interactions among substrates inhibitors nitrogenase · lene, azide, cyanide, methyl isocyanide,...

JouRNAL OF BACTERIOLOGY, Aug. 1975, p. 537-545Copyright 0 1975 American Society for Microbiology

Vol. 123, No. 2Printed in U.S.A.

Interactions Among Substrates and Inhibitors of NitrogenaseJOSE M. .RIVERA-ORTIZ AND R. H. BURRIS*

Department of Biochemistry, College of Agricultural and Life Sciences, University of Wisconsin, Madison,Wisconsin 53706

Received for publication 1 April 1975

Examination of interactions among various substrates and inhibitors reactingwith a partially purified nitrogenase from Azotobacter vinelandii has shown that:nitrous oxide is competitive with N2; carbon monoxide and acetylene arenoncompetitive with N2; carbon monoxide, cyanide, and nitrous oxide arenoncompetitive with acetylene, whereas N2 is competitive with acetylene; carbonmonoxide is noncompetitive with cyanide, whereas azide is competitive withcyanide; acetylene and nitrous oxide increase the rate of reduction of cyanide.The results are understandable if nitrogenase serves as an electron sink andsubstrates and inhibitors bind at multiple modified sites on reduced nitrogenase.It is suggested that substrates such as acetylene may be reduced by a lesscompletely reduced electron sink than is required for the six-electron transfernecessary to reduce N2

Nitrogenase is a versatile enzyme which iscapable of reducing N2, nitrous oxide, acety-lene, azide, cyanide, methyl isocyanide, pro-tons, and analogs of some of these compounds.Nitrogenase consists of two proteins, a molyde-num-iron protein (MoFe protein) and an ironprotein (Fe protein); neither has been demon-strated to have catalytic activity by itself.Under physiological conditions the Fe protein isreduced by ferredoxin or a flavoprotein. The Feprotein specifically binds magnesium adenosine5'-triphosphate (MgATP) (3, 29); upon bindingthe MgATP the potential of the Fe proteinbecomes substantially more negative (7, 27, 31),and it acquires the unique ability to reduce theMoFe protein. The reduced MoFe protein servesas an electron sink capable of reducing all of thesubstrates of nitrogenase; all substrates com-pete for electrons from the reduced MoFe pro-tein. Although reduction of substrates alwayshas required both proteins in experimentalsystems, it is quite possible that reduced MoFeprotein if separated from the Fe protein wouldbe capable of reducing substrates by itself. Thepresent paper extends earlier work (2, 20, 23)and explores the interactions among substratesand inhibitors of nitrogenase in an attempt tolearn more about the substrate and inhibitorbinding sites and the mode of action of nitrogen-ase. The information discussed has been re-corded in more detail by Rivera-Ortiz (M. S.thesis, Univ. of Wisconsin, Madison, 1973).

MATERIALS AND METHODSCylinder gases (Matheson Gas Products) were

passed over BASF catalyst R3-11 (Chemical Dynam-

ics Corp., Hadley Industrial Plaza, South Plainfield,N. J.) at about 120 C to remove oxygen. Nitric oxide(NO) of about 98.5% purity was used without treat-ment. Mixtures were prepared by adding gases to anevacuated vessel and measuring their pressures with amercury manometer. CO and NO were used in verylow concentrations and were diluted serially withargon. Acetylene (C2H2) was generated with calciumcarbide and water. Sodium dithionite (Na2S2O4)solutions were prepared anaerobically immediatelybefore use.

Azotobacter vinelandii strain OP was grown onmodified Burk's medium (24) in 1,500-liter quantitieswith vigorous stirring (100 rpm) and aeration (1,100liters of air/min). Cells recovered (Sharples centri-fuge) in the exponential growth phase were frozen as apaste with liquid N2 and then were stored at -20 C.Thawed and resuspended cells were ruptured with aFrench press and debris was removed by centrifuga-tion at 82,000 x g for 1 h (4). The dark brown super-natant, termed crude extract, contained 50 to 60mg of protein per ml. The crude extract was partiallypurified by the procedure of Hwang and Burris (19).All operations were performed anaerobically. Thecrude extract was dialyzed against N2-sparged water(two changes) for 3 h. The dialyzed extract was heatedto 60 C for 10 min, centrifuged at 37,000 x g for 20min, and then the precipitate was discarded. Thesupernatant was centrifuged at 144,000 x g for 90to 120 min. The nitrogenase, which sedimented in adark brown pellet, was resuspended in tris(hydroxy-methyl)aminomethane-hydrochloride buffer (20 mM,pH 7.3) and was centrifuged at 37,000 x g for 20 min(discard sediment). The supernatant, which contained15 to 30 mg of protein per ml, was stored in liquidN2 until used. Such a partially purified extractreduced 80 to 150 nmol of C2H2/(milligram of proteinx minute); it is a convenient preparation for studyingN, fixation (4), because it is relatively easy toprepare, is reasonably active, is more stable than the

537

on April 3, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/

538 RIVERA-ORTIZ AND BURRIS

separated component proteins, removes the necessityfor recombination of nitrogenase components in theproper proportions and catalyzes all the reactionsreported for nitrogenase.

