDS-polyacrylamide gel patterns, these contaminating proteins are removed from the cytochrome P-45Oj preparation by chromatography on CM-Sepharose if preceded hy the DEAE-Sepharose column. Residual trace contaminant,s are eliminated from the c-ytochrome P-450j preparation hy chromatography on phosphocellulose. Fig. 1 shows an SDS-polyacrylamide slab gel (0.75-mm thick) of the purified hemoprotein that illustrates the apparent, homogeneity of the enzyme preparation even at high prot,ein concentrations. Only one protein-staining band corresponding to cytochrome P-450j is observed; however, a t high concent.rations, a very trace contaminant of low minimum M , is detected in some enzyme preparations. The minimum M , of cytochrome P-450j was determined to he 51,500 in the gel syst.em of Laemmli (21). based on the electrophoretic mobility of the protein relative to cytochromes P-450a-P45Oi, epoxide hydrolase, and NADPH-cytochrome c reductase as shown in Fig. 1. The distinct mobililty of cytochrome P-450j is apparent when a comparison is made between the mobilities of the other purified microsomal enzymes in the first well (labeled M I X ) , and c-ytochrome P-450j. Therefore, cytochrome P-450j is distinct from the nine other purified rat hepatic microsomal hemoproteins (cytochromes P-450a-P450i) previously purified in our laboratory (8, 9) based on minimum Mr. As shown in the last well (labeled M I X ) , however, when c-ytochrome P-45Oj is included in the mixture of proteins, cytochromes P-450h ( M , = 52,000), P-45Oj, and P-450f/P-450h ( M , = 51,000) appear as a single broad proteinstaining hand. In contrast to the report of Guengerich et al. (42), t.he relative mobilities of the 10 cytochromes P-450 relative to each other were unaffected by the substitution of lithium dodecyl sulfate for SDS in this gel system (data not shown). Results of preliminary experiments (data not shown) suggest that cytochrome P-450j, like cytochromes P-450c and P-450d (43), tends to streak in two-dimensional isoelectric focusing-SDS gel electrophoresis. Fig. 2 shows the electrophoretic profiles of purified cyto-

REO--
MIX chrome P-45Oj and hepatic microsomes from untreated and isoniazid-treated adult male rats in two SDS-polyacrylamide slah gels. In the experiment on the h f t , electrophoresis was stopped when the tracking dye reached the bottom of the gel; in the experiment on the right, electrophoresis was continued for an additional 30 min. The samples and all other conditions were identical for the two experiments. As illustrated on the left, the electrophoretic patterns of hepatic microsomes from untreated and isoniazid-treated rats are indistinguishahle except for a marked broadening of a protein-staining band at -51,000 in the profile from the treated animals. 'I'he upper region of this broadened hand has the same electrophoretic mohility as purified c-ytochrome P-4.50j. N'hen electrophoresis was continued for a longer period of time, as shown on the right, the broadened protein-staining hand can he resolved into two hands. Purified c-ytochrome P-45Oj co-migrates with the higher M , hand. Hepatic microsomes from untreated rats appear to have only a trace amount of the higher M , protein as evidenced by the lack of a distinct protein-staining hand of that mohility. Within the inherent limitations to the interpretation of SDS-polyacrylamide gel profiles of hepatic microsomes, these observations suggest that isoniazid is a selec-  2. SDS-polyacrylamide slab gels of rat hepatic microsomes and cytochrome P-45Oj. The purified enzyme was applied to each of the wells laheled "j" at 0.3 pg. Hepatic microsomes ( 6 pg) from untreated and isoniazid-treated adult male rats were applied to the wells designated U N and ISN, respectively. The protocol for hoth slah gels shown was identical except electrophoresis was stopped when the tracking dye (bromphenol blue) reached the hottom of the gel in the experiment on the k f t , whereas electrophoresis was continued for an additional 30 min in the experiment on the rizht. + WAVELENGTH ( n m ) total cytochrome P-450 content occurs when rats are treated with isoniazid. Therefore, one might predict that an induction of cytochrome P-450j occurs with a concomitant decrease in another form(s) of cytochrome P-450. However, no marked decrease in any protein-stained hand in the electrophoretic pattern of microsomes form isoniazid-treated rats relative to the protein profile of microsomes from untreated animals is observed in the M , region of c-ytochrome P-450 (45,000).
