Amino Acid Sequence of an Analogous Peptide from Two Forms of Cytochrome P-450*

Two cytochrome P-450 preparations, a constitutive isozyme, form 3b, and a phenobarbital-induced iso- zyme, form 2, were isolated from rabbit liver micro- somes and compared by peptide mapping following digestion with trypsin and by partial sequence analysis. The NHz-terminal sequence of form 3b differed from form 2 in 15 out of 18 amino acids, but both forms have an NHz-terminal methionine residue followed by an acidic residue. Comparisons many the tryptic peptides of the two forms high pressure acid peptides from these forms, with one exception, are different. A tridecapeptide, differing only in a methionine (form 3b)/isoleucine (form 2) replacement isolated from both forms. The amino acid sequence of this peptide is as follows: The observation that these two cytochromes share an analogous peptide suggests that this tridecapeptide may contribute structural information necessary for common functional properties.

responsible for the metabolism of a wide variety of endogenous and exogenous substrates, and the various forms of P-450 show overlapping specificity wit.h regard to substrate metabolism (cf. review in Ref. 1). The existence of spectral and catalytic similarities between the different forms of cytochrome P-450 would be most readily explained by regions of structural homology. However, aU structural comparisons to * 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 United States Public Health Service Grant GM 26351. To whom correspondence should be addressed. date have suggested a complete lack of homology between the different cytochrome P-450 forms, e.g. they display different peptide maps following limited proteolytic digestion (2-6) as well as distinct NHZ-terminal amino acid sequences (7,8).
In this communication, we report the isolation and primary structure of an analogous peptide from rabbit microsomal cytochrome P-450 form 2, the principal form of the cytochrome inducible by phenobarbital in the rabbit liver (9, 10) as well as a major constitutive form in pulmonary microsomes (5,6) and from form 3b, a major constitutive form in rabbit liver (11, 12).

EXPERIMENTAL PROCEDURES
The manner of preparation and characterization of the cytochromes, form 2 (13) and form 3b' ( l l ) , are given in the references indicated. The cytochromes used in this study had the following specific contents expressed as nanomoles/mg of protein: form 2, 18 Prior to trypsin digestion or NH2-terminal amino acid sequence analysis, cytochrome P-450 samples were dialyzed against 0.2 M NH4acetate, pH 7.4, and lyophilized. Trypsin digestion of cytochromes P-450 was performed as previously described (14). The lyophilized proteins were dissolved in 8 M urea, 2 mM NH4HC03 buffer, pH 8.1, and the urea concentration was reduced to 2 M by the addition of water. Trypsin (~-l-tosylamido-2-phenylethylchloromethyl ketonetreated, Worthington) was added as a 2 mg/ml solution in 1.2 mM HC1 to give a 1:14 molar ratio of trypsin to cytochrome, and digestion was allowed to proceed 12 to 16 h at room temperature.
Reverse phase chromatography was performed using a pBondapack C-18 column (0.4 X 30 cm, Waters Associates). Alkyl-phenyl columns which were used to identify phenylthiohydantoin derivatives of amino acids were also obtained from Waters Associates. Resolution of peptide mixtures was achieved by gradient elution from the octadecyl alkyl silane (C-18) column using 20 m~ potassium phosphate, pH 2.2, as the aqueous solvent and acetonitrile (Burdick and Jackson) as the organic solvent (15). The additional details concerning the conditions for the separation of the peptides are described in Fig. 1. Peptides isolated in phosphate buffers were desalted on a Bio-Gel P-2 column (0.9 X 12 cm) equilibrated with 50% aqueous acetic acid prior to sequence analysis. In some cases, peptides were purified further by rechromatography on high performance liquid chromatography using IO mM ammonium acetate, pH 6.0, as the aqueous buffer. Peptides isolated in ammonium acetate buffers were lyophilized and applied to the sequenator cup without desalting. Lyophilized proteins or peptides were transferred to the cup with 0.5 ml of 88% formic acid. The presence of ammonium salts did not interfere with the determination of amino acid composition or sequence analysis at the concentrations used.
The amino acid compositions of the peptides and proteins were determined on acid hydrolysates of the samples using an updated single-column Beckman 121 automatic amino acid analyzer. Tryptophan content was estimated fluorometrically as described by Pajot (16), and cysteine residues were quantitated as cysteic acid following performic acid oxidation of the cytochrome. Automated sequence analyses of proteins and peptides were performed on a Beckman 890 C sequenator, using dimethyldylamine peptide Program No.
102074, or 0.1 M Quadrol Program No. 030176, in the presence of 3 mg of Polybrene which together with 100 nmol of glycyl glycyl glycine, was subjected to 3 precycles of Edman degradation (17). Conversion of the thiazolinone products from the 0.1 M Quadrol program was carried out in-line with a Sequamat P-6 autoconverter using 1.5 N acetylchloride in methanol (18). Reagents for sequencing were ob- and 2 (bottom) using reverse phase chromatography. In each instance, the peptides were eluted with a gradent of 0-70% acetonitrile in 20 mM potassium phosphate, pH 2.2, over a 90-min period. The flow rate tained from Beckman. Phenylthiohydantoin derivatives of amino acids were identified by reverse phase chromatography on a C-18 column using a 5-45% linear gradient of 0.1% aqueous acetic acid/ methanol. High performance liquid chromatography on an alkylphenyl column using a 1844% linear gradient of 0.06% aqueous propionic acid/methanol was used to identify those phenylthiohydantoin derivatives not resolved on the '2-18 column. Seryl and threonyl phenylthiohydantoins were identified by amino acid analysis after back conversion to alanine or a-aminobutyric acid by hydrolysis in hydroiodic acid vapor ( 14).

