Vol. 59, No. 12APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Dec. 1993, p. 4136-41420099-2240/93/124136-07$02.00/0Copyright © 1993, American Society for Microbiology

Cloning and Mutagenesis of a Cytochrome P-450 Locus fromBradyrhizobium japonicum That Is Expressed

Anaerobically and SymbioticallyRAYMOND E. TULLY AND DONALD L. KEISTER*

Soybean and Alfalfa Research Laboratory, USDA Agricultural Research Service, Building 011, HH-19,Beltsville Agricultural Research Center-West, Beltsville, Maryland 20705

Received 7 June 1993/Accepted 13 September 1993

Cytochromes P-450, which in many organisms participate in the metabolism of a variety of endobiotic andxenobiotic substances, are synthesized by symbiotic bacteroids of Bradyrhizobium japonicum. Polyclonalantibodies were raised against two cytochromes P-450 (CYP112 and CYP114) purified from bacteroids. Alambda gtll expression clone ofB.japonicum USDA 110 DNA that reacted with the anti-CYP112 antibody wasobtained and was used to screen a library ofUSDA 110 genomic DNA in pLAFRi for a clone of the P-450 locus.Forced expression of subclones of the P450 locus in Escherichia coli produced polypeptides that reacted witheither the anti-CYP112 antibody or the anti-CYP114 antibody; no cross-reactivity was evident. A Western blot(immunoblot) analysis showed that neither protein was present in free-living aerobically grown B. japonicumcells, but that both proteins were present in cells grown anaerobically, as well as in bacteroids. A mutant straindisrupted in the CYP112 locus produced neither CYP112 nor CYP114, indicating that the mutation was polarfor CYP114. The mutant produced effective nodules on soybeans, even though the bacteroids contained nodetectable P-450. This suggests that the cytochromes P-450 which we examined are not involved in an essentialsymbiotic fimction.

Cytochromes P-450 mediate a wide variety of chemicalreactions in animals, plants, and microbes, including enzy-matic hydroxylation, demethylation, N oxidation, and re-duction of some azo and nitro compounds (9). Part of thechemical function of these P-450s is to bind and activate 02for subsequent transfer to a substrate. Bradyrhizobiumjaponicum, the soybean microsymbiont, produces cy-tochromes P-450 both in the symbiotic state and in anaero-bically cultured cells (1, 3, 7, 12). The function of thesebradyrhizobial P-450s is unknown, and no enzymatic activityhas been discovered (3). However, P-450 may play a role inan efficient, oxyleghemoglobin-facilitated pathway of oxida-tive phosphorylation (4, 5, 26).Among the other N2-fixing bacteria, an Anabaena sp.

contains a P-450-like sequence which interrupts the nifDgene (15), and Agrobacterium tumefaciens contains twoP-450-like open reading frames in a plant-inducible locus(11), although none of the actual products of these geneshave been identified or shown to be present during N2fixation.As the first step in elucidating the function of bradyrhizo-

bial cytochromes P-450, cloning and mutagenesis of a locusof anaerobically and symbiotically expressed cytochromeP-450 genes from B. japonicum USDA 110 were performed,and the results are described in this paper. A mutant with aP-450- phenotype was symbiotically competent, indicatingthat P-450s are not involved in an essential symbiotic func-tion.

MATERIALS AND METHODS

Bacterial strains, plasmids, phages, and growth. Thestrains, phages, and plasmids used in this study are listed in

* Corresponding author.

Table 1. Escherichia coli was grown in Luria-Bertani me-dium or M9 minimal medium (16) supplemented with tetra-cycline (25 pg/ml) or ampicillin (50 ,ug/ml) for plasmidmaintenance or with 40 ,ug of X-Gal (5-bromo-4-chloro-3-indolyl-p-D-galactopyranoside) per ml to indicate expressionof lacZ reporter genes. An expression library of B. japoni-cum DNA in phage lambda gtll and a cosmid library invector pLAFR1 have been described previously (23).

B. japonicum was grown in AlE-gluconate medium (13)supplemented with kanamycin, streptomycin, and tri-methoprim (100 ,ug/ml [each]) when needed for selection ofthe TnS mutant. Anaerobic cultures were grown in the samemedium supplemented with 6 mM KNO3 (6).