The mixture for measurement of nitrogenase activ-ity contained in 1 ml: 40 Mmol of creatine phosphate,0.1 to 0.25 mg of creatine kinase (2.7.3.2; SigmaChemical Co., 50 to 100 U/mg of protein), 5 gmolof ATP, 5 qmol of MgCl2, 50 umol of tris(hydroxy-methyl)aminomethane at pH 7.3, 15 Amol of Na2S2O4,enzyme preparation with 0.5 to 2.0 mg of protein(protein determined by the method of Gornall et al.[14]). The components (minus Na2S2O4 and theenzyme preparation) were mixed in a 21-ml serumbottle closed with a rubber serum stopper; the bottlewas evacuated and refilled four times with argonthrough a hypodermic needle. Substrate (often gase-ous) was added, and the pressure was brought to 1 atmwith argon. The bottle was shaken for 5 min in a 30 Cwater bath before the Na2S2O4 and enzyme prepara-tion in anaerobic solution were injected to start thereaction. After 15 min, the reaction was terminated byinjecting 1 ml of 1.25 M trichloroacetic acid (whenfollowing reduction of C,2H12 or CN -) or 1 ml ofsaturated K2CO3 (when measuring N2 fixation).

Fixation of N2 was evaluated by measuring NH3formed; after microdiffusion of NH, (6), it was deter-mined with the indophenol method of Chaykin (10).C2H4 and CH4 produced (13, 15, 17, 26) from C2H2and CN -, respectively, were measured by gas chroma-tography with a flame ionization detector (150 cmlong, 2 mm ID column of Porapak R, 50 C; N, carriergas; 0.5-ml gas samples). H2 evolution was deter-mined in Gilson constant pressure all glass volumom-eters or by the appearance of mass 2 as analyzed in anisotope ratio mass spectrometer.Data used to calculate Michaelis constants were

treated by the computer program described by Cle-land (11, 12). Inhibition patterns were obtained ini-tially with double reciprocal plots (1/velocity versus1/substrate concentration); these patterns defined theappropriate rate equations conforming to fortranprograms of Cleland (11, 12). Analysis by theseprograms yielded values for inhibition constants (Ks,K,,, relative to intercept; KAS, relative to slope) max-imum velocities (V), Michaelis constants (Kin), andstandard error of the estimates. The computer plotsincluded all data and assumed equal variances for allvelocities; lower velocities carry a greater intrinsicerror. Data showing competitive inhibition were fittedto the equation: v = VA/[K(1 + i/K/) + A] in whichA = substrate concentration, I = inhibitor concentra-tion, and v = velocity. Data showing noncompetitiveinhibition were fitted to the equation: v = VA/[K(1 + I/K,,) + A(1 + IIK,,) ].

RESULTS AND DISCUSSION

Measurements of the time course of NH3formation at 0.05 and 0.40 atm N2 indicated noinitial lag and linear production of NH3 withtime for 25 min. Likewise, production of C2H4from C2H2 at 0.005 and 0.125 atm showed no lag

and was linear for 25 min. Production of CH4from 1.04 and 2.96 mM CN- was linear forabout 20 min and showed no lag. H2 evolutionwas linear without lag for 35 min. A 15-minreaction period was adopted for all experimentsto avoid any nonlinear responses with time.Apparent Michaelis constants. Measure-

ments of the Km for N2 gave a value of 0.136 +0.003 atm (7.75 x 10-5 M); this is in goodagreement with other reported values (18, 19,28). The Km for C2H2 was 0.012 ± 0.002 atm(4.33 x 10-4 M); the value did not differ greatlyfrom most others in the literature (13, 15, 19, 21,26), but was quite different from the value of0.001 atm reported by Hardy and Knight (18).Hardy et al. (16) later revised their estimatesupward to 0.002 to 0.009 atm for A. vinelandii.Determining the Km for CN- poses special

problems, because above 4 mM concentrationCN- inhibits its own reduction markedly (seeFig. 5). By assuming that self-inhibition byCN- is negligible at low concentrations and byusing data up to 4 mM CN-, one arrives at anapparent Km of 1.23 + 0.22 mM for CN-.Hwang and Burris (19) reported 1.28 mM, butBiggins and Kelly (1) reported 0.19 mM fornitrogenase of Klebsiella pneumoniae. Hardyand Knight (18) found that CN- inhibited ATPhydrolysis and they pointed out that theirapparent Km of 4 mM for an extract from A.vinelandii, "is susceptible to error because ofsubstrate inhibition which occurs above 5 mM."Later Hardy et al. (16) revised the Km down-ward to 0.4 to 1.0 mM.Evolution of H2. Reduced nitrogenase plus

MgATP reduces H+ to H2; the reaction istermed ATP-dependent H2 evolution. Becausethis reaction draws electrons from the commonelectron sink of nitrogenase, evolution of His inhibited by the presence of other nitro-genase substrates competing for electrons fromthe sink. As indicated by Hardy et al. (16),C 2H2 is particularly effective in suppressingevolution ofH 2.

