JOURNAL OF BACTERIOLOGY, Feb. 1981, p. 743-751 Vol. 145, No. 20021-9193/81/020743-09$2.00/0

Molybdenum Accumulation and Storage in Klebsiellapneumoniae and Azotobacter vinelandii

PHILIP T. PIENKOSt AND WINSTON J. BRILL'

Department ofBacteriology and Center for Studies of Nitrogen Fixation, University of Wisconsin, Madison,Wisconsin 53706

In Klebsiella pneumoniae, Mo accumulation appeared to be coregulated withnitrogenase synthesis. 02 and NW, which repressed nitrogenase synthesis, alsoprevented Mo accumulation. In Azotobacter vinelandii, Mo accumulation didnot appear to be regulated. Mo was accumulated to levels much higher than thoseseen in K. pneumoniae even when nitrogenase synthesis was repressed. Accu-mulated Mo was bound mainly to a Mo storage protein, and it could act as asupply for the Mo needed in component I synthesis when extracellular Mo hadbeen exhausted. When A. vinelandii was grown in the presence of WO4 ratherthan MoO24r, it synthesized a W-containing analog of the Mo storage protein. TheMo storage protein was purified from both NW and N2-grown cells of A.vinelandii and found to be a tetramer of two pairs of different subunits bindinga minimum of 15 atoms of Mo per tetramer.

Nitrogenase, the enzyme which catalyzes thereduction of N2 to NH3, is composed of twodissociable proteins, components I and II. Com-ponent I contains the transition element molyb-denum, which is thought to be responsible forsubstrate binding and reduction (24). A cofactor(FeMo-co), containing Mo, Fe, and acid-labile S,has been isolated from component I (21), andthere is evidence that it is the active site ofnitrogenase (7, 17, 22). FeMo-co is distinct froma similar cofactor (Mo-co) found in nitrate re-ductase, xanthine oxidase, and other molyb-doenzymes (15).

In addition to its catalytic role, Mo seems tocontrol synthesis of nitrogenase in several N2-fixing organisms (2, 6, 14). However, little isknown about the metabolic pathway from extra-cellular Mo to the synthesis of active componentI. In Clostridium pasteurianum, Mo is seques-tered by a Mo-binding protein (13). Althoughthe role of this protein is not known, it mayserve as a Mo storage mechanism or as a pre-cursor of component I.

Species of Azotobacter have a high affinityfor Mo and an ability to accumulate high intra-cellular levels of Mo (11). This high affinity istaken advantage of in the use of A. paspali as abiological assay for available Mo in soil samples(9).This work represents a comparative study of

the regulation and capacity for Mo accumulationin Klebsiella pneumoniae and A. vinelandii.Preliminary accounts ofthis work have appeared

t Present addres: Department of Microbiology, Universityof Texas at Austin, Austin, TX 78712.

(P. T. Pienkos and W. J. Brill, Abstr. Annu.Meet. Am. Soc. Microbiol. 1976, K165, p. 164; P.T. Pienkos and W. J. Brill, Abstr. Annu. Meet.Am. Soc. Microbiol. 1977, K77, p. 199).

MATERIALS AND METHODSGrowth of organisms. A. vinelandii OP (4) was

grown at 300C with vigorous aeration in modified Burkmedium (25). Cultures were inoculated with cells thathad been depleted of intracellular Mo through at leasttwo cycles of growth in Mo-free medium containingammonium acetate (400 ,ug of N/ml). Mo accumula-tion studies were performed by growing cells in 200-mlbatches in 1-liter side-arm flasks with fluted bottomswith medium containing NW and lacking MoO4-.These cells were harvested by centrifugation and sus-pended in fresh medium in the presence or absence ofNH! with various levels of MoOr or WOi4. At se-lected intervals, 1.0-ml samples were removed forwhole-cell acetylene reduction assays (23), 20.0-milsamples were removed, and the cells were harvested,washed with 25mM Tris-hydrochlorrde, pH 7.4 (whichhad been sparged with Ar), and suspended in glass-distilled water for analysis of Mo or W. To obtainenough cell material for purification of the Mo storageprotein, cells were grown with vigorous sparging in 15liters of medium in 5-gallon (ca. 19-liter) vessels.

K. pneumoniae M5al was grown in medium de-scribed by Yoch and Pengra (26) anaerobically at300C. Because this organism does not accumulatemeasurable levels of Mo when grown on NH!, it wasunnecessary to starve cells for Mo before inoculatingexperimental cultures. Mo accumulation studies withK. pneumoniae were identical to those with A. vine-landii except that K. pneumoniae was grown in 250-ml bottles, and 1.5 h after suspension of cells into freshmedium, L-serine was added to a level of 50 jg/ml toshorten the lag period before appearance of nitrogen-

743

on May 4, 2018 by guest

http://jb.asm.org/

Dow

nloaded from

744 PIENKOS AND BRILL

ase activity (19). Also, 50-mi samples of cultures wereremoved to provide cells for Mo determination.