Bacteroid extraction. Nodules were collected from field-grown soybean plants (Glycine max L. Merr.). Bacteroidswere extracted as described previously (23), and a yield ofabout 350 g (fresh mass) of bacteroid cells was obtained from2.5 kg of nodules. The cells were resuspended in 2 volumesof 0.2 M Tris (pH 8.0) containing 1 mM dithiothreitol, 1 mMEDTA, 0.5 mM phenylmethylsulfonyl fluoride, 4 ,ug ofleupeptin per ml, and 4 ,ug of pepstatin per ml. The cells weredisrupted by three passages through a French pressure cellat 18,000 lb/in2, and the cell debris was removed by centrif-ugation. The proteins that precipitated between 25 and 50%ammonium sulfate saturation were collected and redissolvedin a minimal volume of column buffer (10 mM sodiumphosphate [pH 7.0] in a solution containing 20% [vol/vol]glycerol, 0.1 mM EDTA, 1 mM dithiothreitol, 0.5 mMphenylmethylsulfonyl fluoride, 1 ,g of leupeptin per ml, 1 ,ugof pepstatin per ml, and 0.02% NaN3). Four separate por-tions were chromatographed on a Fractogel TSK HW-55column (2.5 by 85 cm), and the peaks containing P-450, asdetected by difference spectra ([reduced plus CO] minusreduced), were collected. The P-450 solutions were broughtto 1 M ammonium sulfate and applied to a column (2.5 by 17cm) containing C4 hydrophobic interaction medium (Toyo-

4136

on May 28, 2018 by guest

http://aem.asm

.org/D

ownloaded from

B. JAPONICUM CYTOCHROME P-450 LOCUS 4137

TABLE 1. Bacterial strains, plasmids, and phages used in this work

Strain, phage, or Description Reference or sourceplasmid

E. coli strainsDH5a Plasmid host for lacZ expression; F- X- +80d lacZ M15 Bethesda Research Laboratories, Inc.,

d(IacZYA-argF)U169 endAl hsdRl7 supE44 thi-1 recAl gyrA96 Gaithersburg, Md.reLU1

SE5000 Plasmid host; F- araD139 dlacU169 rpsL150 reLA1 flbBS301 20deoCi ptsF25 rbsR recAS6

Y1089 Host for Agtll lysogens; dlacU169proA+ dlon araDl39 strA Promega Corp., Madison, Wis.hflA150 (chr::TnlO) (pMC9)

Y1090 Host for Xgtll; dlacU169proA- dlon araD139 strA supF Promega Corp., Madison, Wis.(trpC22::TnlO) (pMC9)

S17-1 Host for pSUP202 conjugation; RP4(Tc::Mu) (Km::Tn7) dTnl 21pro- hsdR hsdM+ recA- thi-

B. japonicum strainsUSDA 110 Wild-type strain USDA-ARS Rhizobium Culture

Collection'BJ1005 USDA 110 CYP112::Tn5 cassette This study

PhagesXgtll diac5 ninS cI857 S100 Promega Corp., Madison, Wis.Xgt11.108A Agtll with in-frame fusion to CYP112 gene from USDA 110 This study

PlasmidspLAFRl Mobilizable cosmid vector; Tcr 8pSUP202 Mobilizable suicide vector; Apr Cmr Tcr 21pUC18 Cloning vector with lacZ reporter gene containing multiple cloning 18

site; Apr6A pLAFR1 clone from USDA 110 genomic library containing the This study

P-450 regionpRET61 pUC18 clone of CYP112 gene in forward orientation in lacZ gene This studypRET60 Reverse orientation of pRET61 This studypRET70 TnS cassette insertion into CYP112 on pRET60, in pUC18 This studypRET72 Same as pRET70, but in pSUP202 This studypRET74 pUC18 clone of CYP114 gene in forward orientation in lacZ gene This study

a USDA-ARS, U.S. Department of Agriculture Agricultural Research Service, Beltsville, Md.

pearl TSK-gel butyl-650 M) equilibrated with column buffercontaining 1 M ammonium sulfate. The column was elutedstepwise with 50-ml aliquots of column buffer containingdecreasing concentrations of ammonium sulfate (1 to 0 M),and the fractions containing two different P-450s were col-lected. These P-450s were designated CYP112 and CYP114by using standard nomenclature (17). At this stage thepartially purified P-450s obtained from the four separateextractions were pooled for final purification. The pooledfractions were rechromatographed with a column containinghydrophobic interaction medium.CYP112 purification. The CYP112-containing fractions

