Appl Microbiol Biotechnol (2006) 71: 522–532DOI 10.1007/s00253-005-0190-8

APPLIED MICROBIAL AND CELL PHYSIOLOGY

Barbara Brezna . Ohgew Kweon . Robin L. Stingley .James P. Freeman .Ashraf A. Khan . Bystrik Polek . Richard C. Jones .Carl E. Cerniglia

Molecular characterization of cytochrome P450genes in the polycyclic aromatic hydrocarbondegrading Mycobacterium vanbaalenii PYR-1Received: 29 July 2005 / Revised: 1 September 2005 / Accepted: 9 September 2005 / Published online: 30 November 2005# Springer-Verlag 2005

Abstract Mycobacterium vanbaalenii PYR-1 has the abil-ity to degrade low- and high-molecular-weight polycyclicaromatic hydrocarbons (PAHs). In addition to dioxygenases,cytochrome P450 monooxygenases have been implicatedin PAH degradation. Three cytochrome P450 genes,cyp151 (pipA), cyp150, and cyp51, were detected andamplified by polymerase chain reaction from M. vanbaa-lenii PYR-1. The complete sequence of these genes wasdetermined. The translated putative proteins were ≥80%identical to other GenBank-listed mycobacterial CYP151,CYP150, and CYP51. Genes pipA and cyp150 werecloned, and the proteins partially expressed in Escherchiacoli as soluble heme-containing cytochrome P450s thatexhibited a characteristic peak at 450 nm in reduced carbonmonoxide difference spectra. Monooxygenation metabo-lites of pyrene, dibenzothiophene, and 7-methylbenz[α]anthracene were detected in whole cell biotransformations,with E. coli expressing pipA or cyp150 when analyzed by

gas chromatography/mass spectrometry. The cytochromeP450 inhibitor metyrapone strongly inhibited the S-oxida-tion of dibenzothiophene. Thirteen other Mycobacteriumstrains were screened for the presence of pipA, cyp150, andcyp51 genes, as well as the initial PAH dioxygenase (nidAand nidB). The results indicated that many of the Myco-bacterium spp. surveyed contain both monooxygenasesand dioxygenases to degrade PAHs. Our results providefurther evidence for the diverse enzymatic capability ofMycobacterium spp. to metabolize polycylic aromatichydrocarbons.

Introduction

Polycyclic aromatic hydrocarbons (PAHs) are ubiquitousenvironmental pollutants. Chemically, they constitute aclass of organic compounds containing two or more fusedbenzene rings in linear, angular, or cluster arrangement.Because of their human and ecotoxicity, there is a con-siderable interest to determine the fate of these compoundsin the environment and to consider possible use ofmicroorganisms for remediation of polluted sites (Cernigliaand Sutherland 2001; Kanaly and Harayama 2000; Muelleret al. 1996).

M. vanbaalenii PYR-1, isolated from a petrogenicchemical polluted site, can utilize or biotransform a widerange of PAHs (Khan et al. 2002). Studies of PAH metab-olites showed that this bacterium uses both dioxygenase(s)and cytochrome P450 monooxygenase(s) to metabolizePAHs (Heitkamp et al. 1988; Kelley et al. 1990; Khan et al.2001; Kim et al. 2004a,b, 2005; Moody et al. 2001, 2002,2003, 2004, 2005). While cis-dihydrodiols produced bythis strain are typical metabolites of aromatic ring-hy-droxylating dioxygenases, trans-dihydrodiols (Heitkampet al. 1988; Kelley et al. 1990; Kim et al. 2005; Moodyet al. 2001, 2003, 2004, 2005) are presumably formed bycytochrome P450 catalyzed epoxidation of the aromaticnucleus with enzymatic hydration by epoxide hydrolase.

B. Brezna . O. Kweon . R. L. Stingley .A. A. Khan . C. E. Cerniglia (*)Division of Microbiology,National Center for Toxicological Research,US Food and Drug Administration,3900 NCTR Road,Jefferson, AR 72079, USAe-mail: ccerniglia@nctr.fda.govTel.: +1-870-5437341Fax: +1-870-5437307

B. Brezna . B. PolekInstitute of Molecular Biology, Slovak Academy of Sciences,845 51 Bratislava, Slovakia

J. P. FreemanDivision of Biochemical Toxicology, National Centerfor Toxicological Research, Food and Drug Administration,Jefferson, AR 72079, USA

R. C. JonesDivision of Systems Toxicology, National Centerfor Toxicological Research, Food and Drug Administration,Jefferson, AR 72079, USA

Among other suspected cytochrome P450 metabolitesformed by M. vanbaalenii PYR-1 are 4-hydroxybiphenyl(Moody et al. 2002), dibenzothiophene (DBT) sulfoxide (un-published results), and a 7-hydroxymethyl-12-methylbenz[α]anthracene (Moody et al. 2003). Despite the metabolicevidence that implicated cytochrome P450, it has not beenidentified in M. vanbaalenii PYR-1. However, there aregenetic data on several cytochrome P450 families in dif-ferent Mycobacterium spp. that were not associated withPAH degradation (Aoyama et al. 1996, 1998; Bellamine et al.1999; Cole and Barrell 1998; Jackson et al. 2003; Kellyet al. 2003; Lepesheva et al. 2001; McLean et al. 2002;Mowat et al. 2002; Poupin et al. 1999a; Trigui et al. 2004).In this study, we report the detection and molecular char-acterization of three CYPs from M. vanbaalenii PYR-1,two of which were cloned and expressed in Escherchia coliand assessed for their ability to oxygenate PAHs. In addi-tion, 13 other Mycobacterium strains were screened for thepresence of cytochrome P450 and aromatic ring-hydrox-ylating dioxygenase genes.

Materials and methods

Chemicals

Pyrene, DBT, and piperidine hydrochloride were pur-chased from Sigma Chemical Company (St. Louis, MO,USA). Solvents were purchased from J.T. Baker, Inc.

(Miamisburg, OH, USA). 7-Methylbenz[α]anthracene (7-MBA) was synthesized by Dr. Peter Fu at the NationalCenter for Toxicological Research, (Jefferson, AR, USA).

Strains and media

Mycobacterium strains used in this study are listed inTable 1. For cloning, E. coli host strains EPI300 (Epicentre,Madison, WI, USA), DH5α (Promega, Madison, WI,USA), Novablue (Novagen, Madison, WI, USA), andvector pET-17b (Novagen) were used. Middlebrook 7H11medium was purchased from Remel (Lenexa, KS, USA).Mineral salts medium (MBS) with nutrients and MBS agaramended with phenanthrene by using a spray-plate tech-nique were prepared as described by Heitkamp et al.(1988).

DNA preparation

Mycobacterium strains were cultivated for 7 days onMiddlebrook agar 7H11, or on MBS amended withphenanthrene, if they were capable of phenanthrene PAHutilization. The cells were scraped from the plates, and theirtotal genomic DNAwas isolated with the DNeasy tissue kit(Qiagen, Valencia, CA, USA). Plasmid and fosmid DNApreparations from E. coli strains were made by using theQiaprep Spin Miniprep kit (Qiagen).

