Biochem. J. (1965) 94, 40

The Reduction of Nitrate, Nitrite and Hydroxylamine to Ammonia byEnzymes from Cucurbita pepo L. in the Presence of Reduced Benzyl

Viologen as Electron Donor

By C. F. CRESSWELL,* R. H. HIAGEMAN,t E. J. HEWITT AND D. P. HUCKLESBYUniver8ity of Briatol Long Ashton Research Station, Bri8tol

(Received 10 March 1964)

1. lEnzyme systems from Cucurbita pepo have been shown to catalyse the reduc-tion ofnitrite andhydroxylamine to ammonia inyields about 90-100%. 2. Reducedbenzyl viologen serves as an efficient electron donor for both systems. Activity ofthe nitrite-reductase system is directly related to degree of dye reduction whenexpressed in terms of the function for oxidation-reduction potentials, but appearsto decrease to negligible activity below about 9% dye reduction. 3. NADH andNADPH alone produce negligible nitrite loss, but NADPH can be linked to an

endogenous diaphorase system to reduce nitrite to ammonia in the presenceof catalytic amounts of benzyl viologen. 4. The NADH- or NADPH-nitrate-reductase system that is also present can accept electrons from reduced benzylviologen, but shows relationships opposite to that for the nitrite-reductase systemwith regard to effect of degree of dye reduction on activity. The product of nitratereductionmay be nitrite alone, or nitrite and ammonia, or ammonia alone, accordingonly to the degree of dye reduction. 5. The relative activities of nitrite-reductaseand hydroxylamine-reductase systems show different relationships with degree ofdye reduction and may become reversed in magnitude when effects ofdegree ofdyereduction are tested over a suitable range. 6. Nitrite severely inhibits the rate ofreduction ofhydroxylamine without affecting the yield ofammonia as a percentageof total substrate loss, but hydroxylamine has a negligible effect on the activity ofthe nitrite-reductase system. 7. The apparent Km for nitrite (1 KtM) is substantiallyless than that forhydroxylamine, for which variable values between 0 05 and 0 9mM(mean 0*51 mM) have been observed. 8. The apparent Km values for reducedbenzyl viologen differ for the nitrite-reductase and hydroxylamine-reductasesystems: 60 and 7-5 ,um respectively. 9. It is concluded that free hydroxylaminemay not be an intermediate in the reduction ofnitrite to ammonia by plants, and a

possible mechanism for reduction of both compounds by the same enzyme systemis discussed in the light of current ideas relating to other organisms.

The biological reduction of nitrate to ammoniahas been assumed to proceed in stages each involvingtwo electrons. Nitrite is often observed as an endproduct of nitrate reduction by plant enzymes invitro although ammonia is ultimately formed invivo. Hydroxylamine has been regarded as a likelyintermediate in the reduction of nitrite to ammoniawhere this process has been demonstrated. Thissubject has been extensively reviewed, e.g. byNason & Takahashi (1958), Nicholas (1957, 1958,1959), Webster (1959), McKee (1962) and Nason(1962).In an investigation of nitrate reduction by en-* Present address: Department of Botany, University of

Witwatersrand, Johannesburg, South Africa.t Present address: Department ofAgronomy, University

of Illinois, Urbana, Ill., U.S.A.

zymes in leaf extracts of vegetable marrow (Cucur-bita pepo L.) we sought to use reduced benzylviologen as a substitute for NADH as the electrondonor. The results at first were difficult to interpretbecause of inconsistency in the behaviour of thesame, or identically treated, preparations, used onthe same or on successive days. Thus separatepreparations having identical nitrate-reductase(NADH-nitrate oxidoreductase, EC 1.6.6.1) acti-vities as determined by measurements ofproductionof nitrite as the sole end product with NADHwould produce nil to large amounts of nitrite fromnitrate when assayed with reduced benzyl viologen,and sometimes small amounts ofammonia. Further,these same preparations might, when used on differ-ent occasions, either fail to reduce added nitrite orreduce it to ammonia with great rapidity. These

An0~

NITRATE REDUCTION BY CUCURBITA PEPO ENZYMESresults indicated the presence of an active nitrite-reductase system that had hitherto remainedundetected in plants (Hageman, Cresswell &Hewitt, 1962). As is shown below, the differenttypes ofbehaviour noted above were found to be allrelated to the probable degree of benzyl viologenreduction, when expressed in terms used to calculatethe oxidation-reduction potential of the system.In view ofthe recent descriptions (Tagawa & Arnon,1962; Valentine, Brill, Wolfe & San Pietro, 1962;Valentine, Jackson & Wolfe, 1962; Whitely &Woolfolk, 1962; Losada, Paneque, Ramirez & DelCampo, 1963; Mortenson, Valentine & Carnahan,1962; Paneque, Del Campo & Losada, 1963) ofsystems in which benzyl viologen and ferredoxins,the iron-containing proteins obtained from severalsources, have been shown to exhibit similar electron-transferring functions, we think our results are ofinterest and are relevant to studies in progress inseveral Laboratories (Hewitt & Betts, 1963; Losadaet al. 1963; Paneque et al. 1963) on systems thatutilize ferredoxins in nitrite and hydroxylaminereduction by plants. Experience obtained withreduced benzyl viologen may be useful in helping toelucidate the mechanism of action of the naturalsingle-electron carrier in nitrite and hydroxylaminemetabolism.Our preliminary results (Hageman et al. 1962) and

further observations reported briefly (Cresswell,Hageman & Hewitt, 1962) suggested a reappraisalof the status of hydroxylamine as an intermediatein nitrite reduction to ammonia. This problem isconsidered in relation to some kinetic aspects ofnitrite and hydroxylamine reduction to ammonia.

MATERIALS AND METHODS

Plants. Seedlings of Cucurbita pepo var. Bushy Type(Suttons Seeds Ltd., Reading) were grown in sand culture ina controlled environment chamber with a day length of18 hr. and alight-intensity of 1100 lumens/ft.2 at plant levelat 26.60 in the light and at 200 in the dark. The nutrientsolutions, including those deficient in micronutrients, wereprepared as described by Hewitt (1952); nitrogen was sup-plied solely in the nitrate form.

Reagent8. The following chemicals and enzyme prepara-tions were obtained from sources indicated: FMN, FAD,GSH, the sodium salts of NAD+ and NADP+, enzymicallyreduced NADH and NADPH, D-glucose 6-phosphate,the enzyme glucose 6-phosphate dehydrogenase (yeastZwischenferment) (EC 1.1.1.49), NADP+ and NADPH weregiven by Sigma Chemical Co.; L-cysteine hydrochloride,EDTA (disodium salt), diphenylthiocarbazone, NN'-dibenzyl-4,4'-bipyridylium dichloride (benzyl viologen),sulphanilamide, N-1-naphthylethylenediamine dihydro-chloride, tris and all other inorganic chemical reagents usedwere from British Drug Houses Ltd.; palladized asbestos(5% Pd) was from Hopkin and Williams Ltd. Ammoniumsulphate was heavily contaminated by heavy metals, andsaturated solutions were purified by extraction at room

temperature with dithizone at pH 7.5, and 8-hydroxy-quinoline at pH 5 as described by Hewitt & Nicholas(1964). ThepH was then raised by adding ammonia solutionuntil free ammonia could be detected.