were pooled, pressure concentrated to a small volume byusing an Amicon Diaflo YM30 ultramembrane filter, andbrought to a volume of 10 ml with DEAE column buffer,which contained 8mM Tris in 20% [vol/vol] glycerol-0.1 mMEDTA-1 mM dithiothreitol and was adjusted to pH 7.5 withMES [2-(N-morpholino)ethanesulfonic acid]. The solutionwas applied to a DEAE-cellulose column (2.5 by 15 cm)equilibrated with the same buffer. The column was washedwith buffer containing 0.1 M NaCl, and CYP112 (a brownband) was eluted by increasing the NaCl concentration in theeluent stepwise up to a concentration of 0.15 M. The solutionwas pressure concentrated to a volume of 1 ml and thensubjected to size exclusion chromatography on a FractogelTSK HW-55 column (0.9 by 161 cm) equilibrated with 10mM sodium phosphate (pH 7.0) in 20% (vol/vol) glycerol-1mM dithiothreitol-0.02% NaN3. The brown fraction contain-ing CYP112 was eluted and applied to a column (0.7 by 16.5cm) containing Prissm ceramic hydroxyapatite (Enprotech)

buffered with 8 mM sodium phosphate (pH 7.0) in 20%(vol/vol) glycerol-80 ,uM CaCl2. CYP112, which only slightlyadsorbed to the substrate, was eluted by increasing theconcentration of phosphate in the column buffer to 30 mM.The eluted CYP112 was pressure concentrated to a smallvolume and stored at -20°C in 50% (vol/vol) glycerol. Thepreparation was judged to be essentially homogeneous onthe basis of the results of sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis (PAGE) followed bysilver staining (data not shown).CYP114 purification. The crude CYP114 preparation was

chromatographed on the Fractogel column as describedabove for CYP112. The CYP114 fractions were applied to aDEAE column (1.5 by 44 cm) and eluted with NaCl byincreasing the concentration of NaCl in the eluent stepwiseup to a concentration of 0.25 M. The brown fraction con-taining CYP114 was desalted by pressure concentration andsubjected to preparative isoelectric focusing by using aBio-Rad Rotofor isoelectric focusing cell and Bio-Lyte 3/10ampholytes, as recommended by the manufacturer.To further purify the CYP114 for antibody production, the

protein was subjected to preparative gel electrophoresis andstained with Coomassie blue, and the blue-stained CYP114band was excised and electroeluted.

Immunological procedures and Western blot (immunoblot)analysis. Polyclonal rabbit antisera were prepared against 1mg of purified CYP112 and about 0.5 mg of CYP114 byHazleton Research Products, Inc., Denver, Pa. Non-immu-noglobulin-containing contaminants were removed from theserum by caprylic acid precipitation, followed by precipita-

VOL. 59, 1993

on May 28, 2018 by guest

http://aem.asm

.org/D

ownloaded from

4138 TULLY AND KEISTER

A

97-

66

45-

29-

20.1-

97-

66-

A RB C D E F G H I

B C D E F G H I

45-

2920.1-

FIG. 1. Western blots of bacterial extracts and P-450 prepara-tions reacted with anti-CYP112 (A) or anti-CYP114 (B) antibody.Lanes A and B, purified CYP112 and CYP114, respectively; lanes Cthrough E, extracts of B. japonicum USDA 110 grown aerobically,anaerobically, and symbiotically, respectively; lane F, extract ofstrain BJ1005 grown symbiotically; lanes G through I, extracts of E.coli containing pRET61 (CYP112 gene, forward orientation),pRET60 (CYP112 gene, reverse orientation), and pRET74 (CYP114gene, forward orientation), respectively. All B. japonicum lanescontained between 31 and 36 pg of total protein; all E. coli lanescontained the equivalent of 20 pll of cells from overnight Luria-Bertani broth cultures. The numbers on the left indicate the posi-tions molecular weight standards (x 103). The bands at approxi-mately 45 kDa were presumed to be P-450 bands, since theycomigrated with the purified protein bands in lanes A and B. Theantibody used for the experiment shown in panel A containedcontaminants that reacted with several constitutively expressedproteins in B. japonicum and E. coli strains and were apparently notrelated to the P-450 constructions. The bands in panel B at the lower

tion of the immunoglobulin with ammonium sulfate at 50%saturation (22). The antibodies were used to screen lambdagtll plaque lifts and to process Western blots of SDS-polyacrylamide gels as described previously (23).Recombinant DNA procedures. E. coli plasmids were iso-

lated, digested with restriction enzymes, and used for clon-ing by using standard procedures (16). The E. coli plasmidtransformations were performed by the method of Hanahan(10).