Spectral Propartips- Fig. 3 illustrates the CO-and ethvlisocyanide-difference spectra of ferrous cytochrome P-45Oj ( A ) as well as the ahsolute spectral characteristics of the purified hemoprotein ( R ) . The protein was diluted in a buffer mixture containing Emulgen 911 and sodium cholate for all spectral measurements, since detergents have a protective influence on the enzyme. The CO-reduced difference spectral maximum of the enzyme is a t approximately 451-452 nm, and the ahsence of a shoulder a t 420 nm indicates that the c.ytochrome P-450j preparation does not contain a significant amount of cytochrome P-420 (Fig. 3 A ) . Rice and Talcott ( 5 ) had reported that isoniazid treatment of rats resulted in an upward shift in the hepatic microsomal CO-reduced difference spectral maximum t.o 451 from 450 nm observed in the spectrum from untreated animals. This shift is most likely the consequence, a t least in part, of the induction of cytochrome P-450j by isoniazid. Of the other rat cytochromes P-450 previously purified in our laboratory (8,9). the Soret maximum of the ferrous carhonyl complex of five show a shift downward to 447-449 nm (cytochromes P-450c, P-450d. P-450f, P-450g. P-450i), one is at 450 nm (cytochrome P-45oh), and three are shifted upward to 451-452 nm (cytochromes 1'-45Oa, P-450e.
Two spectral maxima are generated at 458 and 430 nm when the ethylisocyanide difference spectrum of ferrous c-ytochrome P-450j is recorded a t pH 7.4, as shown in Fig. 3A

Rat
430 peak ratio at pH 7.4 was calculated to be 0.7 which is similar to the ratio observed in the ethylisocyanide difference spectrum of ferrous cytochrome P-450b (8). The spectrum of ferrous cytochrome P-450j bound with ethylisocyanide is distinct from the corresponding spectra of cytochromes P-450a-P-450i in that an absorption peak is observed at 458 nm. A Soret maximum a t 452-455 nm is observed in the ethylisocyanide difference spectra of the nine other hemoproteins. A peak or shoulder a t 430 nm is characteristic of the spectrum of all of these hemoproteins. The ethylisocyanide difference spectral characteristics of cytochromes P-450a-P-450i have been previously described (8,9). Unlike cytochromes P-450b, P-450e, and P-450i (8, 9), the ligand metyrapone does not bind to ferrous cytochrome P-450j (data not shown). The absolute spectral properties of cytochrome P-450j are shown in Fig. 3B. Cytochrome P-450j is primarily a high spin ferric hemoprotein as indicated by the absorption maximum a t 395 nm in the absolute oxidized spectrum of the enzyme ( E = 90 mM" cm"). The extinction coefficient and absorption maximum (395 nm) of oxidized cytochrome P-450j is not influenced by the addition of detergents or glycerol or affected by protein concentration from 0.5-6.0 pM. Cytochrome P-450d is a high spin ferric hemoprotein, cytochrome P-450f is predominately high spin, and cytochrome P-450e contains a relatively minor high spin component, whereas cytochromes P-450a, P-450b, P-450c, P-450g, P-450h and P-450i are low spin ferric hemoproteins as determined by their absolute oxidized spectral characteristics (8,9,17). When the hemoprotein is reduced with sodium dithionite, the Soret maximum of cytochrome P-450j shifts downward to approximately 412 nm. The CO-reduced absolute spectral maximum of cytochrome P-450j is a t 452 nm. E'[(:. 1. Ouchterlong double diffusion analysis. Antit~odies to cytochromes P-450a-1'45Of were prepared and purified as previously reported (24-26).' The components of the immunodiffusion plate have been described elsewhere ( 2 5 ) . Purified cytochromes P-450j (center well), P-450b and P-45Oc (bottom well) were each present at 4 PM. Antibodies to cytochromes P-450a-P-450f at 25 mg/ml were applied to the peripheral wells, as indicated.
