(Bacillus megaterium) Hosts - Semantic Scholar

Vol. 262, No. 14, Issue of May 15, pp. 6676-6682,1987 Printed in U. S. A.

THEJOURNAL OF BIOLOGICAL CHEMISTRY 01987 by The American Society of Biological Chemists, Inc.

Cloning of the Gene Encoding a Catalytically Self-sufficient Cytochrome P-450Fatty Acid Monooxygenase Induced by Barbiturates inBacillus megateriumand Its Functional Expression and Regulation in Heterologous (Escherichia coli)and Homologous (Bacillus megaterium)Hosts* (Received for publication, September22, 1986)

Long-Ping Wen and ArmandJ. FulcoS From the Department of Biological Chemistry, School of Medicine,and the Laboratory ofBiomedical and Enuironmental Sciences, University of California, L o s Angeles, California 90024

In a previous publication (Narhi, L. O., and Fulco, A. roid hormone andbile acid synthesis, fattyacid monooxygenJ. (1986) J. Biol. Chem. 261, 7160-7169) we de- ation, xenobioticdetoxification, theactivation of certain scribed the characterization of a 119,000-dalton P- polycyclic aromatic hydrocarbons to potentcarcinogens, and 450 cytochrome that is strongly induced by barbitu- the oxidation of other hydrophobic substances in numerous ratesin Bacillus megaterium. Inthe presence of peripheral catabolic pathways (1, 2). NADPH and 02,this single polypeptide can catalyze For the past decade, we have been studying a soluble Pthe hydroxylation of long-chain fatty acids without the 450-dependent monooxygenase from Bacillusmegaterium aid of any other protein. The gene encoding this unique ATCC 14581 that catalyzes the w-2-hydroxylation of satumonooxygenase (cytochrome P-450BM-3) has now been cloned by an immunochemical screening technique. rated long-chain fatty acids, amides, and alcohols (3-5) and The Escherichia coli clone harboring the recombinant the epoxidation and/or hydroxylation of unsaturated fatty acids (6). This monooxygenase, the most catalytically active plasmid produces a 119,000-dalton protein that appears to be electrophoreticallyand immunochemically P-450 cytochrome yet reported, has recently been purified to identical to the B. megaterium enzyme and contains homogeneity and extensively characterized in our laboratory the same N-terminal amino acid sequence. The recom- (7). It exists as a single polypeptide with a molecular weight binant DNA product also exhibits the characteristic of 119,000 5000 daltons and, in the presence of NADPH can catalyze the oxygenation of long-chain fattyacids cytochrome P-450 spectrum andis fully functional as and 02, a fatty acidmonooxygenase. In E. coli, the synthesisof without the participation of any other protein. This cytois directed by its own promoter included in chrome, designated P-450BM.3to distinguish from P-450BM-3 it two much the DNA insert and proceeds constitutively at a very smaller P-450 proteins(BM-1 and BM-2)previously isolated How- from the same B. megaterium strain (8), contains 1 mol each high rate but is not stimulated by pentobarbital. ever, when thecloned P-450BM-3 gene, either intact or of FAD and FMN/molof heme andalso functions as an active in a truncated form, is introduced back intoB . mega- cytochrome c reductase in thepresence of NADPH. We have terium via an E. coli/Bacillus subtilis shuttle vector, speculated (7) that the catalytically self-sufficient P - 4 5 0 ~ ” ~ its expressionis constitutively repressed butis induced contains two domains analogous to the two-protein P-450by pentobarbital. This finding demonstrates that the regulatory region of the P ‘ 4 5 0 B M - 3 gene that responds dependent monooxygenase complexes of mammalian liver by the fusion of genes to barbiturates is included in the cloned DNA. The microsomes (9) and perhaps arose two components evidence also indicates that pentobarbital cannot di- encoding proteins functionally similar to the rectly act on the gene to cause induction but presum- of the microsomal systems. Another unique featureof P-450BM.3is our finding that this ably interacts with another component such as a repressor molecule that is present in B. megateriumbut monooxygenase is highly inducible by phenobarbital (10) and by many other barbiturates (11, 12) and barbiturate analogs is absent in the E. coli clone.

*

(13, 14). None of these inducers served as substrates, activators, or inhibitors of the fatty acid oxygenase system i n vitro nor did they appear to interact with it in any way. Although Cytochrome P-450 proteins have been isolated from a large there have been no reports of other barbiturate-inducible Pvariety of eukaryotic and prokaryotic sources and areinvolved 450-dependent monooxygenases inprokaryotes,there is, in many importantbiochemical processes. These include ste- again, an analogy between P - 4 5 0 ~ ”and ~ certain liver microsomal P-450 cytochromes that are also strongly barbiturate* This work was supported by National Institutes of Health Re- induced (1,9). In phenobarbital-mediated induction of specific search Grant GM 23913 and by Contract DE-AC03-76-SF00012 between the Department of Energy (Director of the Office of Health P-450s in the liver, the rate of transcription from the P-450 and Environmental Research) andthe University of California. The gene is stimulated, probably by an increase in transcription initiation. However, the mechanismby which the barbiturate costs of publication of this article weredefrayedin part by the payment of page charges. This articlemusttherefore behereby exerts thiseffect remains unknown(15). We propose that the marked “aduertiement” in accordance with 18 U.S.C. Section 1734 elucidation of the mechanismof barbiturate inductionof a Psolely to indicate this fact. 4 To whom all correspondence should be addressed University of 450 cytochrome in a relatively simple prokaryote would not California, Laboratory of Biomedical and Environmental Sciences, only be interesting in itself but might also provide insight into the analogous liver microsomal systems. As a first step 900 Veteran Ave., Los Angeles, CA 90024.