Table 1 Mycobacterium strains used in this study that were screened for the presence of cytochrome P450 genes and nidA and nidB genes

Strain Substrate or other characteristic Detection of

nidAc nidBd pipAe cyp150f cyp51g

M. aurum (ATCC 23366) Type strain (−) (−) − + +M. austroafricanum (ATCC 33464) Type strain, related to M. vanbaalenii (−) (−) − + +M. austroafricanum GTI-23a PAHs + + + − −M. chlorophenolicum PCP-1 (ATCC 49826) Polychlorinated phenols (−) (−) − − +M. flavescens PYR-GCK (ATCC 700033) PAHs (−) (+) + + −M. frederiksbergense FAn9T (DSM 44346) PAHs (−) (+) + + +M. gilvum (ATCC 43909) Type strain − − − + −M. gilvum BB1 (DSM 9487) PAHs (+) (+) + + −M. petroleophilum (ATCC 21497) n-Paraffins (−) (−) − + +M. smegmatis mc2155 (ATCC 700084) Transformation host − − − + +M. vaccae JOB-5 (ATCC 29678) Gaseous, long chain, cycloparaffinic

and monoaromatic hydrocarbons(−) (−) − + +

M. vanbaalenii PYR-1(DSM 7251) PAHs (+) (+) + + +Mycobacterium sp. 7E1B1W (ATCC 29676) Gaseous and long chain hydrocarbons (−) (−) − + −Mycobacterium sp. PAH 2.135 (RJGII-135)b PAHs (+) (+) + − +

“+” PCR product of expected size was present, “−” PCR product of expected size was not obtained. In brackets are the cumulative PCR andSouthern hybridization results from the previous study as follows: “(+)” the studied gene is present (Brezna et al. 2003), “(−)” the studiedgene is not present (Brezna et al. 2003)aObtained from Dr. B.W. Bogan at the Gas Technology Institute in Des Plaines, ILbFrom Dr. D. Warshawsky at the University of CincinnaticPCR primers nidAf and nidAr were useddPrimers nidBf and nidBrePrimers RP1F1 and RP1R2fPrimers FM10F1 and FM10R2gPrimers Cyp51F and Cyp51R

523

PCR reactions

Polymerase chain reaction (PCR) primers used in this studyare listed in Table 2. Regular PCR was performed with TaqDNA polymerase and supplied PCR solutions according tothe manufacturer’s instructions (Qiagen). The PCR regimeconsisted of 3 min preincubation at 95°C; 30 cycles of 30-sdenaturation at 94°C, 30-s annealing at 55°C, and 1-minextension at 72°C; followed by a final hold at 72°C for7 min. In each PCR reaction, the concentration of primerswas 0.5 μM each, and the template DNA was added at afinal concentration of 10 pg μl−1.

Proofreading PCR was performed with PCR SupermixHigh Fidelity (Invitrogen, Carlsbad, CA, USA). Theconditions were the same as for regular PCR, but theprimers were added at a final concentration of 0.2 μM andthe extension step of the PCR cycle was 90 s.

Southern hybridization

pipA-specific digoxigenin (DIG)-labeled DNA probe wasprepared using PCR DIG-labeling kit (Roche Diagnostics,Indianapolis, IN, USA), primers RP1F1 and RP1R2, andthe total genomic DNA from M. vanbaalenii PYR-1 as aPCR template. cyp150-specific DIG-labeled DNA probewas prepared in the same way, using PCR primers FM10F1and FM10R2. DNA was transferred from colonies grownon Petri dishes to positively charged nylon membranes(Roche) using the procedures described in the GeniusSystem User’s Guide (Roche). Hybridization and detectionwas performed according to the DIG DNA Labeling andDetection Kit instruction manual (Roche). The results werevisualized using the chemiluminiscent substrate CSPD(Roche).

Detection of cytochrome P450

Primers RP1F1 and RP1R2 for detection of pipA weredesigned according to conserved regions in GenBanksequences AF102510 (pipA from Mycobacterium smeg-matis mc2155) and AJ310142 (morA from Mycobacteriumsp. RP1). Primers FM10F1 and FM10R2 for cyp150 detec-tion were designed from the conserved regions in se-quences AF107047 (probable cytochrome P450 from M.smegmatis mc2155) and AF107046 (probable cytochromeP450 from Mycobacterium sp. FM10). To design primersMSCYP51F1 and MSCYP51R2 for detection of sterol 14α-demethylase cytochrome P450 (cyp51), sequence cover-ing cyp51 and surrounding regions from Mycobacteriumtuberculosis H37Rv (BX842574) was aligned with se-quences from unfinished genome sequencing projects ofM. smegmatis mc2155 and Mycobacterium avium 104available at http://www.tigr.org, the web site of theInstitute for Genomic Research. Total genomic DNAfrom M. vanbaalenii PYR-1 was used as a PCR template.

PCR screening of pipA, cyp150, nidA, and nidB genesin 14 Mycobacterium strains

Total genomic DNA of each strain was used as a PCRtemplate at a final concentration 50 pg μl−1. For screeningof pipA, the primer pairs RP1F1 and RP1R2 were used. Forscreening of cyp150, FM10F1 and FM10R2 were used;Cyp51F and Cyp51R were used for cyp51. In strains wheregenes nidA and nidB were not screened previously (Breznaet al. 2003), the nidA detection using nidAF and nidARprimers and the nidB detection using nidBF and nidBRwere performed.

Table 2 PCR primers used in this study

Primer name Primer sequence Reference microorganism Reference sequence Position

RP1F1 agctggatcctcaacaag Mycobacterium sp. RP1 AJ310142 1,161–1,178RP1F2 tcatcgcgatcatgctc 1,354–1,370Fm10F1 ccctacttcgatcacctgcgc Mycobacterium sp. FM10 AF107046 489–509Fm10R2 ccgaacgcgatgtgctcgcg 1,504–1,485MSCYP51F1 gggccgatgttccagccg M. smegmatis mc2155 contig3312 1,289,527–1,289,544MSCYP51R2 tcgccgagacgccgcgcg 1,291,356–1,291,339PipAclonF acgccatatgtcgtcggccactgtcggttctgtca,b M. vanbaalenii PYR-1 AY485998 1–27PipAclonR agctaagcttcaatggtgatggtgatggtgggaa

gcgggcgtgaagccgaa,c1,200–1,181

Cyp150clonF acgccatatgagcgacttcgacacgatcgactaca,b M. vanbaalenii PYR-1 AY496703 322–348Cyp150clonR agctaagcttcaatggtgatggtgatggtgtcga

accggggtgaacgtgaa,c1,589–1,608

Cyp51F cgacggcctgcctgatcg M. vanbaalenii PYR-1 AY575951 507–524Cyp51R tcctcggggatccggttg 1,137–1,120aItalic denotes parts of primers not aligning to target sequencebUnderlined are NdeI restriction sitescUnderlined are HindIII restriction sites, double underlined are 6xHis-tagged codons