Figure 1 indicates that H2 evolution continuesat about 35°k of its maximal rate even in thepresence of 1 atm of N,. If the distance from lineA to the curve is plotted versus pN2, a curve isobtained corresponding to the mirror image ofthe inverted curve shown (appearance is like asubstrate concentration versus rate curve).Analysis of this derived curve by a 1/v versus1/pN2 plot (insert curve is calculated with theHypero program of Cleland [11] which givesapparent maximum velocity and the standarderror of the estimate) defines the maximumvelocity (y intercept) in relative terms at infi-nite pN . Values for V from three experiments

J. BACTERIOL.


http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/

INHIBITORS OF NITROGENASE 539

were 0.749 + 0.018, 0.808 + 0.025 and 0.760 ±0.016; average 0.77 (23% residual H2 evolution).If one uses measured velocities of H2 evolutionand produces a double reciprocal plot by visualinspection, the velocity of H2 evolution at infi-nite pN2 appears to be 13 to 15% of its velocityin the absence of N2. The results with eitheranalysis indicate that the pN2 cannot be raisedto a point where it will block evolution of H2completely; at infinite pN2 the evolution of H2still continues at a rate 13% to 23% its maximumvalue.

Figure 2 indicates the influence of the partialpressure of C2H2 on evolution of H,2 Treatingthe data as in Fig. 1 yielded the following valuesfor V (Maximum velocity from Hypero com-puter program of Cleland, 11): 0.95 ± 0.01, 1.01

0.07 and 1.04 A 0.02; average 1.00. In contrastto the response to N2, evolution of H2 would besuppressed completely by infinite pC2H2. Evo-lution of H2 responds in a qualitatively differentfashion to N2 and to C2H2. The effect of CN - onevolution of H2 is similar to the effect of C2H2 inthe sense that an infinite concentration ofcyanide would suppress evolution of H2 com-pletely. However, CN - in addition exhibitspowerful substrate inhibition.N2 is unable to block evolution of H2 com-

pletely, whereas C2H2 or CN- at infinite con-centration is able. These observations empha-

PN2 (otrrfFIG. 1. Inhibition of H2 evolution by N2. Reaction

mixtures contained in 1 ml: 40 Aimol of creatinephosphate, 0.25 mg of creatine kinase, 50 MAmol ofTris-hydrochloride (pH 7.3), 5 ;imol of MgCl2, 5 1smolof ATP, 15 Mmol of NaS,O4, 0.10 ml of enzymepreparation. Temperature, 30 C. N2 was varied asindicated. The insert is a derived plot from the data,and v as plotted represents 100 minus the observedpercentage ofH2 evolved (see text).

pC2l2 (atm)FIG. 2. Inhibition of H2 evolution by C2H2. The

experiment was run as described in Fig. 1 except thatC2H2 was used instead of N2.

size that there are differences in the reductionof N2 and its alternative substrates, henceobservations on reduction of C 2H2 or CN -should be related to N2 reduction with caution.

Inhibition of reduction of N2. Carbonmonoxide. Nitrogenase was incubated for 15min in the presence of 0.01 or 0.02 atm CO.When the CO was pumped off, the activity wasrestored completely; this confirms that theinhibitory effect of CO at these levels is com-pletely reversible. In contrast, inhibition byNO was only partially reversible, and itsinhibitory action was not studied further.CO generally is accepted as a noncompetitive

inhibitor of N2 fixation, but there have beenambiguities in some experiments (2, 8, 23). Thedata obtained with the nitrogenase preparationsfrom A. vinelandii showed unequivocal noncom-petitive inhibition of N2 reduction by CO. TheK&8 = 0.0004 + 0.00008 atm and the Kii = 0.0015± 0.0002 atm CO.Acetylene. Reports have differed on the na-

ture of C2H2 inhibition of N2 reduction (20, 26).Figure 3 shows clearly that the inhibition isnoncompetitive in agreement with Hwang et al.(20). The Ki, = 0.0052 ± 0.0019 atm and the Kii= 0.0064 4 0.0021 atm C2H2.Nitrous oxide. There have been few studies

of N 20 as an inhibitor of N2 fixation, but it hasbeen concluded that N20 is competitive with N2(25). The present studies yielded the datasummarized in Fig. 4; the inhibition is competi-tive in these preparations. The inhibitor con-stant was Ki = 0.108 + 0.015 atm N20. Al-though it is evident that N2 and N 20 bind to thesame enzyme site, it should be pointed out that