6676

Cloning and Expression

of

a Cytochrome P-450 f r o m B. megaterium

toward this goal, we have now cloned the P-45oBM.3 gene, using an immunochemical screening technique. Moreover, by comparing the expression of the cloned gene in Escherichia coli a n d in B. meguterium we have been able to elucidate several fundamental characteristicsof the induction process. EXPERIMENTALPROCEDURES

Materials-Equipment and reagents for polyacrylamide gel electrophoresis were from Bio-Rad as were goat anti-rabbit IgG conjugated with horseradish peroxidase and its color-development substrate, 4chloro-1-naphthol. Hybond-N nylon membranes were purchased from Amersham Corp. whilenitrocellulose membrane filters and discs were from Schleicher & Schuell. Restriction enzymes, T4 DNA ligase, 8-IPTG' and bacterial alkaline phosphatase were from International Biotechnology, Inc. T4 polynucleotide kinase, ultra-pure phenol and agarose were obtained from Bethesda Research Laboratories. E. coli strain JM 103 and plasmid pUC13 were purchased from Pharmacia was P-L Biochemicals. 5-Bromo-4-chloro-3-indal-~-~-galactoside from Boehringer Mannheim. The Bacillus subtilis plasmid vector, pC194, was obtained from the American Type Culture Collection. Preparation of Antibodies and Oligonucleotides-The preparation of rabbit antibodies to cytochrome P-45oBM.3 from B. megaterium ATCC14581 has been described indetail previously (7, 8). The antibody against p-galactosidase was a generous gift from Dr. Audree Fowler of the Department of Biological Chemistry, UCLA Medical School. A mixture of 16 oligonucleotides (17-mer), deduced from a portion of the N-terminal sequence (-Lys-Glu-Met-His-Gln-Pro) of P-450BM.3(7) was synthesized by Dr. ThomasSutherland of the Molecular Biology Institute, UCLA (16). Deprotected oligonucleotides were purified by electrophoresis in 20% polyacrylamide-7 M urea gels. The purified oligonucleotide with the sequence 3'TTzCTzTACGTgGnGG-5' was labeled at the 5' end with [y3'P] ATP (3000 Ci/mmol, Amersham Corp.) using T4 polynucleotide kinase (17). A preparation with a specific activity of 5 X lo6 dpm/ pmol was used as a probe in Southern blotting. Preparation of DNA-Plasmid DNA was isolated by the procedure of Birnboim and Doly (18) and purified through two successive CsCl centrifugations. The totalcellular DNA was prepared by the method of Miura (19). The isolation of restriction fragments from agarose gels wasaccomplished with a unidirectional electroeluter (model UEA from International Biotechnology, Inc.). Cloning Procedures-The total cellular DNA isolated from B. meguterium ATCC 14581 grown in medium containing 4 mM pentobarbital (7, 13) was partially digested with Sau3AI and then ligated with BamHI-digested, dephosphorylated pUC13. The ligation preparation was then used to transform competent E. coli JM103 (17). When a small portion of the transformed culture was plated out on a Luria-Bertani plate containing 8-IPTG and 5-bromo-4-chloro-3-indal-0-D-galactoside as well as 50 pg/ml of ampicillin, about 70% of the colonies were white, indicating that they had inserts.The remaining cells were plated onto Luria-Bertani agar containing ampicillin and replica-plated onto nitrocellulose membrane discs. The membranes with bacterial colonies were suspended in a chloroform vapor chamber for 45 min and then incubated overnight in TBS (Trisbuffered saline containing 50 mM Tris-C1, pH 7.5, and 150 mM NaCl) containing 1%gelatin, 100 pg/ml lysozyme (Sigma), 10 mM MgCl,, and 1 pg/ml DNase I (Worthington). After arinseinTBS, the membranes were incubated with antiserum to P-450BM.3 (1:lOO dilution in TBS) for 1 h, washed twice in TBS plus 0.1% Tween-20 and twice in TBS, and incubated with goat anti-rabbit IgG conjugated with peroxidase (1:3000 dilution) according to the instructions provided by Bio-Rad. Visualization was achieved by using 4-chloro-lnaphthol as the color-development substrate for the peroxidase conjugate. BlotAnalyses-For hybridization tests with the oligonucleotide probe (17-mer), the recombinant plasmid BM3.1 was first digested with the restriction enzymes and thensubjected to electrophoresis on a 0.7% agarose gel. The separated DNA fragments on the gel were denatured with NaOH, neutralized, and transferred to a Hybond-N nylon membrane (17). After fixation of the DNA fragments by UV light for 5 min, the nylon membrane was treated with prehybridization fluid and then hybridized with 32P-labeledoligonucleotide. Visualization was achieved by autoradiography, using Kodak X-Omat

