JouRNAsL OF BACTERIOLOGY, Feb. 1980, p. 470-4750021-9193/80/02-0470/06$02.00/0

Vol. 141, No. 2

Roles of nifF and nifJ Gene Products in Electron Transport toNitrogenase in Klebsiella pneumoniae

SUSAN HILL* AND EUGENE P. KAVANAGHAgricultural Research Council, Unit of Nitrogen Fixation, Universit) ofSussex, Fa*ner, Brighton, BNI

9RQ, United Kingdom

Crude extracts of the wild-type Klebsiella pneumoniae reduced C2H2 witheither pyruvate or formate as reductant (specific activity, 3 nmol min-1 mg ofprotein-l), whereas crude extracts of nifF mutant were almost inactive (specificactivity, 0.05). However, activity in the latter extracts was stimulated by addingAzotobacter chroococcum flavodoxin (specific activity, 10). Thus, nifF mutantsmay lack an electron transport factor. Crude extracts of nifJ mutants had about20% of the wild-type level of active MoFe protein, and thus nifJhas a presumptiverole in maintaining active MoFe protein. Studies on pyruvate or formate asreductants for nitrogenase in extracts of the nifJ mutants suggest in addition arole in electron input to nitrogenase for the following reasons. (i) Nitrogenaseactivity with these reductants was very low (specific activity, 0.06) and was notstimulated by extra MoFe protein or the flavodoxin. (ii) Activity was increasedby adding a crude extract of a mutant lacking the structural nif genes (specificactivity, 1) or a crude extract of the nifF mutant (specific activity, 4).

Nitrogenase from all sources so far examinedrequires ATP, a source of reducing power, andanaerobic conditions to reduce N2 or other sub-strates. The pathways of electron transport tonitrogenase are only partly understood. In thoseorganisms studied, either a flavodoxin or a fer-redoxin transfers electrons from a donor to ni-trogenase; such electron transport proteins candonate electrons nonspecifically to nitrogenasesfrom different sources (20, 22). In Clostridiumpasteurianum and Bacilluspolymyxa, pyruvateis a major source of electrons, but other metab-olites such as formate, NADH, or NADPH canalso provide electrons (20, 22). In Klebsiellapneumoniae, which fixes N2 anaerobically, py-ruvate, and to a lesser extent formate, malate,or an NADPH-generating system, supportedC2H2-reducing activity in crude extracts (21).Yoch (21) isolated from K. pneumoniae a fla-vodoxin which transferred electrons from imlu-minated heat-treated spinach chloroplasts to ni-trogenase; in addition, he obtained a brown pro-tein which was also active but less effective.The products of two genes, nifF and nifJ,

have been implicated in electron transport tonitrogenase in K. pneumoniae (16, 17). Acety-lene-reducing activity in cultures of these mu-tants is low but that in crude extracts withsodium dithionite and ATP is high. The latteractivity was not stimulated by adding either ofthe purified components of nitrogenase (16).

In this paper we present some biochemicalproperties of nifF and nifJ mutants and tenta-

tively assign steps in the electron transport tonitrogenase which these nif genes may control.

MATERIALS AND METHODSBacterial strains and plasmids. The strains of

K. pneumoniae used are listed in Table 1.Growth of strains. Strain M5al was grown in

large-scale cultures and harvested by the method ofEady et al. (4). All other strains were grown as follows.A 4-h aerobic culture in Luria broth was used as a 5%inoculum for either 100 ml (for toluene treatment) or1.5 liters (for crude extract preparation) of nitrogen--free medium (NFDM) (1) containing 1 and 1.5% glu-cose, respectively, and 1.3 mg of Casamino Acids perml. After 20 h of incubation at 280C with gentlesparging with 1% C02 in N2, organisms had grown toan absorbance at 540 nm of about 32 in a 0.5-cmcuvette and were harvested by centrifugation underN2.C2H2 reduction in cultures. Samples (1 ml) of

culture were removed before centrifugation and in-jected into Suba-seal-capped conical flasks (35 ml)containing 0.2 ml of 1.1 M glucose under 10 kPa ofC2H2 in Ar. After 10 min of shaking at 300C, thereaction was terminated by injecting 0.1 ml of 40%KOH, and gas samples (0.5 ml) were removed foranalysis by gas chromatography.