Immunochemical Reactiuities-Ouchterlony double diffusion analysis was used to detect potential reactivity of cytochrome P-450j with antibodies prepared against cytochromes P-450a-P-450f with the results shown in Fig. 4. There is no detectable recognition of cytochrome P-450j by these antibodies. Although the Ouchterlony plate shown in Fig. 4 contained 0.2% Emulgen, the same results were obtained in the absence of detergent (data not shown). Although antibody to cytochrome P-450e was not included in the immunodiffusion plate, both cytochromes P-450b and P-450e are immunologically indistinguishable (8) when tested against polyclonal antibody to cytochrome P-450b. Therefore, cytochrome P-450j is immunochemically distinct from cytochromes P-450a-P-450f previously purified in our laboratory. Antibody prepared against cytochrome P-450f cross-reacts strongly with the heterologous proteins, cytochromes P-450g, P-450h, and P-45Oi.' Since cytochrome P-450j was not recognized by antibody to cytochrome P-450f, cytochrome P-450j also differs immunochemically from cytochromes P-450g, P-450h, and P-450i, Preliminary results also indicated that cytochrome P-450j does not react with several monoclonal antibodies (44) to cytochrome P-45Oc in an enzyme-linked immunosorbent assay.
Peptide Mapping- Fig. 5 shows the comparative SDS-polyacrylamide gel profiles of cytochromes P-450a-P-450j after limited proteolysis in the presence of SDS by the method of Cleveland et al. (22). The peptides of each hemoprotein generated by cleavage with chymotrypsin ( Fig. 5A) or Staphylococcus aureus V8 protease (Fig. 5B) under identical conditions were analyzed electrophoretically in a SDS-polyacrylamide slab gel (1.5-mm thick) that was composed of 12.5% acrylamide. The wells are labeled by the subscript of the cytochrome P-450, and each protease alone was applied to the gel as indicated. As illustrated in Fig. 5A, cytochrome P-450j is relatively resistant to digestion by chymotrypsin compared to most of the other isozymes. The peptide map of cytochrome P-450j contains two major peptides of relatively large minimum M , as well as the undigested hemoprotein a t a M , of 51,500. The chymotryptic digest of cytochrome P-450j is markedly different from the peptides generated by cleavage of cytochromes P-450a-P-450i. Fig. 5B shows the gel patterns of peptides of cytochromes P-450a-P-450j after digestion with S. aureus V8 protease which cleaves adjacent to aspartic and glutamic acid residues. The protease cleaves cytochrome P-450j a t several sites yielding a peptide map clearly distinguishable from the maps of cytochromes P-450a-P-450i. The results presented in Fig. 5 indicate that cytochrome P-450j differs structurally from cytochromes P-450a-P-450i.
Amino Acid Composition and Amino-terminal Sequence-The amino acid content of cytochrome P-450j was analyzed with the results shown in Table I. The composition of the hemoprotein is similar to those of cytochromes P-450a-P-450i (23) since all of these enzymes are composed of 40-50% hydrophobic residues (Pro, Ala, Val, Met, Ile, Leu, Phe) although significant differences in content among the cytochromes are also apparent. Cytochrome P-450j contains five Cys residues/molecule which is fewer than any of the other proteins except cytochrome P-450a (23).
