Biochem. J. (1972) 129, 1063-1070 1063 Printed in Great Britain
Mossbauer Effect in Rubredoxin DETERMINATION OF THE HYPERFINE FIELD OF THE IRON IN A SIMPLE IRON-SULPHUR PROTEIN
By K. K. RAO, M. C. W. EVANS, R. CAMMACK and D. 0. HALL Department of Botany, King's College, London SE24 9JF, U.K. and C. L. THOMPSON, P. J. JACKSON and C. E. JOHNSON Oliver Lodge Laboratory, University ofLiverpool, Liverpool L69 3BX, U.K. (Received 17 May 1972)
1. Rubredoxin isolated from the green photosynthetic bacterium Chloropseudomonas ethylica was similar in composition to those from anaerobic fermentative bacteria. Amino acid analysis indicated a minimum molecular weight of 6352 with one iron atom per molecule. 2. The circular-dichroism and electron-paramagnetic-resonance spectra of Ch. ethylica rubredoxin showed many similarities to those of Clostridium pasteurianum, but suggested that there may be subtle differences in the protein conformation about the iron atom. 3. Mossbauer-effect measurements on rubredoxin from Cl. pasteurianum and Ch. ethylica showed that in the oxidized state the iron (high-spin Fe3+) has a hyperfine field of 370+3kG, whereas in the reduced state (high-spin Fe2+) the hyperfine field tensor is anisotropic with a component perpendicular to the symmetry axis of the ion of about -200kG. For the reduced protein the sign of the electric-field gradient is negative, i.e. the ground state of the Fe2+ is a d.z orbital. There is a large non-cubic ligand-field splitting (s/k=900'K), and a small spin-orbit splitting (D-+4.4cm'1) of the Fe2+ levels. 4. The contributions of core polarization to the hyperfine field in the Fe3+ and Fe2+ ions are estimated to be -370 and -300kG respectively. 5. The significance of these results in interpretation of the Mossbauer spectra of other iron-sulphur proteins is discussed.
The Mossbauer effect is becoming increasingly The present paper describes measurements on the useful in studying the nature ofiron atoms in proteins non-haem iron protein rubredoxin by using Moss- (see Lang, 1970; Johnson, 1971). Mossbauer spectra bauer spectroscopy. Rubredoxin was first isolated may show a chemical shift (relative to a standard from Clostridiumpasteurianum by Lovenberg & Sobel iron sample), a quadrupole splitting and a magnetic (1965). Subsequently rubredoxins have been isolated splitting, each of which is characteristic of the iron from many species (see Meyer et al., 1971; Tsibris & in the protein molecule. Thechemical shift in principle Woody, 1970). The 'labile' sulphur which is found in allows the oxidation state ofthe iron to be determined, other types of non-haem iron proteins such as although in practice it is not generally very sensitive ferredoxins has not been detected in rubredoxins. when the iron has strong covalent bonds. An excep- All rubredoxins, except that isolated from Pseudo- tion to this is high-spin Fe2+ which has a large monas oleovorans, have a molecular weight of about positive shift (of the order of lmm/s relative to 6000, contain one iron atom per molecule, accept one metallic iron). The quadrupole splitting may also electron on reduction and have an oxidation-reduc- provide an indication of the oxidation state, but only tion potential of about -0.06V at pH7. P. oleovorans indirectly, so that it is usually not possible to draw rubredoxin, however, has a molecular weight of unambiguous conclusions from it. Most powerful 19000, contains one or two iron atoms per molecule and useful is the measurement of magnetic hyperfine depending on the method of preparation and accepts splitting. This is different for each charge and spin one electron per iron atom on reduction (Lode & state of iron, and may therefore be used to determine Coon, 1971). Rubredoxin from the aerobe P. oleo- the valence state of the metal ion. It may also provide vorans is part of a hydroxylase system involved in the more detailed information such as the degree of oxidation of long-chain hydrocarbons and fatty covalent bonding (i.e. chemical coupling to the acids (Boyer et al., 1971; Peterson et al., 1966). All neighbours) and magnetic interaction (i.e. exchange the other rubredoxins are isolated from anaerobic bac- coupling). teria and no function has yet been ascribed to them, Vol. 129 1064 K. K. RAO AND OTHERS
