rabbit liver microsomes - Europe PMC

Download PDF
2 downloads 0 Views 1014KB Size Report
Comment
Cleavage at tryp- tophan by BNPS-skatole (13) was performed on pyridyleth- ylated protein that was also carboxyl-blocked as follows: to the protein dissolved in ...
Proc. Nati. Acad. Sci. USA

Vol. 81, pp. 4260-4264, July 1984 Biochemistry

On the amino acid sequence of cytochrome P-450 isozyme 4 from rabbit liver microsomes (cysteine-containing peptides/NH2-terminal residues/homology of P-450 cytochromes/sequence analysis/membrane protein)

VALERIE S. FUJITA, SHAUN D. BLACK, GEORGE E. TARR, DENNIS R. Koop, AND MINOR J. COON Department of Biological Chemistry, Medical School, The University of Michigan, Ann Arbor, MI 48109

Contributed by Minor J. Coon, March 22, 1984

ABSTRACT Isozyme 4 of rabbit liver microsomal cytochrome P-450 was shown earlier in this laboratory to contain multiple NH2-terminal residues, whereas isozymes 2, 3a, 3b, and 3c have single, unique NH2-terminal sequences. Similar results were obtained with isozyme 4 obtained from animals that were untreated, treated with phenobarbital (which does not induce this isozyme), or induced with f8-naphthoflavone or isosafrole. With the use of selective chemical blocking at seryl or at nonprolyl residues, the complexity of the NH2-terminal sequence has now been shown to be due to the presence of three forms of the cytochrome differing only in the absence of the first or the first two residues: NH2-Ala-Met-Ser-Pro-AlaAla-Pro-, NH2-Met-Ser.Pro-Ala-Ala-Pro-, and NH2-Ser-ProAla-Ala-Pro-. These forms may result from variable biological processing. Peptides containing the seven cysteine residues were sequenced and compared with similar peptides reported for other P-450 cytochromes; homology was extensive with respect to two of the cysteine regions in isozyme 4, and a third cysteine region showed partial identity. The sequence of peptides representing about two-thirds of the amino acids in isosafrole-induced cytochrome P-450 isozyme 4 was determined. Comparison with phenobarbital-induced rabbit cytochrome P450 isozyme 2 indicated about 25% homology. In contrast, comparison of isozyme 4 with rat cytochrome P-450d, which is also induced by isosafrole and for which the sequence has recently been deduced from cDNA [Kawajiri, K., Gotoh, O., Sogawa, K., Tagashira, Y., Muramatsu, M. & Fujii-Kuriyama, Y. (1984) Proc. NatI. Acad. Sci. USA 81, 1649-1653], showed about 70% homology.

In the present paper, evidence is provided that the variable NH2 terminus results from the partial loss of the first and second residues of the protein. In addition, the cysteinecontaining peptides have been identified and compared with similar peptides from other P-450 cytochromes. With most of the protein sequenced, a close similarity is evident between the structure of isozytne 4 and that recently deduced for rat liver cytochrome P-450d from cloned cDNA (9).