C2H2 reduction in crude extracts. Organismsharvested from 1.5-liter cultures were suspended in 10ml of N2-sparged 33 mM Tris-hydrochloride (pH 8.0)containing 1 mg of dithiothreitol, and the pH wasadjusted to 7.2 with 2 M NaOH. All bacteria weredisrupted by passage through a French pressure cellas described by Eady et al. (4). After centrifugation at38,000 x g for 30 min at 10°C, the supematants wereeither kept at 100C for use on the same day or stored

470

on August 27, 2017 by guest

http://jb.asm.org/

Dow

nloaded from

ELECTRON TRANSPORT TO NITROGENASE 471

TABLE 1. Bacterial strainsRelevant geno- Source or refer-

Strain Plasrnid0 type or pheno- encetype

M5al Nif+ P. W. WilsonUNF107 Anif 13UNF3001 pRD1 Nif+ 3UNF1624 pRD264 nifF2664::Tn7 M. J. MerrickUNF1202 pRD183 nifF2183::Tn7 13UNF1229 pRD213 nifH2213::Tn7 M. J. MerrickUNF3707 pMF107 nifJ2016::TnlO 13UNF3702 pMF102 nifJ2009.:TnIO M. J. Merrick

a All plasmids were in strain UNF107.

in liquid nitrogen. The latter treatment did not affectthe pyruvate- or formate-supported C2H2-reducing ac-

tivity. The supernatants, referred to as crude extracts,contained about 30 mg of protein per ml for derivativesof strain UNF107 and about 55 mg of protein per mlfor strain M5al.

Acetylene reduction assays were performed inSuba-seal capped serum bottles (8 ml) containing 10kPa of C2H2 in Ar and the following in a total aqueousvolume of 1.5 ml: 33 mM Tris-hydrochloride (pH 8.0),5.3 mM ATP, 13.3 mM MgCl2, 26.7 mM creatinephosphate, 50 ,ug of creatine kinase, and, where indi-cated, either 26.4 mM sodium dithionite, 33.3 mMsodium pyruvate, or 66.7 mM sodium formate. Withthe latter two substrates, 0.087 mM coenzyme A and0.029mM thiamine pyrophosphate were included. Theassay, performed at 30°C with shaking, was started byinjecting crude extract (0.1 ml) and terminated byinjecting 0.1 ml of 40% KOH. Gas samples (0.5 ml)were then taken for analysis by gas chromatography.In some experiments gas samples (0.5 ml) for gaschromatography were removed at intervals during theincubation and replaced by 0.5 ml of Ar.

C,H2 reduction in toluene-treated organisms.Harvested organisms were suspended in one-quarterthe original culture volume of N2-sparged, glucose-freeNFDM. After further incubation for 4 h at 28°C withgentle N2-sparging to remove residual glucose, samples(2 ml) of culture were injected into Suba-seal-cappedconical flasks (35 ml), which contained, under Ar, 0.2mg of dithiothreitol in 0.1 ml of water and 4 drops ofcresol red indicator. The indicator color was changedfrom yellow to orange (about pH 7.7) by injecting a

few drops of N2-sparged 2 M NaOH. N2-sparged tolu-ene (0.5 ml) was then injected, and the flask wasshaken at 30°C for 4 min. Portions (0.6 ml) of theaqueous phase were removed and immediately in-jected into serum bottles to start the assay for C2H2reduction. This was performed in 8.3 mM Tris-hydro-chloride (pH 7.4) as described for crude extracts; thesodium pyruvate concentration was 66.7 mM, andcoenzyme A and thiamine pyrophosphate were omit-ted.

Analyses. Protein was estimated by the procedureof Lowry et al. (11), and C2H4 (8) and H2 (4), by gaschromatography.