6677

film in combination with an intensifying screen. Genomic Southern blots were performed in essentially the same manner as described above. Total cellular DNA from B. megaterium 14581, either uncut ordigested with various restriction enzymes, was subjected to electrophoresis on a 0.7% agarose gel, transferred to a Hybond-N nylon membrane, hybridized with a nick-translated, gelpurified 1.8-kb PstI fragment of the insert from clone BM3.1, and visualized by autoradiography (17). Western blot analyses were carried out as described by Towbin et al. (20) using aTrans-Blotapparatus(Bio-Rad). Visualization of antibody-binding bands was achieved as described above for cloning procedures. Other Methods-The preparation of cell-free soluble protein extracts from B. meguterium by sonication (lo), the assay for fatty acid monooxygenase activity utilizing [l-14C]myristateas a substrate ( l l ) , and the spectrophotometric determination of cytochrome P-450 in the cell-free extracts (21) have all been described in detail previously. Protein was assayed by the Biuret method (22, 23). Polyacrylamide gel electrophoresis was performed as described by Laemmli (24). Transformation of B. megaterium protoplasts by plasmids and subsequent regeneration was carried out as described by Imanaka et al. (32). RESULTS

Cloning and Expression of the Cytochrome P-450sM.3Gene in E. coli-In planning a protocol for cloning the P-45oBM.3 gene, we were aware of the strong possibilitythat the associ-

ated promoter from B. meguterium might not be functional in E. coli (25). We thereforechose as a cloningvector a plasmid, pUC13, that harbored a strong, regulatable P-galactosidase gene promoter and that also permitted us t o easily distinguish recombinants from nonrecombinants by 2-complementation (26). In theory, an insert, cloned into any one of the multiple cloning sites in the plasmid, in the correct orientation and translational frame, could be expressed as a fusion protein.The synthesis of such a protein would be under control of the P-galactosidase promoter which, in turn, could be activated by P-IPTG. As we will show later, however, this feature of pUC13 t u r n e d o u t t obe unnecessary. Fragments of B. meguterium DNA (size range, 200-15,000 base pairs) were generated by partial digestion with Suu3AI and then cloned into pUC13. Screening of 50,000 colonies (70% of which contained inserts) revealed one clone, BM3.1, that reacted strongly with antibody to P-450BM.3. When the recombinant plasmid prepared from clone BM3., was linearan approximately 15ized with the restriction enzyme EcoRI, kb band was seen on agarose gel, indicating an insert size of about 12 kb. To determine whethera fusion protein had been produced, the gene product of clone BM3.1was subjected to Western blot analysis. As Fig. 1 (lune 4 ) shows, a protein band with a molecular weight of about 119 kDa (Le. with an electrophoretic mobility identical to of that P-450eM.3)reacted strongly with P-450BM.3 antibody. The synthesis of this protein in E . coli is dependent upon the B. megaterium-derived DNA insert present in clone BM3.1,since E. coli transformed by the parental plasmid pUC13 produces no antibody-positive polypeptides (lune 1 ). It may also be notedthat the antibody to its affinity forP-450BM.3, also preparation used, in addition B. reactswith a 125-kDaproteinfoundconstitutivelyin meguterium (lanes 2 and 3 ) . However, this 125-kDa protein isnotinducedbypentobarbital;incontrastthe119-kDa monooxygenaseisalmostundetectablein B. megaterium grown normally (lane 2 ) but is present as the major protein component in pentobarbital-induced cells (lane 3). T h e p r o tein bands in lune 3 smaller than 119 kDa are degradation products of P-450BM.3. These data suggest that clone BM3.1 directs the synthesis of authentic cytochromeP-450BM.3 rather The abbreviations used are: 8-IPTG, isopropyl-8-D-thiogalacto- t h a n a fusion protein. Additional evidence that we had cloned the P-450BM.3 gene side; kb, kilohase pair(s).


6678 4

3

2

1

of a Cytochrome P-450 from B. megaterium

kDa

kb

1

2

3

-

4

-1 25 -1 19

FIG. 1. Detection by Western blotting of cytochrome P450eM.3apoprotein synthesized from clone BM3.1. Bacterial cells were harvested in late log phase, washed, and then ruptured by sonication. The sonicatefrom each batch of cells was centrifuged a t 40,000 X g and the supernatant samples, equivalent20 to pg of protein each, were taken for assay. The samples were subjected to electrophoresis on a sodium dodecyl sulfate-10% polyacrylamide gel, transferred to a nitrocellulose membrane by electroelution, and exposed to antibody against P - 4 5 0 ~ ~ Immunoreactive .~. bands were visualized by a peroxidase-catalyzed color reaction (see “Experimental Procedures”) and are shown in the figure above. The source of protein in each lane is as follows: lane I , E. coli JM103 transformed by pUC13; lane 2, B. megateriurn 14581 grown in the absence of pentobarbital; lane 3, B. megateriurn 14581 grown in the presence of 4 mM pentobarbital; and lane 4, E. coli JM103 transformed by plasmid BM3+