Reduction of NN'-dibenzyl-4,4'-bipyridylium dichloride.The dye (3.65 mM) dissolved in 0 05 M-potassium phosphatebuffer, pH 7*5, was reduced by hydrogen and palladizedasbestos in the apparatus described by Hageman et al. (1962)and Hewitt & Nicholas (1964), and the extent of reductionwas determined by titration into potassium permanganatein an evacuated Thunberg tube as previously described. Theextent of dye reduction is expressed either as % of BVH*or as the logarithm of the ratio of reduced to oxidized dyeor as the concentration ofBVH which is independent of theexpressions above.Enzyme preparation. Rapidly expanding leaves from

plants between 3 and 5 weeks old were used in preparing theenzymes. The leaves were harvested, washed with deionizedwater, dried on clean filter paper and chilled in a cold roomat 20 for 15-30 min. One part (by wt.) of tissue and 3 vol.of 0 1 M-tris buffer, pH 7-8, containing disodium EDTA(0-4 mm) and L-cysteine hydrochloride (5 mm) were homo-genized in an MSE Ato-Mix blender for 90 sec. at full speed.All operations were carried out so that the enzyme prepara-tions remained cold (0-2°). The homogenate was squeezedthrough two layers of muslin, and centrifuged at 15MO0g^..for 15 min. Cold saturated (NH4)2SO4 solution containing1 mM-L-cysteine hydrochloride and 1 mm-KOH was addedto the supernatant solution to give 66% saturation and leftfor 30 min. before centrifuging at 3000 g for 30 min. Thesupernatant solution was discarded and the precipitate wasresuspended in cold 0.1 M-potassium phosphate buffer, pH7*5, containing L-cysteine hydrochloride (1 mM), in a volumeusually equivalent to the original weight of fresh tissue.The enzyme solution was placed in Visking dialysing

tubing that had previously been soaked in 10 mM-GSH forseveral hours, and dialysed against 1 mM-cysteine in 0-03M-phosphate buffer, pH 7 5, with three changes each of 31.for a total period of at least 15 hr. The enzyme solution wasthen centrifuged at 15000g^.y for 30 min. and decanted intoseveral chilled small polythene bottles. Some preparationswere separated from (NH4)2SO4 by passage through Sepha-dex G-50 (Pharmacia, Uppsala, Sweden) in a column or byadding solid ground Sephadex and then collecting theenzyme solution by centrifugation. If the enzyme prepara-tion was required immediately it was placed in ice; otherwiseit was stored at - 15°. Nitrate-reductase, nitrite-reductaseand hydroxylamine-reductase activities were stable forseveral weeks at -15°, but activity deteriorated signifi-cantly in a few hours at 00. During long experiments, frozenpreparations were crushed in the bottles and small amountswere thawed at intervals.Assay ofenzyme activity. Enzyme assays were done at 270

in Thunberg tubes under anaerobic conditions for periodsbetween 5 and 25 min. according to circumstances, loss ofsubstrate being a function of time. Assays were made atpH 7.5 in 0 05 M-potassium phosphate buffer. Between 0.1and 0 5 ml. of enzyme solution containing 4-15 mg. ofprotein/ml. was used according to activity and the objectsof the experiment; activity was proportional to enzymeprotein. Assays were made in duplicate and showed satis-factory agreement. In all cases the inclusion of boiled

* Abbreviations: BVH, reduced benzyl viologen; BV,benzyl viologen.

Vol. 94 41

C. F. CRESSWELL AND OTHERS

enzyme preparations showed that the reactions consideredhere were enzymic under the conditions specified.

Nitrate-reductase activity was measured by both nitriteand ammonia formation. Nitrite-reductase and hydroxyl-amine-reductase activities were measured both as substrateloss and ammonia formation until satisfactory evidence oftheir stoicheiometry was obtained.

Nitrate, when used, was present at a concentration of6 mm. The usual concentrations of nitrite or hydroxylaminewere 0-28 mm, except where otherwise stated. Nitritereductase was saturated by substrate under these conditionsbut hydroxylamine reductase was not.When NADPH (or NADH) was used as electron donor

at concentrations of 1 mm or more, the nucleotides wereremoved with zinc acetate and ethanol as described byHewitt & Nicholas (1964) before estimating substrate losses.Preparations contained an active glucose 6-phosphate-NADPHdehydrogenaseandan NADPH-specificdiaphorase(Avron & Jagendorf, 1956) that Hageman et al. (1962)showed readily reduced BV. These enzymes, supplementedon some occasions by Zwischenferment from yeast, formedan efficient system linking glucose 6-phosphate to nitritereduction or hydroxylamine reduction in the presence ofcatalytic quantities of NADP and BV as described in theResults section. When glucose 6-phosphate (20 ,Lmoles)was added and NADPH concentration was decreased to0-1 mm, usually in the presence of small amounts of BVH,as shown in the Results section, nitrite and hydroxylaminewere determined without prior removal of NADPH. Theseassays were carried out in a total volume of 2-5-3-0 ml.When partially reduced benzyl viologen was used as thesole electron donor in substrate amounts, 6-0 ml. of dye wasusually addedinthe manner described (Hageman et al. 1962).The assays were terminated by opening the Thunberg tubesand shaking for a few seconds to oxidize all the BVH. Theactual volumes of reaction mixture were determined to+ 0-05 ml. after withdrawing portions for various deter-minations. No reaction occurred between the substratesand hydrogen under the conditions used with precautionsto exclude any palladized asbestos from the assay mixtures.

Nitrite was determined by a slightly modified method ofEvans & Nason (1953) as described by Candela, Fisher &Hewitt (1957). Ammonia was determined in portions(1 0 ml.) by the microdiffusion method ofConway (1957) andcolorimetric measurement by the alkaline phenate-hypo-chlorite procedure of Russell (1945). Light-extinction wasmeasured at 625 mlu. The estimation of hydroxylamine wasat first carried out by the method of Csaiky (1949), but seriousinterference occurred when BV was present, apparentlybecause of pyridinium iodide formation. This reaction wasdecreased to negligible importance by using iodine at pH8-5+ 0-2, whereas hydroxylamine was oxidized nearlystoicheiometrically to nitrite provided that the concentra-tion of BV did not exceed 0-25 mm. The method used wasbased on that of Yamafuji & Akita (1952) and has beendescribed in a modified form suited to the present work(Hewitt & Nicholas, 1964). The extinction coefficient at540 m,u for nitrite was the same in either the above methodor the usual Griess-Ilosvay method. The same extinctioncoefficient was obtained for nitrite with the Griess-Ilosvaymethod in the presence or absence of hydroxylamine. Theamounts of nitrite and hydroxylamine present in a mixturewas therefore determined by first estimating nitrite bythe Griess-Ilosvay method and subtracting this extinction

value from that produced by the two together in the alka-line iodine method. The appropriate factor for hydroxyl-amine derived from calibration data obtained in thepresence of BV and nitrite (Hewitt & Nicholas, 1964) wasthen used to calculate changes in hydroxylamine concen-trations. When the effects of pyruvic oxime formation onenzyme activity were investigated a further modification ofthe method was found necessary. This consisted in adding0 3 ml. of a solution of iodine [1.25% (w/v) dissolved in thepresence of 9.5% (w/v) KI] to the test solution made up to6-7 ml. with 0-16 M-potassium phosphate, pH 8&5, and thenadding 0 05 ml. of N-KOH, which partially bleached theiodine colour. The mixture was heated in a water bath to800 and left to cool at room temperature before addition ofthe other reagents (Hewitt & Nicholas, 1964). By thismethod hydroxylamine alone yielded 90% of the extinctionat 540 m,u ofan equivalent amount of nitrite; in the presenceof pyruvate the yield was 100% and diacetyl monoximealso yielded 100% of the extinction of nitrite. Yamafuji &Akita (1952) observed that oximes yielded more nitrite thanfree hydroxylamine by their method. The method of Frear& Burrell (1955), which would have provided an independentcheck on hydroxylamine alone, could not be used in thepresence of BV.

Protein estimation. Protein was determined by Folin-Ciocalteu method as described by Lowry, Rosebrough, Farr& Randall (1951) and was calibrated with standard casein.Protein was precipitated in 5% (w/v) trichloroacetic acidand stored until convenient, when it was redissolved in0 1 N-NaOH before estimation.