RESULTS AND DISCUSSION

As has been found previously (16, 17), C2H2

reduction in vivo by nifF and nifJ mutants wasnot limited by active nitrogenase. C2H2-reducingactivity in cultures of the nifF mutants(UNF1624 and UNF1202) and the nifJ mutants(UNF3707 and UNF3702) was 4% or less of thatin the Nif+ strain (UNF 3001), whereas theactivity in crude extracts, supplied with sodiumdithionite as reductant, was between 22 and 60%of the level in comparable assays of the Nif+strain (Table 2). The higher dithionite-sup-ported activity in the nifJ mutant, UNF3702,compared to strain UNF3707, may be a conse-quence of the extended growth time because theculture was grown for 22 h instead of 20 h, beforethe crude extract was prepared (Table 2). In-creasing the growth time of strains UNF3707and UNF1624 from 20 to 25 h only increased thedithionite-supported activity in crude extracts ofthe former strain, from 22 to 30% of the activityin comparable assays of the Nif+ strain (datanot shown).Although neither purified Fe protein nor pu-

rified MoFe protein, when added to crude ex-tracts of the nifFmutant (UNF1624) or the Nif+strain (UNF3001), caused a significant stimula-tion in dithionite-supported C2H2 reduction, theaddition of MoFe protein to crude extracts ofthe nifJ mutants (UNF3707 and UNF3702) pro-duced a four- to sixfold stimulation, whereasthere was no significant stimulation when the Feprotein was added (data not shown). These re-sults for the nifJ mutants are contrary to theobservations of Roberts et al. (16). The stimu-lation that they observed by adding nitrogenasecomponents to crude extracts of nifJ mutants,prepared after 48 h of growth under helium, wasvariable and nonspecific for MoFe protein.Clearly, further investigations are needed to de-termine whether the nifJ product is involved inregulation or in the stabilization or activation ofMoFe protein. Nevertheless, C2H2 reduction bycultures of nifJ mutants appeared to be limitedby some components other than active MoFeprotein.

If nifF and nifJ gene products are involved inprocesses providing electrons for nitrogenase ac-tivity, then C2H2 reduction supported by naturalelectron donors might be lower in crude extractsof the nifF and nifJ mutants than in the Nif+strain.Pyruvate and formate as reductants. Py-

ruvate or formate, both key intermediates inanaerobic glucose metabolism in the Enterobac-teriaceae, provides electrons for C2H2 reductionin crude extracts of K. pneumoniae M5al (21).In crude extracts of the Nif+ strain UNF3001,C2H2-reducing activity supported by either py-ruvate or formate was about 10% of that sup-

VOL. 141, 1980

on August 27, 2017 by guest

http://jb.asm.org/

Dow

nloaded from

TABLE 2. C2H2-reducing activity in cultures and in crude extracts of wild-type and mutant strainsSp acta

Relevant

Strain no. genotype Crude extracts with reductants:orpheno- Dithionite Pyruvate Pyruvate Fornate Formate plus Nonetype Dithionite ~~~~~~~~plusFldc om Fldc Nn

UNF3001 Nif+ 48.0 + 6.1 25.8 i 1.8 2.9 + 0.3 24.7 + 4.3 3.9 + 0.6 20.9 ± 2.9 0.3 ± 0.1(4) (12) (13) (3) (13) (3) (7)

UNF1624 nifF 1.0 ± <0.1 15.4 ± 1.7 0.04 ± 0.01 16.5 + 3.1 0.05 ± 0.01 15.2 ± 2.3 0.02 ± 0.01(4) (9) (9) (6) (9) (6) (6)

UNF1202d nifF 1.2 10.6 0.05 9.7 0.03 7.9 NDUNF3707 nifJ 1.1 ± 0.2 5.7 ± 0.5 0.03 ± 0.01 0.04 ± 0.01 0.06 ± 0.01 0.03 ± 0.02 0.01 ± <0.01

(4) (17) (11) (4) (14) (4) (7)UNF3702d nifJ 1.8 10.9 0.01 0.04 0.06 0 NDUNF3707 nifJ e 3.6 ± 0.1 0.8 ± <0.1 ND 1.0 ± 0.2 ND 0.02UNF1229 nifKDH J (4) (4) (4)UNF3707 nifJ e 8.2 ± 2.0 3.7 ± 1.5 ND 5.5 ± 1.6 ND 0.1UNF1624 nifF | (3) (3) (3)