came from analysis with anoligonucleotide probe. Treatment of plasmid BMB.I with the restriction endonucleaseEcoRI generated a single 15-kb fragment which hybridized with the oligonucleotide (17-mer) probe corresponding to a sequence in the N-terminal portion of P - 4 5 0 ~ ” (Fig. ~ 2, lanes 2 and 4). On the other hand, HindIII digestion produced seven fragments (only three of which are seen in Fig. 2., lane 1; the other four have run off the gel) but only one, a 4.2-kb fragment, hybridized with the oligonucleotide probe (Fig. 2, lane 3 ) . A separate experiment (data not shown) indicated that none of the four smaller HindIII fragments hybridized with the 17-mer. Since the synthetic oligonucleotide had a sequence deduced from a portion of the N-terminal amino acid sequence of P-450RM.n (7), thehybridization results also indicated that clone BMs., included the N-terminal portion of the P-450eM.3structural gene. We should note,however, that the 17-mer probe we constructed is now known to contain one or two adjacent mismatched bases. Recently, the P-45oRM.3 protein from E. coli clone BM3., was purified to homogeneity.* Upon determination of its N-terminal amino acid sequence, a discrepancy was discovered in the published N-terminal sequence for the same protein purified from B. megaterium (7). The E. coli protein contained prolineat position 6 rather than histidine as shownin the publishedsequence. When cytochrome P-450RM.3 purified from B. megaterium was again sequenced,3 the results were unequivocal. Proline, not histidine, was also recovered a t position 6, indicating that the

* L. 0. Narhi, L.-P. Wen, and A. J. Fulco, unpublished results. L. 0. Narhi and A. J. Fulco, unpublished observation.

15

-

4.2

-

3.0

-

2.6

-

FIG. 2. Hybridization of restriction fragments from plasmid BM3-1with the oligonucleotide (17-mer) probe. The plasmid (1 pg) was digested with either EcoRI (lanes2 and 4 ) or HindIII (lanes 1 and 3 ) and separated ona 0.7% agarose gel. The developed gel was then visualized by UV after treatment with ethidium bromide (lanes I and 2) and by autoradiography of the hybridized bandsafter treatment with the“P-labeled oligonucleotide probe (lanes 3 and 4 ) .

publishedsequencewas erroneous for this residue. Since position 6 was in the middle of the sequence utilized for the synthesis of the oligonucleotide probe, the resulting 17-mer had one or two mismatched bases, depending on the exact nucleotidesequence of the gene encoding cytochrome P 4 5 0 ~ ” ~In. spiteof this mismatch,however, the hybridization results (Fig. 2) constituted additional evidence (along with identicalelectrophoretic mobility andantigenicity)that BMs., gene product was the apoprotein of cytochrome P4 5 0 ~ ~Indeed, . ~ . in view of a report by Unger et al. (27) on the functional expression of the Pseudomonas putida P-450,, gene in E. coli, we thought itpossible that the119-kDa product from the transformedE. coli JM103 might also contain P-450 heme and perhaps could function asa fatty acid monooxygenase. As the data in Table I reveal, this is clearly the case. While there is no detectable cytochrome P-450 or monooxygenase activity in preparation from E. coli transformed by the insert-freepUC13 vector, preparations from E. coli clone BMs.l show both a high cytochrome P-450content anda level of myristate hydroxylase activity comparable to that routinely observed in preparations from B. megaterium grown in the presence of 3 mM pentobarbital. Functionality of the P - 4 5 0 8 ~ Gene . ~ Promoter in E. coliTable I also indicates that the expression of the P - 4 5 0 ~ ~ . ~ holoenzyme by clone BM3.1 is independent of P-IPTG. This

Cloning and Expression of a Cytochrome P-450 from B. megaterium

6679

TABLEI Functional expression of the cytochrome P - 4 5 0 ~ ~gene . 3 cloned into E. coli (plasmid BMR-I) Cell-free, soluble protein extracts from the bacterial cells were assayed for cytochrome P-450 content and for

myristate hydroxylase activity as described under "Experimental Procedures." -~ Bacterial strain