RESULTS

Nature and products of reaction

Over the range of dye reduction corresponding tothe points in Figs. 1 and 3 the reactions resultingin substrate disappearance were strictly enzymic.When about 75% of the dye was reduced, i.e. whenlog ([BVH]/(BV]) = 0 5, a chemical reduction ofnitrite occurred but yields of ammonia were low.Within the range ofdye reduction in Figs. 1 and 3,

the conversion of nitrite or hydroxylamine intoammonia was nearly complete (Tables 1 and 2).The recovery of ammonia from nitrite alone was90 + 2 9% for 12 experiments, that from hydroxyl-amine was 85 + 4 0% for 17 experiments, and thetotal recovery of ammonia when nitrite and hy-droxylamine were both present was 91 + 2-4% for20 experiments that involved values of log ([BVH]/[BV]) between 1.15 and 0 08.Ammonia was also produced from nitrite in

quantitative yield when NADPH and catalyticamounts of BVH were used with glucose 6-phos-phate in place of substrate amounts of BVH, asshown in Table 3.

Effect of BVH concentration and percentage dyereduction on enzyme activity

The effect on enzyme activity of different con-centrations ofBVH in a partially reduced dye may

42 1965

NITRATE REDUCTION BY CUCURBITA PEPO ENZYMESTable 1. Los8 of nitrite or hydroxylamine and production of ammonia when 8ubStrateW were pre8ent

either 8eparately or together in different concentrations in the presence of benzyl viologen

Assays were carried out in duplicate under anaerobic conditions in a volume of 7.2 ml. at 270 for 25 min. with7*8 mg. of protein as described in the Materials and Methods section, the same enzyme preparation being usedin both experiments. Values are given as m,umoles.

Initial substrate given

Nitrite Hydroxylamine2000 02000 4002000 2000400 2000

0 2000400 0

0 400

log ([BVH]/[BV]) T-81

Substrate lossIit Ammonia

Nitrite Hydroxylamine gained.1050 1005998 63 903960 105 1153391 268 633

514 527395 303

154 107

log ([BVH]/[BV]) 1-64

Substrate lossAmmonia

Nitrite Hydroxylamine gained606 - 486568 75 532627 119 771395 166 553- 478 375400 391

158 110

Table 2. Lo88 of nitrite or hydroxylamine and production of ammonia when sub8trate8 were present eitheralone or together with different values for benzyl viologen reduction such that the relative activitiem of nitrite

reducta8e and hydroxylamine reductase are rever8ed.

Assays were carried out under anaerobic conditions with 13*3 mg. of protein in 7-2 ml. at 270 for 25 min.,as described in the Materials and Methods section.

Substrate given(m,umoles)

log ([BVH]/[BV]) Nitrite HydroxylamineTf31 2000 0

0-08

2000400

0

20002000400

0

200020002000

0

200020002000

Substrate loss(m,umoles) Ammonia gained

Nitrite Hydroxylamine (m,umoles) (% recovery)245 - 244 100225 20 248 101212 48 231 89- 296 288 971173 - 1078 921105 101 1187 98391 204 568 95- 415 407 98

Table 3. Production of ammoniafrom nitrite with NADPH as electron donor and catalytic amounts of reducedbenzyl viologen, in comparison with the sy8tem with reduced benzyl viologen alone or NADH

The NADPH was placed in the side arm of the Thunberg tube and mixed with the contents of the tube (buffer,substrate and enzyme) after evacuation. The benzyl viologen oxidized (BV) or partially reduced (BVH) dyewas admitted as described in the Materials and Methods section. No glucose 6-phosphate was added in theseexperiments.

Ammonia gained

Expt. no.

Electron donors(,umoles)

1 BVHHonly, 9.5NADPH, 6-8, +BV, 0-6NADPH, 13*6, + BV, 0 3NADPH only, 6-8

2 BVH only, 12-6BVH only, 1-3NADPH, 3 0, +BV, 1*3NADPH 3 0, + BVH, 1-3NADH, 3-0, +BV, 1-3

Nitrite loss

(m,umoles)1180510540

0

7940

313439

0

(m,fmoles)957600506

0

7570

2824640

(yield %)8111794

95

90105

Percentageinhibition of

hydroxylaminereductase

9384

7651

Vol. 94 43

C. F. CRESSWELL AND OTHERS

reflect the combined effects ofthree possible factors,namely the redox potential ofthe BV-BVH system,the actual concentration of reduced dye in relationto the saturation of the enzyme, and possiblecompetition between BV and BVH for the enzymesite.

Effect of BVH concentration on nitrite reductame.The effect of concentration of BVH on nitrite re-

duction was studied in two ways. In Fig. 1 constantvolumes of BV+BVH present at a total concen-

tration of 3-65 rm but of varying degrees of reduc-tion (% ofBVH) were introduced into the Thunbergtube containing nitrite, buffer and enzyme as

described in the Materials and Methods section.The results ofseveral experiments with preparationsof comparable activity showed a linear relationshipbetween extent of nitrite loss in a given time anddegree of dye reduction when expressed as

1:es

0

aqGo:0)

*zv

:._

1965

log([BVH]/[BV]). Since this expression is related tothat for the standard electrode potential in thefollowing manner: E=Eo-0.061og([BVH]/[BV])(e = 1), the relationship indicates that nitrite-reductase activity as measured by nitrite loss andammonia formation was directly proportional tothe redox potential of the BV-BVH system over

the range tested. The results suggest also that,when dye reduction was less than about 10%, i.e.log([BVH]/[BV]) < 1, nitrite-reductaseactivitywasnegligible (Figs. 1, 2 and 3). Further evidence to

100

804

0

.5600

xocq1

Go

)00

0

0)- Af

2001

01.5log (CBVH]/[BV])

Fig. 1. Relationship between activity of nitrite reductaseand proportion of benzyl viologen reduced expressed as

log([BVH]/[BV]). Results are shown for a random collec-tion of four preparations (represented by four symbols)having comparable activity, and containing between 6 and13 mg. of protein/ml. Each was assayed at several differentvalues oflog ([BVH]/[BV]) calculated from titration againstpermanganate of a 3-65 mm solution of the dye reducedcatalytically to various extents and used as electron donorin substrate amounts as described in the Materials andMethods section.

To 1 5 0

log ([BVH]/[BV])

Fig. 2. Relationship between activity of nitrite reductaseand ratio of constant concentration ofBVH to varying con-

centrations of BV. Dye was reduced catalytically to about50% and a constant amount (8 ,umoles of BVH) was addedas described in the Materials and Methods section toThunberg tubes containing different amounts of oxidizedBV to produce different values of log ([BVH]/[BV]). A ando represent results of separate experiments.