MATERIALS AND METHODS Materials. Cytochrome P-450 isozyme 4 was purified to electrophoretic homogeneity from liver microsomes of male New Zealand White rabbits treated with isosafrole (2, 8). Acetyl chloride, CNBr, unsym-dimethylethylenediamine, 1(3-dimethylaminopropyl)-3-ethylcarbodiimide, CF3COOH, and 4-vinylpyridine were obtained from Aldrich; o-phthalaldehyde (OPA) was from Sigma; hexafluoroacetone sesquihydrate was from PCR Research Chemicals (Gainesville, FL); and 5,5'-dithiobis(2-nitrobenzoate) and 2-(2-nitrophenylsulfenyl)-3-methyl-3-bromoindolenine (BNP$-skatole) were from Pierce. Endoproteinase Lys-C was from Boehringer Mannheim, and Staphylococcus aureus V8 protease was from Miles. All HPLC solvents were obtained from Burdick and Jackson (Muskegon, MI). S-Pyridylethylation of Cysteine. A solution of the protein in 0.1 M potassium phosphate buffer (pH 7.4) was made 6 M in guanidine hydrochloride by the addition of 1 g of the solid per ml of solution. Five equivalents of dithiothreitol were added per cysteine residue, and the solution was allowed to stand for 4 hr at 220C. Total thiols were alkylated with 1.5 equivalents of 4-vinylpyridine overnight (10). Excess reagents were removed by dialysis against 10 mM phosphate buffer (pH 7.4) or by 1:3 dilution of the reaction mixture and precipitation of the protein with 50% acetone. Protein Cleavage. The digestion of native protein by endoproteinase Lys-C (6) was quenched by the addition of guanidine hydrochloride to a final concentration of 6 M, followed by the S-pyridylethylation of cysteinyl residues as described above. Pyridylethylated protein was submitted to limited acid cleavage in aqueous 0.1% CF3COOH at 1000C (6) or to CNBr cleavage in 70% HCOOH followed by boiling in 0.1% CF3COOH for 30 min to promote cleavage at homoserine (11). In one instance a peptide from the CNBr cleavage was then digested With S. aureus protease (12). Cleavage at tryptophan by BNPS-skatole (13) was performed on pyridylethylated protein that was also carboxyl-blocked as follows: to the protein dissolved in 4 vol of hexafluoroacetone sesquihydrate, 2 vol of 4 M dimethylethylenediammonium chloride (pH 4) and 1 vol of this buffer saturated with ethyldimethylaminopropyl carbodiimide were added. After 2 hr at 220C, the reaction mixture was acidified with 1 vol of 88% formic

Following the demonstration by enzyme fractionation that liver microsomal cytochrome P-450 (cytochrome P-450LM) occurs as isozymes with distinct chemical, physical, and catalytic properties (1), this laboratory reported that phenobarbital-induced isozyme 2 and 5,6-benzoflavone-induced isozyme 4 of rabbit cytochrome P-450LM (cytochromes P-45OLM2 and P-45OLM4, respectively) differ in amino acid composition and COOH-terminal amino acid residues (2). Isozyme 2 has a single NH2-terminal amino acid sequence highly similar to the "signal peptides" ordinarily removed from preproteins during biological processing (3); the complete sequence was subsequently determined (4-7). Isozymes 3a, 3b, and 3c also have unique NH2-terminal sequences (8). In contrast, multiple NH2-terminal residues were found with isozyme 4, including alanine, methionine, and serine, and it was suggested that the preparation might either have contained a mixture of cytochromes or undergone proteolysis at the NH2 terminus (3). The latter explanation was favored by evidence for homogeneity obtained by electrophoretic, immunochemical, and peptide mapping procedures and by the apparent identity of preparations from animals either untreated or treated with phenobarbital, 5,6-benzoflavone, or isosafrole. The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact.

Abbreviations: cytochrome P-45OLM, liver microsomal cytochrome P-450; OPA, o-phthalaldehyde; BNPS-skatole, 2-(2-nitrophenylsul-