Transforming plasmid

E. coli JM103

None B. megaterium 14581

Cytochrome P-450 content

Myristate hydroxylation

nrnol/ny:protein

nrnollrninlrng protein

None 20 mM 8-IPTG 4 mM pentobarbital

BM3-1

None

0.00 0.00 0.00 0.115

BM3-I BM3.1

20 mM 8-IPTG 4 mM pentobarbital

0.118 0.122

None

4 mM pentobarbital

0.196

0 0 0 205 210 220 5 260

puc13 puc13 puc13

None

Additions to standard growth medium -

by the chromosome of E. coli JM103 (Fig. 3B). To confirm that the cloned P-45oRM.3 gene promoter was functional in E. coli, we subcloned the 9.2-kb XbaI fragment of the BM3., insert into an E. coli/B. subtilis shuttle vector -1 19 (pUCC) that was simply the fusion product of pUC13 and "B-Gal pC194 (28) as shown in Fig. 423. Theresulting plasmids, of the termed BM3.2A or BM3.2, depending on the orientation 9.2-kb insert with respect to thelac Z promoter of the vector, were used to transform E. coli JM103. As shown in Table 11, both clones BM3.2A and BM3.,, produce levels of cytochrome P-450 and monooxygenase activity that are 3-4-fold greater thanthose of clone BM3.1. More importantly,thesedata confirm that the synthesis of P"i5oRM.3 in E. coli did not depend ona promoter from the vector since theexpression of the insertwas not affected by its orientation. The implication, of course, is that the promoter for t h e P - 4 5 0 ~ " ~gene of B. megaterium can function efficiently in E. coli. Induction of the Cloned P-450~"~Gene by PentobarbitalAlthough the cloned P-450BM.3gene was under the controlof its own promoter in E. coli, it was expressed constitutively at a very high level but was unaffected by pentobarbital in this strain (see Tables I and I1 and Fig. 3A). There were several possible explanations for this finding. It was conceivable, for example, that, despite the presence of the active promoter, A B FIG.3. Effects of 8-IPTC or pentobarbital on the synthesisthe regulatory sequences involved in the barbiturateresponse of cytochrome P-450sm.s apoprotein i n E. coli containing were not included in thecloned DNA. We thought thisexplaplasmid BMaSl.Sample preparation and the Western blotting pro- nation unlikely, however, since we had located the P'450RM.3 cedure are described in the legend for Fig. 1. A shows the results for gene on a 5-kb BglII fragment (see Fig. 4A) and had inferred E. coli containing plasmid BM3.1in the absence of inducers (lane I ) , was at least a3.8-kb in the presence of 4 mM pentobarbital (lane 2 ) or 20 mM 8-IPTG from the restriction map that there (lane 3 ) . all immunoblottedwithantibody against cytochrome P- sequence upstream from the translation initiation siteof the 4 5 0 8 ~ . ~B. shows a duplicate run in which immunoblotting was cloned P-450BM.3gene. A second, more probable, explanation performed with anti-8-galacosidase. was that the regulatory sequences were present but that one or more components that interact with the regulatoryseunexpectedfinding thus strongly suggested that P-45oRM.3 quences were not produced in E. coli. To test thishypothesis, synthesis in E. coli was not under the control of the lac Z we introduced either the shuttlevector, pUCC, or BMR.2A (see promoter present on the vector. To test this possibility, WestFig. 4B) into B. megaterium 14581. B. megaterium, as transern blot analyses for P-450BM.3were performed on preparations from E. coli clone BM3., grown in thepresence of either formed by pUCC, behaves exactly like the nontransformed bacterium; it produces only a trace amount of P - 4 5 0 ~ " ~in @-IPTGor pentobarbital or in the absence of both. As Fig. the absence of inducer and a high level in the presence of 4 3A shows, thereappearedto be nosignificant difference amongthethreepreparationsintheir levels of 119-kDa mM pentobarbital (Fig. 5A, lanes 1 and 2). On the other hand, protein reacting with anti-P-450~"~. On the otherwhen hand,B. megaterium transformed by plasmid BMB.2A synthesizes a the same preparations were tested with anti-P-galactosidase, significant amount of P-45OR"3 protein in the absence of only that from theculturetreated with @-IPTG showed inducer (Fig. 5A, lane 3 ) and an extremely high level in the expression of the defective @-galactosidase protein4 encoded presence of 4 mM pentobarbital (Fig. 5A, lane 4 ) . This result was interpreted as indicating that P-450BM.3 the gene residing The a-donor polypeptide could notbe detected by Western blot- on the plasmid vector was also subject to induction by penting in any of the samples including that derivedfrompUC13we contransformed E. coli grown in the presence of 8-IPTG. In view of the tobarbital. T o conclusively establishthispoint, structed a deletion derivative of plasmid BM3.2A, designated reported instability of the a-polypeptides (26). this result wasnot I (see Fig. 4B). The portiondeleted (about plasmid BMJ.,,.,, surprising. 1

2

3

kDa

1

I.

2

..

3


6680

ofCytochrome a

P-450from B. megaterium

TABLE I1 Functwnul expression of the cytochrome P - 4 5 0 ~ ~gene . 3 cloned into E. coli (plnsmids BM,., and BM3.2R) Cell-free, soluble protein extractsfrom E. coli JM103, transformed as indicated, were assayed for cytochrome P-450 content and myristate hydroxylase activity as described under "Experimental Procedures."

A

POLYLINKER

"_

"_

" " " " " " " " "

f"H"'"

H

H

Transforming plasmid

Additions to standard growth medium

" " "

H

XH

Sa ~

0

2

4

6

a

10

12

1.8 kb

Cytochrome P-450 content

Myristate hydroxylation

nrnoll mg protein

nrnollrninl mg protein

0.00 0.00 0.00 0.579 0.576 0.524 0.542 0.525 0.536

0.000 0.000 0.000 1028 1008 932 1050 1020 1022

None 20 mM B-IPTG 4 mM pentobarbital None 20 mM 6-IPTG 4 mM pentobarbital None 20 mM 6-IPTG 4 mM pentobarbital