44

c

I

0

m

NITRATE REDUCTION BY CUCURBITA PEPO ENZYMES

700

z

03

0

z

01.4~ ~ ~ ~

o

1400t £w i

0

1o T-5 0 0-2log ([BVH]/[BV])

Fig. 3. Differential effects of degree of reduction of benzylviologen on nitrite or hydroxylamine reduction. Results are

shown for two different preparations represented by , Eand A, A respectively when assayed for nitrite-reductase(El, ) and hydroxylamine-reductase (U, A) activities atthe same time for different values of log([BVH]/[BV]) withsubstrate quantities of 3-65 mM-benzyl viologen solution asdescribed in the Materials and Methods section.

support the conclusion drawn from the results ofthe experiments of Fig. 1 was obtained in anothermanner, namely by observing the effects on nitritereduction ofa constant concentration ofBVH whensupplied in the presence of different concentrationsof oxidized BV. For this purpose 9-65 ,equiv. ofBVH, present in 6 ml. of partially reduced dye, i.e.with log([BVH]/[BV]) = 1.91, as described in theMaterials and Methods section, was introduced intoThunberg tubes to which enzyme, substrate, bufferand varying amounts, or none at all, of oxidizedBV had already been added. In this manner a rangeof values of log ([BVIH]/[BV]) from 1-13 to 1 91 wasobtained similar to those shown in Fig. 1, but witha constant BVH concentration comparable withthe highest values present in the previous experi-ments. The maximum total concentration of dye

was about 4 mg./ml., which was not of itself inhi-bitory to the enzyme. Results of two experimentspresented in Fig. 2 show that there was a linearrelationship between quantity of nitrite reduced ina given period of time and the initial value oflog ([BVH]/[BV]). Extrapolation to zero nitrite re-duction indicated as before that activity tended toa negligible value when dye reduction was less thanabout 10%, in spite of the relatively high concen-tration ofBVH present.The assumption that initial activity is linearly

related to log([BVH]/[BV]) in a system where suchactivity causes a progressive decrease with time inthe ratio ([BVH]/[BV]) is in accordance with theobservation that the mean activity measured overan interval of time tends to be linearly related tothe initial value of the ratio. We are indebted toMr G. M. Clarke of the Statistics Section at LongAshton for confirming the validity of this con-clusion. It does, however, fail to hold when initialactivity is so great that most of the available BVHis utilized during the reaction, as occurred for thetwo highest points in the upper line of Fig. 2.The value of log([BVH]/[BV]) as distinct from

the actual concentration of BVH seems to be theprincipal rate-limiting factor at low values ofnitrite-reductase activity (Fig. 1). This was con-firmed in another manner. As described in theMaterials and Methods section, our preparationscontained both a highly active diaphorase (Avron& Jagendorf, 1956) and a glucose 6-phosphatedehydrogenase. These systems, in the presence ofadded glucose 6-phosphate and NADP+, maintainedcatalytic amounts of dye in the reduced state, theconcentration of which could be calculated by themethod explained in the section describing themeasurement of apparent Km for BVH. The resultsin Table 4 show that when the concentration ofreduced dye in this system (B) was maintainedas low as 0 35 mm, i.e. that corresponding tolog(([BVH]/[BV]) = 1'15 for the conditions used inFig. 1, the reduction of nitrite was as rapid asthat which occurred in system A when the concen-tration of BVH was between 1 2 and 1 7 mM andthe total dye was approximately 50% reduced, i.e.log ([BVH/BV]) = 1.95 to 0-02. When less-reduceddye was used in system A, i.e. log([BVH]/[BV])= i-31 to 1s51, the rates were much less than insystem B in spite of the higher concentrations andquantities of BVIH still present in system A.As the dye in system B was not reduced to the

maximum extent possible until a few minutes afterstarting the assays by tipping NADP+ or glucose6-phosphate from the side arm, the average ratesshown in Table 4 were in fact somewhat less than themaximum possible under steady-state conditions.In system A also the average rates shown for thenominal values of log([BVH]/[BV]) were less than

Vol. 94 45

Table 4. Effects on the activity of nitrite-reductase and hydroxylamine-reductase systems with partiallyreduced benzyl viologen either alone in substrate quantities (A) or in catalytic amounts in the presence of a

reducing system (B)

The amounts of BVH used are shown as log ([BVH]/[BV]), as concentration and as total quantity. The assayswere carried out under anaerobic conditions at 270 as described in the Materials and Methods section. In thereducing system (B) glucose 6-phosphate (5 mM), and NADP (0-3 mm), glucose 6-phosphate dehydrogenase andNADPH diaphorase were also present.

Benzyl viologen*

Preparation(i) Fe- and Mn-deficient

plants(ii) Mn-deficient plants

(iii) Fe-deficient plants

(iv) Full nutrient plants

Conen. ofBVH Total BVHSystem log ([BVH]/[BV]) (mM) (,moles)A 1-31A 0-02B 007A 1-98B 0-07A 1-95B 0-07A 1-90A I-51B 007

0-621-660-341-600-361-560-601-230-650-35

3-4511-90-95

11-50-95

11-21-908-854-680-95

Enzyme activity (as substrate loss)(m/imoles/min./mg. of protein)

Nitrite Hydroxylaminereductase reductase

1-87-78-1

14-812-16-96-8

20-82-4

17-4

9.912-513-18-77-36-06-28-57-57-3

* Values of log ([BVH]/[BV]) and corresponding values for concentration of BVH and total BVH are calculated eitherfrom titration with permanganate for system A or from data obtained for equilibrium conditions in the NADP-glucose6-phosphate system B as described in the text in connexion with the determination of Km for BVH.

the initial rates which would have been expectedunder steady-state conditions for the reasons ex-plained above in connexion with the results in Fig. 2.

Effect of BVH concentration on hydroxylaminereductase. The data of Fig. 3 show the effect ofconcentration ofBVH on hydroxylamine-reductaseactivity as compared with that on nitrite reductaseassayed simultaneously on the same preparations.Hydroxylamine reductase was much less sensitiveto changes in amount of BVH than nitrite reduct-ase. It was observed frequently, as shown in Fig. 3and Table 2, that whereas nitrite was reduced at aslower rate than hydroxylamine when the concen-tration of BVH gave log ([BVH]/[BV]) valuesbelow 1-4, the reverse relationship existed whenlog ([BVYH]/[BV]) was between about 1-7 and 0-1.At intermediate values around 1-5 the two sub-strates were reduced at similar rates. As statedabove, yields of ammonia from either substratewere around 90%, and it therefore follows thatnitrite could be reduced to ammonia either moreor less rapidly than hydroxylamine, depending onlyon the degree of dye reduction.

Effects of BVH concentration on nitrate reductase.Although the present paper is principally concernedwith the characteristics of nitrite-reductase andhydroxylamine-reductase systems, it is relevant todraw attention to the effects of concentration ofBVY on nitrate reductase, present in the samepreparations as those used above, in view of theconfusing results mentioned in the introduction.

The results of three experiments similar to thosedescribed for the effects of concentration ofBVH onthe other two enzymes are shown in Fig. 4. Therewas a striking and entirely opposite type of effectof concentration of BVH, expressed as log([BVH]/[BV]), on nitrate-reductase activity which was ob-viously independent of any possible effects on en-zyme saturation by BVH. When the degree of dyereduction exceeded a value of 1-5 for log([BVH]/[BV]) nitrate-reductase activity was negligible. Atvalues between 1 4 and 1- 1 nitrite accumulated andammonia was also formed. At values below 1-1no ammonia appeared but nitrite production wasmaximal for the range of dye-reduction values used.Errors in ammonia estimations where amountsproduced were small probably decreased theirprecision more than those of nitrite, and thus therange of dye reduction over which actual ammoniaproduction was detected varied in the three experi-ments. The general conclusion, however, thatnitrate reductase and nitrite reductase present inthe same preparation may produce only nitrite, ornitrite and ammonia together, or ammonia only, ormay fail entirely to produce either, according onlyto the degree of dye reduction seems justified andexplains completely the confusing situation en-countered in earlier stages of the work. Inspectionof Figs. 3 and 4 and Tables 1 and 2 together suggeststhat there is a relatively narrow range of dye-reduc-tion values log ([BVH]/[BV]) = 1-2 to 1-4 withinwhich nitrate willform nitrite, nitrite will be reduced

46 C. F. CRESSWELL AND OTHERS 1965

NITRATE REDUCTION BY CUCURBITA PEPO ENZYMESlog ([BVH]/[BV])

1-0 1-5 1-8 [N02-1 (PM)log ([BVH]/[BV]) Fig. 5. Effect of hydroxylamine on nitrite-reductase acti-

vity. Hydroxylamine, when present, was at an initial con-elationship between nitrate reduction to nitrite or centration of 0-3 mM. Substrate quantities of BVH wereand degree of benzyl viologen reduction expressed used as electron donor as described in the Materials andTH]/[BV]). A 3-65 mm solution ofdye was reduced Methods section. The results are plotted according to onefly to different extents as shown by titration of the methods described by Dixon & Webb (1958). V wasermanganate and used as electron donor in sub- calculated as mumoles of nitrite lost/min./mg. of protein inounts as described in the Materials and Methods 20 min. andS as concentration ofN02- (PM). e, Hydroxyl-Che left-hand part of the Figure shows production amine present; 0, hydroxylamine absent.

ofnitrite (0, 0, A) and ammonia (, A,A) measured simul-taneously in three separate experiments represented by thedifferent symbols. The right-hand part of the Figure showsthe calculated nitrate loss as sum of nitrite and ammoniaproduction.

and form ammonia, and hydroxylamine will bereduced and form ammonia at comparable rates.