RESULTS A N D DISCUSSION
It is now generally accepted that there are multiple forms of microsomal cytochrome P-450. The multiplicity of the purified cytochromes is plainly evidenced by their discrete mobilities on polyacrylamide gel electrophoresis in the presence of sodium dodecyl sulfate, specific immunoprecipitation, analysis of their terminal amino acid sequences, and peptide mapping on polyacrylamide gels following limited proteolysis in the presence of sodium dodecyl sulfate (cL review in Ref. These contrasting properties of the electrophoretically distinct forms of cytochrome P-450 could conceivably arise as the result of post-translational modifications of a single gene product or through the expression of specific genes coding for the individual cytochromes. Evidence supporting the latter hypothesis is provided by the translation of heterologous liver mRNA in uztro. In experiments conducted in several laboratories (19-22), antibodies to a phenobarbital-inducible form of rat cytochrome P-450 precipitate nascent protein with a similar mobility to the purified cytochrome, indicating that this cytochrome is not derived from a higher molecular weight precursor and that the monospecific antibody recognizes the nascent protein chain. T h i s was corroborated in one case by NHz-terminal amino acid sequence analysis of the protein synthesized in vitro (19). On the other hand, it has been reported that antibody to a form of rat cytochrome P-450 which is inducible by 3-methylcholanthrene precipitates a higher molecular weight precursor for this cytochrome (20). Thus, it may be somewhat premature to infer from a limited number of studies that post-translational processing does not occur for the microsomal cytochromes such as that seen for mitochondrial forms of cytochrome P-450 (23).
Peptide mapping provides additional evidence that the cytochromes represent distinct gene products. The cytochromes are clearly distinguished by one-dimensional maps employing polyacrylamide gel electrophoresis in the presence of sodium dodecyl sulfate following limited proteolysis (2)(3)(4)(5)(6). However, the extent of the differences in amino acid sequence which exist between forms cannot be gauged from peptide mapping studies of this nature.
In this report, we describe the application of reverse phase chromatography to the resolution and isolation of tryptic peptides derived from each of two highly purified forms of rabbit microsomal cytochrome P-450, forms 2 and 3b. Comparison of chromatograms of tryptic peptides from both fonns showed extensive differences in the retention times of the individual peptides as shown in Fig. 1. Amino acid analysis followed by sequence analysis of all major peaks again indicated a lack of structural similarity. This is illustrated in Table I which displays the fist three amino acid residues of the NHz-terminal sequences of several major tryptic peptides isolated from form 3b with peptides of similar mobility from form 2. However, peptides 19A of form 3b and 23C of form 2, despite different retention times, are very similar in structure as indicated in Table 11. The yields of these two peptides were both greater than 60%. The amino acid sequences of these peptides are homologous except for a methionyl residue at position 7 of the peptide from form 3b and an isoleucyl residue at that position for form 2. Although this lack of homology may contribute to the differences in retention times observed for the two peptides, the oxidation states of the methionyl residues could also affect the retention times of these peptides. In the example shown in Fig. 1, the NHz-terminal methionine of peptide 19A derived from form 3b occurred as the sulfoxide form. However, this was not observed in all cases examined.
The NHz-terminal amino acid sequences of forms 2 and 3b as obtained by automated sequence analysis of the intact cytochrome are also compared in Table 11. There was no significant homology between the NHz terminus of form 3b and the NHz-terminal sequence reported by Haugen et al. for form 2 (8) or by Botelho et al. (7) for three forms of rat cytochrome P-450. The one common feature is that both forms 2 and 3b have methionyl residues at the NHz terminus of the molecule followed by an acidic residue and a segment of hydrophobic residues.
Interestingly, some variation in the end-group was noted with form 3b in as much as one preparation completely lacked the NHz-terminal methionine. This preparation was isolated from microsomes of a pool of three rabbits in the usual manner and yield. The lack of NHz-terminal methionyl in this preparation was unexpected. Methionine serves as the end-group for form 2 (8) as well as for rat cytochrome P-450 form a ( 7 ) . It has been pointed out that there is significant homology between NHz-terminal sequences of rat form b and rabbit form 2 provided that the NH2 terminus of the rat form b is aligned with the second residue of rabbit form 2 ( 7 ) . Both of these cytochromes are induced by phenobarbital in the livers of their respective species and share similar enzymatic properties. In light of the apparent lability of the NH-terminal methionine of form 3b, a similar loss of methionine may have occurred with rat form b.
The amino acid composition of form 3b and three analyses reported for form 2 are shown in Table 111. The composition of form 3b is similar to but not identical with that reported for form 2 by Haugen and Coon (24), Imai et al. (25), and Wolf et al. (5). Differences among the analyses for the two forms occur in several amino acids, most notably lysine, argi-  Glu-Gly-Glua Each peptide is numbered according to Fig. 1. The peptide was collected and purified further by rechromatography at pH 6.0 as described under "Experimental Procedures." When more than one peptide is resolved using this procedure, they are denoted by an additional letter designation.
' Methionine sulfoxide was observed in some cases.