Specific activities are eXpresed in nanomoles of C2H2 reduced minute- milligram of protein-' (reaction time, 10 min).Values for more than one estimate are means ± the standard error, the number of estimates is shown in parentheses. ND, Notdetermined.

b Culture assays were performed just before harvesting for crude extract preparation.CA. chroococcum flavodoxin (Fld; 0.25 mg) was added to the assaysd These strains were grown for 22 h, whereas all other strains were grown for 20 h before crude extracts were prepared.Equal volumes of crude extract of the two strains indicated were mixed under N2 just before performing the assay, only in

the mixture UNF3707 plus UNF1229 was the latter crude extract's protein component ignored for calculating the specificactivity.

ported by sodium dithionite (Table 2). The spe-cific activities with pyruvate or formate weresimilar to those obtained by Yoch (21) withcrude extracts of strain M5al, but with dithio-nite, C2H2 reduction was approximately fivetimes greater in our experiments.With pyruvate or formate as reductant, the

C2H2-reducing activity in crude extracts of thenifF and nifJ mutants was 2% or less of theactivity in comparable assays of the Nif+ strain(Table 2). Extending the growth time of culturefrom 20 to 25 h before crude extracts were pre-pared did not significantly alter the pyruvate- orformate-supported activity in the nifF mutant(strain UNF1624), the nifJ mutant (strainUNF3707), or the Nif+ strain (data not shown).To check that the activity in the nifJmutants

with these reductants was not limited by activeMoFe protein, dithionite-free purified MoFeprotein was added to the crude extract of strainUNF3707 prepared after 20 h of growth. Thisaddition failed to stimulate the activity withpyruvate or formate (data not shown). To con-firn that the MoFe protein present in crudeextracts of strain UNF3707 was active in theabsence of dithionite, a crude extract of a polarnifH mutant, strain UNF1229 (which containsneither the MoFe protein nor the Fe protein),was mixed with the crude extract of strainUNF3707. Significant activity was then obtainedwith either pyruvate or formate as reductant(Table 2).These results suggest that both the nifF and

the nifJ mutants might lack components of theelectron-donating system from either pyruvateor formate. Since neither the concentration northe composition and structural integrity of pro-teins in crude extracts may adequately representthe environment of nitrogenase in vivo, we in-vestigated the ability of pyruvate or formate toprovide electrons for C2H2 reduction in toluene-treated organims.Toluene treatment of cells from glucose-lim-

ited batch cultures of the Nif+ strain (UNF3001)abolished their ability to reduce C2H2 but ren-dered them permeable to ATP and dithionite.The C2H2-reducing activity supported by thesesubstrates was then 10% of the activity in intactorganism supplemented with glucose (Table 3).The activity supported by pyruvate in toluene-treated organisms was similar to that supportedby dithionite (Table 3). Apparently pyruvatewas principally providing reducing power, sincein assays where ATP was omitted the activitywas 10-fold lower (data not shown). With for-mate there was no C2H2 reduction unless theATP was added; the activity was then about10% of that in comparable assays with pyruvate(Table 3). This low activity was not due to theinability ofthe toluene-treated organisms to takeup formate, since formate supported H2 evolu-tion (data not shown). Hence, pyruvate, but notformate, was as efficient a source of reducingpower as dithionite in toluene-treated organismsof the Nif+ strain, possibly reflecting the impor-tance of pyruvate as a reductant in vivo.

472 HILL AND KAVANAGH J. BACTERIOL.

on August 27, 2017 by guest

http://jb.asm.org/

Dow

nloaded from

ELECTRON TRANSPORT TO NITROGENASE 473

TABLE 3. C2H2-reducing activity in toluene-treated cultures of wild-type and mutant strainsaRelevant geno- Reductant

Strain no. type or pheno-type Dithionite Pyruvate Formate None

UNF3001 Nif+ 17.7 ± 2.0 (8) 25.8 ± 6.7 (7) 1.6 ± 0.5 (7) 0.3 ± 0.1 (6)UNF1624 nifF 9.5 0.3 0.3 NDUNF1202 nifF 11.6 2.8 1.0 0.1UNF3707 nifJ 4.2 0.9 0.3 <0.1UNF3702 nifJ 6.4 1.0 0.2 0.1

aThe activities are expressed in arbitrary units, nanomoles of C2H2 reduced mmnute-' milliliter of culture-'(reaction time, 30 min). The activity, in the same arbitrary units, of glucose-supplemented cultures of strain3001 before toluene treatment was 191 ± 62.6 (7). Values for more than one estimate are means ± the standarderror; the number of estimates is shown in parentheses. ND, Not determined.