Plant growth and bacteroid isolation. Seeds of soybean (G.max cv. Williams) were sown in modified Leonard jars andinoculated with B. japonicum strains (23). The plants weregrown in a greenhouse under natural daylight conditions for38 days. The plants were harvested, and their shoots wereanalyzed for dry mass and N content (25). (Since shoot Ncontent represents fixation integrated over a period of time,it is considered a better measure of N2 fixation than acety-lene reduction, which is measured at only one or a fewisolated time points.) Nodule number and mass also weredetermined. For Western blot analysis, bacteroids wereextracted by grinding six nodules from each Leonard jar in 1ml of 0.2 M sodium phosphate-1 mM EDTA (pH 7.0). Thebacteroids were recovered by centrifugation at 16,000 x gwith a microcentrifuge, boiled in 0.9 ml of Laemmli SDS-mercaptoethanol treatment buffer, and subjected to SDS-PAGE (14). A larger-scale isolation procedure to obtain apreparation for spectrophotometry was performed by using10 g of nodules. The bacteroids were ruptured with theFrench pressure cell and ultracentrifuged for 1.5 h at 171,000x g. The dark red membrane layer was resuspended inphosphate buffer. Soluble or membrane preparations werereduced with a few grains of sodium dithionite and treatedwith CO. Difference spectra ([reduced plus CO] minusreduced) were recorded with a Shimadzu model UV-3000spectrophotometer operated in the dual-beam mode.

RESULTS

P-450 isolation. The CYP112 and CYP114 proteins werefirst resolved by hydrophobic interaction chromatography.They were judged to be separate and distinct proteins on thebasis of their slightly different mobilities on SDS-polyacryl-amide gels (although this is not clear in the Western blotsshown in Fig. 1, because the two gels were not run simulta-neously). Thus, we inferred that these proteins probably areencoded by two separate genes. The fact that they aredistinct proteins was confirmed by the specificities of theantibodies.Antibody specificity. The anti-CYP112 antibody reacted

with purified CYP112 (Fig. 1A, lane A) and with a polypep-tide having the same mobility and presumed to be CYP112 ina cell-free bacteroid extract from USDA 110 (lane E) on aWestern blot. A weak reaction with an extract obtained fromanaerobically grown cells also was observed (lane D), but noreaction was seen with an extract obtained from aerobicallygrown cells (lane C). Similarly, the anti-CYP114 antibodyreacted with purified CYP114 (Fig. 1B, lane B) and with apolypeptide having the same mobility and presumed to beCYP114 in extracts of bacteroids and anaerobically growncells, but not with an extract obtained from aerobically

molecular weight (lanes B and I) may be bands that resulted frombreakdown of P-450; these bands were observed especially duringforced expression in E. coli.

APPL. ENVIRON. MICROBIOL.

on May 28, 2018 by guest

http://aem.asm

.org/D

ownloaded from

B. JAPONICUM CYTOCHROME P-450 LOCUS 4139

screen lambda library200-bp

P450:: lacZ fusion

A gt1l1l.1 08A 1d3E(E)

screen cosmid libCYP112

11BS ES V

; subolone CYP112CYP112

Po

I i ii

E 1.1 kbp E 1.0 kbp

t insert Tn5 cassettTn5

, 11E rE

pRET74 H(pUC18) V

(E)E

)ranyCYP114

E P S BS

region

E'e

subclone CYP114region

ECYP114

lI -f/-A8.4 kbp

FIG. 2. Restriction endonuclease maps showing steps in the cloning, expression, and mutagenesis of the P-450 locus of B. japonicumUSDA 110. Restriction site abbreviations: B, BamHI; E, EcoRI; P, PstI; Pv, PvuII; S, SalI; V, EcoRV; X, XhoI. The arrowheads indicatethe direction of transcription, and the arrow lengths indicate the estimated sizes of the polypeptides. The exact positions of the genes,

however, were not determined in this work. Constructions pRET61 and pRET74 have the same orientation as the pUC18 lacZ promotor forforced expression of CYP112 and CYP114, respectively; pRET60 is a control having CYP112 in the reverse orientation. The EcoRI sites inparentheses on the Xgt11.108A map are artificial sites made during library construction and were derived from HaeIII sites.

grown cells (Fig. 1B, lanes C, D, and F, respectively). Theanti-CYP112 antibody did not react with purified CYP114(Fig. 1A, lane B), and the anti-CYP114 antibody did notreact with purified CYP112 (Fig. 1B, lane A). This confirmsthat CYP112 and CYP114 are different proteins that areexpressed simultaneously in B. japonicum. The next stepwas to screen a library for one of the genes. The anti-CYP112 antibody was judged to be sufficiently specific toscreen the lambda gtll expression library of USDA 110genomic DNA for the prospective CYPJ12 gene.