VOL. 123, 1975


http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/


X o in preliminary experiments there was a devia-0 tion from linearity for points corresponding to a8 _ _ low pN2 in the presence of a higher pN2O-

Nevertheless, analysis confirmed the competi-

v / tive nature of the inhibition.Inhibition of reduction of cyanide. The use6_ of cyanide as a substrate poses problems, be-

cause it is inhibitory to its own reduction (Fig.5), it is volatile under acid conditions, and therelative effectiveness of HCN and CN- as

4 ~ /)- _5'substrate species is not established. LJones (22)has shown that CN - inhibits electron transfer inthe nitrogenase system. All our experiments

2> < _ were performed with concentrations of CN - lessthan 4 mM to minimize any inhibition by theCN-. Rates of reduction of CN- by nitrogenaseare lower than its rates of reduction of most

0 other substrates (16. 30).5 10 15 20 Carbon monoxide. CO inhibits all reductions

I by nitrogenase except the reduction of protons(itm-H2 to H2. CO proved to be a noncompetitiveinhibitor of CN - reduction (Fig. 6). A somewhat

FIG. 3. Inhibition of N2 reduction by C2H2. The higher level of CO is required to give 507reaction mixture was described in Fig. 1. All lines inhibition of CN- than of N2 reduction. Thewere derived by computer analysis. Velocity v =micromol of NH4+ produced in 15 min. The pC2H2 inatm was as follows: (V) 0.0, (0) 0.002, (A) 0.004, (0) l lX0.006. 20

0 v~~~~2

V 10

340.0.3-

023

0.2I

°2 4 6 0 2 4 6 8 10NaCN (mM)

pN2 atm) FIG. 5. Effect of cyanide on methane formationfrom cyanide. The reaction mixture was described in

FIG. 4. Inhibition of N2 reduction by N2O. The Fig. 1. The reaction was run at 30 C for 15 min underreaction mixture was described in Fig. 1. All lines argon. Velocity v = micromoles of CH4 produced inwere derived by computer analysis. The pN2O in atm 15 min. The insert is a Lineweaver-Burk plot of thewas as follows: (V) 0.0, (0) 0.02, (A) 0.04, (0) 0.08. first six points of the main curve.

J. BACTERIOL.


http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/


I 2

CN'(mM)FIG. 6. Inhibition of cyanide reduction by carbon

monoxide. The reaction mixture was described in Fig.1. Reactions were run under argon at 30 C, and thecyanide concentration was varied as indicated. Veloc-ity, v = micromoles CH4 produced in 15 min. ThepCO in atm was as follows: (V) 0.0, (0) 0.001, (A)0.003, (0) 0.005.

computer analysis indicated that K = 0.0026+ 0.0008 atm and Kis = 0.0028 0.0010 atmCO.

Azide. Azide reduction is competitively in-hibited by CN- (20). As shown in Fig. 7, azidealso is a competitive inhibitor of CN - reduction.Apparently azide, CN- and methyl isocyanideall bind at the same site, as they are mutuallycompetitive. The K, for azide as an inhibitor ofCN- reduction was found to be 2.38 + 0.28 mM.The Km for azide has been reported as 1.15 mM(19) and 1.0 mM (16).Acetylene. Biggins and Kelly (1) reported

that C2H2 enhanced formation of CH4 fromCN- by nitrogenase. We also observed thatC2H2 enhanced CH4 formation from CN-; theconcentration of C2H2 giving greatest enhance-ment varied with the level of CN - present. CH22at a level of 0.2 atm enhanced production ofCH4 from 0.5 mM CN- by about 40%. When0.05 atm C2H2 was present, CN- inhibition ofCN- reduction to CH, (up to 6 mM CN-) wassuppressed as shown in Fig. 8 (compare withFig. 5). Apparently CN- inhibits its own reduc-

tion by binding at a second site (this couldinvolve nonspecific inhibition of electron trans-fer) distinct from the site of its reduction; C2H2may relieve CN- inhibition by binding or pro-tecting this second site from CN- binding. Asthe time course of CN- reduction is linear, thebinding must occur immediately. CN- reactswith a number of iron-sulfur proteins, andBiggins and Kelly (1) reported nonspecific bind-ing of CN - to both the MoFe protein and Feprotein of nitrogenase. Bui and Mortenson (3)

±V

2

CN- (mM)FIG. 7. Inhibition of cyanide reduction by azide.