Apparent Michaeli8 con8tants for nitrite andhydroxylamine

The apparent Michaelis constants were deter-mined as shown in Figs. 5 and 6 by the graphicalmethods of Lineweaver & Burk (1934) as describedby Dixon & Webb (1958). A value about 1 ,Mwas consistently obtained for nitrite. The nitrite-reductase system was practically saturated at a

concentration of 0-05 mm, and concentrations above0-5 mm produced substrate inhibition. Valuesobtained for hydroxylamine in Fig. 6 were lessconsistent than for nitrite, and for ten experimentsgiving linear reciprocal plots they ranged from0-05 to 0-9 mm (mean 0-51 + 0-24 mm). The appar-ent Km for hydroxylamine nevertheless appearedto be much greater than that for nitrite.

-1/Km

0

1/[NH2OH] (mnT1)

Fig. 6. Lineweaver-Burk plots for hydroxylamine-reduct-ase preparations obtained on different occasions or fromplants grown under different nutritional conditions: E,o, separate full nutrient plants; A, Cu-deficient plants;*, Mn-deficient plants. Substrate quantities of BVH were

used as described in the Materials and Methods section. V

was calculated as m,umoles of hydroxylamine lost/min./mg. of protein in 20 min.

Vol. 94 47

.66

c1Oa)0

4

00

A-

6

.4)

z

15001-

.1000[

500-

0

Fig. 4. R(ammonia i

aslog([BNcatalyticaagainst I(strate amsection. I

2000r

C. F. CRESSWELL AND OTHERSEffects of nitrite and hydroxylamine presenttogether on hydroxylamine-reductase and

nitrite-reductase activitiesAs our preliminary work (Hageman et al. 1962)

had raised some doubt about the status ofhydroxyl-amine as an intermediate in nitrite reduction toammonia, we investigated in more detail the inter-action between the two compounds with respect toammonia production and kinetic behaviour. Rele-vant data are presented in Tables 1, 2 and 5 and inFig. 5. Nitrite reduction to ammonia occurred atrates that were practically independent of thepresence or absence of hydroxylamine; in 13 experi-ments covering a range of hydroxylamine/nitriteratios between 10:1 and 0*2:1, under conditionswhere nitrite concentrations were approachingsaturation, the observed mean inhibition of nitritereductase by hydroxylamine was only 3 + 2.90%, i.e.it did not differ significantly from zero. As shownin Fig. 5 hydroxylamine had an almost negligibleeffect on the apparent Km for nitrite, raising it from1.0 to 1*4 /uM. By contrast, however, the presence ofnitrite sharply decreased the rate of ammoniaproduction from the mixture of substrates com-pared with the expected rate for the sum ofthe separate nitrite-reductase and hydroxylamine-reductase activities. Simultaneous estimations ofsubstrate loss and ammonia formation showed thathydroxylamine reductase was severely inhibited bythe presence of nitrite; in 14 experiments the meaninhibition was 81 + 7.1%. The degree of inhibitionby nitrite appeared to be independent of either theinitial nitrite concentration or of the rate of nitritereduction, whether expressed on a protein basis oras controlled by percentage dye reduction, providedthat (i) some nitrite (about 20 m,tmoles in 7-2 ml.)still remained by the end of the assay period and(ii) nitrite-reductase activity was appreciable.However, as shown in Table 5, when percentage dye

reduction was decreased to values such thatnitrite-reductase activity was negligible, hydroxyl-amine-reductase activity, which was still consider-able, was inhibited by nitrite much less than atconcentrations of BVH which also permitted rapidnitrite reduction. This was clear not only in termsofpercentage inhibition but also ofactual hydroxyl-amine loss. One other point arising from Table 1 isthat the percentage inhibition of hydroxylaminereduction in the presence of nitrite decreased as theconcentration ofhydroxylamine was decreased from2000 to 400 m,umoles in 7-2 ml. Although threeexperiments (not presented here) were performed toinvestigate this relationship in more detail, theinconsistent kinetic behaviour of the hydroxyl-amine-reductase system noted above precluded thedetermination with any confidence of the effect ofnitrite on apparent values of Km or Vmax. for thehydroxylamine-reductase system.

Apparent Michaelis constants for BVH

The relationship between log([BVH]/[BV]) andenzyme activities made impracticable the deter-mination of Km merely by increasing the concen-tration of BVH without corresponding increase inthat of BV. The preparation of several solutionsproviding different concentrations of BVH withthe ratio [BVH]/[BV] practically constant wasachieved as an approximation by using the endo-genous NADPH-diaphorase and glucose 6-phos-phate-dehydrogenase systems (supplemented byadding Zwischenferment) in the presence ofNADP+and glucose 6-phosphate to reduce the dye to aconstant equilibrium value. This was reachedwithin 15 min. under anaerobic conditions as shownby measurements at 780 mp, of light-extinction ofthe solution in a 1 cm. cuvette attached to theThunberg tube; the maximum extinction when no

Table 5. Inhibition of hydroxylamine reductase by nitrite in relation to the presence orabsence of nitrite-reductase activity

Assays were carried out under anaerobic conditions for 25 min. using substrate quantities of benzyl viologenas described in the Materials and Methods section. The degree of dye reduction was adjusted in accordance withthe relationship shown in Fig. 1 to produce either negligible or considerable nitrite-reductase activity in twopreparations showing either intrinsically fairly high (A) or relatively low (B) nitrite-reductase activity and fairlyhigh hydroxylamine-reductase activities on a protein basis. Each system contained 2000 m,tmoles of hydroxyl-amine and the same quantity of nitrite when added.

Substrate loss (mumoles/mg. of protein)

Preparation log ([BVH]/[BV])A i*71

1.03B 1.84

1.11

Nitrite1950

230

Hydroxylamine26413217797

Hydroxylaminenitrite61805171

Percentageinhibitionby nitrite

77407226

48 1965

NITRATE REDUCTION BY CUCURBITA PEPO ENZYMESfurther change occurred was proportional to theconcentration of BV provided initially. It wastherefore possible to obtain a range of values forBVE concentration each with practically the samevalue oflog ([BVH]/[BV]), which was determined asfollows. On the basis of the increase ih light-extinc-tion at 320 m,u for the cyanide-addition reaction atpH 10-0 (Colowick, Kaplan & Ciotti, 1951), andfreedom from NAD+ as detected chromatographic-ally (Sigma Chemical Co. assay sheet), it was foundfrom change in extinction at 340 mp in the absenceof BV that the NADP+ was 90% reduced whenequilibrium was attained at pH 7-5. Taking E' forNADPH at pH 7-5 (270) as -0-332v (Mansfield-Clark, 1960) and as -0-363v for 90% reduction, andEo ofbenzyl viologen as - 0-359v (pH-independent)(Michaelis & Hill, 1933a,b), we calculate that theBVshouldhave been 54% reducedwhen equilibriumwas reached in the NADPH system; this value wasused to calculate the concentration of BVH pro-duced for each amount ofBV originally taken.When it was considered that equilibrium had

been established for dye reduction in a series ofinitially differing amounts of BV, the substrate(nitrite or hydroxylamine) was tipped in from theside arm of the Thunberg tube and the assay wascontinued for 20 min. It was assumed on the basis ofobservations ofdye colour and light-extinction thatBVE oxidized by nitrite or hydroxylamine wasreduced again fast enough to prevent the occur-rence of appreciable changes in the initial concen-