fenyl)-3-methyl-3-bromoindolenine. 4260

Biochemistry: Fujita et aL acid, diluted 1:3 with aqueous 0.1% CF3COOH, and submitted to reversed-phase HPLC to remove excess reagents. Peptide Purification. Reversed-phase HPLC was used to separate the peptides in the cleavage mixtures. A Beckman 340 system equipped with a model 165 variable-wavelength detector was used to facilitate identification of cysteine-containing peptides (6). The standard mobile phase was water vs. CH3CN/2-propanol, 3:1 (vol/vol), with 0.1% CF3COOH present throughout (14). A flow rate of 0.5 ml/min and a gradient rate of 1%/ml were generally used. Columns used for the initial fractionation of the mixtures were Beckman Ultrapore RPSC C3 (5 Am, 4.6 x 75 mm), Brownlee RP-300 (10 Am, 4.6 x 250 mm), Vydac C18 (5 Am, 4.5 x 250 mm), or Waters ,uBondapak CN (10 Aim, 3.9 x 300 mm). Rechromatography of the collected fractions was with a Waters ABondapak phenyl column (10 pm, 3.9 x 250 mm) or an Ultrapore RPSC column. Sequence Analysis. Peptides were sequenced manually (15); effective sequencing of carboxyl-modified peptides required the addition of 10 Mg of Ponceau S (a tetrasulfonic acid) to 20 pug of the usual carrier (Polybrene) and incubation for 2 min at 50'C with methanol after cleavage and drying in order to convert anilinothiazolinones to the more easily extracted phenylthiocarbamyl amino acid methyl esters. An extended NH2-terminal sequence analysis of isozyme 4 was done by the usual procedure except that cleavage was performed at 22°C for 2 min during cycles 2 and 3 and at 50°C for 10 min on cycle 4. Phenylthiohydantoin amino acids were identified and quantified as reported earlier (16). Analysis of sequence results was aided by various computer programs (6). NH2-Terminal Sequence. Blocking of nonproline NH2 termini was achieved with the use of OPA. To the protein, 10 Al of pyridine/triethylamine/water, 5:2:3 (vol/vol), was added, and the solution was brought to dryness in vacuo. This was followed by addition of 10 ,ul of OPA reagent [prepared from 10 mg of OPA, 0.45 ml of absolute methanol, 10 ,ul of ethanethiol, and 50 Al of 0.1 M sodium borate buffer (pH 9.5)], 20 ,ul of 0.1 M borate buffer, and 50 ,ul of absolute methanol. After incubation at 22°C for 30 min, the reagents were removed by decantation. One "pseudocycle" of Edman degradation with the coupling step omitted removed residual OPA reagent, and sequence analysis was continued in the usual manner. The isoindole thioether blocking group (17) was stable to the extreme acid-base conditions of the manual Edman degradation, in agreement with results from solid phase (18) and Sequenator analysis (19). Selective blocking of NH2-terminal seryl residues was performed by O-acetylation of the hydroxyl group under acidic conditions, followed by the use of alkaline conditions to bring about an O-to-N shift of the acetyl group (20, 21). About 1 nmol of protein was dissolved in 30 Al of 5% acetyl chloride in anhydrous CF3COOH (vol/vol), and the mixture was incubated at 22°C for 15 min under N2 (21). The protein was dried to a thin film under reduced pressure and washed three times with 0.1% acetic acid in benzene/acetonitrile, 1:1 (vol/vol), to remove residual reagent. Incubation in 15,ul of triethylamine/water/ethanol, 1:1:7 (vol/vol), for 10 min at 50°C under N2 was required to effect the O-to-N shift. Manual sequence analysis of the protein was then performed. Cysteine Determination. Amino acid compositions were determined by the phenylthiocarbamyl amino acid method (8), with and without performic acid oxidation. The number of cysteine residues was also determined by incubating the cytochrome (10,uM) for 10 min at 50°C in 0.2 M phosphate buffer, pH 7.4/0.1% NaDodSO4, with 5,5'-dithiobis(2-nitrobenzoate) added to a final concentration of 0.9 mM. Protein was omitted in control experiments. A value of 13,600 M-l cm-l for AE412 was used to quantify the number of cysteines modified (22).

Proc. NatL Acad Sci. USA 81 (1984)