Reactions were run as described in Figure 6 exceptthat azide was the inhibitor. The millimolar azideconcentrations were: (V) 0.0, (0) 0.98, (A) 1.94, (0)3.83.

0.8

0.6 -

1 2 3 4 5 6NaCN (mM)

FIG. 8. Effect of acetylene in relieving cyanideinhibition of cyanide reduction. Conditions were asdescribed in Fig. 5 except that 0.05 atm of acetylenewas present.

VOL. 123, 1975

I on April 3, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/


suggested that CN - was bound specificallyby the MoFe protein. Hardy and Knight (18)found that when the reduction of CN - isblocked by CO, CN - still inhibits evolution ofH2 (CO does not inhibit evolution of H2)-Nitrous oxide. N20 like C2H2 enhances the

rate of CN- reduction (Fig. 9). Enhancementwas seen when 0.05. 0.1 and 0.2 atm N2O wasadded to CN- at concentrations of 0.30, 0.50,0.99, and 1.28 mM. The enhancement increasedwith increasing pN2O up to 1.0 atm, the highestconcentration of N2O used (Fig. 9), and thecurve still was rising at this point. It is interest-ing that C2H2 and N2O have similar effects onreduction of CN-, because N2O is a noncom-petitive inhibitor of the C2H2 reduction, andhence they do not appear to bind at the samesite for their reduction.

Dinitrogen. N2 had little effect on reductionof CN-. When the concentration of' CN- was1.96 mM, 1.0 atm of N2 decreased CN- reduc-tion to CH4 by 14%. When concentrations of'CN- as low as 0.1 mM were used, the highestinhibition by N2 was 22%7(. Under these condi-tions, the pattern of N2 inhibition could not bedetermined.

It would be interesting to determine theproduction of H2 simultaneously with produc-tion of CH, to obtain a more complete picture ofreductive processes when CN- is furnished assubstrate and alternative substrates also arepresent. It is evident that CN- must be used asa nitrogenase substrate with caution because ofits self-inhibitory action and the enhancementof its reduction by C2H2 or N2O.

Inhibition of reduction of acetylene. Reduc-tion of C 2H2 has acquired importance because ofthe extensive use of C2H2 reduction as an indexof N2 fixation. The assumption often is made

1.0 _

.ECL 0.8 0

0.,

E0.4_

v~' Q2 0.4 0.6 0.8 1.0pN20 (atm)

FIG. 9. Effect of N20 in relieving cyanide inhibi-tion of cyanide reduction. Conditions were as de-scribed in Fig. 5 except that N20 was added asindicated. The concentration ofNaCN was 0.99 mM.

that N2 and C2H2 are strictly equivalent assubstrates for nitrogenase. However, there aredemonstrable differences between them, otherthan the requirement of six electrons f'or reduc-tion of N2 to 2 NH3 and 2 electrons for reductionof C2H2 to C2H4. Hwang et al. (20) reported thatazide and CO are noncompetitive inhibitors of'C2H2 reduction, and that H2 does not inhibitthe reduction.Carbon monoxide. When 0.1 atm of' C2H2

was present, 0.005 atm CO gave about 50%/(inhibition. The inhibition was noncompetitiveas reported by Hwang et al. (20).Nitrous oxide. C2H2 was f'urnished as sub-

strate at 0.1 atm, and its reduction was inhib-ited about 50% by 0.3 atm N2O. The inhibitionis noncompetitive (Fig. 10). The observationsthat H2 competitively inhibits N2 fixation buthas no effect on reduction of C2H2 and that N2Ois a competitive inhibitor of N2 fixation but anoncompetitive inhibitor of reduction of C2H2give f'urther support to the suggestion that thebinding sites for N2 and C2H2 are not equiva-lent. The f'act that C2H2 is a noncompetitiveinhibitor of N2 reduction reinforces the conceptof' distinguishable sites.Cyanide. CN- strongly inhibits reduction of

C2H2; in the presence of'0.05 atm C2H2, 0.5 mMCN- effects about 50%S. inhibition. The inhibi-tion pattern (Fig. 11) is noncompetitive. CN-has been reported to be noncompetitive towardN2 and competitive toward azide reduction(20). Whereas CN- is a powerful inhibitor ofC2H2 reduction, C2H2 enhances reduction ofCN -.