524

Cloning and sequencing

A fosmid genomic library of M. vanbaalenii PYR-1 wasconstructed previously (Stingley et al. 2004a,b). Coloniesof library clones were transferred from the petri dishes topositively charged nylon membranes. pipA- or cyp150-containing clones were identified by colony hybridizationwith pipA- or cyp150-specific DIG-labeled DNA probes,respectively. One positive fosmid clone was selected foreach cytochrome gene. The fosmid DNAwas digested withEcoRI in the case of the PipA-containing clone and withSacI in the case of the CYP150-containing clone. Therestriction fragments were subcloned into pGEM-11zf(+)(Promega), and the resulting subclones were rescreenedby colony hybridization with pipA- or cyp150-specificDIG-labeled DNA probes. pipA- and cyp150-containingsubclones were named pGEM-PIP and pGEM-CYP, respec-tively, and were sequenced. For subcloning into expressionvector pET-17b (Novagen), the genes were amplified withproofreading PCR. In the forward PCR primers, an NdeIrestriction site was incorporated; in the reverse primers, the6-His-tag codon and a HindIII restriction site were incor-porated. Primers PipAclonF and PipAclonR for amplifi-cation of pipA and Cyp150clonF and Cyp150clonR foramplification of cyp150 are listed in Table 2. As the PCRtemplate, the plasmid DNA of pGEM-PIP and pGEM-CYPwas used. The PCR amplicons were subcloned into pET-17b, resulting in plasmids pET-17b-PIP and pET-17b-CYP.Recombinant plasmids were transformed into NovaBlue E.coli host strain and subsequently retransformed into BL21(DE3)pLysS host strain (Novagen).

DNA sequencing was performed on an AppliedBiosystems Model 377 DNA sequencer at the Universityof Arkansas for Medical Sciences, Little Rock, AR, USA.Sequences were compiled, translated, and analyzed usingLasergene software (DNASTAR, Madison, WI, USA) andcompared to similar genes and proteins using onlinedatabase searches (http://www.ncbi.nlm.nih.gov/BLAST/).

Protein expression, purification, and identification

A single colony of BL21(DE3)pLysS E. coli cells ex-pressing cytochrome PipA or cytochrome CYP150, re-spectively, were grown overnight on LB plates with 100 μgampicillin ml−1. A single colony of each E. coli clone wastransferred into 10 ml of liquid LB medium containing100 μg ampicillin ml−1. After overnight incubation withshaking, these starter cultures were added to 250 ml of LBmedium with 100 μg ampicillin ml−1. Cultures wereincubated with shaking at 20°C until the OD600 reached0.5. Subsequently, 2 mM of heme precursor 5-aminolevu-linic acid (ALA), 10 μg ml−1 of FeCl3, and 1 mM of theinducer isopropylthiogalactoside (IPTG) were added. Forapo-P450 synthesis analysis, ALA and FeCl3 were notadded. Cells were incubated overnight at 20°C and thenspun at 4,000×g. The cells were lysed by boiling for 3 minor by sonication of six 10-s bursts at 300 W, with a 10-s

cooling period between each burst depending on its usage.The lysates were centrifuged at 8,400×g to pellet thecellular debris.

The 6xHis-tagged proteins were purified from the totalsoluble protein fraction using the Ni-NTA resin, asdescribed in the Qiaexpressionist handbook (Qiagen). Allthe buffers were adjusted to pH 7.4.

PipA and CYP150 containing covalently attached hemewere identified by heme-staining with dimethoxybenzidine(Francis and Becker 1984) after separation via denaturingpolyacrylamide gel electrophoresis (PAGE). The stainedbands were excised, digested robotically with trypsin, andanalyzed using nano liquid chromatography (LC)–tandemmass spectrometry (MS/MS) on an LCQ Deca XP Plus iontrap mass spectrometer (Thermo, San Jose, CA, USA)(Edmondson et al. 2002). Samples (40 μl) were loadedusing an Endurance autosampler (Micro-Tech Scientific,Vista, CA, USA) onto an IntegraFrit (New Objective,Woburn, MA, USA) vented column (Licklider et al. 2002)(75 μm×3 cm) packed with 1 cm Jupiter C12 material(Phenomenex, Torrance, CA) at 14 μl min−1 and elutedwith a 50-min gradient (0.1–30% B in 35 min, 30–50% Bin 10 min, and 50–80% B in 5 min, where A=99.8% H2O,0.1% acetonitrile, 0.1% formic acid; and B=80% acetoni-trile, 19.9% H2O, 0.1% formic acid) at 200 nl min−1

(generated with a split tee) using an UltraPlus II capillaryHPLC pump (Micro-Tech Scientific) over a 75 μm×15 cmIntegraFrit analytical column packed also with Jupiter C12material. The column was coupled to a stainless steelemitter (30 μm ID×3 cm; Proxeon, Odense, Denmark).MS/MS was performed on the top four ions in each MSscan using the data-dependent acquisition mode. Normal-ized collision energy was set at 35%, and three microscanswere summed following AGC implementation (targetvalues for MS and MS/MS were 2×108 and 6×107 counts,respectively). Dynamic exclusion and repeat settingsensured each ion was selected only once and excludedfor 30 s thereafter. Product ion data were searched againstthe NCBInr protein database using a locally stored copy ofthe Mascot search engine (Matrix Science, London, UK).

Spectrophotometric analysis of cytochrome P450

Total soluble protein fractions were prepared from E. colicells expressing cytochromes PipA and Cyp150 or thosecontaining pET-17b vector without insert, as describedearlier in the protein expression and purification section.Reduced CO difference spectra were measured in solubleprotein extracts as described previously (Omura and Sato1964; Schlenk et al. 1994).

Biotransformation experiments

The transformed E. coli BL21 (DE3)pLysS(pET-17b-PIP),BL21 (DE3)pLysS(pET-17b-PIP)(pBRCD), BL21 (DE3)pLysS(pET-17b-CYP150), BL21 (DE3)pLysS(pET-17b-CYP150)(pBRCD), and control E. coli BL21 (DE3)pLysS

525

(pET-17b) were cultivated analogously as in the proteinexpression experiment. The total culture volumes were50 ml. After the addition of IPTG, FeCl3, and ALA forholo-P450 and IPTG only for apo-P450, the cultures weregrown for 8 h at 20°C with vigorous shaking. Afterwards,the cells were spun at 4,000×g, washed with 50 ml of50 mM sodium–phosphate buffer (pH 7.4), and resus-pended in 20 ml of the same buffer, with the final cellsuspension adjusted to OD600=4.3. To each flask, 7.5 μl ofthe prepared substrate stock solutions, which were 10%DBT, 7-MBA, or pyrene in dimethylformamide, wasadded. For cytochrome P450 inhibition studies, metyr-apone was added to a final concentration of 0.27 mM. Aftera 16-h incubation with shaking at 20°C, the transformationreaction was stopped by the addition of an equal volume ofethyl acetate. The cell suspensions were extracted threetimes with 70 ml ethyl acetate. Combined ethyl acetatefractions were evaporated at 25°C on a vacuum rotaryevaporator, redissolved in 3 ml ethyl acetate, and dried in avacuum evaporator.