In toluene-treated cultures ofthe nifFand thenifJ mutants the C2H2-reducing activity, withdithionite as reductant, was between 24 and 66%of that in the Nif+ strain (Table 3), which issimilar to the results obtained with crude ex-tracts. With pyruvate as reductant, the activitycompared with the Nif+ strain was 4% or less inthe nifJ mutants (UNF3707 and UNF3702) andin one nifF mutant (UNF1624), and 11% in theother nifF mutant (UNF1202). Although thelatter percentage is significantly higher than thatobtained in crude extracts, these results suggestthat electron transport from pyruvate is im-paired in the nifF and the nifJ mutants.Possible role for the nifF gene product.

Since nifF mutants might lack a factor requiredto donate electrons to nitrogenase, we investi-gated the effect of adding purified Azotobacterchroococcum flavodoxin to assays for C2H2 re-duction in crude extracts with pyruvate or for-mate as reductant.The addition of purified A. chroococcum fla-

vodoxin to crude extracts of the Nif+ strain(UNF3001) stimulated pyruvate- and formate-supported C2H2-reducing activity 8.5-fold and5.4-fold, respectively (Table 2), which broughtthe specific activity up to near the level achievedwith dithionite. Yoch (21) found that a compo-nent of the pyruvate electron-donating systemin crude extracts of strain M5al was apparentlyunstable, because the addition of either a crudeextract of anaerobically grown Escherichia coli(21) or the flavodoxin from E. coli (22) stimu-lated C2H2 reduction with pyruvate. Thus, theconcentration of an active electron transportfactor probably limits C2H2 reduction supportedby pyruvate or formate in crude extracts of K.pneumoniae.The addition ofA. chroococcum flavodoxin to

crude extracts of the nifF mutants stimulatedthe activity more than 100-fold with either py-ruvate or formate as reductant (Table 2). Whenthis degree of stimulation is compared with thatof less than 10-fold in crude extracts of the Nif+

strain (Table 2), it strongly suggests that thesenifF mutants lack an electron transport proteinof similar redox potential to A. chroococcumflavodoxin. Such a protein may either be underthe control of the nifF product or be the nifFproduct itself, which has tentatively been iden-tified by sodium dodecyl sulfate-polyacrylamideelectrophoresis as a polypeptide of molecularweight 17,000 (16).A flavodoxin has been isolated from K. pneu-

moniae strain M5al, which transferred electronfrom illuminated heat-treated chloroplasts to K.pneumoniae nitrogenase (21). However, sinceeither a flavodoxin or a ferredoxin can transferelectrons to nitrogenase nonspecifically in thechloroplast assay (20, 22), the flavodoxin isolatedby Yoch (21) may not necessarily be involved inthe transfer in vivo. Repeated attempts in thislaboratory with the chloroplast assay to isolatean electron transfer protein from K. pneumo-niae, strain M5al, capable of transferring elec-trons to nitrogenase have failed, but experimentsare now under way to isolate such a protein withcrude extracts of a nifF mutant.Possible role for the nifJ gene product.