[BVH] (p)

Fig. 7 Relation between nitrite-reductase (A\) or hydroxyl-amine-reductase (o) activities and concentration of BVHfor constant degree of dye reduction shown as reciprocalplots by one of the methods described by Dixon & Webb(1958). V was calculated as m,umoles of nitrite or hydroxyl-amine lost/min./mg. ofprotein in 20 min.; actual rates werelow enough to be rate-limiting in the overall system in whichglucose 6-phosphate, NADPH, glucose 6-phosphate de-hydrogenase and the diaphorase were used to maintaincatalytic amounts ofBVH in the required concentrations asdescribed in the Materials and Methods section. S is shownas concentration of BVH (juM).

tration of BVE at any given value. The apparentKm for BV in the NADPH-diaphorase system wasdetermined in separate experiments and found tobe 0-44mM; saturation occurred at about 4 mM-BV.However, the rate of BV reduction in this systemwas still rapid enough in selected preparations toensure that this reaction was not a rate-limitingfactor in the reduction of nitrite or hydroxylamineto ammonia. The results with the procedure justdescribed are shown for the nitrite-reductase andhydroxylamine-reductase systems in Fig. 7. Thevalues for the apparent Km were 60 p,u. for nitriteand 7-5 um for hydroxylamine. The values ofBYEin Figs. 1, 2 and 3 for which nitrite-reductase acti-vity tended to zero were therefore still six times thatof the apparent Km.

Effects of some inhibitors

Results of inhibitor tests are given in Table 6.Cyanide at pH 7-5 in concentrations between 1and 5 mm was severely inhibitory to both systems.Effects of lower concentrations suggested that thenitrite-reductase system was the more sensitive ofthe two. Inhibition by azide at pH 7-5, whencorrections were introduced for chemical reactionbetween azide and nitrite (Hewitt & Hallas, 1959),was negligible at concentrations up to 0-2 mm fornitrite-reductase and 0-7 mu for hydroxylamine-reductase systems. Higher concentrations cannotbe used reliably when nitrite or hydroxylamineare estimated colorimetrically (Villanueva, 1959;Hewitt & Hallas, 1959); the manometric proceduredescribed by Hewitt & Hallas (1959) was not usedin view of the negative results obtained with lowerconcentrations. Several other inhibitors of metal-enzyme systems, i.e. 1,1'-bipyridyl, 8-hydroxy-quinoline, sodium diethyldithiocarbamate, o-phenanthroline and EDTA (disodium salt), werepractically inert and no evidence for the presence ofmetal components was obtained in these studies.The effects of cyanide are not conclusive evidencein this respect, and it is relevant to note that Senez& Pichinoty (1958c) made the remarkable andemphatic statement that 1 mM-potassium cyanideseverely inhibited the non-enzymic reoxidation ofBVE by nitrite. It is possible that a cyanide addi-tion compound analogous to that formed withNADP+ is produced (Colowick et al. 1951). Senez &Pichinoty (1958a) found that hydrazine at 10 and1-0 mm inhibited hydroxylamine reduction 65 and23% respectively in cells of Desulfovibrio desulfuri-cans. Walker & Nicholas (1961) reported thathydrazine acted as a competitive inhibitor ofhydroxylamine reductase of Pseudomonas aeru-ginosa; concentrations of 1-7 and 0-33 mm inhibited82 and 37% respectively. Several tests were madewith nitrite-reductaseandhydroxylamine-reductase

Vol. 94 49

Table 6. Effects of soMc inhibitors on nitrite-reductase and hydroxylamine-reductase activities

Assays were carried out at pH 7.5 under anaerobic conditions at 270 for 25 min. with reduced benzyl viologenas the electron donor in substrate quantities as described in the Materials and Methods section. Further detailsof the use of azide, thiol reagents and pyruvate are given in the text. Values indicate percentage inhibition com-pared with control systems, except where stimulation occurred as shown, for inhibitors given at the concentration(mM) shown for each system.

Nitrite-reductase system Hydroxylamine-reductase system

InhibitorPotassium cyanide

Sodium azidep-Chloromercuribenzoate

Phenylmercuric acetate

Hydrazine sulphate

Sodium pyruvate

o-Phenanthroline1,1'-Bipyridyl8-HydroxyquinolineSodium diethyldithiocarbamateEDTA (disodium salt)

(mM)510-10-015

0-250-2

0-2 (+ 1 mM GSH)0-2

0-2 (+ 1 mM-GSH)41-53-312511111

(% inhibition)1001008510

07814090000001090

18% stimulation0

(mM)52-20-510-220-070-70.5

0 5 (+ 2-5 mm-GSH)

0.50 5 (+ 2-5 mM-GSH)

41.53-51*2511111

(%/ inhibition)100786855170

4724

732210-1502070

10% stimulation5% stimulation

912

Table 7. Effects of soMe micronutrient deficiencies on activity of nitrite-reductase and hydroxylamine-reductasepreparations

Severity of the deficiency condition of the plants is indicated as slight, moderate or severe in terms of visiblesymptoms. Necrotic leaves were omitted from the preparations. Assays were carried out under anaerobic condi-tions at 270 for 25 min., as described in the Materials and Methods section with substrate quantities of partiallyreduced benzyl viologen as shown.

Origin of preparation and severity ofdeficiency(i) Plants 5 weeks old; preparation stored

10 days at -15°Full nutrient plantsSeverely Fe-deficient plantsSeverely Mn-deficient plantsModerately Zn-deficient plantsSlightly Cu-deficient plants

(ii) Plants 7 weeks old; preparation stored20 days at -15°

Full nutrient plants

Moderately Cu-deficient plants

(iii) Plants 4 weeks old; preparation stored24 hr. at -15°

Full nutrient plantsModerately Fe-deficient plantsSlightly Mn-deficient plantsModerately Fe- and Mn-deficient plants

Protein(mg.fg.

fresh wt.)

16-15-66-06-18-4

9*0

7-3

10-490

11-710-4

log ([BVH]/[BV])

i-66T-661-6666

1-66

1 88_-641.881.64

-881-881.881-88

Nitritereductase

(m,umoles/min./mg. of protein)

2-90-42-73.93-5

5.77-0

21-014-714-813-8

Hydroxylaminereductase

(m,[moles/min./mg. of protein)

1-336-526-312-72-0

3.938-7

8-515-28-7

18-4

C. F. CRESSWELL AND OTHERS 196550

NITRATE REDUCTION BY CUCURBITA PEPO ENZYMESsystems in the present work, but no evidence ofmarked inhibition by hydrazine present at 4 mMwas obtained with either. Sodium pyruvate atconcentrations up to 3*3 m made no difference tothe activity ofnitrite reduction, and hydroxylaminereduction was inhibited to amaum of 20%.The inhibition by phenylmercuric acetate and

p-chloromercuribenzoate suggested the presence offunctional SH groups. Phenylmercuric acetate wasthe more effective of the two. Inhibition of nitritereductase was fully reversed by 1 m -GSH, butthat ofhydroxylaminereductase was less completelyreversed by 2-5 mM-GSH even though inhibitionwas less severe with phenylmercuric acetate thanfor nitrite reductase.

Effects of micronutrient elements on enzyme activity

The addition of iron, manganese, copper, zinc ormolybdenum at concentrations between 0-01 and0-1 mM to extensively dialysed preparations ob-tained after precipitation by ammonium sulphatehad no effect on the activity of either nitrite orhydroxylamine reduction when BVH was used.The requirement for any dissociable metal cofactorin either system therefore seems unlikely.The results of separate experiments obtained

when plants were grown with full nutrient or metal-deficient treatments under comparable conditionsare shown for activities on a protein basis in Table 7.