fragment

FIG. 4. Construction of cloning vectors and probes. A shows a restriction map of plasmid BMa.,. The direction of the lac Z gene transcription and translation is shown by the arrow. The bold line shows the 5-kb BglII fragment within which the P-450BM.3 gene is located. The 1.8-kb PstI fragment which hybridized with the oligonucleotide (17-mer) probe deduced from the N-terminal sequence of P - 4 5 0 ~ ~ (7) . 3 was itself used as a probe for Southern blotting. The dashed arrow shows the beginning and endof the deleted segment of DNA in plasmid BM3.2A.De, 1. The abbreviations used in thisfigure for restriction enzyme sites are: B, BamHI; Bg, BglII; E, EcoRV; E,, EcoRI; H, HindIII; P, PstI; S, SmaI; Sa, SalI; Sc, SacI; and X, XbaI. R shows the construction of plasmids BM3.2A, BM3.2~, and BMa.z~.kl1 (the deletion derivative of BM3.2~).In step 1 , HindIIIdigested pUC13 wasligated with HindIII-digested pC194 (a E. subtilis vector) and the ligated preparation was used to transform E. coli to both ampicillin and chloramphenicolresistance. In step2, the resulting fusion vector, designated pUCC,wasdigested with XbaI and ligated with the gel-purified 9.2-kb XbaI fragment from the 12-kb BM,., insert (see A , above). E. coli transformants were screened for the presence of the 9.2-kb insert by gel analyses of plasmid preparations. Twocolonies, BMa.2A (with the orientationof the insertrelative to the lac Z promoter on the vector shown in A above) and BM3.2~ (with theopposite orientation), were isolated. In step3, plasmid BM3. 2A was double-digested with SacI and BamHI (both corresponding to unique restriction sites present in the polylinker region of pUC13), digested with exonuclease 111 to a limited extent, and then digested with mung bean nuclease. The products were recircularized with T, DNA ligase and utilized to transform E. coli. One of the deletion derivatives, BM3.2A.kl1 (with a deleted portionof 2.8 kb) was used for the studies described in Fig. 5.

2.8 kb) encoded the C-terminal portion of P-450BM+ In E. coli, plasmid BMR.2A.Del I directed the synthesis of a 60-kDa protein that reacted with antibodyto P-450BM.3and exhibited a typical P-450 heme absorption spectrum but lacked monotheintact Poxygenase activity. As was the casewith 4 5 0 ~ " gene, ~ the truncatedgene was expressed at a high level constitutively in E, coli but its synthesis was unaffected by pentobarbital (data not shown). When plasmid BM3.2A.Del I was introduced into B. megaterium 14581, the same 60-kDa polypeptide was produced in only a trace amount constitutively (Fig. 5B, lane I ) but was stronglyinduced inthe presence of 4 mM pentobarbital (Fig. 5B, lane 2 ) . It should be noted, however, that the level of the deleted gene product, as detected by Western blotting, appeared to be lower than that

t

1

2

3

4

1

2

FIG. 5. Expression of the cloned intact and truncated P4 5 0 gene ~ ~in ~B. megaterium ~ 14581. Transformed cells of B. megateriurn were disrupted by sonication and centrifugation a t 40,000 X g, and 20 pg of protein from each samplewere electrophoresed and immunoblotted with antiserumto P-450BM.3. The source of each sample was as follows. A , B. megateriurn transformed by plasmid pUCC in the absenceof inducer (lane 1 ) or in the presence of 4 mM pentobarbital (lane 2 ) ; B. megaterium transformed by plasmid BM3.= in the absenceof inducer (lane 3 ) or in the presence of 4 mM pentobarbital(lane4). B, B. megaterium transformed by plasmid BMa.pA.&l in the absence of inducer (lane 1 ) or in the presence of 4 mM pentobarbital (lane 2). The arrow ( E ) locates the 60-kDa protein expressed by the truncatedP - 4 5 0 ~ ~gene. .3

of intact P - 4 5 0 ~ ~This . ~ . may have reflected a lower affinity of anti-P-450sM.3for the 60-kDa fragment or perhaps a more rapid degradation of the truncated protein. Nevertheless, i t was clear from these results that the cloned P - 4 5 0 ~ gene ~.~ contained theregulatory sequences sufficientfor barbituratemediated induction. It is also evident that, in B. megaterium transformed by BM3.2A.Del 1, the expression of the intact P4 5 0 ~ encoded ~ . ~ by the host's chromosome remains tightly repressed in the absence of inducer (Fig. 5B, lane I). Southern Genomic Blot Analysis-In B. megaterium 14581 we had detected at least six distinct plasmids by agarose gel electrophoresis (data notshown). We were also aware that P450,.,, the prototypical bacterial P-450 cytochrome, is encoded on an extremely large plasmid in Pseudomonas putida (27, 29-31). We were therefore interested in determining whether t h e P - 4 5 0 ~ ~gene . 3 of B. megaterium also resided on a plasmid or was a part of the bacterial chromosome. To distinguish between these possibilities, we performed a genomic Southern blot hybridization usingthe 1.8-kb PstI frag-