Library screening for the CYP112 gene. Eight anti-CYP112antibody-positive clones were selected from the gene li-brary. DNA from one of these clones (Xgtll.108A) (Fig. 2)hybridized with the EcoRI inserts of six of the other clonesin a Southern hybridization analysis (data not shown), indi-cating that all but one of the eight clones were from thesame region of DNA. Thus, the EcoRI insert of clone Agtll.108A was judged to be suitable as a probe for the prospec-tive CYP112 gene from the cosmid library, after it was

first subcloned into pUC18. Seven cosmids, each of whichshared one or more EcoRI restriction fragments with theothers, were identified, and one of these cosmids (cosmid6A) (Fig. 2) was used for restriction site mapping andsubcloning.

Expression of the CYP112 and CYP114 loci in E. coli. A2.1-kb partial EcoRI fragment from cosmid 6A (Fig. 2) wassubcloned into pUC18 downstream of the constitutivelyexpressed lacZ promotor. One clone (pRET61) expresseda polypeptide in host E. coli SE5000 cells that comigratedwith authentic CYP112 (Fig. 1A, lane G). Another clone

(pRET60) did not express the polypeptide (Fig. 1A, lane H)and was found by restriction analysis to be in the reverseorientation in pUC18. A spectral analysis of an E. colisoluble extract revealed no active P-450, however (data notshown).An 8.4-kb EcoRV fragment from cosmid 6A was sub-

cloned into the SmaI site of pUC18. A clone in the forwardorientation, as determined by restriction analysis (clonepRET74) (Fig. 2), was chosen for expression in E. coli DH5aof putative genes downstream of CYP112. Expression of a

polypeptide that comigrated with authentic CYP114 was

demonstrated by performing a Western blot analysis withanti-CYP114 antibody (Fig. 1B, lane I); the results indicatedthat pRET74 contains the putative CYP114 gene. The prod-uct of the CYP112 locus did not react with anti-CYP114antibody (Fig. 1B, lane G), and the product of the CYP114gene did not react with anti-CYP114 antibody (Fig. 1A, laneI).

Mutagenesis of the P-450 region and plant tests. In order todisrupt the CYP112 locus and downstream genes, a TnScassette (a 5.3-kb HpaI fragment from transposon Tn5encoding Kmr Smr [19]) was inserted into an EcoRI restric-tion site in the presumptive CYP112 gene in pRET60.Plasmid pRET60 was partially digested with EcoRI, and thefragments were separated by preparative agarose gel elec-trophoresis. A partially digested fragment having the mobil-ity expected for a fragment with a single cut (4.8 kb) was

isolated (cuts at three EcoRI sites were possible [Fig. 2]).The 3' recessed EcoRI termini were filled in by usingKlenow fragment DNA polymerase (16), and the TnS HpaI

cosmid 6A(pLAFRI)

S PvP X P E

pRET60/pRET61(PUC18)

pRET70/pRET72(pUC18/pSUP202)

V

s ~~~I

VOL. 59, 1993

on May 28, 2018 by guest

http://aem.asm

.org/D

ownloaded from

4140 TULLY AND KEISTER

TABLE 2. Effect of the B. japonicum P-450 mutation on soybeanplant growth and nodulationa

Nodules Plant topsPrepn No Wet mass Dry mass Total N

No.

(g) (g) (g)

USDA 110 55 1.74 5.55 0.239BJ1005 62 1.73 5.39 0.228Controib 14 0.16 0.48 0.017

a All values are means. Seven jars were used for USDA 110 and BJ1005,and four jars were used for controls. Each jar contained two plants. Themeans for USDA 110 and BJ1005 for all measurements were not significantlydifferent at the 0.05 level as determined by the F test.

b Uninoculated plants.

cassette was ligated into the site. Restriction analysis of theresulting clones was used to screen for one clone having aninsertion in the desired EcoRI site. The final constructionwas pRET70 (Fig. 2). To introduce this mutation into the B.japonicum genome, the 7.4-kb EcoRI insert containing theTnS cassette was excised from pRET70 and ligated into theEcoRI site of broad-host-range suicide vector pSUP202. Theresultant plasmid (pRET72) (Fig. 2) was transformed intodonor host E. coli S17-1 for mating into USDA 110. Platemating was performed for 23 days (although a shorter timewould have been adequate) on AlE-gluconate agar withoutantibiotic selection. To select cells that had acquired the Tn5cassette, cells were suspended in H20 and plated on thesame medium supplemented with kanamycin, streptomycin,and trimethoprim. Southern blot analysis of DNA fromselected transconjugants was used to screen for a strain thathad acquired the TnS cassette through homologous ex-