Gas chromatography and mass spectrometry

After collection, the metabolites were analyzed by gaschromatography (GC)/electron ionization mass spectrom-etry (EI-MS) with a TSQ 700 or TSQ 7,000 tandem quad-rupole mass spectrometer (ThermoFinnigan, San Jose, CA,USA). The mass spectrometer was operated in the singlequadrupole mode with 70 eV electron ionization (EI) en-ergy and 150°C ion source temperature. AVarian 3,400 gaschromatograph was employed for the GC/EI-MS analyses.Separation was achieved with a 30m×0.25mm×0.25 μmDB-5 ms capillary column (J&W Scientific, Folsom, CA,USA). The column was heated from 70°C to 280°C at20°C min−1. The helium carrier gas flow rate was con-trolled at 15 psi. In order to estimate the relative amountsof the detected metabolites, extracted ion chromatogramsand base peak ions were generated for the molecular ionsand for the metabolites, respectively, and the resultingchromatographic peaks were integrated electronically withthe chromatographic software. Ratios were calculated forthe resulting peak areas and averaged for each metabolite.

Results

Detection, cloning, and sequence analysisof cytochrome P450 in M. vanbaalenii PYR-1

Carbon monoxide difference spetra of cellular lysates ofM.vanbaalenii PYR-1 grown in the presence of PAHsindicated that the 100,000-g supernatant fraction containedtrace levels of cytochrome P450. As a strategy for a moresensitive detection of cytochrome(s) P450 in M. vanbaa-lenii PYR-1, we used a genomic approach and consideredmost likely that member(s) of the CYP150 or the CYP151family would be present since they originate from fast-growing environmentalMycobacterium strains RP1 (Trigui

et al. 2004) and FM10 (AF107046). The presence ofCYP51 in M. vanbaalenii PYR-1 was also hypothesizedbecause of its conservation among several mycobacteriaand even in different biological kingdoms (Aoyama et al.1996).

PCR screening ofM. vanbaalenii PYR-1 genomic DNAwith two primer pairs designed from strain RP1 for pipAgene and from strain FM10 for cyp151 gene (Table 2) gaveexpected PCR products sizes, 0.25 and 1.0 kb, respectively.The preliminary sequencing of PCR products confirmedthat they were indeed parts of the targeted cyp isogenes.Afterwards, the DIG-labeled versions of these PCR prod-ucts were used as probes to screen M. vanbaalenii PYR-1genomic library. Complete sequences of both cyp isogeneswere obtained from positive library subclones. The PCRproduct of expected size 1.8 kb of the third cyp isogene,cyp51, in M. vanbaalenii PYR-1 resulted from the primercombination MSCYP51F1 and MSCYP51R2. Since thisPCR product covered the complete cyp51 gene, no sub-sequent library screening was necessary. The 1.8-kb PCRproduct was sequenced by primer-walking.

The sequence ofM. vanbaalenii PYR-1 pipA region wassubmitted to GenBank under accession number AY485998.This sequence contains a gene for cytochrome P450PipA, a ferredoxin, and a partial sequence of glutaminesynthetase (Fig. 1). This locus organization is identicalto M. smegmatis mc2155 (GenBank accession numberAF102510) and similar toMycobacterium sp. RP1 (GenBankaccession number AJ310142). However, strain RP1 has agene for a putative ferredoxin reductase located between aferredoxin and a putative glutamine synthetase gene unlikeM. vanbaalenii PYR-1. The cytochrome P450 and theferredoxin genes were identified based on ≥86% and≥69% identity, respectively, of proposed proteins to theircounterparts in M. smegmatis mc2155 (AF102510) andMycobacterium sp. RP1 (AJ310142) and by using theconserved domain database search at http://www.ncbi.nlm.nih.gov. Partial sequence of the putative glutamine synthe-tase gene covers only 33 aminoacids at the N terminus ofthe proposed protein, where no conserved domains arepresent. However, this N-terminal sequence is 75%identical to putative protein of M. smegmatis(AF102510), where beta-grasp domain of glutamine syn-thetase (pfam03951) is present.

The sequence of the M. vanbaalenii PYR-1 cyp150region was submitted to GenBank under accession numberAY496703. In addition to the open reading frame (ORF)for cytochrome P450 (cyp150), the sequence contains anincomplete ORF coding for a possible protein that containsa conserved domain of bacterial regulatory TetR-familyproteins, as well as two hypothetical proteins (Fig. 1).There is a similar locus organization in Mycobacterium sp.FM10 (AF107046) and in M. smegmatis (AF107047).CYP150 from M. vanbaalenii PYR-1 is 96 and 88%identical to its counterparts in Mycobacterium sp. FM10and M. smegmatis, respectively, at the protein level.

The sequence of the cyp51 region of M. vanbaaleniiPYR-1 (GenBank accession number AY575951) containsan incomplete ORF coding for a probable oxidoreductase,

526

cytochrome CYP51, and a ferredoxin, and another incom-plete ORF coding for a hypothetical protein of unknownfunction (Fig. 1). This ORF organization is conserved inseveral other Mycobacterium spp. (Jackson et al. 2003).CYP51 from M. vanbaalenii is 80–92% identical to thoseof M. smegmatis, M. avium M. tuberculosis, and Myco-bacterium bovis subsp. bovis (GenBank accession numbersBX842574, AE006970, BX248336, and AE017229 andunfinished genomes of M. smegmatis mc2155 and M.avium 104 at http://www.tigr.org).

A phylogenetic tree was constructed by the neighbor–joining (NJ) approach of the collection of 29 alignedcytochrome P450 protein sequences including three cyto-chrome P450s from M. vanbaalenii PYR-1 (Fig. 2). Thetree shows six distinct groups for the 29 cytochrome P450s.The three cytochrome P450s from M. vanbaalenii PYR-1fall into three different groups but belong to the same classI. PipA of M. vanbaalenii PYR-1 belongs to the PipA(morA) group that shows over 86% identity to each other.CYP150 of M. vanbaalenii PYR-1 belongs to the CYP150group displaying very high identity (≥82%) to CYP150sfrom other Mycobacterium spp. but with very low identity(≤39%) to the remaining CYP150s. CYP51 of M. vanbaa-lenii PYR-1 is placed in CYP51 group and shows very highidentity (≥79%) to CYP51s from Nocardia farcinica

IFM10152 and other Mycobacterium spp. but with verylow identity (≤36%) to other CYP51s. The pairwisedistance values obtained by using Gonnet weight matrixwere less than 0.700 within each group, with the exceptionof the group CYP150. The CYP150 group can be dividedinto two subgroups by using the pairwise distance value,0.700. Within each group, the pairwise distance valueswere less than 0.700.

Expression, purification, and identification of PipAand CYP150

Recombinant His-tagged protein PipA heterologously ex-pressed in E. coli was soluble. After a passage through Ni-NTA resin, partial purification was achieved, i.e., His-taggedPipAwas a predominant protein in the resin eluate (Fig. 3a,lane 5). His-tagged CYP150 was localized mainly in theinsoluble fraction when expressed in E. coli (data notshown), probably due to the formation of inclusion bodies.However, some of the protein was also produced in thesoluble form and was partially purified from the solublefraction using the Ni-NTA resin.

