488

BIOCHEMICAL SOCIETY TRANSACTIONS

I am grateful to Dr. R. C. Bray, Dr. M. C. W. Evans, Dr. R. N. Mullinger and Dr. K. K. Rao for providing samples of iron-sulphur proteins and Miss V. K. O'Brien for technical assistance. This work was supported by grants from the Science Research Council.

Adman, E. T., Sieker, L. C. & Jensen, L. H. (1973) J. Biol. Chem. 248,3987-3996 Anderson, R. E., Anger, G., Peterson, L., Ehrenberg, A., Cammack, R., Hall, D. O., Mullinger, R. & Rao, K. K. (1975) Biochim. Biophys. Acru 376,63-71 Blumberg, W. E. & Peisach, J. (1974) Arch. Biochem. Biophys. 162,502-512 Cammack, R. (1973) Biochem. Biophys. Res. Commun. 54,548-554 Cammack, R., Rao, K. K. & Hall, D. 0. (1971) Biochem. Biophys. Res. Commun. 44,s-14 Coffman, R. E. & Stavens, B. W. (1970) Biochem. Biophys. Res. Commun.41,163-169 Fee, J. A. & Palmer, G. (1971) Biochim. Biophys. Actu 249,175-195 Genonde, Ic,Schlaak, M. E., Brietenbach, M., Parak, F.,Eicher, H., Zgorzalla, W., Kalvius, M. G. & Mayer, A. (1974) Eur. J. Biochem. 43,307-317 Hall, D. O., Cammack, R. &Rao, K. K. (1973)Pure Appl. Chem. 34,553-577 Hall, D. O., Cammack, R. & Rao, K. K. (1974) in Iron in Biochemistryand Medicine (Jacobs, A. & Worwood, M., eds.), pp. 279-334, Academic Press, London Kimura, T. (1971) Biochem. Biophys. Res. Commun. 43,1145-1149 Lowe, D. J., Lynden-Bell, R. M. & Bray, R. C. (1972) Biochem. J. 130,239-249 Mathews, R., Charlton, S., Sands, R. H. & Palmer, G. (1974) J. Biol. Chem. 249,4326-4328 McDonald, C. C., Phillips, W. D., Lovenberg, W. & Holm, R. H. (1973) Ann. N. Y. Acud. Sci. 222,789-799

Orme-Johnson, W. H. & Beinert, H. (1969) Biochem. Biophys. Res. Commun. 36,337-344 Orme-Johnson, W. H. & Sands, R. H. (1973) in Iron-Sulfur Proteins (Lovenberg, W., ed.), vol. 2, pp. 195-238, Academic Press, New York Padmanabhan, R. & Kimura, T. (1970) J. Biol. Chem. 245,2469-2475 Petering, D. H. &Palmer, G. (1970) Arch. Biochem. Biophys. 141,456-464

Nitrogenase in Azotobacter chroococcum and Klebsiella pneumoniae ROBERT R. EADY, CHRISTINA KENNEDY, BARRY E. SMITH, ROGER N. F. THORNELEY, GEOFFREY YATES and JOHN R. POSTGATE A R C Unit of Nitrogen Fixation, University of Sussex, Brighton BN1 9QJ, U.K. Nitrogenase, the enzyme system responsible for biological nitrogen fixation, has been the subject of numerous review articles, among which Eady & Postgate (1974) and Eady & Smith (1975) are recent examples and may be consulted for primary references. Nitrogenase from all sources examined so far consists of two oxygen-sensitive ironsulphur proteins; the larger tetrameric, Mo-Fe protein, contains at least three types of iron-sulphur centres (Smith & Lang, 1974) as well as molybdenum, and the smaller dmeric iron protein apparently contains only one type of iron-sulphur centre and is particularly oxygen-labile (Table 1). For enzymic activity, Mg2+and substratequantities of ATP are needed; Nz is reduced to NH3 but other triply bonded substrates (CN-, N3-, CH3NC, CZHJ can be reduced in place of Nz. CO inhibits their reduction. The enzyme also reduces H+ to Hz, a reaction which is not inhibited by CO but is largely inhibited by other reducible substrates. Experimental observations on the mechanism of nitrogenase action are consistent with Scheme 1. In the presence of NazSzOl a semi-reduced form of the Mo-Fe protein exists which has a distinctive e.p.r. (electron-paramagnetic-resonance)spectrum ( g l 4.3, g23.65, g3 2.01), associated with one type of its iron-sulphur clusters, and which can bind Nz or other substrates. In step 1 the reduced e.p.r.-active (g = 1.94) form of the iron protein interacts with MgATP to give a complex with a modified e.p.r. signal and changed redox potential (-250mv 3 -390mv) (Zumft et al., 1974); in step 2 this species forms the active enzyme complex by reacting with the semi-reduced form of the Mo-Fe protein. Electron transfer from the iron-sulphur centre of the Fe protein to the e.p.r.1975