4261

RESULTS AND DISCUSSION NH2-Terminal Sequence. In view of the multiple NH2-terminal residues reported for isozyme 4 (3), the possibility was investigated that the preparation might consist of several proteins and, therefore, that various sources might yield cytochromes in relatively different amounts. However, isozyme 4 isolated from rabbits that had been treated with isosafrole, phenobarbital, or 5,6-benzoflavone or had no treatment were identical as judged by NaDodSO4/PAGE, Ouchterlony double-diffusion studies, reversed-phase HPLC tryptic maps, and electrophoretic maps of peptides generated by S. aureus protease, chymotrypsin, or papain. Furthermore, NH2-terminal sequences were determined for preparations of isozyme 4 from pooled livers of outbred rabbits and from single livers of three strains of inbred rabbits (IIIVO/J, ACEP/J, and B/J from The Jackson Laboratory), all untreated or treated with isosafrole or phenobarbital, and from a single liver of an outbred rabbit treated with isosafrole. The results consistently showed alanine, methionine, and serine on the first cycle of Edman degradation with similar alanine/methionine ratios. Yields of phenylthiohydantoin-serine varied according to the extent of dehydration during manual sequence analysis, so quantitative comparisons could not be made with this residue. The second cycle consistently showed methionine, serine, and proline. The occurrence of serine and proline in the first two cycles permitted selective chemical blocking to unravel the sequence of the NH2 terminus as indicated in Table 1. After one cycle of Edman degradation, OPA was used to block all peptides but those with NH2-terminal prolyl residues, and a single amino acid sequence following the proline on cycle 2 was clearly established. The sequences following the proline on cycles 3 and 4 were determined in an analogous manner. These sequences were identical, suggesting that the complexity of the NH2-terminal sequence was due to a mixture of the n, n - 1, and n - 2 forms of the protein. However, this technique could not establish the sequence preceding the proline residue. In another set of experiments, NH2-terminal seryl peptides were specifically blocked by acetylation. The 0-to-N shift of the acetyl group occurred much less rapidly than expected (21), and elimination of the incubation step left the amino group of the serine unacetylated. Thus, when O-acetyl seryl residues became NH2-terminal during sequence analysis, blocking was optional. N-Acetylation of the serine on cycle 1 eliminated only proline from cycle 2, which established that the NH2-terminal serine must be followed uniquely by proline and that the proline on cycle 2 must be preceded uniquely by serine. The residues following this proline had already been established by OPA-blocking, and the entire sequence beginning with serine was confirmed in a peptide isolated from the endoproteinase Lys-C digest. Further analysis of the sample blocked at serine showed that the other two sequences were identical except for an apparent offset of one residue. This offset was confirmed after serine on cycles 1 and 2 was blocked. Methionine was the only residue to appear on cycle 2, which permitted subsequent residues to be determined unambiguously. These results account for all of the residues observed during NH2-terminal sequence analysis of the intact cytochrome. The presence of n and n - 1 forms of the NH2 terminus has been found for rat cytochrome P-450b (23), rabbit isozyme 3b (24), and porcine adrenocortical cytochrome P-450 21-hydroxylase (25). The most probable explanations are variable biological processing or cleavage by contaminating proteases during enzyme isolation. The reason for the more extensive processing of isozyme 4 is not evident. A slightly different strategy was adopted to obtain an extended NH2-terminal sequence analysis. Though blocking two of the three chains with OPA allowed the order of the

4262

Biochemistry: Fujita et aL

Proc. NatL Acad Sci USA 81 (1984)

Table 1. NH2-terminal sequence data for rabbit cytochrome P-45OLM4

Cycle number

Protein sample Form n Form n - 1 Form n - 2

INTACT

2

1

A M S

M S P

3

6 4 5 Residues in deduced sequence A A S P P A A P L A A P Residues and experimental yield, pmol A 984 A 600 A 855 P 633 P 705 P 444 A 528 P 516 L 375 S 171 P 405 L 303 A 441 A 384

7

P L S

8 L S V

P 681 M 642 L 612 A 855 L 474 P 525 V 240 M 684 S 78 S 117 S 63 S 157 S 96 A 921 P 384 OPA 1 M 522 S 231 M 543 P 486 A 402 A 381 P 363 L 291 S 96 A 1014 OPA 2 P 519 M 582 S 303 S 228 A 70 L 61 P 88 A 73 P 80 A 89 M 77 P 74 OPA 3 P 70 M 60 A 70 S 23 S 34 S 61 P 102 L 25 V 64 S 42 P 120 A 75 A 84 X LYS-C S 60 P 182 A 335 M 239 SER 1 A 287 S 51 P 139 A 146 M 163 P 179 M 178 SER 2 A 280 S 35 M 207 A 476 A 492 P 418 L 320 P 473 PHASE A 122 M 148 S 82 M 168 S 95 S 88 Sequences deduced for the processed forms of the NH2 terminus of isozyme 4 are designated at the top of the table by n, n - 1, and n - 2. In the experimental data, INTACT refers to unblocked protein. OPA 1, 2, and 3 refer to protein blocked with OPA after 1, 2, and 3 cycles of manual Edman degradation, respectively. SER 1 refers to protein blocked at serine at cycle 1 and SER 2 to protein blocked at cycles 1 and 2. LYS-C is the peptide purified from digestion of the protein with endoproteinase Lys-C, and PHASE refers to unblocked protein brought into phase at proline with mild acid cleavages on cycles 2 and 3. X, unidentified residue.