The overexpressed PipA and CYP150 were stained withdimethoxybenzidine, and the two stained bands were ana-

Fig. 1 Physical maps and conserved sequence alignments of thecytochrome P450 monooxygenases and ferredoxins from M.vanbaalenii PYR-1 with those from other sources. The amino acidresidues involved in binding to heme (FX2GX3CXG) and to the[3Fe-4S] cluster (CX5CXnC) are indicated by highlighted char-acters. Designations: CYP151 (PipA), CYP150, and CYP51—cytochromes P450; Fdx—ferredoxins; GlnA—putative glutaminesynthetase; orf2 and orf3—hypothetical proteins; orf4—probableregulatory protein from TetR-family. GenBank accession numbersare as follows. PipA of M. vanbaalenii PYR-1, AY485998; PipA ofM. smegmatis mc2155, AF102510; morA of Mycobacteriumtokaiense THO100, AY816211; morA of Mycobacterium sp. RP1,

AJ310142; morA ofMycobacterium sp. HE5, AY816211; CYP51 ofM. vanbaalenii PYR-1, AY575951; CYP51 of M. avium subsp.paratuberculosis str. k10, NC_002944; CYP51 of M. tuberculosisCDC1551, AE000516; CYP51 of M. tuberculosis H37Rv,NC_000962; M. bovis AF2122/97NC_002945; CYP150A2 of M.smegmatis mc2155, ; CYP150 of M. vanbaalenii PYR-1,AY496703; CYP150 of Mycobacterium sp. FM10, AF107046;CYP of Burkholderia fungorum LB400, NZ_AAAJ03000005;CYP150 of Arthrobacter sp. FB24, NZ_AAHG01000018;CYP107L2 of Streptomyces avermitilis MA-4680, BA000030;CYP of Pseudomonas aeruginosa PA01, NC_002516

527

lyzed by LC–MS/MS. The resultant product ion data weresearched against the public NCBI protein database. The44.8-kDa band for CYP150 matched to cytochrome P450from M. vanbaalenii PYR-1 (AY496703) with 46 uniquepeptides and 75% sequence coverage. The 48.7-kDa bandfor PipA (Fig. 3a, lane 4) also matched to cytochrome P450sfrom the same species with accession number AY485998showing 31 peptides and 78% sequence coverage.

Reduced CO-difference spectra of soluble protein cellextracts from E. coli expressing PipA and CYP150 showeda typical peak at 450 nm, confirming the cytochrome P450-like character of these proteins (Fig. 3b). A negative con-trol, E. coli containing a vector without insert, showed nopeak at 450 nm. (Fig. 3b).

Biotransformation of DBT, 7-MBA, and pyreneby PipA and CYP150

The expression of pipA and cyp150 in E. coli lacking theelectron-transport components produced one metabolite

from DBT, which had a retention time (12.44 min) (Fig. 4a,b)and mass spectral fragmentation pattern (m/z 200 and m/z184) (Fig. 4f) identical to DBT 5-oxide (Schlenk et al. 1994).Moreover, supplementation of pBRCD, containing the cis-trons encoding [3Fe-4S] ferredoxin (phdC) and ferredoxinreductase (phdD) component from Nocardioides sp. KP7(Saito et al. 2000) in pipA and cyp150, increased DBT 5-oxide formation twofold.

To confirm PipA DBT S-oxidase activity, the effects ofPipA inhibitor metyrapone and ALA and FeCl3 weretested. As shown in Fig. 4c, metyrapone markedly inhib-ited DBT-sulphoxide production in E. coli containing pET-17b-PIP (43% decrease), but was less efficient in E. colicontaining pET-17b-PIP (pBRCD) (11.8% decrease) (datanot shown).

To support more direct evidence of the role of PipA inDBT oxidation, apo-PipA lacking heme was expressedwithout the addition of ALA and FeCl3 and confirmed byheme-staining with dimethoxybenzidine (Fig. 3a, lanes 8and 9). Apo-PipA and the negative control showed no DBT5-oxidation activity (Fig. 4d,e). This result indicates that

Fig. 2 Phylogenetic tree obtained from the alignment of threecytochrome P450s from M. vanbaalenii PYR-1 with relatedproteins. The protein sequences of the 29 cytochrome P450s areclassified. The amino acid sequences were aligned with the ClustalX package (version 1.83), and the tree was constructed by the NJmethod and displayed with the program TreeView X (1.6.6). Scalebar indicates the percentage divergence. The pairwise distancematrix was obtained by using Clustal X (Gonnet 250). Class Icytochrome P450s are three-component systems comprising of aflavin adenine dinucleotide (FAD)-containing reductase, an iron–sulfur protein (ferredoxin), and a cytochrome P450. The eukaryoticclass I enzymes are associated with the mitochondrial membrane.Class II cytochrome P450s are two-component systems, and bothclass III and class IV are a single polypeptide. In a class II system,the cytochrome P450 is partnered with a diflavin (FDA/FMN)

reductase, whereas in the class III systems, the diflavin (FDA/FMN)reductase is fused to the cytochrome P450. Class IV system is madeup of FMN-containing reductases with a ferredoxin-like centerlinked to a cytochrome P450 (Roberts et al. 2002). GenBankaccession numbers are as follows (refer to Fig. 1 for the remainingprotein sequences): P450 of Ralstonia metallidurans CH34,NZ_AAAI00000000; P450RhF of Rhodococcus. sp. NCIMB9784, AF459424; P450 of Rhodococcus rubber DSM 44319, ;P450cam of Pseudomonas putida, M12546; P450 Novosphingo-bium aromaticiviorans DSM 12444, NZ_AAAV02000002; CYP505of Fusarium oxysporum MT-811, AB030037; P450BM-3 of Bacil-lus megaterium Fulco PB85, J04832; P450 of Bacillus cereus ATCC14579, AE017008; P450 of B. cereus E33L, NC_006274; CYP51 ofSorghum bicolor SS1000, U74319; CYP51 of Homo sapiens,CH236949; CYP51 of N. farcinica IFM 10152

528

PipA was responsible for S-oxygenation of DBT andneeded heme as a prosthetic group for enzyme activity. Asshown in Fig. 3a (lanes 6 and 8), the addition of ALA andFeCl3 did not increase the expression level of PipA.

When the substrate was 7-MBA, one metabolite wasobserved in each sample eluting at 17.7 min. The mass spec-trum consisted of an apparent molecular ion at m/z 258 and amajor fragment ion at m/z 229. The mass spectrum and re-tention time are consistent with an authentic 7-hydroxymethyl-benz[α]anthracene (Fig. 5b) (Cerniglia et al. 1982).

GC/MS analysis of the pyrene extracts produced threechromatographic peaks with apparent molecular mass of218 that contained a major fragment ion at m/z 189. Thesepeaks eluted at 16.6, 16.8, and 19.5 min. The mass spectralfragmentation data were identical to pyrenols (Cerniglia

et al. 1986). Authentic 1-hydroxypyrene eluted at the sameretention time (19.5 min) and produced the same massspectrum as the third peak. Since pyrene is a symmetricalmolecule, the only isomers that could be formed are 1-hy-droxy-, 2-hydroxy-, or 4-hydroxypyrene (Fig. 5c).