556th MEETING, LONDON

489

Table 1. Properties of nitrogenase proteins from Klebsiella pneumoniae and Azotobacter chroococcum The maximum specific activity is expressed as nmol of the ethylene formed/min per mg of protein (complementary protein in excess).

A. chroococcum

K . pneumoniae ____

r

Mol.wt. Subunits Molybdenum (g-at omlmol) Iron (g-atom/mol) Sz- (g-atom/mol) Oz-sensitivity ( t ~ ,min) ~, Maximum specific activity

Mo-Fe protein (KPl)

Fe protein WP2)

Mo-Fe protein (A4

Fe protein (Ac2)

218000 51 300 59 600 1.85-2.15

66 700 34000

227 OOO Tetrameric

64OOO

0

1.9

0

4 3.85 0.75

21-25 18-22 10

3.9 3.9 0.50

1200

2300

2350

2200

31 OOO

Fe proteinMgADP complex

reduced ferredoxin, Ravodoxin

Reduced Fe proteinMgADP complex

I

T

+d

MgADP + PI

\

Mo-Fe protein-Nz reduced; Fe protein oxidized ADP Mg ATP

l2

Semi-reduced Mo-Fe protein 1

/A

N

2 or other substrate and H

I

MgA+

+

Reduced Fe proteinMgATP complex

Semi-reduced Mo-Fe protein-N,

-__ .-

Scheme 1. A model for nitrogenase action consistent with current experimental dam

VOl. 3

490

BIOCHEMICAL SOCIETY TRANSACTIONS

active iron-sulphur centre of the Mo-Fe protein precedes substrate reduction, the Mo-Fe protein becoming fully reduced and e.p.r.-inactive in the steady state. At some point during turnover, after formation of the active enzyme complex, MgATP becomes hydrolysed to MgADP; ultimately MgADP and P, are released (step 4). At least two ATP molecules are hydrolysed for every electron transferred. The reductive cycle is initiated as oxidized iron protein becomes re-reduced (step 3), and the substrate-binding cycle is completed by release of the semi-reduced form of the Mo-Fe protein. This outline description of nitrogenase action, necessarily brief, has been arrived at as a synthesis of ultracentrifuge, Mossbauer and e.p.r. studies (see reviews cited) supported by stopped-flow spectrophotometric studies (Thorneley, 1975). It is clearly oversimplified and will require elaboration at almost every step. In particular the sequence of the reactions is not as definite as Scheme 1 would suggest and there is no compelling reason why the active complex should be regarded as dissociating at any point during turnover; both features have been introduced into Scheme 1 arbitrarily for clarity of exposition. This contribution presents new data from our laboratory bearing in particular on the tertiary structure of the Mo-Fe protein and on the steps 1 4 in Scheme 1. For this work the nitrogenase proteins from K. pneumoniae and A . chroococcum were prepared in a high state of purity and homogeneity by a number of techniques (Eady et al., 1972; Smith et al., 1975; M. G. Yates & K. PlanquB, unpublished work). Table 1 gives data for the proteins used and indicates the code (Kpl, Kp2, Acl, Ac2) used in this communication; notice that the specific activity of protein Kpl is greater than that of Eady et al. (1972), an improvement matched by an increased content of molybdenum and iron (Smith et al., 1975). Subunits of Mo-Fe protein