first 19 amino acids to be determined, the background from nonspecific chain cleavage of the whole protein obscured the contribution from the one processed form being sequenced. To improve the yields over the background, the processed forms were brought into phase and degraded together. It is well known that phenylthiocarbamyl-Pro-X bonds are slow to cleave (26); phenylthiocarbamyl-Ser-Pro and phenylthiocarbamyl-Met-Ser were cleaved to completion with anhydrous CF3COOH within 2 min at 220C with no appreciable cleavage of phenylthiocarbamyl-Pro-Ala. With these mild conditions on cycles 2 and 3, the phenylthiocarbamyl-prolyl residues were left intact on the peptide chain until the proline on cycle 4 of the n form was coupled and the sequences of all processed forms were aligned as shown in Table 1. Thus, the sequence could be determined for at least an additional 24 cycles, sufficient to overlap a peptide cleaved at tryptophan. The yields of phenylthiohydantoin-serine and phenylthiohydantoin-methionine were higher with the gentler cleavage conditions (Table 1, PHASE experiment), and cleavage/extraction with gentle and then rigorous conditions gave the most quantitative results for acid-labile residues while insuring that the peptide remained in phase. So far we have done this only in special cases, because the postcleavage methanol treatment described for COOH-modified peptides has been adopted as a general technique and gives much improved yields of phenylthiohydantoin-serine and threonine. Two of the methods adapted to determine the NH-terminal sequence of isozyme 4 were originally aimed at reducing. the background, although use of OPA for analysis of peptide mixtures has been suggested (18). O-Acetylation of hydroxylamino acids to prevent the N-to-O shift that occurs during the cleavage step of the Edman degradation (21) was unsatisfactory because the conditions used for spinning-cup sequenators caused NH2-terminal serine and threonine resi-

dues to be blocked. For our purposes, blocking was an advantage and was easily controlled in combination with manual sequence methods. Mild cleavage conditions to bring NH2-terminal proline residues into phase was a novel application of known Edman chemistry. These approaches should be useful in the study of other proteins with processed NH2-termini and for the sequence analysis of peptide mixtures that may be difficult to fractionate because of similarities in their amino acid sequences, including mixtures of partial cleavage products that differ only in length at the NH2 terminus. Cysteinyl Peptides. The total number of cysteine residues present in isozyme 4 was determined to be seven by amino acid analysis, in agreement with earlier results (2), and with the use of 5,5'-dithiobis(2-nitrobenzoate). The equivalence of these values indicated the absence of disulfide bridges, as also found for isozyme 2 (6). Sequence analysis of the NH2 terminus of the entire protein revealed the first cysteine residue, and further information on the carboxyl side of this residue was obtained with a peptide cleaved at tryptophan. A limited endoproteinase Lys-C digest provided peptides with four cysteine residues, while acid cleavage and CNBr cleavage gave the remaining two. The results are shown in Fig. 1. In view of the possibility that a highly conserved cysteine residue might be the source of the ligand to the heme iron atom, homology of these peptides with other known P450 sequences was assessed, with results as shown in Fig. 2. Two of the peptides show a high degree of similarity to the two highly conserved cysteine peptides reported for other P450 cytochromes (6, 7, 27, 28)-cysteine-152 and cysteine436 in isozymes 2 and b, with the third exhibiting partial identity. The peptide containing cysteine-152 is favored by this laboratory to provide the fifth ligand to the heme (6). However, the most striking homology is seen in the align-

Biochemistry: Fujita et aL

Proc. NatL Acad Sci USA 81 (1984)