Iron deficiency, when severe, caused a markeddecrease in nitrite-reductase activity. Less acutedeficiency ofiron or ofiron and manganese togetheralso caused appreciable decreases in specific activity.None of the other deficiencies, manganese, copperor zinc, alone caused marked decreases in speci-fic activity of the nitrite-reductase system.The effects of metal deficiencies on the hydroxyl-

amine-reductase system were entirely different.With few exeptions, specific activities of prepara-tions from metal-deficient tissues were substantiallyincreased compared with those fromplants receivingfull nutrient.

DISCUSSION

The relationships between the extent ofreductionof BV and nitrite-reductase activity which wasassociated with practically stoicheiometric produc-tion of ammonia may be most simply explained byassuming that the overall rate of the reaction:

HN02+6H++6e- NH3+2H20 (1)is directly proportional to the redox potential ofthe benzyl viologen system over a certain range.The standard electrode potential for the BV/BVHcouple, which involves a single-electron reaction,has been given as Eo = - 0-359 v (Michaelis & Hill,

1933a,b; Homer, Mees & Tomlinson, 1959; Homer&Tomlinson, 1959) and is independent of pH. Lessnegativevalues between - 0 31 and - 0 32v quotedby the manufacturers are regarded as incorrect andthe result of failure to exclude oxygen entirelyduring measurement. When dye reduction wasless than about 9%, for which the expressionE = BZo-0-061og ([BVH]/[BV]) has a value of- 0 30 v, nitrite-reductase activity was negligible,whereas for values exceeding about - 0 395 v (78%reduction) chemical reduction ofnitrite was inferredand ammonia production was no longer stoicheio-metrically related to nitrite loss (Hageman et al.1962). Senez & Pichinoty (1958c) reported thatnitrite reacted non-enzymically with BVH thathad been produced by extensive reduction by ahydrogenase system, and was therefore probablyreduced to a more negative potential than that forwhich only an enzymic reaction has been observedby us. Hydroxylamine also, according to Senez &Pichinoty (1958b), was reduced chemically toammonia by BVH that had been produced bydithionite or a hydrogenase system.The conditions for the enzymic reduction of

nitrate to nitrite contrasted remarkably with thosefor nitrite and hydroxylamine in that activitydecreased as log([BVH]/[BV]) increased and wasnegligible for values greater than 1*5 for whichthe electrode potential would be about - 0 33 v.Whereas nitrate-reductase systems that containmolybdenum accept electrons freely either fromtwo-electron donors, e.g. the NADPH-FADsequence, or from a single-electron donor (BVH),the enzymic reduction of nitrite in C. pepo requiresthe mediation in an obligatory manner of a single-electron donor. This can be provided by BVHor alternatively, as stated in the introduction, byferredoxin, which is also a single-electron donor(Tagawa & Arnon, 1962). If nitric oxide (Fewson& Nicholas, 1961) and/or a nitroxyl radical areintermediates in nitrite reduction to ammonia atleast two single-electron steps would then be in-volved and a single-electron mediator might wellbe necessary in accordance with the Michaelisprinciple (cf. Mansfield-Clark, 1960, pp. 340-342).The enzymic production of ammonia from nitrite

in nearly stoicheiometric quantities regardless ofa fourfold difference in the rate of the reaction(Table 2) suggests either that the initial reductionof nitrite is the rate-limiting step in the overallreaction or that an intermediate product is severelyinhibitory to nitrite reduction, and that the furtherreduction of this intermediate is the rate-limitingstep. The results of experiments in which nitriteand hydroxylamine were present together excludehydroxylamine as such an inhibitory intermediate.Hydrazine may similarly be excluded from thiscategory with respect to either nitrite-reductase or

Vol. 94 51

C. F. CRESSWELL AND OTHERS

hydroxylamine-reductase systems in Cucurbita.Hyponitrite may, however, require consideration inthe present context. Nicholas (1959) reported thatthis compound, but not hydroxylamine, severelyinhibited the nitrite-reductase system of Neuro-8pora. We have obtained evidence for the formationin some experiments of a compound, which yieldsnitrite after treatment with the alkaline iodinereagent as reported for hyponitrite by Yamafuji &Osajima (1961), that occurs in amounts between 5and 10% of the nitrite reduced (E. J. Hewitt &D. P. Hucklesby, unpublished work).The status of hydroxylamine as a possible inter-

mediate in the reduction of nitrite to ammoniarequires consideration. If free hydroxylamine isinvolved its rate of reduction when present inconcentrations that could be produced from nitriteshould not be less than the rate of the overallreaction in which ammonia yields attained 90% ofnitrite loss. Our results show that this requirementis not necessarily fulfilled when BVH is the electrondonor in an enzymic system since nitrite loss andammonia production can occur more rapidly thanthe corresponding changes for hydroxylamine pre-sent at equivalent concentrations when the sameenzyme preparation is used on the same occasion.The apparent Km for hydroxylamine is substantiallygreater than that for nitrite, so the nitrite-reductasesystem is saturated at a concentration that, ifconverted instantaneously into hydroxylamine,would be barely half-saturating for this substrate.The presence of nitrite severely inhibits hydroxyl-amine reduction, probably in a reversible manner,since hydroxylamine loss was much greater whenpractically all the nitrite present initially had dis-appeared during the assay period (Tables 1 and 2).Senez & Pichinoty (1958a) observed a similar re-versible inhibition of hydroxylamine reduction byD. de8ulfurican8.

Lazzarini & Atkinson (1961) could not detect theproduction ofhydroxylamine by an isotope-dilutiontechnique when unlabelled hydroxylamine wasadded to, and later recovered from, the system inE8cherichia coli during the reduction of 15N-labellednitrite. Further, Mager (1960), Lazzarini & Atkin-son (1961) and Kemp, Atkinson, Ehret & Lazzarini(1963) have collectively concluded on good evidencethat in E. coli the reduction of nitrite or hydroxyl-amine to ammonia and of sulphite to hydrogensulphide are manifestations of a single enzyme sys-tem in which transfer of up to six electrons occurswithout the production of any free intermediates.In C(ucurbita the independent effects of micro-nutrient deficiencies, the differences in apparent Kmvalues for BVH and the differential effects of per-centage reduction of BV suggest, but (as indicatedbelow for the second and third points) may not infact require, that separate enzymes are concerned

in the reduction of nitrite and hydroxylamine toammonia in plants.

In the mechanism proposed by Kemp et al. (1963)nitrite combines via the ionized oxygen atom with ahydroxyl group in the enzyme, an OH- ion beingreleased. The intermediate complexes produced ateach stage ofreduction remain bound via an oxygenatom until ammonia is released. Hydroxylamineforms hydroxamic acids with certain configurationscontaining hydroxyl groups, e.g. hemiacetal, andwe suggest that it might combine thus with thesame prosthetic group as nitrite but initially itwould be combined via the nitrogen instead of anoxygen atom. A rearrangement would then berequired to produce the same structure as an inter-mediate formed from nitrite reduction, before reduc-tion to ammonia. This scheme would account fordirect inhibition of hydroxylamine reduction bynitrite and for increased inhibition when nitrite isundergoing reduction (Table 5), because still lessfree enzyme would be available. Hydroxamateformation and rearrangement might account for anapparently large Km for hydroxylamine and theirregular kinetic behaviour. The scheme alsoaccommodates the differential or independenteffects of percentage dye reduction in nitrite-reductase and hydroxylamine-reductase activitiesif it is assumed that the first step in nitrite reductionis rate-limiting for this activity, whereas the rate ofrearrangement of the hydroxamic acid is the rate-limiting step in the hydroxylamine reduction. It isconceivable that different configurations in thesuccessive intermediate complexes correspondingto successive stages ofreduction would be associatedwith different affinities of the enzyme for BVH,and in this way different apparent Km values forBVH might occur for the two activities. The ob-served relative magnitude of the apparent Kmvalues would be compatible with that for nitritebeing associated with the first step and that for hy-droxylamine with the last. The different effects ofmicronutrient deficiencies on the two activities anddifferent ratios of these activities for comparablevalues of log ([BVH]/[BV]) (Table 4) are not sosimply explained by the scheme presented above.Mager (1960) suggested that a carbonyl group was

present at the active site. A tautomerism betweenan enolic and carbonyl configuration would becompatible with most ofour observations and wouldadditionally accommodate the effects of cyanide,which could compete with hydroxylamine if oximeformation were involved and could also shift theequilibrium from the enolic form reactive withnitrite to the inactive cyanhydrin with which thenitrite could not compete. The relative sensitivitiesof the two systems to cyanide would be compatiblewith this suggestion.