Rubredoxin from Micrococcus aerogenes was chemistry. As in previous work with the ferredoxins, studied by Bachmayer et al. (1967), who suggested the iron in the native rubredoxins was replaced by on the basis of the e.p.r.* spectrum, position of the 57Fe isotope to increase the sensitivity of the Moss- cysteine residues in the amino acid sequence, and the bauer measurements. The reconstituted proteins were reactivity of the protein towards thiol reagents, freed from contaminating iron by gel filtration and that the iron atom in the oxidized rubredoxin is chromatography on DEAE-cellulose. The purity and high-spin Fe3+, with rhombic symmetry, bonded to conformational intactness of the proteins were deter- four cysteine residues of the protein. An X-ray mined by measuring the optical absorption, electron- crystallographic analysis of Cl. pasteurianum rubre- paramagnetic - resonance and circular-dichroism doxin to 0.5nm (5A) resolution (Watenpaugh et al., spectra. 1971), confirmed that the iron atom of the protein is bound to four cysteine sulphur atoms to form an Experimental approximately tetrahedral complex. M6ssbauer-effect measurements on C. pasteur- Cl. pasteurianum cells grown with NH3 as nitrogen ianum rubredoxin were reported by Phillips et al. source were obtained from the Microbiological (1970) in a study which also included proton magnetic Research Establishment, Porton Down, Wilts., U.K. resonance and magnetic susceptibility. In the oxidized Ch. ethylica was grown in a modified Pfennig's form of rubredoxin a hyperfine interaction with an medium with N2 gas as nitrogen source as described effective field at the nucleus of 375±5kG was previously (Evans & Smith, 1971). Protein was observed. It was concluded that the iron atom in determined by the phenol method as described by rubredoxin is high-spin Fe3+ in the oxidized form Rabinowitz & Pricer (1962). Performic acid oxida- and high-spin Fe2+ in the reduced form. The e.p.r. tion of aporubredoxin was done as described by Hirs spectrum of the oxidized protein (Lovenberg, 1966) (1967). Amino acid analysis of native and oxidized has a signal at g = 4.3, which shows that the Fe3+ rubredoxin was carried out with the assistance of is in a strongly distorted non-axial ligand field which Mrs. R. Hynes by using a Beckman model B amino splits the 6S ground state by several cm-'. acid analyser. The present paper describes more extensive Moss- M6ssbauer spectra were observed by using a con- bauer studies on rubredoxin over a wider range of ventional velocity drive and a monochromatic source temperature and making use of applied magnetic of 57Co in palladium, and were accumulated on a fields to study the hyperfine splitting (Johnson, multichannel analyser. Chemical shifts are expressed 1967a). One objective of the investigations was to relative to metallic iron at room temperature. determine the effect of covalent bonding to sulphur Other materials used, and the methods of spectro- atoms on the hyperfine interaction ofiron in a protein scopic measurement and analysis were as described environment. In general it is known that covalency previously (Rao et al., 1971). implies a strong overlap and admixing of s-electrons from the ligand atoms with the d-electrons of the Preparation ofrubredoxins iron (and vice versa), and this decreases the hyperfine interaction compared with that of a free ion. The The Ch. ethylica cells were processed in such a way values of the hyperfine fields in the Fe3+ and Fe2+ as to isolate the nitrogenase, rubredoxin and the atoms in a sulphur environment would be valuable in ferredoxin-type proteins. Unless otherwise specified obtaining a more quantitative interpretation of the the buffer used was 20mM-tris-HCI, pH8.5. All Mossbauer results in the ferredoxins. eluting solutions contained the buffer and were In this study we have used the rubredoxins from flushed with N2. Cl. pasteurianum andthegreen photosynthetic bacter- Frozen cells (stored in liquid N2) were broken into ium Chloropseudomonas ethylica. The isolation of small pieces and suspended in four times their rubredoxin from the latter organism and from the volume of buffer containing 50mM-MgCl2. The related green photosynthetic bacterium Chlorobium suspension was passed through