Cell densities were measured by viable cell countsor with a Klett-Summerson colorimeter.

Quantitation of Mo, W, and Fe. Mo analysis ofwhole-cell suspensions or protein solutions was per-formed with a Perkin-Elmer atomic absorption spec-trophotometer equipped with a graphite furnace. Col-orimetric assays were used to determine W (5) and Fe(3).Preparation and handling of crude extracts.

Crude extracts were prepared anaerobicaJ1y by theosmotic shock technique (18). Acetylene reductionassays (23), anaerobic column chromatography (20),and in vitro activation of inactive component I fromA. vielandii mutant strain UW45 and of inactivenitrate reductase from Neurospora crassa mutantstrain Nit-1 (15) have been described. Protein wasquantitated by the biuret reaction (10). DEAE-cellu-lose columns were washed with linear gradients ofNaCl in 25 mM Tris-hydrochloride, pH 7.4, and thefractions were analyzed for NaCl concentration witha conductivity meter.Polyacrylamide gel electrophoresis. The tech-

nique for performing nondenaturing slab gel electro-phoresis has been described (20). Sodium dodecylsulfate (SDS)-gel electrophoresis used a 3% polyacryl-amide stacking gel and a 10% polyacrylamide separat-ing gel (12).

RESULTSMo accumulation in A. vinelandii and K

pneumonsa& Early attempts to elucidate therole of Mo in N2 fixation led to the observationthat Mo is accumulated in A. vinelandii regard-less ofthe nitrogen source (11). This implies thatMo accumulation in this organism is not co-regulated with nitrogenase synthesis as it ap-pears to be in C. pasteurianum (8). We com-pared the levels of intracellular Mo in cells thathad been derepresed for nitrogenase synthesiswith those of repressed cells. NW-grown, Mo-starved cells ofA. vinelandii were suspended inmedia containing increasing concentrations ofMoO4 in the presence and absence of NW.After 3 h ofincubation, samples ofthese cultureswere removed for whole-cell nitrogenase assaysand Mo analysis. Nitrogenase activity was de-tectable only in cells suspended in media lackingNW: but containing MoO3r, and Mo accumula-tion occurred in both the presence and absenceof NW (Fig. 1). A concentration of 0.2 uMMoO3r was sufficient for full expresaon of nitro-genase activity, but Mo accumulation increasedfar beyond that level. Raising the concentrationof MoO3- in the growth medium to 1.0 mM ledto an intracellular Mo concentration 25 timeshigher than was required for the synthesis of afull complement of component L.When the same experiment was done with K.

pneumoniae, two differences were immediately

0

05 -lo Ur.

*1~~~~~~1.0.0 2.5 50

Mo in medium (WM)FIG. 1. Mo accumulation and nitrogenase activity

in A. vinelandii incubated with increasing levels ofMoO-. Mo-starved cells weregrown to late logphasein medium lackingMo but containingNH . The ceUswere coUected by centrifugation and suspended inmedia containing increasing levels ofMoO3- in thepresence or absence of NH4. After 3 h of vigorousshaking at 30°C, 1.0-mi culture samples were re-moved for acetylene reduction assays and 20.0-misamples were centrifuged; the peUets were washedwith 25 mM Tri-hydrochloride, pH 7.4, centrifuged,and suspended in 1.0 ml of glass-distilled water forMo analysis by atomic absorption spectroscopy. Sym-bols: 0, Mo accumulation in media with NH,; 0, Moaccumulation in media withoutNH+; I nitrogenaseactivity in media with NH4; [, nitrogenase activityin media without NH4+.

evident (Fig. 2). First, Mo accumulation wasobserved only in celLs that were suspended inNW-free media. Second, the level of intracel-lular Mo was much lower in this organism thanin A. vinelandii. Like A. vinelandii, K. pneu-moniae did not exhibit nitrogenase activity inthe ?resence of NW or when starved forMoOl..Another method for correlating the control of

Mo accumulation with the control ofnitrogenasesynthesis was to compare the time courses ofthese two processes. Mo-starved cells were sus-pended in medium containing 0.2 ,uM MoOr4 inthe presence and absence ofNHW. At intervals,samples were removed for nitrogenase assaysand Mo quantitation. Figure 3 shows the time

J. BACTIERIOL.

on May 4, 2018 by guest

http://jb.asm.org/

Dow

nloaded from

MOLYBDENUM ACCUMULATION 745

Again, it was observed that the level of Moaccumulation was lower in this organism than in

04 cOlO A. vinelandii; only 25% of the extracellular4MoO was accumulated by K. pneumoniae

en / compared with 100% for A. vinelandii. Mo ac-cumulation was not rapid in K. pneumoniae;

o 1 o rather, it closely followed the development ofo r = nitrogenase activity.