Regardless of the mechanism which can be pro-

52 1965

Vol. 94 NITRATE REDUCTION BY CUCURBITA PEPO ENZYMES 53

posed, our results appear contrary to the classicalidea that free hydroxylamine occurs as an inter-mediate in nitrite reduction by higher plants. Thisview is supported by the results of physiologicalexperiments with algae described by Syrett (1954)and Kessler (1957a,b). The requirement for asingle-electron mediator and the doubtful evidencefor the participation of metals suggest that themechanism differs in important respects from thatinferred by Nicholas (1957, 1958, 1959) to occur infungi, where a number of metal-dependentNAD(P)H-flavine enzymes have been described.

The tenure by R. H. H. of a year's sabbatical leave fromthe University of Illinois and the award of the BootsScholarship to C. F. C. that enabled them to participate inthis work are gratefully acknowledged. We are also gratefulto Mr E. F. Watson and Mr B. A. White for their help ingrowing the plants used in this work, to Dr A. Couper of theDepartment of Biological Chemistry for the helpful dis-cussions in connexion with the redox systems involved andto the Sigma Chemical Co. for generous gifts ofNADP+ andNADPH.

REFERENCES

Avron, M. & Jagendorf, A. T. (1956). Arch. Biochemn. Bio-phy8. 65, 475.

Candela, M., Fisher, E. G. & Hewitt, E. J. (1957). PlantPhy8iol. 32, 280.

Colowick, S. P., Kaplan, N. 0. & Ciotti, M. M. (1951). J.biol. Chem. 191, 447.

Conway, E. J. (1957). Microdiffusion Analyi8 and Volu-metric Error, 4th ed., pp. 90-1 10. London: Crosby, Lock-wood and Co. Ltd.

Cresswell, C. F., Hagemati, R. H. & Hewitt, E. J. (1962).Biochem. J. 83, 38P.

Csaky, T. Z. (1949). Acta chem. 8cand. 2,450.Dixon, M. & Webb, E. C. (1958). Enzymes, pp. 19-22.London: Longmans, Green and Co. Ltd.

Eadie, G. S. (1942). J. biol. Chem. 146, 85.Evans, H. J. & Nason, A. (1953). Plant Physiol. 28, 233.Fewson, C. A. & Nicholas, D. J. D. (1961). Nature, Lond.,

188, 794.Frear, D. S. & Burrell, R. C. (1955). Analyt. Chem. 27,1664.Hageman, R. H., Cresswell, C. F. & Hewitt, E. J. (1962).

Nature, Lond., 193, 247.Hewitt, E. J. (1952). Tech. Commun. Bur. Hort., E. Malling,

no. 22.Hewitt, E. J. & Betts, G. F. (1963). Biochem. J. 89, 20P.Hewitt, E. J. & Hallas, D. G. (1959). Nature, Lond., 184,

1485.Hewitt, E. J. & Nicholas, D. J. D. (1964). In Modern

Methods of Plant Analys8i, vol. 7, pp. 67-172. Ed. byLinskens, H. F., Sanwal, B. D. & Tracey, M. V. Heidel-berg: Springer-Verlag.

Homer, R. F., Mees, G. C. & Tomlinson, T. E. (1959). J. Sci.Fd Agric. 11, 309.

Homer, R. F. & Tomlinson, T. E. (1959). Nature, Lond., 184,2012.

Kemp, J. D., Atkinson, D. E., Ehret, A. & Lazzarini, R. A.(1963). J. biol. Chem. 238, 3466.

Kessler, E. (1957a). Arch. Mikrobiot. 27, 166.Kessler, E. (1957b). Planta, 49, 505.Lazzarini, R. A. & Atkinson, D. E. (1961). J. biol. Chem.

236, 3330.Lineweaver, H. & Burk, D. (1934). J. Amer. chem. Soc. 56,

658.Losada, M., Paneque, A., Ramirez, J. M. & Del Campo, F. F.

(1963). Biochem. biophy8. Re8. Commun. 10, 298.Lowry, 0. H., Rosebrough, N. J., Farr, A. L. & Randall,

R. J. (1951). J. biol. Chem. 193, 265.McKee, H. S. (1962). Nitrogen Metabolism in Plants.

Oxford: The Clarendon Press.Mager, J. (1960). Biochim. biophys. Acta, 41, 553.Mansfield-Clark, W. (1960). Oxidation Reduction Potentials

of Organic Systems. Baltimore: Williams and Wilkins Co.Michaelis, L. & Hill, E. S. (1933a). J. gen.Physiol. 16,859.Michaelis, L. & Hill, E. S. (1933b). J. Amer. chem. Soc. 55,

1491.Mortenson, L. E., Valentine, R. C. & Carnahan, J. E. (1962).

Biochem. biophys. Res. Commun. 7, 448.Nason, A. (1962). Bact. Rev. 26, 16.Nason, A. & Takahashi, H. (1958). Annu. Rev. Microbiol.

12, 203.Nicholas, D. J. D. (1957). Ann. Bot., Lond., N.S., 21, 587.Nicholas, D. J. D. (1958). Colloq. 4th int. Congr. Biochem.,

Vienna, vol. 13, p. 307. London: Pergamon Press Ltd.Nicholas, D. J. D. (1959). Symp. Soc. exp. Biol. 13, 1.Paneque, A., Del Campo, F. F. & Losada, H. (1963). Nature,

Lond., 198, 90.Russell, J. A. (1945). J. biol. chem. 156, 457.Senez, J. C. & Pichinoty, F. (1958a). Biochim. biophys. Acta,

27, 569.Senez, J. C. & Pichinoty, F. (1958b). Biochim. biophys. Acta,

28, 355.Senez, J. C. & Pichinoty, F. (1958c). Bull. Soc. Chim. biol.,

Paris, 48, 2099.Syrett, P. J. (1954). Symp. Soc. exp. Biol. 4, 126.Tagawa, K. & Armon, D. I. (1962). Nature, Lond., 195,537.Valentine, R. C., Brill, W. J., Wolfe, R. S. & San Pietro, A.

(1962). Biochem. biophys. Res. Commun. 10, 73.Valentine, R. C., Jackson, R. L. & Wolfe, R. S. (1962).

Biochem. biophys. Res. Commun. 7, 453.Villanueva, J. R. (1959). Nature, Lond., 184, 549.Walker, G. C. & Nicholas, D. J. D. (1961). Biochim. biophys.

Acta, 49, 361.Webster, G. C. (1959). Nitrogen Metabolism in Plants.

Evanston, Ill., and New York: Row, Peterson and Co.Whiteley, H. R. & Woolfolk, C. A. (1962). Biochem. biophys.

Res. Commun. 9, 517.Yamafuji, K. & Akita, T. (1952). Enzymologia, 15, 313.Yamafuji, K. & Osajima, Y. (1961). Nature, Lond., 190,534.