Cloning and Expression of a Cytochrome P-450 from B. megaterium

kb

1 2 3 4

- 23 - 9.4

6681

chromosomal DNA (lane 2 ) but not with any of the six B. megaterium plasmid bands visible on the original agarose gel nor with E. coli DNA (lane 1 ). When the B. megaterium chromosomal DNA was digestedwith PstI, a 1.8-kb fragment hybridized with theprobe (lane 4 ) . These results support the hypothesisthat, in B. megaterium 14581, cytochrome P4.50~"~ isencoded on the chromosome (although the possibility still exists that P - 4 5 0 ~ ~ is . 3 encoded on an extremely large plasmid that comigrates with chromosomal DNA). Finally, examination of the hybridization results after treatment of B. megaterium chromosomal DNA with PstI (Fig. 6, lane 4 ) reveals that a second band of about 6.4 kb in size also interacts with the probe although much less strongly than does the 1.8-kb fragment. The larger homologous DNA fragment is not part of the BMR.l(12-kb) insert but we have not yet further characterized it. DISCUSSION

Three lines of evidence indicate that we have cloned the cytochrome P-450BM.a gene of B. megaterium 14581. First of all, E. coli clone BM,., produces a protein that has the same size as and cross-reactswith the antibody to the119-kDa P450BM.3apoprotein of B. megaterium (Fig. 1). Second, the BMs., insert, as well as a 4.2-kb Hind111 fragment derived I fromit, hybridized with an oligonucleotide probe, the sequence of which was based on a portion of the N-terminal amino acid sequence of P - 4 5 0 ~ (Fig. ~ . ~ 2). In this regard, we have recently determined that the sequence of at least the first 18 N-terminal residues from the purified 119-kDaprotein produced in E. coli is identical to the N-terminal aminoacid sequence obtained for P-450B~.3 from B. megaterium.2 Perhaps the most convincing evidence that we have cloned the P - 4 5 0 ~ " ~gene is the finding that the 119-kDa protein produced in E. coli from either clone BMR.,(Table I) or clones BM3.2A,2~ (Table 11) not only exhibited a typical cytochrome P-450 (reduced CO versus reduced) difference spectrum but was fully active as a fatty acid hydroxylase. Indeed, the mean turnover number (1848 f 89 nmol of myristate hydroxylate/ nmol of P-450) calculated for the six enzyme samples from clones BM3.2A and BM,.,, (Table 11) is essentially the same as the turnover number of 1931 f 200 obtained for purified B. megaterium 119-kDa monooxygenase assayed by exactly the same procedure (see Ref. 7 , Table I). It should also be noted that, by either P-450 content or specific activity, the twice soluble protein from these E. coli clones contained about as much119-kDaP-450 holoenzyme as did an equivalent preparation from B. megaterium induced with 5 mM pentobarbital ( 7 ) . Thefactthatthepromoter for the B. megaterium P 450BM.3gene functioned with high efficiency in E. coli (Table 11) was a surprising result and one that may be significant for the study of the regulation of expression of this gene. In E. coli, the expression of the cloned P - 4 5 0 ~ gene ~ . ~ was constiFIG.6. Genomic Southern hybridization. Total cellular DNA tutively high and unresponsive to pentobarbital (see Table I1 (10 pg/sample) from either B. megateriurn 14581 (lanes 2 and 4 ) or 3). In B. megaterium, the gene, even as a component and Fig. from E. coli JM103 (lanes I and 3 ) and either uncut (lanes 1 and 2 ) or digested withPstI (lanes 3 and 4 ) was subjected to electrophoresis of a multicopy plasmid, exhibited low constitutive expression but was expressed a t high levels in response to pentobarbital on a 0.7% agarose gel, transferred to a Hybond-N nylon membrane and hybridized under high stringency conditionswith the nick-trans- (Fig. 5). Thus, the cloned gene not only contained its own lated 1.8-kb PstI fragment of plasmid BMa.l (see Fig. 44 ). functionalpromoterbut also retainedthe regulatorysequences involved in the induction of cytochrome P-450BM.3 ment of the BM3., insert (see Fig. 4A). This 1.8-kb fragment synthesis by barbiturates. It is apparent from the comparative hybridized with the oligonucleotide probe (17-mer; data not results in E. coli and B. megaterium, however, that pentobarshown) and thus presumably contained a portion of the P- bital could not simply interact with these regulatory sequences 4 5 0 ~ " ~gene encoding the N-terminal region of the P-450 to cause induction. Rather, as a working hypothesis, we poscytochrome. As Fig.6 shows,the PstI fragment alsohybridized tulate the existence of a specific repressor for the P - 4 5 0 ~ ~ . ~ with the electrophoretic band correspondingB. to megaterium gene that is produced in B. megaterium but not in E. coli.