Electrophoresis of various Mo-Fe proteins with sodium dodecyl sulphate gives evidence for sometimes one, sometimes two, classes of subunit (see reviews). In our hands either an impurity (probably dodecanol) in the sodium dodecyl sulphate or different gel-making procedures determines whether one or two bands are seen with protein Kpl or Acl, findings which raise the possibility that the presumptive subunits may be artifacts. Presumptive subunits of protein Kpl had mobilities corresponding to molecular weights of 51000 and 60000 and were present in equivalent amounts (Eady et al., 1972). The subunits have been isolated, digested with trypsin and 'maps' of the resulting peptides prepared by two-dimensional electrophoresis and chromatography on paper. Their peptide 'maps' were highly distinctive and the peptide 'map' of undissociated protein was a clear hybrid of both, confirming our earlier report that protein Kpl is a heteromeric tetramer of two of each type of subunit. Complex-formation

Evidence bearing on step 2 has been obtained by sedimentation-velocity studies of mixtures of protein Kpl +Kp2, Acl +Ac2 and the heterologous mixtures of proteins Kpl +Ac2 or Acl +Kp2. All showed the formation of a tight 1:1 molar complex in 25m-Tris-HC1 buffer, pH7.4, containing 10m-MgC1,. Binding was weakened by 5m-NazSz04in the case of protein mixture Acl+Ac2 and suppressed completely in the case of protein mixture Kpl +Kp2. All such mixtures gave enzymically active species which showed decreased specific acetylene-reducingactivity at high dilution ;association constants (assuming the formation of a 1:1 protein complex) derived from steady-state kinetic studies of such mixtures centred on 3 x 10-7~-1.With protein mixture Kpl +Kp2 the temperature-dependence of the association constant showed a sharp break from &€=O above 17°C to bH=418KJ*rnol-' below 17°C. An Arrhenius plot for the activity of the complex formed by the Klebsiella proteins was linear between 10°C and 40°C (AH:= 8OKJ*mol-'). 1975

556th MEETING, LONDON

491

Role of ATP The mode of action of ATP during nitrogenase function (reactions 1 and 2) has been studied by using the ATP analogues: 0 1 I oI

0

II

s II

A- A-

5’-adenosine-P-O-P-O-P-O

0

-

0

0

I/ I/ 1I 5’-adenosine-P-O-P-N-P-O I I l l 0-

-

0-H 0-

A W B , PNH)

ATP(7-S)

Both analogues in the presence of ATP inhibited acetylene reduction by protein Kpl f Kp2 and permitted enhanced hydrogen evolution, though hydrogen evolution was inhibited by 35-55 % in the absence of acetylene (Table 2). ADP inhibited both acetylene reduction and Hz evolution. The electron transfer induced by ATP (3mh.r) from reduced protein Kp2 to semioxidized protein Kpl has a half-time (tl12) of 6ms according to e.p.r. rapid-reaction studies and stopped-flow spectrophotometry at 425nm, in which ATP was mixed rapidly with proteins Kpl +Kp2+ MgZ++NazSz04. Replacement of ATP by equimolar ATP (y-S) gave a slow absorption change (tllz = 1s) with a smaller amplitude. Equimolar ATP(B,y-NH) induced a very slow change (tllz= 200s) accompanied by loss of the e.p.r. signal of protein Kp2 but no concomitant change in the e.p.r. signal of protein Kpl. The half-lives of these changes are in the order of the labilities of the y-terminal groups of ATP and the two analogues. These observations collectively suggest that ATP has at least two effects which the analogues mimic to different extents: one concerned with an oxidation of reduced protein Kp2, another related to acetylene reduction but possibly not to Hz evolution. Interaction with sodium dithionite (steps 3 and 4) Like protein Kp2 and the analogous protein from Azotobacter uinelandii and Clostridium pasteurianum (see the reviews), purified protein Ac2 shows an e.p.r. signal in the reduced form which changes from rhombic to axial symmetry when Mg-ATP is added (step 1). Purified protein Ac2 can be oxidized with only marginal loss of activity with phenazine methosulphate, yielding an e.p.r.-inactive species. Stopped-flow spectrophotometry at 430nm in 25m~-Tris-HCl buffer, pH 7.4, containing IOmM-MgZ+was used to study its interaction with Na2S204(step 3) and showed three distinct phases, the