Fig. 3 Expression and purification of recombinant PipA of M.vanbaalenii PYR-1 at 20°C. aLane M molecular size marker, lane 1cell extract from E. coli (BL21)(pET-17b), lane 2 cell extract fromE. coli (BL21)(pET-17b-PIP) prepared by glass bead cell disruption,lane 3 cell extract from E. coli (BL21)(pET-17b-PIP) prepared byboiling, lane 4 heme-stain of the same cell extract as lane 2, lane 5partial purification of 6xHis-tagged PipA on Ni-NTA resin(Qiagen), lane 6 Coomassie blue stained cell extract from E. coli(BL21)(pET-17b-PIP) grown in the presence of ALA and FeCl3,lane 7 heme-stain of lane 6, lane 8 Coomassie-blue-stained cellextract from E. coli (BL21)(pET-17b-PIP) grown in the absence ofALA and FeCl3, lane 9 heme-stain of lane 8. b Reduced COdifference spectra of total soluble E. coli protein extracts fromcultures of expressing transgenic PipA (solid line), CYP150 (dashedline), and E. coli containing pET-17b vector without insert (dash anddotted line). The protein concentrations were 6 mg ml−1

Fig. 4 GC/EI-MS extracted ion chromatograms of DBT extracts for(a) CYP150, (b) PipA, (c) PipA+metyrapone, (d) apo-PipA, (e)negative control sample, and (f) mass spectrum of DBT 5-oxide

Fig. 5 Reactions catalyzed by E. coli cells expressing cytochromesP450 PipA and CYP150 from M. vanbaalenii PYR-1

529

Detection of cytochrome P450 genes pipA, cyp150,and cyp51 and of dioxygenase genes nidA and nidBin Mycobacterium strains

The amplification of 0.25-kb PCR product indicating pipApresence, 1.0-kb product indicating cyp150, and 0.63-kbproduct indicating cyp51 presence varied among the strains,as documented in Table 1. nidA and nidB genes codingfor the large and small subunits of an aromatic ring-hy-droxylating dioxygenase were detected by PCR in Myco-bacterium austroafricanumGTI-23.Mycobacterium gilvumATCC 43909 and M. smegmatis mc2155 did not producethe PCR products (Table 1). For the rest of the strains, theresults of nidA and nidB screening from the previous study(Brezna et al. 2003) are summarized in Table 1.

Discussion

M. vanbaalenii PYR-1 was the first organism known toproduce both cis-dihydrodiol and trans-dihydrodiol me-tabolites of high-molecular-weight PAHs such as pyrene(Heitkamp et al. 1988; Kelley et al. 1990; Kim et al. 2005;Moody et al. 2001), indicating that both dioxygenase(s)and cytochrome P450 monooxygenase(s) can initiate PAHdegradation in this bacterium. One of these enzymes, aro-matic ring-hydroxylating dioxygenase, is encoded by nidAand nidB genes and has been cloned and characterizedpreviously (Khan et al. 2001; Kim et al. 2004a; Stingleyet al. 2004b). This study complements the previous in-formation by identifying three genes encoding alternativePAH-oxidative enzymes, cytochromes P450, in this organ-ism. To our knowledge, this is the first study proving thatfunctional cytochrome P450 genes can coexist with aro-matic ring-hydroxylating dioxygenase in a high-molecular-weight PAH utilizer. Three CYPs were detected in M.vanbaalenii PYR-1 using the PCR approach that were>80% identical to other mycobacterial CYP151, CYP150,and CYP51, respectively. However, considering the highnumber of CYP isozymes in complete genomes of somemycobacteria, i.e., 20 CYPs in M. tuberculosis (Cole andBarrell 1998), 18 CYPs in M. bovis (Garnier et al. 2003),42 CYPs in M. avium ssp. paratuberculosis (NC_002944),and approximately 40 CYPs inM. smegmatis (Jackson et al.2003) and M. avium 104 (Kelly et al. 2003), it is quitelikely that M. vanbaalenii also has more than three CYPs.Only the complete genome sequence of M. vanbaaleniiPYR-1 will tell the total number of CYPs present in thisbacterium.

Studies on the activity of bacterial CYPs towards PAHsare scarce (Carmichael and Wong 2001; England et al.1998; Harford-Cross et al. 2000; Joo et al. 1999; Li et al.2001; Taylor et al. 1999), and none of these previouslyassayed CYPs originate from a Mycobacterium or from aPAH utilizing strain. In our study, two CYPs from M.vanbaalenii PYR-1, PipA and CYP150, were heterolo-gously expressed in E. coli, and whole cell biotransforma-tion experiments were performed to prove their ability tooxygenate PAHs.

Cytochrome P450s are multicomponent enzymes con-sisting of two separated functional classes, electron transferand oxygenation. Interaction and complementation be-tween two functional classes are necessary for the fullcatalytic function. The expression in E. coli of both PipAand CYP150 from M. vanbaalenii PYR-1 with functionalactivity suggests that the electron transport system for PipAand CYP150 can be complemented by the unidentifiedelectron transport systems of E. coli used as a host. Thisresult has been observed in many cytochrome P450s anddioxygenases (Joo et al. 1999; Khan et al. 2001; Kurkela etal. 1988; Laurie and Lloyd-Jones 1999; Simon et al. 1993).Moreover, the supplementation of phdCD electron-trans-port system of a nonheme dioxygenase from Nocardioidessp. KP7 (Saito et al. 2000) increased the transformingactivity of PipA and CYP150. This suggests the relativelylow specificity of the electron-transport systems towardPipA and CYP150. The PhdC protein used in our studywas the first example of the [3Fe-4S] type of ferredoxin innonheme dioxygenases (Saito et al. 2000) and has 38%identity to the [3Fe-4S] type of ferredoxin component ofPipA from M. vanbaalenii PYR-1. The three cysteine res-idues that serve as ligands for the iron–sulfur cluster areconserved in both PhdC and ferredoxin of PipA (Fig. 1).This tolerance between cytochrome P450s and other elec-tron transport systems gives clues for understanding therelationships in the number and the specificity of a redoxpartnership between cytochrome P450 and electron trans-port systems.

In order to obtain direct evidence of the oxidation ofPAHs by cytochrome P450 inM. vanbaalenii PYR-1, PipAthat was expressed mainly as a functionally active formin the cytosol was chosen and tested for the inhibition ofcytochrome P450 and the oxidation activity of apo-PipA.Based on the previous investigation (Schlenk et al. 1994),DBT and metyrapone were used as a substrate and in-hibitor, respectively. S-oxidation of DBT in E. coli that canexpress active PipA was decreased by metyrapone. In ad-dition, apo-PipA lacking heme could not formDBT 5-oxide.These conclusive results support the metabolic evidence ofthe involvement of cytochrome P450s in PAH metabolismin M. vanbaalenii PYR-1. According to Richardson et al.(1995), the influence of ALA on cytochrome P450 synthe-sis in E. coli depends on the form of cytochrome P450. Theaddition of ALA and FeCl3 did not increase the expressionlevel of PipA, based on the intensity of the bands corre-sponding to holo-PipA and apo-PipA in the SDS-PAGE gel(Fig. 3a, lanes 6 and 8), indicating that heme synthesis wasnot a rate-limiting step in PipA production.