The NH,-terminal sequence of cytochrome P-450j, determined by Edman degradation for the first 19 residues, is also listed in Table I. The purity of the hemoprotein preparation was confirmed by the presence of a single NH2-terminal sequence. The NH2-terminal residue of cytochrome P-450j is Ala, and the enzyme possesses a hydrophobic leader sequence; 15 of the first 19 amino acid residues are hydrophobic. Very little, if any, homology is observed in the NH2-terminal se-   I  TABLE I1 Amino acid composition and amino-terminal sequence of cytochrome Catalytic activity of purified rat hepatic cytochrome P-450j P-450;

Rat Hepatic Cytochrome P-450 Isozymes
The catalytic activity of purified cytochrome P-450j was deter- The amino acid composition of cytochrome P-450j was determined by duplicate analyses and expressed based on an approximate M, of 52,000. Trp content was not determined (ND).
The NH2-terminal sequence of cytochrome P-450j was analyzed in duplicate by the manual Edman method on the acetone-precipitated enzyme (23). The initial yields were 30-5056, and the numbers in parentheses are the recoveries (pmol) of each amino acid residue. quence of cytochrome P-450j compared to cytochromes P-450a-P-450i previously purified in our laboratory (23) which establishes the isozymic nature of this hemoprotein. Based on available NHz-terminal sequence data, cytochrome P-450j also differs markedly from purified rat liver microsomal cytochromes P-450 PB-1 (45), PB-2 (46), RLM5 (47) -450 (40, 48) show little or no homology with the NH2 terminus of cytochrome P-45Oj; however, 13 of the first 19 residues of cytochrome P-450j are identical to rabbit liver cytochrome P-450 LM-3a (40). Catalytic Actiuity- Table 11 lists the catalytic activity of cytochrome P-450j toward several substrates and various types of reactions. The hemoprotein has no measurable activity toward the metabolism of hexobarbital or benzo[a]pyrene nor does it participate in the hydroxylation of the steroids, testosterone, or 5a-androstane-3a,l7~-diol-3,17-disulfate. Low but measurable metabolism of benzphetamine, zoxazolamine, 7-ethoxycoumarin, estradiol-17P (2-hydroxylation), and p-nitroanisole is catalyzed by cytochrome P-45Oj. The

Influences of several effectors on p-hydroxylation of aniline catalyzed by purified cytochromes P-450j and P-450d
Cytochrome P-450 (0.10 nmol), NADPH-cytochrome c reductase (1200 units), and dilauroylphosphatidylcholine (15 pg) were mixed, incubated for 2 min at 37 "C, and returned to ice. The other assay components were then added (0.1 M potassium phosphate buffer, pH 7.4, 2.0 mM aniline, and various effectors), the reaction was initiated with the addition of 1.0 mM NADPH, and the samples were incubated for 10 min at 37 "C. Product formation was measured as previously described (39). The turnover numbers (nmol p-aminophenollminl nmol cytochrome P-450) for cytochromes P-450j and P-450d in the absence of effectors are listed in parentheses. addition of cytochrome bs to the reconstituted system at 0.8 nmol/nmol cytochrome P-450j has no effect on the O-demethylation of p-nitroanisole (data not shown). In contrast to the results with the other substrates, cytochrome P-450j effectively catalyzes p-hydroxylation of aniline with a turnover of 12.7 nmol/min/nmol cytochrome P-450. Of the isozymes purified in our laboratory, cytochrome P-450j has the highest catalytic activity toward the metabolism of aniline. Cytochrome P-450d is also a relatively effective catalyst of aniline p-hydroxylation (9.6 nmol/min/nmol cytochrome P-450); cytochromes P-450b, P-450c, and P-450h have low but measurable activity (1.0-2.0 nmol/min/nmol cytochrome P-450), whereas cytochromes P-450a, P-450e, P-450f, P-450g, and P-450i are very poor catalysts of this reaction (c0.5 nmol/ min/nmol cytochrome P-450). Therefore, the substrate selectivity of cytochrome P-450j is clearly distinct from the enzymatic activities of cytochromes P-450a-P-450i (8,9,38).