Dinitrogen. Reduction of C2H2 is inhibitedby N2 (Fig. 12); when 0.02 atm C2H2 waspresent, 0.6 atm N2 caused about 33% inhibi-tion. With 0.10 atm C2H2, no measurableinhibition was demonstrable with 0.80 atm N2.The inhibition is competitive, and the Ki valuesdetermined in five experiments were: 0.389 40.059, 0.395 4 0.060, 0.572 + 0.118, 0.381 +0.025 and 0.262 , 0.023 atm N2; the averagewas 0.40 atm N2. The competitive nature ofthis inhibition is not reciprocal in the sensethat C2H2 is a noncompetitive inhibitor of N2reduction.Conclusion. Nitrogenase is a versatile en-

zyme system which can reduce a variety ofsubstrates. It is clear from the current ob-servations and the earlier work of others (16,18-20, 23) that the interactions are complexamong these substrates and among inhibitorswhich are not reduced by nitrogenase. Table 1summarizes observations which have been madeto date on substrate and inhibitor interac-tions. Although some blanks remain in the

J. BACTERIOL.


http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/


chart, notably N20 and CH,NC as substratesand NO as inhibitor, a number of conclusionscan be drawn from the available data.

In interpreting the interactions among sub-

50 100

pC H,(atm)FIG. 10. Inhibition of acetylene reduction by ni-

trous oxide. The reaction mixture was described inFig. 1. Velocity v = micromoles of C2H4 produced in15 min. The pN2O in atm was as follows: (V) 0.0, (0)0.1, (A) 0.2, (0) 0.4.

;.V I *vvpC2H2(atm)

FIG. 11. Inhibition of acetylene reduction by cya-nide. The conditions were as described in Fig. 10except that cyanide was the inhibitor. The millimolarcyanide concentrations were: (V) 0.0, (0) 0.498, (A)0.990.

I

pC2H2(atm)FIG. 12. Inhibition of acetylene reduction by N2.

The conditions were as described in Fig. 10 exceptthat N2 was the inhibitor. The pN2 in atm was asfollows: (V) 0.0, (0) 0.2, (A) 0.4, (0) 0.6.

strates and inhibitors, we will assume that theMoFe protein binds substrates and that thereduced MoFe protein of nitrogenase serves asthe electron sink which reduces all nitrogenasesubstrates. Any blockage of electron transportbefore the electrons reach the MoFe proteinwould block all reductions by nitrogenase, al-though a partial inhibition in electron flow tothe MoFe protein could affect reduction ofsubstrates differentially.The fact that the presence of one substrate

decreases the reduction rate of other substratessupports the concept that all are bound to thereduced MoFe protein. However, there is anonequivalence of the binding sites. As pointedout by Hwang et al. (20), the responses of thesubstrates and inhibitors suggest modified sitesto accommodate:

(i) N2, H2, N20 (H2 and N20 competitiveinhibitors of N2);

(ii) CN-, N3-, CH3NC (mutually competitivebut noncompetitive with N 2);

(iii) C2H2 (noncompetitive with the N2 andN3-; C2H2 enhances CN - reduction; C2H2 canblock H2 evolution completely; N2 is competi-tive with C2H2);

(iv) CO (noncompetitive with substrates anddoes not inhibit H+ reduction);

(v) H2 evolution (not identical to N2 site as H2does not inhibit; unique in escaping CO inhibi-tion).New elements which have been added by the

present work are the demonstration that N2

VOL. 123, 1975


http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/


TABLE 1. Nature of inhibition by some electron acceptors and inhibitors of nitrogenasea

InhibitoraSubstrate

N, N3- C,H2 | HCN NO0 H, Co CH,NC

N2 NC NCb C C NCN3- NCb Cb NI NCb CbC 2H2 C NCb NC NC NI NCHCN I C E E E NCHJ+I lb I I Ic NI NIb I

aSymbols: C, competitive inhibition; NC, noncompetitive inhibition; NI, does not inhibit; I, inhibits; E,enhances.

bReference 20.c Reference 5.

competitively inhibits reduction of C2H2 (it hadbeen known that C2H2 is noncompetitive withN2); that CN-, N20 and CO are noncompetitivewith C 2H2; that N20 enhances reduction ofCN-; and that N2- is a competitive and CO anoncompetitive inhibitor of CN- reduction.

Assimilating most of the data into the earlierscheme of Hwang et al. (16) presents no diffi-culty. However, the escape of H2 evolution fromCO, which inhibits all other reductions, andexplaining the nonreciprocal responses of N2and C2H2 pose problems.The escape of H2 evolution, which requires

electrons from reduced MoFe protein, from COinhibition indicates that CO cannot block elec-tron transfer before the electrons reach theMoFe protein; rather, CO blocks electron trans-fer from the MoFe protein to the various sub-strates. The site for reduction of H+ is unique,for steric or other reasons, in escaping the COblock (note that other substrates decrease therate of H+ reduction, because all draw from thesame electron sink).The nonreciprocal response of N2 and C2H2