400 500

Mvelength, nm

600

change. First, labeled pRET61 was used as a probe ofEcoRI-digested genomic DNA. In one strain (BJ1005) the1.1- and 1.0-kb EcoRI fragments observed in pRET60 andpRET61 (Fig. 2) had been replaced by a single fragment thatwas more than 7 kb long (data not shown) and was found inpRET72 (Fig. 2), which was the expected pattern if the twofragments were joined by the TnS cassette. This fragmentalso hybridized with labeled TnS probe pSUP1011 (21),verifying the TnS cassette insertion, but did not hybridizewith labeled pSUP202, verifying the loss of the suicidevector. Thus, this strain, BJ1005, was judged to be asuccessful recombinant.To determine the effect of the mutation in the CYPJ12

locus on symbiosis, plant tests were performed with soy-beans by using wild-type USDA 110 and mutant BJ1005. Nodifferences in nodule number, nodule weight, or the totalamount of N2 fixed per plant were observed (Table 2).Nodule occupancy was determined by using Western blotanalysis to demonstrate the presence (USDA 110) or ab-sence (BJ1005) of the P-450 gene products. Nodules from allseven jars inoculated with USDA 110 produced CYP112(Fig. 1A, lane E) and CYP114 (Fig. 1B, lane E), but thenodules in the seven jars inoculated with BJ1005 weremissing both gene products (Fig. 1A and B, lanes F). Thisalso confirmed that the mutation in CYPJ12 was a polarmutation for CYPJ14. CO difference spectra of solubleextracts of bacteroids confirmed that little or no cy-tochromes P-450 were present in BJ1005 compared withwild-type USDA 110 (Fig. 3A), verifying that the TnScassette insertion was indeed in the correct location in theP-450 locus. CO difference spectra of membrane prepara-tions from the same bacteroids (Fig. 3B) revealed the pres-

400 500Wavelength, nm

600

FIG. 3. CO difference spectra ([reduced plus CO] minus reduced) of soluble (A) and membrane (B) preparations of B. japonicum USDA110 and BJ1005 bacteroids. AOD, change in optical density.

A

TAOD= 0.01

B

USDA 110

\|-/- -w vV BJ1005

551

J,&OD=

10.005

a 1 a a

APPL. ENvIRON. MICROBIOL.

on May 28, 2018 by guest

http://aem.asm

.org/D

ownloaded from

B. JAPONICUM CYTOCHROME P-450 LOCUS 4141

ence of small amounts of P-450 in membranes from USDA110 but no P-450 in membranes from BJ1005.

DISCUSSION

A Western blot (Fig. 1) revealed the presence of at leasttwo P-450 proteins coded for by genes in cosmid 6A and itssubclones (Fig. 2). Each of the two antibodies was suffi-ciently specific so that it did not cross-react with the other'santigen (Fig. 1, lanes A and B), and yet both antibodiesreacted with polypeptides from bacteroids and anaerobicallygrown cells (Fig. 1, lanes F and E, respectively). The twogenes were isolated on subclones pRET61 (gene CYP112)and pRET74 (gene CYP114), as determined by forced ex-pression of the respective polypeptides in E. coli (Fig. 1,lanes F through I). The fact that a transposon insertion in theCYP112 region prevented synthesis of any detectable P-450(Fig. 3) indicates that there are probably no P-450 genesexpressed upstream of the putative CYP1J2 gene, althoughthe presence of more downstream genes (besides CYP114),whose products would not be expressed if they are on thesame operon because of the polar effect of the TnS cassetteinsertion, cannot be ruled out on the basis of this evidence.Expression of spectrally active P-450 from a lacZ pro-

moter in E. coli has been demonstrated for P. putida P450.m(24). Despite the presence of the polypeptide translationproducts in the E. coli clones, we observed no spectrallyactive P-450 in the E. coli extracts (data not shown).A comparison with the genetic organization of other

bacterial cytochrome P-450 loci showed that the closestrelative of B. japonicum that contains P-450 (or P-450-like)genes isA. tumefaciens, which has the plant-induciblepinFland pinF2 genes (11). It is noteworthy that the two pinFgenes occur in tandem on the same transcript, just like theCYPJ12 and CYP114 genes.