2.. c K. pneumoniae is a facultative aerobe, but itx1/ _ ~~~~~~~~~canonly fixN2 anaerobically (19). We examined

CD Mo accumulation under aerobic and anaerobicconditions. Oxygen not only repressed nitrogen-ase synthesis, but also completely inhibited Mo

Q2; 0.05 accumulation (data not shown). Adding chlor-FN amphenicol to inhibit protein synthesis had a

i similar effect on K. pneumoniae (Table 1).I Ur Chloramphenicol-treated cells were unable too synthesize nitrogenase and unable to accumulate

Mo. When cells of A. vinelandii were treatedwith chloramphenicol, however, synthesis of ni-trogenase was completely inhibited but Mo ac-cumulation was unaffected. In fact, the intracel-lular Mo concentration was even higher in the

QCie| 10~~~~~~~~.00.0 2.5 5.0 1.0 , 100

Mo in medium (WM)FIG. 2. Mo accumulation and nitrogenase activity

in K. pneumoniae incubated with increasing levels of lMoO2-. Cells were grown to late log phase in Mo-free lmedium containing NH4 and then collected by cen- 1I0trifugation and suspended in media containing in- x

0

creasing levels ofMoO2- in the presence or absenceofNH4. After being incubated anaerobically at 30°Cfor 1.5 h, 1.0 ml of 1% serine was added to each 200- Eml culture, and the cultures were incubated for an Cladditional 4.5 h. At the end of this incubation, 1.0-ml 0.5 l50culture samples were removed for acetylene reduction 3assays, and 50-ml samples of cultures were prepared 0- lfor atomic absorption spectroscopy as described in lthe legend of Fig. 1. Symbols are as in Fig. 1. N

course of development of nitrogenase activityand Mo accumulation in A. vinelandii. Mo ac- -/cumulation was very rapid, having begun before /cells would be harvested in the centrifuge. At 0.2,uM MoO2- in the medium, accumulation wentto completion within 1 h, and the extracellularMoO2r was exhausted a full hour before the first -dOappearance of nitrogenase activity. Cells sus- 0.0 2.0 40pended in medium containing NH4 demon- Time(h)strated an Mo accumulation pattern almost Time (h)identical to that of NW4-starved cells, but no FIG. 3. Time course of appearance of nitrogenaseidenticaltothatofNW-starvedcesbutn activity and Mo accumulation in A. vinelandii. Mo-nitrogenase activity was observed. starved, NHI-grown cells were suspended in mediaWhen Mo-starved cells of K. pneumoniae containing 0.2 pMMoO4- in the presence or absence

were transferred to medium containing 0.2 ,uM of NH4. At intervals, samples were taken for nitro-MoO2, only the cells in NW-free medium ac- genase activity and Mo determination as describedcumulated detectable levels of Mo (Fig. 4). in the legend ofFigure 1. Symbols are as in Fig. 1.

VOL. 145, 1981

on May 4, 2018 by guest

http://jb.asm.org/

Dow

nloaded from

746 PIENKOS AND BRILL

0~~~~~~~

,0.2- 550

3i5CYJ

150 a ~

0.0 2.0 4.0Time(h )

FIG. 4. Time course of appearance of nitrogenaseactivity andMo accurnulation in K.pneumoniae. Mo-starved, NHI-grown cells were suspended in mediacontaining 0.2 WMoO34- in the presence or absenceofNH4'. The cultures uwere treated as described int thelegend ofPig. 2, and samples were taken at intervalsand assayed for nitrogenase activity andMo content.Symbols are as in Fig. 1.

treated cells than in the control cells becauseboth cultures had exhausted the extracellularMoO2', but chloramphenicol stopped cell divi-sion so that the Mo was divided among fewercells.Mo storage protein. The ability of A. vine-

landii to accumulate high levels ofMo suggestedthat this Mo was being stored and could be usedfor synthesis of component I in the event ofexhaustion of extracellular Mo. This hypothesiswas tested by growing cells in 200-ml cultures inmedia containing NW4 in the presence and ab-sence of 1.0 ,uM MoO4'-. The MoO'--rw cellswere resuspended inNW+-free, MoO'--free me-dium, and the MoO42--starved cells were dividedinto two batches and suspended intoNW -freemedia in the presence and absence of MoO2-.