According to PCR results, pipA, cyp150, and cyp51detection varied among the strains and exhibited no corre-lation with the strains’ PAH-degrading ability (Table 1). Incontrast to this, the alternative PAH-oxygenation enzyme,the aromatic ring-hydroxylating dioxygenase encodedby nidA and nidB genes, was consistently present in PAH-utilizing mycobacteria and absent in PAH nonutilizers(Table 1) (Brezna et al. 2003). It seems that the genes nidAand nidB are specialized for the degradation of PAHs,whereas the primary role of pipA, cyp150, and cyp51 is

530

different and the PAH monooxygenation is only a fortu-itous or nonspecific reaction. The true biological functionof PipA (CYP150 family) is degradation of heterocyclicamines like morpholine or piperdine (Poupin et al. 1999a,b; Sielaff et al. 2001; Taylor et al. 1999; Trigui et al. 2004),while the physiological roles of CYP150 and mycobacte-rial CYP51 are obscure (Cole and Barrell 1998). Regard-less of the primary function of the three CYPs, the resultsof the PCR screening showed that it is not unusual forsoil mycobacteria to have both ring-hydroxylating dioxy-genases and cytochromes P450 since all of the strains thatcontained nidAB genes possessed at least one of the studiedcyp isogenes. Thus, several mycobacteria have a potentialto use both dioxygenation and monooxygenation reactionsfor initial biotransformation of PAHs.

Acknowledgements This work was supported in part by anappointment to the Postgraduate Research Program at the NationalCenter for Toxicological Research administered by the Oak RidgeInstitute for Science and Education through an interagency agree-ment between the US Department of Energy and the US Food andDrug Administration.

References

Aoyama Y, Noshiro M, Gotoh O, Imaoka S, Funae Y, Kurosawa N,Horiuchi T, Yoshida Y (1996) Sterol 14-demethylase P450(P45014DM*) is one of the most ancient and conserved P450species. J Biochem (Tokyo) 119:926–933

Aoyama Y, Horiuchi T, Gotoh O, Noshiro M, Yoshida Y (1998)CYP51-like gene ofMycobacterium tuberculosis actually encodesa P450 similar to eukaryotic CYP51. J Biochem (Tokyo) 124:694–696

Bellamine A, Mangla AT, Nes WD, Waterman MR (1999)Characterization and catalytic properties of the sterol 14alpha-demethylase from Mycobacterium tuberculosis. Proc Natl AcadSci U S A 96:8937–8942

Brezna B, Khan AA, Cerniglia CE (2003) Molecular characteriza-tion of dioxygenases from polycyclic aromatic hydrocarbon-degrading Mycobacterium spp. FEMS Microbiol Lett 223:177–183

Carmichael AB, Wong LL (2001) Protein engineering of Bacillusmegaterium CYP102. The oxidation of polycyclic aromatichydrocarbons. Eur J Biochem 268:3117–3125

Cerniglia CE, Sutherland JB (2001) Bioremediation of polycyclicaromatic hydrocarbons by ligninolytic and non-ligninolyticfungi. In: Gadd GM (ed) Fungi in bioremediation. CambridgeUniversity Press, Cambridge, pp 136–187

Cerniglia CE, Fu PP, Yang SK (1982) Metabolism of 7-methylbenz[a]anthracene and 7-hydroxymethylbenz[a]anthracene by Cun-ninghamella elegans. Appl Environ Microbiol 44:682–689

Cerniglia CE, Kelly DW, Freeman JP, Miller DW (1986) Microbialmetabolism of pyrene. Chem Biol Interact 57:203–216

Cole ST, Barrell BG (1998) Analysis of the genome of Mycobac-terium tuberculosis H37Rv. Novartis Found Symp 217:160–172; discussion 172–167

Edmondson RD, Vondriska TM, Biederman KJ, Zhang J, Jones RC,Zheng Y, Allen DL, Xiu JX, Cardwell EM, Pisano MR, Ping P(2002) Protein kinase C epsilon signaling complexes includemetabolism- and transcription/translation-related proteins: com-plimentary separation techniques with LC/MS/MS. Mol CellProteomics 1:421–433

England PA, Harford-Cross CF, Stevenson JA, Rouch DA, WongLL (1998) The oxidation of naphthalene and pyrene bycytochrome P450cam. FEBS Lett 424:271–274

Francis RT Jr, Becker RR (1984) Specific indication of hemopro-teins in polyacrylamide gels using a double-staining process.Anal Biochem 136:509–514

Garnier T, Eiglmeier K, Camus JC, Medina N, Mansoor H, Pryor M,Duthoy S, Grondin S, Lacroix C, Monsempe C, Simon S,Harris B, Atkin R, Doggett J, Mayes R, Keating L, WheelerPR, Parkhill J, Barrell BG, Cole ST, Gordon SV, Hewinson RG(2003) The complete genome sequence of Mycobacteriumbovis. Proc Natl Acad Sci U S A 100:7877–7882

Harford-Cross CF, Carmichael AB, Allan FK, England PA, RouchDA, Wong LL (2000) Protein engineering of cytochrome p450(cam) (CYP101) for the oxidation of polycyclic aromatichydrocarbons. Protein Eng 13:121–128

Heitkamp MA, Freeman JP, Miller DW, Cerniglia CE (1988) Pyrenedegradation by a Mycobacterium sp.; identification of ringoxidation and ring fission products. Appl Environ Microbiol54:2556–2565

Jackson CJ, Lamb DC, Marczylo TH, Parker JE, Manning NL,Kelly DE, Kelly SL (2003) Conservation and cloning ofCYP51: a sterol 14 alpha-demethylase from Mycobacteriumsmegmatis. Biochem Biophys Res Commun 301:558–563

Joo H, Lin Z, Arnold FH (1999) Laboratory evolution of peroxide-mediated cytochrome P450 hydroxylation. Nature 399:670–673

Kanaly RA, Harayama S (2000) Biodegradation of high-molecular-weight polycyclic aromatic hydrocarbons by bacteria. J Bacteriol182:2059–2067

Kelley I, Freeman JP, Cerniglia CE (1990) Identification ofmetabolites from degradation of naphthalene by a Mycobacte-rium sp. Biodegradation 1:283–290

Kelly SL, Lamb DC, Jackson CJ, Warrilow AG, Kelly DE (2003)The biodiversity of microbial cytochromes P450. Adv MicrobPhysiol 47:131–186

Khan AA, Wang RF, Cao WW, Doerge DR, Wennerstrom D,Cerniglia CE (2001) Molecular cloning, nucleotide sequence,and expression of genes encoding a polycyclic aromatic ringdioxygenase from Mycobacterium sp. strain PYR-1. ApplEnviron Microbiol 67:3577–3585

Khan A, Kim SJ, Paine DD, Cerniglia CE (2002) Classification of apolycyclic aromatic hydrocarbon-metabolizing bacterium, My-cobacterium sp. strain PYR-1, as Mycobacterium vanbaaleniisp. Int J Syst Evol Microbiol 52:1997–2002