Rhizobial P-450s have been observed to occur in both thesoluble and membrane fractions of cells (2, 12). As Fig. 3shows, the mutation in the CYP112 gene leads to the loss ofP-450 in both the soluble and membrane fractions of cells.Thus, it is likely that the P-450 observed in membranefractions is soluble P-450 entrained in membrane vesicles.The lack of a plant phenotype for the mutant strain leads

to the conclusion that cytochromes P-450 are not requiredfor nodulation or for symbiotic N2 fixation. This is surprisingin light of the presumed role of P-450 in an efficient pathwayof oxidative phosphorylation (4, 5, 26). One possibility isthat there is another P-450 present that compensates for theloss of CYP112 and CYP114. However, a spectral analysis ofthe BJ1005 mutant (Fig. 3) revealed that little or no P-450was present. A very small amount of P-450 could be maskedby the Soret trough of the CO-reactive cytochrome c whichis present. Other possibilities are (i) that under the condi-tions of the plant tests, plant growth and N2 fixation werelimited by something other than P-450, so that a P-450-phenotype was not apparent and (ii) that P-450 is involved ina bacteroid function that does not affect the plant, such asviability after nodule senescence.Cytochromes P-450 are synthesized under low 02 tensions

in B. japonicum. The levels of expression of CYPJ12 andCYP114 were higher in bacteroids than in anaerobic nitrate-grown cells (Fig. 1, lanes D and E). Thus, it is possible thatsome nodule factor in addition to low 02 tension is involvedin expression. Future work will involve testing for factorsthat induce expression of the operon, such as using plantextracts in combination with low 02 tensions. Also, se-quencing of the operon will be performed in order to identify

other genes that are coexpressed with the P-450 genes. Theidentification of other genes on the operon will help inelucidating the biochemical pathway in which the P-450sparticipate. The fact that the genes are expressed undersymbiotic conditions strongly suggests that they play a rolein some aspect of nodule function or biochemistry, eventhough a symbiotic phenotype was not apparent under theconditions which we tested.

ACKNOWLEDGMENTS

We thank Karrie Lovins and Patricia Gelak for skilled technicalassistance.This work was supported in part by National Research Initiatives

Competitive Research Grant 92-37305-7808 from the USDA Coop-erative State Research Service.

REFERENCES1. Appleby, C. A. 1967. A soluble haemoprotein P 450 from

nitrogen-fixing Rhizobium bacteroids. Biochim. Biophys. Acta147:399-402.

2. Appleby, C. A. 1969. Haemoprotein P450, other CO-reactivepigments, cytochromes and oxidases in bacteroids from N2-fixing root nodules. Biochim. Biophys. Acta 172:71-87.

3. Appleby, C. A., and R. M. Daniel. 1973. Rhizobium cytochromeP-450: a family of soluble, separable hemoproteins, p. 415-528.In T. E. King, S. Mason, and M. Morrison (ed.), Oxidases andrelated redox systems. University Park Press, Baltimore.

4. Appleby, C. A., G. L. Turner, and P. K. Macnicol. 1975.Involvement of oxyleghaemoglobin and cytochrome P-450 in anefficient oxidative phosphorylation pathway which supportsnitrogen fixation in Rhizobium. Biochim. Biophys. Acta 387:461-474.

5. Bergersen, F. J., and G. L. Turner. 1975. Leghaemoglobin andthe supply Of 02 to nitrogen-fixing root nodule bacteroids:presence of two oxidase systems and ATP production at lowfree 02 concentration. J. Gen. Microbiol. 91:345-354.

6. Daniel, R. M., and C. A. Appleby. 1972. Anaerobic-nitrate,symbiotic and aerobic growth of Rhizobium japonicum: effectson cytochrome P450. other haemoproteins, nitrate and nitritereductases. Biochim. Biophys. Acta 275:347-354.

7. Dus, K., R. Goewert, C. C. Weaver, D. Carey, and C. A.Appleby. 1976. P-450 hemeproteins of Rhizobium japonicum.Purification by affinity chromatography and relationship toP-450CAM and P-456LM-2. Biochem. Biophys. Res. Commun.69:437445.

8. Friedman, A. M., S. R. Long, S. E. Brown, W. J. Buikema, andF. M. Ausubel. 1982. Construction of a broad host range cosmidvector and its use in the genetic analysis of Rhizobium mutants.Gene 18:289-296.