a Manton-Gaulin thiosulphatophilum has recently been reported by homogenizer (A.P.V. Co. Ltd., Crawley, Sussex, Meyer et al. (1971), but the physicochemical proper- U.K.) at a setting of 55.2MN/M2 (80001bf/in2) and ties of Ch. ethylica rubredoxin were not described. homogenized with recycling. The homogenate was During our investigation, the chemical and physical centrifuged for 20min at 12000rev./min in the properties of Ch. ethylica rubredoxin were investig- 6x250ml rotor of an MSE 18 refrigerated centri- ated in some detail. Any difference between the fuge. The supernatant was treated with polyethylene rubredoxins of the anaerobic fermentative bacteria glycol (mol.wt. 6000; 10%, w/v) at pH 7.1 to remove and the green photosynthetic bacteria would be of the photosynthetic lamellae and then with protamine interest from the point of view of comparative bio- sulphate (20% by weight on a protein basis) to pre- cipitate the nitrogenase. The supernatant contained *Abbreviation: e.p.r., electronparamagneticresonance. ferredoxin, rubredoxin, and cytochromes. It was 1972 MOSSBAUER EFFECT IN RUBREDOXIN 1065 diluted twofold with buffer and stirred with a Pro6, Gly5, Ala2, CyS4, Val5, Met,, Ile,, Leu2, Tyr4, suspension of DEAE-cellulose (Whatman DE 23; Phe3. This analysis gives a value for the minimum W. and R. Balston Ltd., Maidstone, Kent, U.K.). molecular weight of 6352 including one iron atom The DEAE-cellulose was transferred to a column per molecule. The composition is typical of those (30cm x 8.8cm), washed with 0.2M-NaCl and eluted found in other rubredoxins (see Bachmayer et al., with 0.8M-NaCl. The eluate was diluted fivefold 1968). with buffer and adsorbed on top of a column The extinctioncoefficient ofCh. ethylicarubredoxin (45cm x 2.5cm) of DE 23 cellulose equilibrated with was measured by using the Folin-Lowry protein- 0.1 M-NaCl. Theproteins wereeluted from the column estimation method (Rabinowitz & Price, 1962), by chromatography by using successively 0.25M- and the value obtained was E490 = 6800 litre molP1 and 0.4M-NaCl. Eluates containing rubredoxin were cm-'. Lovenberg & Williams (1969) reported that monitored by measuring E490. The rubredoxin was this method gave the same results with Cl. pasteur- further purified by chromatography on a column ianum rubredoxin as those obtained from quantitative (35cmx2.5cm) of Sephadex G-50 followed by a amino acid analysis. By using this value the iron column (35cm x 2.5cm) of DE 23 cellulose in 0.25M- content of the rubredoxin was calculated to be NaCI. The yield of rubredoxin was about 10mg/kg 1.14g-atoms/mol in the native protein, and 1.12g- wet wt. of cells. atoms/mol in the 57Fe-reconstituted protein, assum- Cl. pasteurianum cells were broken in the Manton- ing a molecular weight of 6352. Gaulin homogenizer; the extract was centrifuged and the supematant was used for isolation of rubredoxin by DE 23 cellulose chromatography as described Optical spectra of the rubredoxin above. The absorption spectra of the native and 57Fe- reconstituted rubredoxins from Ch. ethylica wW Preparation of "Fe-substituted rubredoxins very similar to each other, with peaks at 276, 372, Rubredoxin (approx. 20mg) in Sml of buffer was 492 and 575nm in the oxidized form, and 275, 313 treated with 3 ml of 20% (w/v) trichloroacetic acid and 340nmin thereduced form (Fig. 1). Thisspectrum for 15min at 35°C. The mixture was centrifuged at is similar to those of other rubredoxins (e.g. Atherton 100OOg for 10min. The precipitate was washed with et al., 1966). There is a small difference between the 5 % trichloroacetic acid and with water and then dis- spectra of the oxidized forms of Ch. ethylica and Cl. solved in 3ml of 0.2M-tris-HCl buffer, pH8.5. The pasteurianum rubredoxins, in the region 300-400ip, protein was reprecipitated with trichloroacetic acid, where Cl. pasteurianum shows the presence of at least centrifuged and washed as before. The precipitate two absorption bands (Lovenberg & Sobel, 1965). The was dissolved in 3ml of 0.2M-tris-HCl buffer as circular-dichroism spectra of the two rubredo9* before. This solution of aporubredoxin showed no showed more pronounced differences. Figs. 2(a) "% optical absorption or circular dichroism in the visible 2(b) show the circular-dichroism