The addition of A. chroococcum flavodoxin tocrude extracts of the nifJ mutants did not stim-ulate the activity with pyruvate or formate asreductant (Table 2). Furthermore, the flavo-doxin was not visibly reduced as it was in com-parable assays of crude extracts of the nifFmutants and the Nif+ strain. To confirm thatthe crude extract of the nifJ mutant, strainUNF3707, contained the electron transport fac-tor, it was mixed with the crude extract of thenifF mutant, strain UNF1624. Significant activ-ity was then obtained with either pyruvate orformate as a reductant (Table 2). These resultssuggest that the nifJ product may either controlthe synthesis of, or be an enzyme involved incatalyzing the reduction of, low-redox-potentialelectron transfer proteins. The nifJ product hasbeen identified by sodium dodecyl sulfate-poly-acrylamide electrophoresis as a large poly-

VOL. 141, 1980

on August 27, 2017 by guest

http://jb.asm.org/

Dow

nloaded from

474 HILL AND KAVANAGH

peptide of molecular weight 120,000 (5, 13, 16).The question then arises, does the nifJ prod-

uct influence pyruvate or formate breakdown?The route of pyruvate breakdown in K. pneu-

moniae is not known in detail. Pyruvate sup-ported anaerobic growth ofstrain M5al in chem-ostats with N2 or NH4' as an N source (8) andalso supported N2 fixation in intact organisms ofAchromobacter (now reclassified as K. pneu-

moniae [12]) (6). Thus, anaerobic pyruvate me-tabolism apparently yields ATP. Whether py-

ruvate breakdown to provide electrons for nitro-genase activity is via the coliform-type routeconsisting of pyruvate-formate lyase (EC2.3.1.54, formate acetyltransferase) followed byformate hydrogenlyase (2, 9, 14) or via the clos-tridial type phosphoroclastic reaction (EC1.2.7.1, pyruvate synthase) (14, 22) is unknown.Both types have been implicated from fermen-tation data in N2-fixing Achromobacter (6).The former route would provide reducing

power from formate hydrogenlyase, an enzymecomplex composed of formate dehydrogenaseand hydrogenase. Such an enzyme complex wasapparently present in the Nif+ strain(UNF3001), the nifF mutant (UNF1624), andthe nifJ mutant (UNF3707), since formate sup-ported H2 evolution in crude extracts of thesestrains (Table 4). A hydrogenase is probably notinvolved in the transport of electrons to nitro-genase since H2 did not support C2H2 reductionin crude extracts of strain M5al unless methylviologen was added (S. Hill, unpublished data).If formate dehydrogenase is the enzyme provid-ing electrons for nitrogenase activity, then theactivity supported by formate should be as goodas, or better than, that supported by pyruvate.This was the case in crude extracts of strainUNF3001 (Table 2).However, several lines of circumstantial evi-

dence suggest that formate dehydrogenase is notthe only source of reducing power. Firstly, py-ruvate was a better source of reductant than

TABLE 4. H2 evolution from formate in crudeextracts of wild-type and mutant strains

Relevant geno- Sp act'Strain no. type or pheno- Mnsfr

type Plus formate mate

UNF3001 Nif+ 16.4 0.6 (2) 1.2 (1)UNF1624 nifF 16.5 ± 1.2 (2) 1.3 (1)UNF3707 nifJ 10.7 ± 1.2 (2) 1.5 (1)

a Specific activities, expressed as nanomoles of H2 evolvedminute' milligam of protein-', were estimated under thesame conditions as for C2H2 reduction, except that the MgC12,ATP, creatine phosphate, and creatine kinase were omitted(reaction time, 10 min). Values are means ± the range; thenumber of estimates is shown in parentheses. Crude extractswere prepared from cultures grown for 25 h.

J. BACTERIOL.

formate in toluene-treated cultures of the Nif+strain (UNF3001) (Table 3), and Yoch (21) alsoreported that in some crude extracts of strainM5al the activity with formate was lower thanthat with pyruvate. Secondly, sodium glyoxyl-ate, an inhibitor of pyruvate synthase in Clo8-tridium kluyveri and Clostridiumpasteurianum(18), more stongly inhibited formate-supportedC2H2 reduction than pyruvate-supported C2H2reduction in crude extracts of K. pneumoniaeM5al (Table 5). Therefore, pyruvate breakdowndoes not necessarily go through formate to yieldelectrons for nitrogenase activity. Thirdly, in E.coli the nirA (synonymous with fnr [15]) geneproduct controls the synthesis or assembly of avariety of anaerobic electron transport enzymessuch as formate dehydrogenase and hydrogenase(10). When the Nif+ plasmid pRD1 was trans-ferred into a nirA mutant, the exconjugantsreduced C2H2 (S. Hill, unpublished data). Thus,apparently formate dehydrogenase is not nec-essary for the production of electrons for nitro-genase activity in E. coli.Hence, the route of pyruvate breakdown to