4

Samples were taken periodically and tested foracetylene reduction activity in vivo. The cellsthat went from Mo-free to Mo-free medium wereunable to synthesize active nitrogenase, but thecells that went from medium containingMoOe-(and had therefore accumulated intracelluvls

Mo) to Mo-free medium demonstrated as muchnitrogenase activity as did the cells transferredfrom Mo-free medium to medium containingMoO4 (data not shown).Crude extracts were prepared from cells that

had been grown on various levels of MoO2- inthe presence and absence of NW, and theseextracts were fractionated on anaerobic DEAE-cellulose columns. Figure 5 shows the Mo elutionprofiles of the fractionation of extracts of cellsgrown in the presence of 0.1, 0.2, and 50.0 ,uMMoO4. All of the Mo from cells grown in mediacontaining NW was eluted in a single peak atan NaCl concentration of 0.15 M. Two Mo peakswere detected from cells grown on N2 as nitrogensource. One peak, eluted with 0.2 M NaCl, cor-responded to component I, and it was the onlyMo detected from cells grown in the limitingMoO4 concentration of 0.1 j,M. When the con-centration ofMoO2- in the growth medium wasincreased, a second Mo peak appeared at thesame NaCl concentration as the stored Mofound inNW -grown cells. The increase in extra-cellularMoO4- above 0.1 ,uM had no effect eitheron the level of Mo found in the component Ipeak or on the level of nitrogenase activity.When similar experiments were done with K.pneumoniae, the only Mo peak eluted corre-sponded to component I (S. Klevickis, unpub-lished data). This agreed with the observationthat K. pneumoniae accumulates much less Mothan A. vinelandii and is evidence that thisorganism has no Mo storage capabilities.The DEAE-cellulose fractionation of A. vine-

landii extracts suggested that both NWH- andN2-grown cells store excess Mo in the same form,but it was not clear if the stored Mo was in theform of MoO4 or in a higher-molecular-weightform such as an inorganic polymer or molybdo-protein. This question was answered by fraction-

TABLE 1. Effects of chloramphenicol on Moaccumulationa

Chlor- Nitrogen- Mo (jamol/Organism amphen- asesp aCtb 1012 cells)

icol

A. vinelandii - 0.59 0.59A. vinelandii + 0.00 0.71K pneumoniae - 0.27 0.09K. pneumoniae + 0.00 0.00

a Mo-starved NH4+-grown cells ofA. vinelandii andK. pneumoniae were suspended in media containing0.2 ,iM MoO42 and lacking NH4+ in the presence andabsence of 100 ug of chloramphenicol per ml. Afternormal derepression of nitrogenase synthesis, sampleswere taken for nitrogenase assays and Mo determina-tions.

b Nanomoles of C2H4 formed per minute x 107 cells.

J. BACTrERIOL.

on May 4, 2018 by guest

http://jb.asm.org/

Dow

nloaded from

MOLYBDENUM ACCUMULATION 747

a)4-.I0;

20, 0.1 Jim mo 20- 0.1 iUM Mo |

Olu i... 0s, l -_@ ... *a0.0 0.2 0.4 GO 0.2 0.4

NaCI in eluting buffer (M)FIG. 5. Mo elution profiles from DEAE-cellulose

columns. Cultures ofA. vinelandii were grown to latelog phase in the presence of 0.1, 0.2, and 50.0 ,utM

MoO3- with NH4 or N2 as the nitrogen source. Thecells were harvested by centrifugation, washed withcold, 25 mM Tris-hydrochloride, pH 7.4, which hadbeen sparged with Ar, and then broken anaerobicalyby osmotic shock. Samples ofeach extract (2)00 mg ofprotein) were fractionated anaerobically on DEAE-cellulose columns (17 by 140 mm) eluted with NaCigradients made with 50 ml mM This-hydrochlorideand 50 ml of 0.5 NaCI in 25mM This-hydrochloride.Five-milliliter fractions were collected and analyzedfor Mo by atomic absorption.