poses a different problem. The competitiveinhibition of C2H2 reduction by N2 providesevidence for suggesting that C2H2 and N2 bindat the same site, and a model for reduction ofthese substrates can be based on the concept ofa single site. However, the noncompetitive inhi-bition of' N2 reduction by C2H2 indicates thatthese substrates bind at separate sites, and thecompetitive inhibition of C2H2 reduction by N2shows that C2H2 at high concentrations cancompletely overcome the inhibitory effect of N2.We will postulate a model possessing separatesites for N2 and C2H2. A model which predictsnonreciprocal inhibitory behavior between C2H2and N2 calls for the electron sink of the MoFeprotein to contain six electrons in order toreduce N2, but only two electrons to reduceC2H2, and for the flow of electrons from the Feprotein to be rate limiting for reduction of' good

substrates such as C2H2 and N2. C2H2 and N2each has its own binding site, and their mutualinhibition arises from the fact that they aretapping a common sink of electrons. Whereashigh C2H2 will keep this sink depleted, so thatit never will contain more than two electrons(and thus will be incapable of reducing N2),high N2 never can prevent access to the sinkby C 2H2, with the result that C 2H2 is non-competitive versus N2. If it is assumed thatelectrons leak from the sink when it containstwo or more electrons, this model also ex-plains why H2 evolution is completely sup-pressed C 2H2, but not by N 2. By keepingstorage in the MoFe protein less than twoelectrons, a high pC 2H2 prevents H2 evolu-tion, whereas the residual 23% of H2 evolutionthat cannot be suppressed by N2 results f'romleakage from the electron sink when it containsat least two but fewer than six electrons. Chattet al. (9) have presented evidence that reductionof N2 to 2 NH2, occurs with a six-electrontransfer in model compounds containing molyb-denum or tungsten, but they express doubtsthat this occurs in nitrogenase.

This model also suggests an explanation forthe failure to observe intermediates before NH2in N2 reduction. If the electron sink mustcontain six electrons to start the reduction ofN2, it is likely that the first electron transferwill be the slowest step, and subsequent electrontransfers will be much faster; thus only the finalend product of reduction, NH2, will be observed.This absolute requirement for all the electronsfor reduction to be in the sink before the processcan start may not apply to other substrates;alternate sets of reaction products are observedfor some substrates, e.g., CN- may be reducedwith six electrons to CH4 + NH2 or alternativelymay be reduced with 4 electrons to CH3NH2.

ACKNOWLEDGMENTSThis investigation was supported by the College of

Agricultural and Life Sciences, University of Wisconsin at

J. BACTERIOL.


http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/


Madison; by Public Health Service grant AI-00848 from theNational Institute of Allergy and Infectious Diseases, and byNational Science Foundation grant GB-21422.We thank W. W. Cleland for helpful discussions.

LITERATURE CITED

1. Biggins, D. R., and M. Kelly. 1970. Interaction ofnitrogenase from Klebsiella pneumoniae with ATP orcyanide. Biochim. Biophys. Acta 205:288-299.

2. Bradbeer, C., and P. W. Wilson. 1963. Inhibitors ofnitrogen fixation, p. 595-614. In R. M. Hochster and J.H. Quastel (ed.), Metabolic inhibitors, a comprehen-sive treatise, vol. 2. Academic Press Inc., New York.

3. Bui, P. T., and L. E. Mortenson. 1968. Mechanism of theenzymic reduction of N2: the binding of adenosine5'-triphosphate and cyanide to the N2-reducing system.Proc. Natl. Acad. Sci. U.S.A. 61:1021-1027.

4. Bulen, W. A., and J. R. LeComte. 1972. Nitrogenasecomplex and its components, p. 456-470. In A. SanPietro (ed.), Methods in enzymology 24B. AcademicPress Inc., New York.

5. Bulen, W. A., J. R. LeComte, R. C. Burns, and J.Hinkson. 1965. Nitrogen fixation studies with aerobicand photosynthetic bacteria, p. 261-274. In A. SanPietro (ed.), Non-heme iron proteins: role in energyconversion. Antioch Press, Yellow Springs, Ohio.

6. Burris, R. H. 1972. Nitrogen fixation- assay methods andtechniques, p. 415-431. In A. San Pietro (ed.), Methodsin enzymology 24B. Academic Press Inc., New York.

7. Burris, R. H., and W. H. Orme-Johnson. 1974. Survey ofnitrogenase and its EPR properties, p. 187-209. In J. B.Neilands (ed.), Microbial iron metabolism, a compre-hensive treatise. Academic Press Inc., New York.

8. Burris, R. H., and P. W. Wilson. 1946. Characteristicof the nitrogen-fixing enzyme system in Nostocmuscorum. Bot. Gaz. 108:254-262.

9. Chatt, J., A. J. Pearman, and R. L. Richards. 1975. Thereduction of mono-coordinated molecular nitrogen toammonia in a protic environment. Nature (London)253:39-40.