AMSPAAPLSVTELLLVSAVFCLVFWAVRASRPKVPKGLK SFSIASNPASSSSCYLEEHVSQEAENLISRFQELMAAVG

PYSQLVVSAAXV TTLNGFH

GAMCFG

PKECC IF

VTLFGLGKRRC

NQWQ

NHDPQLWGD

GETLARWEVFLFLA

LLQXLEFSVPPG

4263

I A SNA[SSSCYL E E H V S Q E AE[

P-45OLM4 P-450d P-45OLM2 P-450b OR -e P-450 21-OH

NH L JTSVSSC Y L E E H A L L R K SV E ER I QE EAR C V EEL K S G S V E E R I Q E E A Q C V E E L R K S QGA P

P-450CAM

K L E N R I Q E L A C S L IE S L

P-45OLM4 P-450d P-45OLM2 P-450b oR -e

Y L V V S A A X V I GA M C F 6 VNl Q V WJES V A N V I G A M C F G K N F P R K S

S

S

I A

SD

V C Q LT Q E F C E R

RPEnGQ C JF

KHPRCEHVQARPSF

FIG. 1. Sequence of isozyme 4 peptides containing cysteine residues. The asterisks mark the cysteines. ment with the peptide containing cysteine-436, which has

been favored by others (7, 28). An alternative explanation for this conservation of residues is that this region binds NADPH-cytochrome P-450 reductase. Additional Sequence Information and Comparisons. Rabbit cytochrome P-450 isozyme 4 and rat cytochrome P-450d, the major isozymes induced by isosafrole (29-31), exist largely in the high-spin state and are similar in size as determined by calibrated NaDodSO4/PAGE. On the assumption that the two proteins have the same number of residues, -70%6 of the primary structure of isozyme 4 has been determined. An alignment of the corresponding regions of these cytochromes is given in Fig. 3. Of these residues, 70% are identical and 25% are related by possible point mutations. However, the NH2-terminal region of isozyme 4 shows very little homology to that of cytochrome P-450d. The distinctiveness of this region in P-450 cytochromes (8, 25, 32) is not necessarily indicative of overall structural relatedness. As in other P-450 cytochromes, isozyme 4 contains a hydrophobic region near the NH2 terminus, followed by a cluster of charged amino acids characteristic of halt-transfer signals for membrane insertion (33). Furthermore, the known sequence of isozyme 4

T L L F H S

I I CS I V F FSjLT CS I I C CL

P-45OLM4

P-45OLM2

P-450b OR -e P-450 21-OH

P-450d

r[GH nGS L[&SLG rQS

P-450CAM

RL

FIG. 2. Comparison of the sequence of cysteine peptides in variP450 cylochromes. Homology with respect to isozyme 4 is shown by the enclosed areas. ous

follow a natural alignment with isozyme 2, as shown by the examples in Fig. 4. Stretches of hydrophobic residues occupy corresponding positions, the proline cluster near the NH2 terminus (6) is present, and the highly accessiappears to

48

A M S P A A P L S V T E L L L V SAV F C M A F S Q Y I S L A PIE L L TA I F C

LIA

P-450d

P-45O0

QHE

L V F W A VRA S R P K V P K G L K R L V F W|V L RG TRT QIV P K G L KIS

LPGPS6V PIP G P W LP

-

132

117

106

-G R P D L Y S|S *G R P D L Y S|F

P-45OLM4

K R F D Y K D

E R F DY T D

EKVTLFGLGKRRC I GE T LARWEVFL EK VMLF G L G K R R C G E I PAEK W E V F L E GFMPF G K R CLG EG A R T ELF L EA F M P F S TG KR C LG E G I A R N ELFL VCLG EPLARL E V S H E1T H JI I V