Kim SJ, Jones RC, Cha CJ, Kweon O, Edmondson RD, CernigliaCE (2004a) Identification of proteins induced by polycyclicaromatic hydrocarbon in Mycobacterium vanbaalenii PYR-1using two-dimensional polyacrylamide gel electrophoresis andde novo sequencing methods. Proteomics 4:3899–3908

Kim YH, Moody JD, Freeman JP, Engesser KH, Cerniglia CE(2004b) Evidence for the existence of PAH-quinone reductaseand catechol-O-methyltransferase in Mycobacterium vanbaale-nii PYR-1. J Ind Microbiol Biotechnol 31:507–516

Kim YH, Freeman JP, Moody JD, Engesser KH, Cerniglia CE(2005) Effects of pH on the degradation of phenanthrene andpyrene by Mycobacterium vanbaalenii PYR-1. Appl MicrobiolBiotechnol 67:275–285

Kurkela S, Lehvaslaiho H, Palva ET, Teeri TH (1988) Cloning,nucleotide sequence and characterization of genes encodingnaphthalene dioxygenase ofPseudomonas putida strainNCIB9816.Gene 73:355–362

Laurie AD, Lloyd-Jones G (1999) The phn genes of Burkholderiasp. strain RP007 constitute a divergent gene cluster forpolycyclic aromatic hydrocarbon catabolism. J Bacteriol 181:531–540

Lepesheva GI, Podust LM, Bellamine A, Waterman MR (2001)Folding requirements are different between sterol 14-alpha-demethylase (CYP51) from Mycobacterium tuberculosis andhuman or fungal orthologs. J Biol Chem 276:28413–28420

Li QS, Ogawa J, Schmid RD, Shimizu S (2001) Engineeringcytochrome P450 BM-3 for oxidation of polycyclic aromatichydrocarbons. Appl Environ Microbiol 67:5735–5739

531

Licklider LJ, Thoreen CC, Peng J, Gygi SP (2002) Automation ofnanoscale microcapillary liquid chromatography-tandem massspectrometry with a vented column. Anal Chem 74:3076–3083

McLean KJ, Cheesman MR, Rivers SL, Richmond A, Leys D,Chapman SK, Reid GA, Price NC, Kelly SM, Clarkson J,Smith WE, Munro AW (2002) Expression, purification andspectroscopic characterization of the cytochrome P450 CYP121from Mycobacterium tuberculosis. J Inorg Biochem 91:527–541

Moody JD, Freeman JP, Doerge DR, Cerniglia CE (2001) Degra-dation of phenanthrene and anthracene by cell suspensions ofMycobacterium sp. strain PYR-1. Appl Environ Microbiol67:1476–1483

Moody JD, Doerge DR, Freeman JP, Cerniglia CE (2002) Degra-dation of biphenyl by Mycobacterium sp. strain PYR-1. ApplMicrobiol Biotechnol 58:364–369

Moody JD, Fu PP, Freeman JP, Cerniglia CE (2003) Regio- andstereoselective metabolism of 7,12–dimethylbenz[a]anthraceneby Mycobacterium vanbaalenii PYR-1. Appl Environ Micro-biol 69:3924–3931

Moody JD, Freeman JP, Fu PP, Cerniglia CE (2004) Degradation ofbenzo[a]pyrene by Mycobacterium vanbaalenii PYR-1. ApplEnviron Microbiol 70:340–345

Moody JD, Freeman JP, Cerniglia CE (2005) Degradation of benz[a]anthracene by Mycobacterium vanbaalenii strain PYR-1.Biodegradation 16:513–526

Mowat CG, Leys D, McLean KJ, Rivers SL, Richmond A, MunroAW, Ortiz Lombardia M, Alzari PM, Reid GA, Chapman SK,Walkinshaw MD (2002) Crystallization and preliminarycrystallographic analysis of a novel cytochrome P450 fromMycobacterium tuberculosis. Acta Crystallogr D Biol Crystal-logr 58:704–705

Mueller JG, Cerniglia CE, Pritchard PH (1996) Bioremediation:principles and applications. Cambridge University Press,Cambridge

Omura T, Sato R (1964) The carbon monoxide-binding pigment ofliver microsomes. I. Evidence for its hemoprotein nature. J BiolChem 239:2370–2378

Poupin P, Ducrocq V, Hallier-Soulier S, Truffaut N (1999a) Cloningand characterization of the genes encoding a cytochrome P450(PipA) involved in piperidine and pyrrolidine utilization and itsregulatory protein (PipR) in Mycobacterium smegmatismc2155. J Bacteriol 181:3419–3426

Poupin P, Godon JJ, Zumstein E, Truffaut N (1999b) Degradation ofmorpholine, piperidine, and pyrrolidine by Mycobacteria:evidences for the involvement of a cytochrome P450. Can JMicrobiol 45:209–216

Richardson TH, Jung F, Griffin KJ, Wester M, Raucey JL, KemperB, Bornheim LM, Hassett C, Omienski CJ, Johnnson EF(1995) A universal approach to the expression of human andrabbit cytochrome P450s of the 2C subfamily in Escherichiacoli. Arch Biochem Biophys 323:87–96

Roberts GA, Gideon G, Andy G, Sabine LF, Nicholas JT (2002)Identification of a new class of cytochrome P450 from aRhodococcus sp. J Bacteriol 184:3898–3908

Saito A, Iwabuchi T, Harayama S (2000) A novel phenanthrenedioxygenase from Nocardioides sp. strain KP7: expression inEscherichia coli. J Bacteriol 182:2134–2141

Schlenk D, Bevers RJ, Vertino AM, Cerniglia CE (1994) P450catalysed S-oxidation of dibenzothiophene by Cunninghamellaelegans. Xenobiotica 24:1077–1083

Sielaff B, Andreesen JR, Schrader T (2001) A cytochrome P450 anda ferredoxin isolated from Mycobacterium sp. strain HE5 aftergrowth on morpholine. Appl Microbiol Biotechnol 56:458–464

Simon MJ, Osslund TD, Saunders R, Ensley BD, Suggs S, HarcourtA, Suen WC, Cruden DL, Gibson DT, Zylstra GJ (1993)Sequences of genes encoding naphthalene dioxygenase inPseudomonas putida strains G7 and NCIB 9816-4. Gene127:31–37

Stingley RL, Brezna B, Khan AA, Cerniglia CE (2004a) Novelorganization of genes in a phthalate degradation operon ofMycobacterium vanbaalenii PYR-1. Microbiology 150:3749–3761

Stingley RL, Khan AA, Cerniglia CE (2004b) Molecular character-ization of a phenanthrene degradation pathway in Mycobacte-rium vanbaalenii PYR-1. Biochem Biophys Res Commun322:133–146

Taylor M, Lamb DC, Cannell R, Dawson M, Kelly SL (1999)Cytochrome P450105D1 (CYP105D1) from Streptomycesgriseus: heterologous expression, activity, and activation effectsof multiple xenobiotics. Biochem Biophys Res Commun 263:838–842

Trigui M, Pulvin S, Truffaut N, Thomas D, Poupin P (2004)Molecular cloning, nucleotide sequencing and expression ofgenes encoding a cytochrome P450 system involved insecondary amine utilization in Mycobacterium sp. strain RP1.Res Microbiol 155:1–9

532