ation of crude extract ofNW -grown cells on an

anaerobic Sephadex G-75 column. This columneasily distinguished the stored Mo present in thecrude extract from MoO3- (Fig. 6) and indicatedthat Mo is stored as a molecule of at least 40,000daltons. A sample of A. vinelandii extract was

then treated with protease and fractionated on

the same Sephadex G-75 column. In the pro-tease-treated extract, the stored Mo was brokendown and eluted as a low-molecular-weight com-pound.Samples ofstored Mo collected from a DEAE-

cellulose column were tested for FeMo-co andMo-co activity by determining if they could ac-

tivate the inactive component I obtained fromA. vinelandii mutant strain UW45 or the inac-tive nitrate reductase obtained from N. crassa

mutant strain Nit-1 (15). Neither untreated sam-ples nor samples mixed with dilute HCl couldserve as a source of either molybdenum cofactor.Tungsten storage in A. vinelandii Substi-

tution of W for Mo has been performed fre-quently in efforts to elucidate the role of Mo inmolybdoenzymes (16). A W-containing speciesof component I was partially purified fromWO -grown cells of A. vinelandii, and it wasobserved that the component I contained only asmall proportion of the total intracellularW (1).This suggested that A. vinelandii can accumu-late W just as it does Mo. Extracts of WO4-grown cells were fractionated in the same man-ner as the extracts of MoO4 -grown cells to de-

30

0

0n

EC:

0w

-Jw0

20

10

0

0 10 20

FRACTION NUMBERFIG. 6. Sensitivity of stored Mo to protease. Os-

motic shock extract of cells grown in the presence ofNH+ and 2.0 MMoO3- containing 26.3mg ofproteinper ml was divided into two samples. One samplereceived protease at 0.5 U/mi; the other sample wasuntreated. Both samples were incubated anaerobi-cally at 300C for 12 h, and then 1.0-mi samples werefractionated on a Sephadex G-75 column (17 by 140mm) which was washed anaerobically with Tris-hy-drochloride at 50C. Three-milliliter fractions werecollected and tested for Mo by atomic absorptionspectroscopy. An anaerobic solution of Na2MoO4 inTris-hydrochloride was fractionated similarly on thesame column. Symbols: 0, untreated extract; 0, pro-tease-treated extract; 4,MoO2.

VOL. 145, 1981

on May 4, 2018 by guest

http://jb.asm.org/

Dow

nloaded from

748 PIENKOS AND BRILL

ternine if a W-containing analog of the Mostorage protein was synthesized. Figure 7 showsthe W elution pattem from DEAE-cellulose col-umns used to fractionate extracts of cells grownin medium containing 5.0pMWO-. The cells inone culture were derepressed for nitrogenasesynthesis by allowing them to exhaust allNW;the other culture contained excess NW to re-press nitrogenase synthesis. As was the case withMoO4'-grown celLs, represed cells producedonly one W-containing compound which waseluted with 0.1 M NaCl, and the extract ofderepressed cells yielded two W peaks: oneeluted with 0.1 M NaCl, and the other eluted

0.0 Q2

NodC in eluting buffer (M)

with 0.2M NaCl. It is likely that the latter peakcorresponded to the inactive W-containing spe-cies of component I. Treatment of extract ofW042-grown cells with protease resulted in elu-tion of a low-molecular-weight form of W, indi-cating that W is stored by A. vinelandii in aprotein-bound fonn just as Mo is (Fig. 8).Purification of the Mo storage protein.

The Mo storage protein was purified from cellsofA. vinelandii that had been grown in mediumcontaining NM and 10 ,uM MoO42. The pres-ence of the Mo storage protein was monitoredthroughout purification by following the Moconcentration by atomic absorption spectros-copy and testing samples on a Sephadex G-75column to insure that we were following thehigh-molecular-weight protein and not a break-down product. The Mo storage protein was un-stable throughout purification whether it wasmaintained anaerobically or aerobically, and soattempts were made topu the protein rapidlyand to maintain it at 0 to 50C.A summary of purification of the Mo storage

0.4

FIG. 7. Elution profiles of stored W from DEAE.celulose columns. Two cultures ofA. vinelandii weregrown in thepresence of5pWWO-. One culture wasderepressed for nitrogenase synthesis by growing itin limiting NH4. In the other culture, nitrogenasesynthesis was repressed by excessNH. . Both cultureswere harvested, and the cels were broken by osmoticshock. The extract of repressed cells contained 21.7mg ofprotein per ml; the extract ofderepressed ceUscontained 18.8 mg ofprotein per ml. Four-millilitersamples ofboth extracts were fractionated on DEAE.ceUlulose columns as described in the legend of Fig.5. Quantitation ofWwasperformed colorimetricaly.Symbols: , extract ofderepressed cells; U, extract ofrepressed cells.

0 10 20Fraction number

FIG. 8. Sensitivity ofstored Wtoprotease. Osmoticshock extract (containing 23.6 mg ofprotein per ml)from ceUs grown in medium contaiingNH4 and 1.0p W02- was divided into two samples and treatedas described in the legend of Fig. 6. Symbols: 0,untreated extract; 0, protease-treated extract.