yield electrons for nitrogenase activity in K.pneumoniae and E. coli may not proceedthrough pyruvate-formate lyase. Alternativeroutes of pyruvate breakdown have been de-scribed in E. coli which may also be present inK. pneumoniae. In addition to pyruvate-formatelyase, E. coli produces a pyruvate dehydrogen-ase complex, which is virtually inoperative inanaerobiosis due to the high NADH/NAD+ ra-tio (7) and a pyruvate-redoxin oxidoreductasewhich catalyzes the coenzyme A-dependent re-duction of flavodoxin required for the activationof the pyruvate-formate lyase (19). Thus, thenifJproduct could be involved in the productionof a pyruvate oxidoreductase which catalyzesthe reduction of the electron transport factor toprovide electrons specifically for nitrogenase ac-tivity in K. pneumoniae. Then, electron trans-

TABLE 5. Inhibition by sodium glyoxylate ofpyruvate- or formate-supported C2H2 reduction in

crude extracts of strain M5al% Inibition'

Reductant5 mM 10 mM 20 mM

Pyruvate 9 + 9 (5) 38 ± 13 (4) 87 ± 4 (3)

Formate 86 ± 9 (5) 99 ± 1 (4) 99 + 1 (3)aSodium glyoxylate, at the concentrations indi-

cated, was added to the assay for C2H2 reduction; thepercent inhibition was calculated from the activityobtained in the presence and in the absence of sodiumglyoxylate (reaction time, 10 min). Values are means± the standard error, the number of estimates is shownin parentheses.

on August 27, 2017 by guest

http://jb.asm.org/

Dow

nloaded from

ELECTRON TRANSPORT TO NITROGENASE 475

port to nitrogenase from formate in vitro mightbe via an energy-driven reversal of pyruvate-formate lyase followed by pyruvate oxidoreduc-tase to yield electrons for nitrogenase activity.The characterization of proteins from crude ex-tracts of wild-type K. pneumoniae which com-plement and restore pyruvate- or formate-sup-ported activity to crude extracts of nifJ mutantsshould therefore provide further informationconcerning the coupling of carbon substrate ox-idation to electron transport for nitrogenase ac-tivity in K. pneumoniae.

ACKNOWLEDGMENTWe thank J. R. Postgate, R. R. Eady, M. J. Merrick, and

M. G. Yates for constructive criticism of the manuscript. Wealso thank R. A. Dixon and M. J. Merick for supplying Kpneumoniae strains, J. A. Cole for providing the E. coli nirAmutant, R. R. Eady for supplying purified nitrogenase com-ponents, and R. Miller and M. G. Yates for providing purifiedA. chroococcum flavodoxin. In addition, we thank H. Bothefor suggesting the inhibitor sodium glyoxylate.

LIIERATURE CITED1. Cannon, F. C., R. A. Dixon, J. R. Postgate, and S. B.

Primrose. 1974. Chromosomal integration of Klebsiellanitrogen fixation genes in Escherichia coli. J. Gen.Microbiol. 80:227-239.

2. Cole, J. A. 1976. Microbial gas metabolism. Adv. Microb.Physiol. 14:1-92.

3. Dixon, R., C. Kennedy, A. Kondorosi, V. Krishnapil-lai, and M. Merrick. 1977. Complementation analysisof Klebsiellapneumoniae mutants defective in nitrogenfixation. Mol. Gen. Genet. 157:189-198.

4. Eady, R. R., B. E. Smith, K. A. Cook, and J. R.Postgate. 1972. Nitrogenase of Klebsiella pneumoniaepurification and properties of the component proteins.Biochem. J. 128:656-675.

5. Elmerich, C., J. Houmard, L Sibold, I. Manheimer,and N. Charpin. 1978. Genetic and biochemical anal-ysis of mutants induced by bacteriophage Mu DNAintegration into Kkebsiella pneumoniae nitrogen fixa-tion genes. Mol. Gen. Genet. 165:181-189;

6. Hamilton, I. R., R. H. Bums, and P. W, Wilson. 1965.Pyruvate metabolism by a nitrogen-fixing bacterium.Biochem. J. 96:383-389.