9. Guengerich, F. P. 1991. Reactions and significance of cyto-chrome P-450 enzymes. J. Biol. Chem. 266:10019-10022.

10. Hanahan, D. 1985. Techniques for transformation of E. coli, p.109-135. In D. M. Glover (ed.), DNA cloning, vol. 1. A practicalapproach. IRL Press, Oxford.

11. Kanemoto, R. H., A. T. Powell, D. E. Akiyoshi, D. A. Regier,R. A. Kerstetter, E. W. Nester, M. C. Hawes, and M. P. Gordon.1989. Nucleotide sequence and analysis of the plant-induciblelocus pinF from Agrobacterium tumefaciens. J. Bacteriol. 171:2506-2512.

12. Keister, D. L., and S. S. Marsh. 1990. Hemoproteins ofBradyrhizobiumjaponicum cultured cells and bacteroids. Appl.Environ. Microbiol. 56:2736-2741.

13. Kuykendali, L. D., M. A. Roy, J. J. O'Neill, and T. E. Devine.1988. Fatty acids, antibiotic resistance, and deoxyribonucleicacid homology groups of Bradyrhizobium japonicum. Int. J.Syst. Bacteriol. 38:358-361.

14. Laemmli, U. K. 1970. Cleavage of structural proteins during theassembly of the head of bacteriophage T4. Nature (London)227:680-685.

15. Lammers, P. J., S. McLaughlin, S. Papin, C. Trujillo-Provencio,and A. J. Ryncarz II. 1990. Developmental rearrangement of

VOL. 59, 1993

on May 28, 2018 by guest

http://aem.asm

.org/D

ownloaded from

4142 TULLY AND KEISTER

cyanobacterial nif genes: nucleotide sequence, open readingframes, and cytochrome P-450 homology of the Anabaena sp.strain PCC 7120 nifD element. J. Bacteriol. 172:6981-6990.

16. Maniatis, T., E. F. Fritsch, and J. Sambrook. 1982. Molecularcloning: a laboratory manual. Cold Spring Harbor Laboratory,Cold Spring Harbor, N.Y.

17. Nelson, D. R., T. Kamataki, D. J. Waxman, F. P. Guengerich,R. W. Estabrook, R. Feyereisen, F. J. Gonzalez, M. J. Coon,I. C. Gunsalus, 0. Gotoh, K. Okuda, and D. W. Nebert. 1993.The P450 superfamily: update on new sequences, gene mapping,accession numbers, early trivial names of enzymes, and nomen-clature. DNA Cell Biol. 12:1-51.

18. Norrander, J., T. Kempe, and J. Messing. 1983. Construction ofimproved M13 vectors using oligodeoxynucleotide-directed mu-tagenesis. Gene 26:101-106.

19. Selvaraj, G., and V. N. Iyer. 1984. Transposon TnS specifiesstreptomycin resistance in Rhizobium spp. J. Bacteriol. 158:580-589.

20. Silhavy, T. J., M. L. Berman, and L. W. Enquist. 1984.Experiments with gene fusions. Cold Spring Harbor Labora-tory, Cold Spring Harbor, N.Y.

APPL. ENVIRON. MICROBIOL.

21. Simon, R., U. Priefer, and A. Puihler. 1983. A broad host rangemobilization system for in vivo genetic engineering: transposonmutagenesis in gram negative bacteria. Biotechnology 1:784-791.

22. Tissen, P. 1985. Practice and theory of enzyme immunoassays,p. 96-99. In R. H. Burdon and P. H. van Knippenberg (ed.),Laboratory techniques in biochemistry and molecular biology,vol. 15. Elsevier, Amsterdam.

23. Tully, R. E., M. J. Sadowsky, and D. L. Keister. 1991. Charac-terization of cytochromes c550 and C555 from Bradyrhizobiumjaponicum: cloning, mutagenesis, and sequencing of the C555gene (cycC). J. Bacteriol. 173:7887-7895.

24. Unger, B. P., I. C. Gunsalus, and S. G. Sligar. 1986. Nucleotidesequence of the Pseudomonasputida cytochrome P450ca geneand its expression in Eschenchia coli. J. Biol. Chem. 26:1158-1163.

25. van Berkum, P. 1990. Evidence for a third uptake hydrogenasephenotype among the soybean bradyrhizobia. Appl. Environ.Microbiol. 56:3835-3841.

26. Wittenberg, J. B., F. J. Bergersen, C. A. Appleby, and G. L.Turner. 1974. Facilitated oxygen diffusion. J. Biol. Chem.249:4057-4066.

on May 28, 2018 by guest

http://aem.asm

.org/D

ownloaded from