J. BACTERIOL.

on May 4, 2018 by guest

http://jb.asm.org/

Dow

nloaded from

MOLYBDENUM ACCUMULATION 749

TABLE 2. Purification of the Mo storage protein

Prepn Vol (Dl) Total protein Total Mo nmol of Mo/ %Recovery(mg) (nmnol) mg ofpr%o Rcover

Crude extract 140 1,172 10,220 8.7 100DEAE fraction 70 82.7 4,382 53.0 4220-50% ammonium suifate cut 3 24.3 1,593 65.6 16Sephadex G-100 fraction 22 3.2 516 161.2 5

protein is given in Table 2. Crude extract wasprepared in 25 mM Tris-hydrochloride, pH 7.4,and 140 ml was loaded on a DEAE-cellulosecolumn (2.5 by 25 cm) that had been equilibratedwith Tris-hydrochloride. The column waswashed with 1 bed volume of 50 mM NaCl inTris-hydrochloride, and the Mo storage proteinwas eluted with 50 mM NaCl in 50 mM imidaz-ole buffer, pH 6.5. The peak Mo storage proteinfractions were treated with 20% ammonium sul-fate, and the precipitate was discarded. Thesupernatant solution was made 50% with am-monium sulfate, and the Mo storage protein wasprecipitated. It was resuspended in 3.0 ml ofimidazole buffer and loaded on a Sephadex G-100 column (2.5 by 90 cm) which was washedwith imidazole buffer. The peak fractions werepooled and tested for purity by nondenaturingand SDS-polyacrylamide gel electrophoresis(Fig. 9). One protein band was visible on thenative gel, but two bands of approximately equalintensity were seen on the SDS-gel, indicatingthat the Mo storage protein consists of twodifferent subunits. When SDS-gels were runwith purified Mo storage protein obtained fromN2-grown cells, two protein bands were seenwith the same mobility as those ofNW -growncells (data not shown). Therefore, Mo is storedby the same system in both N2 andNW -growncells of A. vinelandii.The molecular weights of the subunits were

estimated to be 21,000 and 24,00 by comparisonwith the electrophoretic mobilities of proteinstandards. The elution proffies of native Mostorage protein on Sephadex and Sepharose col-umns indicated that the molecular weight wason the order of 100,000 (data not shown), and soit would appear that the Mo storage protein isa tetramer of two different subunits with a mo-lecular weight closer to 90,000. We used thisestimation to calculate that 14.5 atoms of Moare bound to each molecule of Mo storage pro-tein. Attempts to bind more Mo to the storageprotein by incubating the purified preparationwith 1.0 mM MoO4 failed. The peak fraction ofpurified Mo storage protein was tested for thepresence of Fe by reaction with a,a-dipyridyl(3), and the test was negative, indicating thatthe concentration of Fe was less than 5 nmol/ml

FG. 9. ElectrophoresisofpurifiedMostoragepro-tein. Samples ofpurifiedMo storageprotein obtainedas thepeak fraction from the Sephadex G-10() columnwere prepared for electrophoresis. Then 1.4 pg ofprotein was loaded on the nondenaturing gel (a) and2.3 pg ofprotein was loaded on the SDS-gel (b). Afterelectrophoresis was carried out, both SDS-gels werestained with Coomassie brilliant blue R250.

compared with an Mo concentration of 84 nmol/ml in that fraction.

DISCUSSIONIn K. pneumoniae, Mo accumulation is tightly

regulated and quite likely under the control ofnif gene expression. The presence ofNW or 02(both of which repress nitrogenase synthesis)completely inhibits Mo accumulation. Also, thelevel of intracellular Mo correlates well with thelevel of nitrogenase so that the kinetics of Moaccumulation and nitrogenase synthesis are al-

VOL. 145, 1981

Tf%,"'T J.il..

on May 4, 2018 by guest

http://jb.asm.org/

Dow

nloaded from

750 PIENKOS AND BRILL

most identical. These observations are quite sim-ilar to those of Mo accumulation in C. pasteu-rianum (8), but differ with our observations ofA. vinelandii, in which Mo accumulation ap-pears to be unregulated. Cells accumulate highlevels of Mo whether they are grown on N2 orNW. The concentration of intracellular Mo in-creases with the concentration of extracellularMoO3.r and can reach a level more than 25 timeshigher than needed for the synthesis of a fullcomplement of component I. However, Mo ac-cumulation is not simply a passive process be-cause A. vinelandii can rapidly exhaust thegrowth medium ofMoO,2 and thus must expendenergy to hold it against a concentration gra-dient. Unlike K. pneunoniae, A. vinelandii re-quires no de novo protein synthesis for Mo ac-cumulation; essential permeases and storagemechanisms are apparently synthesized consti-tutively, even in cells grown in the absence ofMoO4. This is evidenced by the ability of A.vinelandii to exhaust the growth medium ofMoO2r even after treatment with chloramphen-icol.The Mo-binding protein from C. pasteu-