1

P-45O,,Lw

G G

TIFG

P-45021-OH

P-450d

T SN I I C S I V F I1SO

TFLFQCI[TEN

** WAARRRLAQDS|L K S F S I A SN[A[S]SS S CYL E E H V'S

SF[I]T. TL I

*| W A A R R R L A Q D|A|L K S F S

I A S D JT[VS S C Y L E E H V S 205

216

RFELA VR FDPYSQLVVSAXVIGAMCFG| * *[iD VVRNS SKF JKLMA HF IEGAMC FGl WLN LWKS[S

P450d

NL KEN HL

P-450O,

S KFVET A SS S PVD F F P I V R Y L P

245

263

254

**

266

273

FM* * F REHYED. . K DFVEN V TSG N A|V D F F PIV LIR Y L P - *N P-450d J. LTVE H Y .QJ. 360 339 295 A|N G G L I P Q E K I V N|L|V N D I F 6 A G F D TI [T T AL S WSL M YLVT EN PR R[REQ * * H P-450~ P-450d ANIGGL I PQEK I VNIILVND I FGAGFDE[fV AI FS[ LLL[ KHVR 379

LPrLEAF

D

.

.

1

392

.- .TTL N G F H Y R YT LF -jJS L N G F H

EIL F X X

NQ RFL R N D N V L Sj

I P K E C C I F I N Q W Q I NH P Q L WG I P K E C C I F I N Q W Q V HE K Q K

P-45OM4

Q

P-450d

QLPYLEAF I LE I

P-45OLM4

DE EIF R P E R F L T AD AAINKPLAE T L F G L G K R R C I G EIT LA]RW E V F L F L M L F G L G K R R C I G EI P AJK W E V F L F L GVF V|F R P E R F L T N JN TLA IDKL E

P-450d

I L

496

484

P-45OLM4 P-450d

A I L A I L

XL S V P P G LGQ LIH QLjTL yyVPP G

.

513

511

P R|C E H V Q A RPSF . P R T[C E H V Q A WPJRF-J

. .[]H

-

-

FIG. 3. Known amino acid sequence of rabbit isozyme 4 compared with corresponding regions of rat cytochrome P-450d. Intervening sequences of various length are indicated by dots, and the positions are numbered according to the rat protein. The COOH-terminal residue was determined by carboxypeptidase digestion (2). Due to space limitations, the numerical data supporting the sequences given are not included in this paper but will be made available by the authors upon request.

4264

Biochemistry: Fujita et aL

P-45OLM4 P-45OLM2

A1A (321

1RR RL A Q D S L K

IWRIALIRRIFSLATMR

L121J

P-45OLM4 WV E T A[SkYIPWV P-45OLM2 FS L S[F Q[V P-45OLM4

P-45OLM2

; F D T ITTA MDDN FIFIAGIT ET STTILI

FJG

1296

P-45OLM4 (L3 YL EILEF1IFL EL F P-45OLM2 MIPYIT DIAIV IIHEIEI Q FIG. 4. Comparison of isozyme 4 peptides to corresponding regions in isozyme 2. The positions for isozyme 4 are numbered with respect to rat cytochrome P-450d. The isozyme 2 peptide shown at the bottom of the figure was reported to be homologous to a peptide in isozyme 3b (24).

ble Lys-C cleavage site (34) is also found. All of these findings indicate that P-45OLM proteins are related in sequence and, therefore, presumably in tertiary structure. We are grateful to Dr. Y. Fujii-Kuriyama, who kindly informed us of the sequence of rat liver cytochrome P-450d deduced from cloned cDNA prior to publication of ref. 9. This research was supported by Grant AM-10339 from the National Institutes of Health. 1. Haugen, D. A., van der Hoeven, T. A. & Coon, M. J. (1975) J. Biol. Chem. 250, 3567-3570. 2. Haugen, D. A. & Coon, M. J. (1976) J. Biol. Chem. 251, 79297939. 3. Haugen, D. A., Armes, L. G., Yasunobu, K. T. & Coon, M. J. (1977) Biochem. Biophys. Res. Commun. 77, 967-973. 4. Black, S. D., Tarr, G. E. & Coon, M. J. (1982) J. Biol. Chem. 257, 14616-14619. 5. Heinemann, F. S. & Ozols, J. (1982) J. Biol. Chem. 257, 14988-14999. 6. Tarr, G. E., Black, S. D., Fujita, V. S. & Coon, M. J. (1983) Proc. Natl. Acad. Sci. USA 80, 6552-6556. 7. Heinemann, F. S. & Ozols, J. (1983) J. Biol. Chem. 258, 41954201. 8. Koop, D. R., Morgan, E. T., Tarr, G. E. & Coon, M. J. (1982) J. Biol. Chem. 257, 8472-8480.