Addition
One catalytic parameter that was used by Rice and Talcott (5) to determine if isoniazid is a "unique" inducer of rat hepatic microsomal cytochrome P-450 was aniline p-hydroxylation. A 3.5-fold induction of aniline hydroxylation, expressed per nmol of cytochrome P-450, was observed in microsomes from isoniazid-treated rats relative to control values. No increase (per nmol of cytochrome P-450) was detected after phenobarbital or /3-naphthoflavone treatment of rats. The marked increase in microsomal metabolism of aniline after treatment of animals with isoniazid may be the consequence, at least in part, of the induction of cytochrome P-450j by this compound.
Influences of Effectors onp-Hydroxylation of Aniline- Table  I11 summarizes the effects of hydroxyl radical scavengers (mannitol, dimethyl sulfoxide), iron chelators (EDTA, desferrioxamine), superoxide dismutase and catalase, as well as Fe-EDTA on p-hydroxylation of aniline catalyzed by cytochromes P-450j and P-450d. The mechanism of this reaction was studied using cytochromes P-450j and P-450d which have the highest catalytic activities for this substrate of the rat hepatic isozymes purified in our laboratory. The results listed in Table I11 show that the rate of p-hydroxylation of aniline catalyzed by cytochromes P-450j or P-450d is unaffected by the addition of any of the effectors tested. The maximum effect of any addition is approximately 20% inhibition observed when Fe-EDTA (0.05 mM) is added to an incubation containing cytochrome P-450j as the catalyst. The results presented in Table I11 contrast with the proposed mechanism of aniline p-hydroxylation reported by Ingelman-Sundberg and Ekstrom (49).
In the study of Ingelman-Sundberg and Ekstrom (491, aniline p-hydroxylation catalyzed by purified rabbit liver cytochrome P-450 LM2 was reported to be mediated by hydroxyl radicals that were generated by an iron-catalyzed Haber-Weiss reaction between superoxide anions and hydrogen peroxide. The proposed mechanism was derived from results showing a marked inhibition of aniline hydroxylation in the presence of catalase, superoxide dismutase, and several hydroxyl radical scavengers. If the formation of hydroxyl radicals occurred via the Haber-Weiss scheme, the addition of Fe-EDTA would be expected to enhance the reaction, and desferrioxamine should inhibit hydroxylation. As outlined in Table 111, however, p-hydroxylation of aniline catalyzed by cytochrome P-450j or P-450d is not affected to any significant extent by any of the potential effectors tested. The reason for this discrepancy is unknown but may be related to the specific cytochrome P-450 used as the catalyst.

DISCUSSION
Cytochrome P-450j has been purified from adult male rats and has been shown to be a distinct isozyme from nine other cytochromes P-450 (P-450a-P450i) previously purified in our laboratory (8, 9) based on amino-terminal sequence analysis. Several biochemical properties of cytochrome P-450j are different from the other purified hemoproteins. Additionally, cytochrome P-450j is clearly distinguishable from several other rat hepatic microsomal cytochromes P-450 purified from untreated or induced animals by other investigators based on various characteristics of the purified enzyme. The NHz-terminal sequence of cytochrome P-450j is different from the sequences of rat liver cytochromes P-450 PB-1 (45), PB-2 (46), RLM5 (47), and RLM3 (47). Of the hemoproteins purified by 50), cytochrome P-450 BNF/ISF-G, which corresponds to cytochrome P-450d, has the highest catalytic activity toward p-hydroxylation of aniline and is the only hemoprotein exhibiting high spin characteristics in its absolute ferric spectrum. Therefore, cytochrome P-450j probably does not correspond to any of the hemoproteins purified in that laboratory. A composite of the spectral, electrophoretic, and catalytic properties of forms 1-5 isolated from P-naphthoflavone-treated rats (51), cytochrome P-452 from clofibrate-treated rats (52), and PCN cytochrome P-450 from pregnenolone 16a-carbonitriletreated rats (53) indicate that none of these hemoproteins correspond to cytochrome P-45Oj. Nor do the properties of the rat liver microsomal cytochromes P-450 purified by Waxman and co-workers (45,54,55) resemble cytochrome P-45Oj. Cytochrome P-450 PB-1 (45) has a different NHz-terminal sequence than cytochrome P-450j, whereas cytochromes P-450 P B -~c , PB-3, PB-4, and PB-5 (54,55) correspond to cytochromes P-450h, P-450a, P-450b, and P-450e, respectively. These results support the original proposal of Rice and Talcott ( 5 ) that isoniazid in a "unique" inducer of cytochrome P-450 in the rat, unlike phenobarbital or P-naphthoflavone.