10. Chaykin, S. 1969. Assay of nicotinamide deaminase.Determination of ammonia by the indophenol reaction.Anal. Biochem. 31:375-382.

11. Cleland, W. W. 1963. Computer programmes for process-ing enzyme kinetic data. Nature (London)198:463-465.

12. Cleland, W. W. 1967. The statistical analysis of enzymekinetic data. Adv. Enzymol. 29:1-32.

13. Dilworth, M. J. 1966. Acetylene reduction by nitrogen-fixing preparations from Clostridium pasteurianum.Biochim. Biophys. Acta 127:285-294.

14. Gornall, A. G., C. J. Bardawill, and M. M. David. 1949.Determination of serum proteins by means of the biuretreaction. J. Biol. Chem. 177:751-766.

15. Hardy, R. W. F., R. C. Burns, and R. D. Holsten. 1973.Applications of the acetylene-ethylene assay for mea-surement of nitrogen fixation. Soil Biol. Biochem.5:47-81.

16. Hardy, R. W. F., R. C. Burns, and G. W. Parshall. 1971.The biochemistry of N2 fixation, p. 219-247. In R. F.Gould (ed.), Bioinorganic chemistry. American Chemi-cal Society, Washington, D.C.

17. Hardy, R. W. F., R. D. Holsten, E. K. Jackson, and R. C.Burns. 1968. The acetylene-ethylene assay for N,fixation: laboratory and field evaluation. Plant Physiol.43:1185-1207.

18. Hardy, R. W. F., and E. Knight, Jr. 1967. ATP-depend-ent reduction of azide and HCN by N2-fixing enzymesof Azotobacter vinelandii and Clostridium pas-teurianum. Biochim. Biophys. Acta 139:69-90.

19. Hwang, J. C., and R. H. Burris. 1972. Nitrogenase-cat-alyzed reactions. Biochim. Biophys. Acta 283:339-350.

20. Hwang, J. C., C. H. Chen, and R. H. Burris. 1973.Inhibition of nitrogenase-catalyzed reductions. Bio-chim. Biophys. Acta 292:256-270.

21. Koch, B., and H. J. Evans. 1966. Reduction of acetyleneto ethylene by soybean root nodules. Plant Physiol.41:1748-1750.

22. Ljones, Torbjorn. 1973. Nitrogenase from Clostridiumpasteurianum. Changes in optical absorption spectraduring electron transfer and effects of ATP inhibitorsand alternative substrates. Biochim. Biophys. Acta321:103-113.

23. Lockshin, A., and R. H. Burris. 1965. Inhibitors ofnitrogen fixation in extracts from Clostridium pas-teurianum. Biochim. Biophys. Acta 111:1-10.

24. Newton, J. W., P. W. Wilson, and R. H. Burris. 1953.Direct demonstration of ammonia as an intermediatein nitrogen fixation by Azotobacter. J. Biol. Chem.204:445-451.

25. Repaske, R., and P. W. Wilson. 1952. Nitrous oxideinhibition of nitrogen fixation by Azotobacter. J. Am.Chem. Soc. 74:3101-3103.

26. Schollhorn, R., and R. H. Burris. 1967. Acetylene as acompetitive inhibitor of N2 fixation. Proc. Natl. Acad.Sci. U.S.A. 58:213-216.

27. Smith, B. E., D. J. Lowe, and R. C. Bray. 1973. Studiesby electron paramagnetic resonance on the catalyticmechanism of nitrogenase of Klebsiella pneumoniae.Biochem. J. 135:331-341.

28. Strandberg, G. W., and P. W. Wilson. 1967. Molecular H2and the pN2 function of Azotobacter. Proc. Natl. Acad.Sci. U.S.A. 58:1404-1409.

29. Tso, M.-Y. W., and R. H. Burris. 1973. The binding ofATP and ADP by nitrogenase components from Clos-tridium pasteurianum. Biochim. Biophys. Acta309:263-270.

30. Wong, P. P., and R. H. Burris. 1972. Nature of oxygeninhibition of nitrogenase from Azotobacter vinelandii.Proc. Nat. Acad. Sci. U.S.A. 69:672-675.

31. Zumft, W. G., L. E. Mortenson, and G. Palmer. 1974.Electron-paramagnetic-resonance studies on nitrogen-ase. Investigation of the oxidation-reduction behaviourof azoferredoxin and molybdoferredoxin with potenti-ometric and rapid-freeze techniques. Eur. J. Biochem.46:525-535.

VOL. 123, 1975


http://jb.asm.org/

Dow

nloaded from
http://jb.asm.org/

interactions among substrates inhibitors nitrogenase · lene, azide, cyanide, methyl isocyanide,...

Documents