rianum is a monomer that binds up to 6 atomsof Mo per molecule of 50,000 daltons (13). TheMo storage protein from A. vinelandii is madeup of a pair each of two different subunits withmolecular weights of 21,000 and 24,000. Afterpurification from both NH! and N2-grown cells,the Mo storage protein contains identical sub-unit structures, indicating that A. vimelandiiuses the same mechanism of storage of excessMo regardless of the nitrogen source. PurifiedMo storage protein has approximately 15 atomsof Mo bound to it. No Fe was found in the Mostorage protein. W02--grown cells ofA. vinelan-dii make a W-containing analog of the Mo stor-age protein. This W-containing protein has beendetected autoradiographically by electrophore-sis of crude extracts of 4W0-grown cells ofA.vinelandii (P. T. Pienkos, S. Klevickis, and W.J. Brill, manuscript submitted for publication).The Mo storage protein is not a aource of

FeMo-co to activate inactive component I fromA. vinelandii mutant strain UW45 or Mo-co toactivate inactive nitrate reductase from N.'crassa mutant strain Nit-i. The Mo storageprotein was unstable throughout the purificationprocess, releasing Mo as a low-molecular-weightcompound (possibly MoO2-). This instabilitymay have been an artifact of the purificationtechniques, but it is possible that Mo release isan inherent feature of this protein which may berequired to provide an immediate source of Mofor component I synthesis. Because of this insta-bility, the number of Mo atoms per Mo storage

protein in vivo may be much greater than 15.Mo-starved cells of A. vinelandii are unable

to synthesize component I (14); therefore,MoO' may play a regulatory role in componentI synthesis. However, cells that had stored Mowhile nitrogenase synthesis was repressed wereable to synthesize component I in the absence ofMoO24. This suggests that Mo bound to the Mostorage protein rather than MoO4r- might act ina regulatory role for component I synthesis. Amutant strain of A. vinelandii, UW71, has adefect in Mo accumulation (16). This strain maybe useful in attempts to understand the regula-tion of expression of nifgenes in A. vineladii.A. vinelandii seems unique in its ability to

store high levels of Mo even under conditionswhere Mo is unnecessary for growth. A possibleexplanation for this characteristic is that A. vi-nelandii may have evolved in habitats with largefluxes in the concentration of Mo. Under suchcircumstances, it would be advantageous to storeMo in times of plenty so that it would be avail-able in times of scarcity.

ACKNOWILEDGMENTSThis reearch was supported by the College of Agricultural

and Life Sciences, University of Wisconsin, Madison, and byPublic Health Service grant GM22130 from the NationalInsitute of General Medical SciencesWe thank V. K. Shah for component I and usefil misges

tions and F. Siegel for use of the atomic absorption spectro-photometer.

LITEATURE CITED

1. Benemann, J. R., G. M. Smith, P. J. Kostel, and C. E.McKenna. 1973. Tungsten incorporation into Azoto-bacter vinelandii nitrogenase. FEBS Lett. 29:219-221.

2. Brill, W. J., A. K. Steiner, and V. K. Shah. 1974. Effectof molybdenum starvation and tungten on the synthe-sis of nitrogenase components in KkbsieUa pneumo-niae. J. Bacteriol. 118:986-989.

3. Brill W. J, J. Westphal, A Stieghorst, L C. Davis,and V. KI Shah. 1974. Detection of nitrogenase com-ponents and other non-heme iron proteins in polyacryl-amide gela Anal. Biochem. 60:237-241.

4. Bush, J. A., and P. W. Wilson. 1959. A non-gummychromogenic strain of Azotobacter vinelandii. Nature(London) 184:381.

5. Cardenas, J., and L. E. Mortenson. 1974. Determina-tion of molybdenum and tungsten in biological mate-riah Anal. Biochem. 60:372-381.

6. Cardenas, J., and L E. Mortenson. 1976. Role of mo-lybdenum in dinitrogen fixation by ClostHidiwmpasteu-ranum. J. Bacteriol. 1:23:978-984.

7. Crame, S. P., W. 0. Gilulm, K. 0. lodgeon, L E.Mortenon, E. L StUel, J. R. Chisnell, W. J. Brlll,and V. K. Shah 1978. The molybdenum site of nitro-genas. 2. A comparative stdy of Mo-Fe proteins andthe iron-molybdenum cofactor by i-ray absorptionspecroscopy. J. Am Chem. Soc. 100:3814-3819.