s-i I-

6.6 4.4

- 2.3 - 2.0

I

-

..-

6682

Cloning and Expression Cytochrome aof

P-450 from B. megaterium

Barbiturates may cause induction by binding to the putative 7. Narhi, L. O., and Fulco, A. J. (1986) J. Biol. Chern. 261, 71607169 repressor and thus competingdirectly with the regulatory 8. Schwalb, H., Narhi, L. O., and Fulco, A. J. (1985) Biochim. sequences forthis molecule. Whenthe P-450BM.3gene is Biophy~.Acta 8 3 8 , 302-311 placed on a plasmid, it still binds repressor but perhaps with 9. Nebert, D. W., Eisen, H. J.,Negishi, M., Lang, M. A., Hjelmeland, L. M., and Okey, A. B. (1981) Annu. Reu. Pharmacal. Toricol. to be a slight somewhat less affinitysincethereappears 21,431-462 increase in the constitutive level of P-450BM.3in B. megute10. Narhi, L. O., and Fulco, A. J. (1982) J. Biol. Chem. 2 5 7 , 2147rium after transformation withBM3.2A(Fig. 5A). Presumably, 2150 this incremental increase in constitutive synthesis cannot be 11. Kim, B.-H., and Fulco, A. J. (1983) Biochem. Biophys. Res. Comrnun. 116,843-850 attributed, even in part, to expression by the chromosomal copy of the gene via a titration effect, since, inB. megaterium 12. Narhi, L.O., Kim, B-H., Stevenson, P. M., and Fulco, A. J. (1983) Biochem. Biophys. Res. Commun. 116,851-858 transformed by the plasmid carrying the truncated structural13. Ruettinger, R. T., Kim, B.-H., and Fulco, A. J. (1984) Biochim. gene, there is nodetectable increase of P-450BM.3in the Biophys. Acta 8 0 1 , 372-380 absence of inducer (Fig. 5B). We are currently attempting to 14. Wen, L.-P., and Fulco, A. J. (1985) Mol. Cell. Biochem. 67, 7781 place the promoter region of the P-450BM.3 gene in front of a 15. Whitlock, J. P., Jr. (1986) Annu. Rev. Phurmacol. Toricol. 26, promoterless chloramphenicol acetyltransferasegene to ena333-369 ble us to more precisely monitor regulatory processes. At the 16. Efimov, V. A., Buryakova, A.A., Reverdatto, S. V., Chakhmakhcheva, 0. G., and Ovchnnikov, Y. A. (1983) Nucleic Acids same time, we have begun efforts to clone and characterize Res. 11,8369-8387 the putative repressor protein and to develop a n in vitro 17. Maniatis, T., Fritsch, E. F., and Sambrook, J. (1982) Molecular system to study the induction mechanism. Cloning:A Laboratory Manual, Cold Spring Harbor Laboratory, Cold Spring Harbor, New York 18. Birnboim, H. C., and Doly, J. (1979) Nucleic Acid Res. 6 , 15131523 19. Miura, K.-I. (1967) Methods Enzymol. 12,543-545 20. Towbin, H., Staehelin, T., and Gordon, J. (1979) Proc. Natl. Acud. Sci. U. S. A. 76,4350-4354 21. Stevenson, P. M., Ruettinger, R. T., and Fulco, A. J. (1983) Biochem. Bwphys. Res. Commun. 112,927-934 22. Gornall, A. G., Bardawill, C. J., and David, M. M. (1949) J. Biol. Chern. 177,751-766 23. Cooper, T. G. (1977) The Tools of Biochemistry, pp. 51-52, WileyInterscience, New York 24. Laemmli, U. K. (1970) Nature 2 2 7 , 680-685 REFERENCES 25. Curiel-Quesada, E., Setlow, B., and Setlow, P. (1983) Proc. Natl. Acad. Sci. U. S. A. 80,3250-3254 Ortiz de Montellano, P. R. (ed) (1986) Cytochrome P-450: Struc26. Vieira, J., and Messing, J. (1982) Gene (Amst.) 19, 259-268 ture, Mechanism and Biochemistry, Plenum Press, New York Sato, R., and Omura, T. (eds) (1978) Cytochrome P-450, pp. 137- 27. Unger, B. P., Gunsalus, I. C., and Sligar, S. G. (1986) J. Biol. Chem. 261,1158-1163 208, Academic Press, New York Miura, Y., and Fulco, A. J. (1975) Biochim. Biophys. Acta 388, 28. Horinouchi, S., and Weisblum, B. (1982) J. Bacteriol. 150,815825 305-317 29. Koga, H., Rauchfuss, B., and Gunsalus, I. C. (1985) Biochem. Ho, P. P., and Fulco, A. J. (1976) Biochim. Biophys. Acta 4 3 1 , Biophys. Res. Commun. 130,412-417 249-256 30. Rheinwald, J. G., Chakrabarty, A. M., and Gunsalus, I. C. (1973) Matson, R. S., Hare, R. S., and Fulco, A. J. (1977) Biochim. Proc. Natl. Acud. Sci. U. S. A. 7 0 , 885-889 Biophys. Acta 487,487-494 31. Chakrabarty, A. M. (1976) Annu. Reo. Genet. 10, 7-30 Ruettinger, R. T., and Fulco, A. J. (1981) J. Biol. Chem. 2 5 6 , 32. Imanaka, T., Fujii, M., Aramori, I., and Aiba, S. (1982)J.Bacterol. 149,824-830 5728-5734

Acknowledgments-We thank Dm. Bruce Birren and Robert Andersen for their expert advice and assistance in many phases of this work. We are also grateful to Drs. Linda 0. Narhi and Richard T. Ruettinger who often shared with us their expertise and their reagents, to Dr. Audree Fowler who kindly provided us with antibody to p-galactosidase, to Jon Von Visger for the preparation of bacterial media, to Ellin James for the preparation of figures, and to Janet Ransom for typing the manuscript. Special thanks go to Dr. Kathryn Calame and Dr. Harvey Herschman for many helpful discussions during the course of this work.

1. 2. 3. 4. 5. 6.