7. Hansen, R. G., and U. Henning. 1966. Regulation ofpyruvate dehydrogenase activity in Escherichia coliK12. Biochim. Biophys. Acta 122:355-368.

8. Hill, S. 1976. The apparent ATP requirement for nitrogenfixation in growing Klebsiella pneumoniae. J. Gen. Mi-crobiol. 95:297-312.

9. Knappe, J., J. Schacht, W. Mockel, T. Hopner, H.

Vetter, Jr., and R. Edenharder. 1969. Pyruvate for-mate-lyase reaction in Escherichia coli the enzymaticsystem converting an inactive form of the lyase into thecatalytically active enzyme. Eur. J. Biochem. 11:316-327.

10. Lambden, P. R., and J. R. Guest. 1976. Mutants ofEscherichia coli K12 unable to use fumerate as ananaerobic electron acceptor. J. Gen. Microbiol. 97:145-160.

11. Lowry, 0. IL, N. J. Rosebrough, A. L Farr, and R. J.Randall. 1951. Protein measurement with the Folinphenol reagent. J. Biol. Chem. 193:265-275.

12. Mahl, M. C., P. W. Wilson, M. A. Fife, and W. H.Ewing. 1965. Nitrogen fixation by members of the tribeKlebsiella. J. Bacteriol. 89:1482-1487.

13. Merrick, M., M. Filser, C. Kennedy, and R. Dixon.1978. Polarity of mutations induced by insertion oftransposons Tn5, Tn7, and TnlO into the nif genecluster of Klebsiella pneumoniae. MoL Gen. Genet.165:103-111.

14. Mortenson, L. E., and J.-S. Chen. 1974. Hydrogenase,p. 232-282. In J. B. Neilands (ed.), Microbiol iron me-tabolism, a comprehensive treatise. Academic PressInc., New York.

15. Newman, B. M., and J. A. Cole. 1978. The chromosomallocation and pleiotropic effects of mutations of thenirA+ gene of Escherichia coli K12: the essential roleof nirA' in nitrite reduction and in other anaerobicredox reactions. J. Gen. Microbiol. 106:1-12.

16. Roberts, G. P., T. MacNeil, D. MacNeil, and W. J.Brill. 1978. Regulation and characterization of proteinproducts coded by the nif (nitrogen fixation) genes ofKlebsiella pneumoniae. J. Bacteriol. 136:267-279.

17. St. John, R. T., IL M. Johnston, C. Seidman, D.Garfinkel, J. K. Gordon, V. K. Shah, and W. J.Brill. 1975. Biochemistry and genetics of Klebsiellapneumoniae mutant strains unable to fix N2. J. Bacte-riol. 121:759-765.

18. Thauer, R. K., E. Ruppecht, and K. Jurgermann.1970. Glyoxylate inhibition of Clostridial pyruvate syn-thase. FEBS Lett. 9:271-273.

19. Vetter, H., Jr., and J. Knappe. 1971. Flavodoxin andferredoxin in Escherichia coli. Hoppe-Seyler's Z. Phys-iol. Chem. 352:433-446.

20. Yates, M. G. 1977. Physiological aspects of nitrogen fix-ation, p. 219-270. In W. E. Newton, J. R. Postgate, andC. Rodriguez-Barrueco (ed.), Recent developments innitrogen fixation. Academic Press Inc., London.

21. Yoch, D. C. 1974. Electron transport carriers involved innitrogen fixation by the Coliform, Klebsiella pneumo-niae. J. Gen. Microbiol. 83:153-164.

22. Yoch, D. C. 1979. Electron-transport systems coupled tonitrogenase, p. 605-652. In R. W. F. Hardy, F. Bottom-ley, and R. C. Burns (ed.), A treatise on dinitrogenfixation sections I and II: inorganic and physical chem-istry and biochemistry. John Wiley & Sons, New York.

VOL. 141, 1980

on August 27, 2017 by guest

http://jb.asm.org/

Dow

nloaded from