Comparison of amino acid sequences of analogous peptides and NH2-terminal segments of cytochromes P-450
Amino Some 20-30 nmol of protein in 0.5 ml of 88% formic acid were applied to the spinning cup. The initial yields of forms 3b, 3b desMet, and 2 were between 70 and 75%. and the repetitive yields ranged from 92-96%.

Analogous Peptides from
Two Forms of Cytochrome P-450 nine, and cysteine. However, on the whole the compositions are rather similar. Analysis of amino sugars following mild hydrolysis indicated less than 0.1 residue/mol of form 3b. In contrast, form 2 is reported to contain 1 residue of glucosamine/mol of protein (24).
It is surprising that more similarities in sequence were not observed for forms 2 and 3b, since these heme proteins share several spectral and enzymatic properkies, and all interact with the same flavoprotein reductase during catalysis (1). These particular properties may ultimately depend on the conformation of the protein chain, the location of only a few critical amino acids, or the conservative substitution of other amino acids. This will be more evident when the primary structures of these proteins have been determined. Such studies are currently being pursued in our laboratories for form 3b and in the laboratory of Dr. M. J. Coon in Ann Arbor for form 2. Thus, additional comparisons between forms 2 and 3b can be expected to contribute to our future understanding of the microsomal cytochrome P-450 proteins.
The peptide mapping procedure described in this report provides further evidence that the multiple forms of cytochrome P-450 represent distinct gene products. The analysis of tryptic digests using reverse phase chromatography and comparisons of the amino acid sequences of selected peptides suggests that these two cytochromes differ extensively in primary structure. Of the peptides compared to date, representing about 70% of the structure (data not shown), only one analogous peptide has been detected as reported here.
The observation of a limited region of marked structural homology in two cytochrome P-450 forms which are by other criteria structurally dissimilar, suggests that the homologous segment contains structural information necessary for common functional properties. The presence of a histidyl residue in this unique peptide is significant, since it has been suggested that histidine is one of the axial heme ligands in cytochrome P-450 (26). Thus, the analogous peptide may be involved in the heme binding segment of cytochrome P-450. Of course, much more evidence will be required to prove this hypothesis, since the two proteins have 11-12 histidyl residues. Dus et al. (27) have reported the isolation of heme peptide from CNBr digests of several forms of cytochrome P-450. Although the compositions of these peptides have been determined, their larger size precludes a meaningful comparison with the two analogous peptides described here.
Nevertheless, at a recent meeting in Tokyo, Fujii-Kuriyama et al. (29) reported the amino acid sequence determined from the cloned complimentary DNA of the phenobarbital-inducible cytochrome P-450 mRNA from rat liver. A portion of this sequence is homologous to the tridecapeptide described here for the phenobarbital-inducible rabbit cytochrome P-450 form 2. Presumably, this peptide is conserved between rat and rabbit species as well as between forms 2 and 3 in the rabbit liver.