8. EllHott, B. B., and L E. Mortenson. 1975. Transport ofmolybdate by Clostidium pasteurianwum. J. Bacteriol.124:1295-1301.

9. Franco, A. A., J. R. RI Peres, and N. Nery. 1978. Theuse ofAzotobacterpaspali Nrase (C2H2 reduction abil-

J. BACTROL.

on May 4, 2018 by guest

http://jb.asm.org/

Dow

nloaded from

MOLYBDENUM ACCUMULATION 751

ity) to measure molybdenum deficiency in soils. PlantSoil 50:1-11.

10. Gornall, A. G., C. J. Bardawll, and M. M. David.1949. Determination of senrum proteins by means of theBiuret reaction. J. Biol. Chem. 177:751-766.

11. Keeler, R. F., and J. E. Varner. 1958. The metabolismof molybdate and tungstate in Azotobacter, p. 297-309.In C. A. Lamb, 0. G. Bentley, and J. M. Beattie (ed.),Trace elements. Academic Press, Inc., New York.

12. Maisel, J. V., Jr. 1969. Acrylamide gel electrophoresis ofproteins and nucleic acids, p. 334-362. In K. Habel andN. P. Salxman (ed.), Fundamental techniques in virol-ogy. Academic Press, Inc., New York.

13. Mortenson, L. E., and R. N. F. Thorneley. 1979. Struc-ture and function of nitrogenase. Annu. Rev. Biochem.48:387-418.

14. Nagatani, H. IL, and W. J. Brill. 1974. Nitrogenase. V.The effects of Mo, W, and V on the synthesis of nitro-genase components in Azotobacter vinelandu.Biochim. Biophys. Acta 362:160-166.

15. Pienkos, P. T., V. K. Shah, and W. J. Brill. 1977.Molybdenum cofactors from molybdoenzymes and in

vitro reconstitution of nitrogenase and nitrate reduc-tase. Proc. Natl. Acad. Sci. U. S. A. 74:5468-5471.

16. Pienkos, P. T., V. K. Shah, and W. J. Brill. 1980.Molybdenum in nitrogenase, p. 385-402. In M. P.Coughlan (ed.), Molybdenum and molybdenum-con-taining enzymes. Pergamon Press Ltd., Oxford, Eng-land.

17. Rawlings, J., V. K. Shah, J. R. ChisnelL W. J. Brill,R. Zimmermann, E. Mianek, andW. H. Orme-John-

son. 1978. Novel metal cluster in the iron-molybdenumcofactor of nitrogenase. J. Biol. Chem. 253:1001-1004.

18. Robrish, S. A., and A. G. Marr. 1962. Location ofenzymes inAzotobacter agilis. J. Bacteriol. 83:158-168.

19. St. John, R. T., V. K. Shah, and W. J. Brill. 1974.Regulation of nitrogenase synthesis by oxygen in Kleb-sieua pneumoniae. J. Bacteriol. 119:266-269.

20. Shah, V. K., and W. J. Brill. 1973. Nitrogenase. IV.Simple method of purification to homogeneity of nitro-genase components from Azotobacter vinekadii.Biochim. Biophys. Acta 305:446-454.

21. Shah, V. K., and W. J. BrIll. 1977. Isolation of an iron-molybdenum cofactor from nitrogenase. Proc. Natl.Acad. Sci. U. S. A. 74:3249-3253.

22. Shah, V. K., J. R. Chisnell, and W. J. Brill. 1978.Acetylene reduction by the molybdenum cofactor fromnitrogenase. Biochem. Biophys. Res. Commun. 81:232-236.

23. Shah, V. K., L C. Davis, and W. J. Brill. 1972. Nitro-genase. I. Repression and derepression of the iron-mo-lybdenum and iron proteins of nitrogenase in Azotobac-ter vinelandu. Biochim. Biophys. Acta 305:445-454.

24. Smith, B. E. 1977. The structure and function of nitro-genase: a review of the evidence for the role of molyb-denum. J. Less Common Metals 54:465-475.

25. Strandberg, G., and P. W. Wilson. 1968. Formation ofN2 fixing system in Azotobacter. Can. J. Microbiol. 14:26-31.

26. Yoch, D. C., and R. AL Pengra. 1966. Effect of aminoacids on the nitrogenase system of Klebsiella pneumo-niae. J. Bacteriol. 92:618-622.

VOL. 145, 1981.

on May 4, 2018 by guest

http://jb.asm.org/

Dow

nloaded from