Over the past several years, our laboratory has been studying the regulation of rat liver cytochromes P-450 by immunochemical quantitation of levels of certain isozymes in microsomes from untreated rats as well as animals induced by numerous compounds (24,25,56). That structurally diverse compounds can induce the same cytochrome P-450 isozymes has become increasingly apparent. Interestingly, certain characteristics of microsomal preparations from rats treated with Rat Hepatic Cytochrome ethanol, pyrazole, and acetone as well as diabetic or fasted animals suggest that cytochrome P-450j may be induced by these treatments. Microsomal electrophoretic patterns indicate that ethanol treatment of rats results in the appearance of a protein-staining band (M, = 51,500) of the same mobility as cytochrome P-450j (9,57,58). Several laboratories have shown hepatic microsomes from male rats treated with ethanol, pyrazole, and acetone as well as diabetic (genetic, alloxan-, or streptozotocin-induced) animals or fasted rats exhibit similar SDS-polyacrylamide gel electrophoretic profiles with an increase in a protein-staining band with a M, similar to cytochrome P-45Oj (57)(58)(59)(60)(61)(62)(63)(64). Yang and co-workers (59,63,64) have reported a correlation between induction of a high affinity NADPH-dependent nitrosodimethylamine demethylase and the appearance of this protein-stained band in SDS-gel profiles. Diabetic rats and animals treated with certain of these inducers have also been reported to exhibit an enhanced ability to p-hydroxylate aniline relative to control animals (60,62,65,66). Additionally, defluorination of ether anesthetics is markedly induced in hepatic microsomes from rats treated with either ethanol or isoniazid (5, 67-69). Interestingly, a common metabolic consequence of treatment of animals with certain of these compounds is increased levels of ketone bodies.
In the rabbit, cytochrome P-450 LM3a, an isozyme with high catalytic activity toward aniline p-hydroxylation as well as the oxidation of alcohols (40), is apparently inducible by ethanol, isoniazid, imidazole, pyrazole, trichloroethylene, and rn-xylene as determined by an increase in microsomal aniline hydroxylation following treatment with these agents (70). In fact, Koop and Coon (71) have purified and characterized cytochrome P-450 LM3a from imidazole-treated rabbits and have shown that isozyme to be identical to the corresponding hemoprotein from ethanol-treated animals. The results of Ingelman-Sundberg and Johansson (72) suggest that benzene may also induce the same cytochrome(s) P-450 (LMeb) as ethanol in rabbit liver. A comparison of the NHp-terminal sequence of cytochrome P-450j (Table I) with the sequence reported for rabbit liver cytochrome P-450 LM3a (40) reveals that 13 of the first 19 amino acid residues of cytochrome P-450j are identical to cytochrome P-450 LM3a. The peptide maps of cytochrome P-45Oj and P-450 LM3a (40), especially following chymotrypsin cleavage, also show marked similarities. Both the CO-reduced difference and absolute oxidized spectral characteristics of the two isozymes are very similar. Furthermore, both enzymes catalyze p-hydroxylation of aniline with equal efficiency. Therefore, it is tempting to speculate that cytochrome P-450j may be inducible in rat liver by at least some of the compounds that induce cytochrome P-450 LM3a in rabbit liver. Studies are currently in progress to verify this proposal.