Table 2. Eflect of ATP and anaIogues on nitrogenase function A mixture of 2nmol of each of protein Kpl and protein Kp2 was assayed for acetylene reduction and hydrogen evolution over 5 min at 30°C with ATP and its analogues (see the text) at the concentrations ( m ~given. ) Amount evolved Amount of Contents of vessel with acetylene Hz evolved (nmol) without c acetylene ATP ATP ATP (74) (B,y-NH) ADP Ethylene Hz (nmol) I

Y

0.4 0.4 0.4 0.4

Vol. 3

-

3.3

-

-

-

-

-

32 3 7

3.3

0

3.3

-

16 27 22 0

40 18 26 0

492

BIOCHEMICAL SOCIETY TRANSACTIONS

two fastest of which showed half-order dependence on [Sz042-]. This observation is consistent with the radical SO,-- being the active reductant. Second-order rate 104~-'-s-' which may be compared constants for two phases were 1.5x lo6 and 9 . 0 ~ with values for the other Fe-S proteins of 3 x lo4 for Micrococcus lactilyticus ferredoxin 5.1 x 104for C.pasteuriunumferredoxin and 2.3 x lo5for spinach ferredoxin (Lambert & Palmer, 1973). ATP had no effect on the rate of reduction of dye-oxidized protein Ac2 by 10m~-Na,S,O~but ADP decreased the rate of the fastest reduction phase. We are grateful to Dr. R. C. Bray and Dr. D. Lowe for collaboration work on e.p.r. spectroscopy, to Dr. R. S. Goody for supplies of ATP analogues and advice on their use, to Dr. Eva Kondorosi for developing gel-electrophoresis techniques and to Dr. D. Rekosh for preparing peptide 'maps'. Eady, R. R. & Postgate, J. R. (1974) Nature (London)249,805-810 Eady, R. R. & Smith, B. E. (1975) in Dinitrogen Fixation (Hardy, R. W . F., ed.), chapter 2, Wiley-Interscience, New York, in the press Eady, R. R.,Smith, B. E.,Cook, K. C. & Postgate, J. R. (1972) Biochem.J. 128,655-675 Lambert, D. 0. & Palmer, G. (1973)J. Biol. Chem. 248,6095-6103 Smith, B. E.& Lang, G. (1974) Biochem.J. 137,169-180 Smith,B. E., Thorneley, R. N. F., Yates, M. G.,Eady, R. R. &Postgate, J. R. (1975)Proceedings of the International Symposium on Nitrogen Fixation: Interdisciplinary Discussions (Newton, W. E. & Nyman, C. J., eds.), Washington State University Press, Pullman, Wash., in the press Thorneley, R. N. F. (1975)Biochem.J. 145,391-396 Zumft, W. G., Mortenson, L. E. & Palmer, G. (1974)Eur. J. Biochem. 46,525-535

Iron-Sulphur Proteins in the Photosynthetic Electron-Transport System of Oxygen-Evolving Organisms MICHAEL C. W. EVANS Department of Botany and Microbiology, University College, Gower Street, London WClE 6BT, U.K. The first ferredoxin to be isolated was probably the soluble two-iron ferredoxin involved in NADP+ reduction in oxygen-evolving photosynthetic organisms. This red protein isolated by Davenport et al. (1952) was shown by Arnon (1965) to be an electron carrier between Photosystem I and the flavoprotein NADP reductase. It is the only readily soluble component of the photosynthetic electron-transport chain. It and the a v o protein reductase are the only components of the chain whose functions can be completely defined. This ferredoxin has a redox potential = 420mV. Photosystem I in chloroplasts is able to reduce low-potential dyes under conditions that indicate a mid-point redox potential for the membrane-bound electron carriers of Photosystem I of about -550mV (Black, 1965). Photosystem I will undergo photochemical reactions at cryogenic temperatures which can be monitored by the photooxidation of the reaction centre chlorophyll P700. Under most conditions this lowtemperature photo-oxidation is essentially irreversible indicating that the electron acceptor is closely associated with the P700 and bound to the membrane. The identity of the primary electron acceptor and of the subsequent electron-transfer reactions are of considerable importance to an understanding of the mechanism of the photochemical reactions and in the possible use of chloroplast systems as solar-energy converters. The very low redox potentials required for the primary electron acceptor suggested that iron-sulphur centres might be candidates for this function. Malkin & Bearden (1971) showed that chloroplasts contain iron-sulphur proteins, ('bound ferredoxin') 1975