Proc. NatL Acad Sci USA 81 (1984) 9. Kawajiri, K., Gotoh, O., Sogawa, K., Tagashira, Y., Muramatsu, M. & Fujii-Kuriyama, Y. (1984) Proc. Nati. Acad. Sci. USA 81, 1649-1653. 10. Friedman, M., Krull, L. H. & Cavins, J. F. (1970) J. Biol. Chem. 245, 3868-3871. 11. Gross, E. (1967) Methods Enzymol. 11, 238-255. 12. Drapeau, G. R. (1977) Methods Enzymol. 47, 189-191. 13. Fontana, A., Vita, C. & Toniolo, C. (1973) FEBS Lett. 32, 139142. 14. Tarr, G. E. & Crabb, J. C. (1983) Anal. Biochem. 131, 99-107. 15. Tarr, G. E. (1982) in Fourth International Conference on Methods in Protein Sequencing, ed. Elzinga, M. (Humana, Clifton, NJ), pp. 223-232. 16. Black, S. D. & Coon, M. J. (1982) Anal. Biochem. 121, 281285. 17. Simons, S. S., Jr., & Johnson, D. F. (1976) J. Am. Chem. Soc. 98, 7098-7099. 18. Machleidt, W. & Hofner, H. (1982) in Methods in Protein Sequence Analysis, ed. Elzinga, M. (Humana, Clifton, NJ), pp. 173-180. 19. Bhown, A. S., Cornelius, T. W., Volanakis, J. E. & Bennett, J. C. (1983) Anal. Biochem. 131, 337-340. 20. Previero, A., Barry, L. G. & Coletti-Previero, M. A. (1972) Biochim. Biophys. Acta 263, 7-13. 21. Brandt, W. F., Henschen, A. & von Holt, C. (1982) in Methods in Protein Sequence Analysis, ed. Elzinga, M. (Humana, Clifton, NJ), pp. 101-110. 22. Ellman, G. L. (1959) Arch. Biochem. Biophys. 82, 70-77. 23. Yuan, P. M., Ryan, D. E., Levin, W. & Shively, J. E. (1983) Proc. Natl. Acad. Sci. USA 80, 1169-1173. 24. Ozols, J., Heinemann, F. S. & Johnson, E. F. (1981) J. Biol. Chem. 256, 11405-11408. 25. Yuan, P. M., Nakajin, S., Haniu, M., Shinoda, M., Hall, P. F. & Shively, J. E. (1983) Biochemistry 22, 145-149. 26. Tarr, G. E. (1977) Methods Enzymol. 47, 335-357. 27. Haniu, M., Armes, L. G., Yasunobu, K. T., Shastry, B. A. & Gunsalus, I. C. (1982) J. Biol. Chem. 257, 12664-12671. 28. Fujii-Kuriyama, Y., Mizukami, Y., Kawajiri, K., Sogawa, K. & Muramatsu, M. (1982) Proc. Natl. Acad. Sci. USA 79, 27932797. 29. Ryan, D. E., Thomas, P. & Levin, W. (1980) J. Biol. Chem. 255, 7941-7955. 30. Fisher, G. J., Fukushima, H. & Gaylor, J. L. (1981) J. Biol. Chem. 256, 4388-4394. 31. Delaforge, M., Koop, D. R. & Coon, M. J. (1982) Biochem. Biophys. Res. Commun. 108, 59-65. 32. Botelho, L. H., Ryan, D. E., Yuan, P. M., Kutney, R., Shively, J. E. & Levin, W. (1982) Biochemistry 21, 1152-1155. 33. Sabatini, D. D., Kreibich, G., Morimoto, T. & Adesnik, M. (1982) J. Cell Biol. 92, 1-22. 34. Black, S. D., Tarr, G. E., Fujita, V. S., Yasunobu, K. T. & Coon, M. J. (1983) Fed. Proc. Fed. Am. Soc. Exp. Biol. 42, 1774 (abstr.).