spectra of oxidiz@d region. The apoprotein solution was incubated with 0.2ml of 2-mercaptoethanol for 2h at room temper- ature in N2. Then 20umol of 11FeSO4 was added and the mixture was incubated at 37°C for 10min. The reconstituted rubredoxin was separated from excess ofreagents by passage through a column ofSephadex G-25. It was further purified by chromatography on DE 23 cellulose with 0.30M-NaCl. The protein was finally concentrated by adsorption on a small column of DE 23 cellulose, washing with buffer, and elution with 0.8M-NaCl containing 50mM-tris-HCl buffer, pH 8.5. Theyield ofpurified, reconstituted rubredoxin was 60-70 %. Samples for measurement of Mossbauer spectra were contained in polyethylene cells of 0.5cm path- length, and were stored in liquid N2 before use. 300 400 500 600 Wavelength (nm) Results and Discussion Fig. 1. Optical absorption spectra of Ch. ethylica Composition of Ch. ethylica rubredoxin rubredoxin The amino acid composition of Ch. ethylica For experimental details see the text. , Oxidized rubredoxin was Lys5, Trpl, Asx8, Thr2, Ser1, Glx6, and ----, reduced rubredoxin. Vol. 129 1066 K. K. RAO AND OTHERS
Ch. ethylica and Cl. pasteurianum rubredoxins re- (Lovenberg, 1966; Atherton et al., 1966; Bachmayer spectively. The spectra shown are of the 57Fe- et al., 1967; Newman & Postgate, 1968; Peterson & reconstituted rubredoxins, which were virtually Coon, 1968), in which the signal at g = 4.3 is identical to the spectra of the corresponding native associated with broader absorptions at around g = proteins. Though the spectra of the rubredoxins 4.6 and g = 4.1 (Fig. 3b). These latter absorptions from the two bacteria were similar to each other in were attributed by Peisach et al. (1971) to absorp- general features, some of the bands appear to be tions from the other two principal directions, al- changed in intensity and shifted. In particular, a though it is not clear why these should be so broad shoulder at 410nm in Cl. pasteurianum rubredoxin when the central absorption is very sharp. In the is replaced by a band at 390nm in Ch. ethylica: this e.p.r. spectrum of Ch. ethylica rubredoxin these apparent shift may be associated with the difference associated absorptions were never observed, and in absorption spectrum. The optical absorption and the signal at g = 4.3 was consequently much more circular dichroism in the wavelength region 300- intense. Fig. 3 shows a comparison of the e.p.r. 600nm most probably arise from optically active spectra of oxidized rubredoxins from Ch. ethylica charge-transfer transitions in the iron-sulphur and Cl. pasteurianum. chromophore (see Tsibris & Woody, 1970), and are These chemical and spectroscopic analyses indicate probably very sensitive to changes in conformation. that the rubredoxin of Ch. ethylica is very similar to those of the anaerobic fermentative bacteria such as E.p.r. spectra Cl. pasteurianum, which have molecular weights of about 6000 and contain one iron atom per molecule. The oxidized rubredoxins showed e.p.r. signals at Some minor differences in the circular-dichroism and g = 9.4 andg = 4.3 (Fig. 3). Peisach et al. (1971) have e.p.r. spectra of the proteins may be a consequence pointed out that these two signals arise from the of small differences in the chelate structure of the lowest and middle Kramer's doublets respectively chromophore, which may in turn reflect differences of the high-spin Fe3+ atom. At temperatures between in amino acid sequence. Reconstitution with 57Fe 4.20 and 77°K the signal at g = 4.3 is much more did not appear to alter the structure of the molecule. prominent than that at g = 9.4. The e.p.r. spectrum of Ch. ethylica rubredoxin (Fig. 3a) was significantly different from those reported in other rubredoxins I
(a) I
I0
-l1o X0 20
x 10 (b)
-10 -20 500 600 700 800 ,, 1300 1400 1500 1600 1700 300 400 500 600 Magnetic field (G) Wavelength (nm) Fig. 3. E.p.r. spectra of oxidized rubredoxins from Fig. 2. Circular-dichroism spectra of oxidized rub- (a) Ch. ethylica and (b) Cl. pasteurianum redoxins from (a) Cl. pasteurianum (E490 = 0.88) and The gain setting for the signals centred at g = 9.4 (b) Ch. ethylica (E490 = 0.68) (left-hand side) was 100 times that for the signal at The rubredoxins were enriched with s7Fe isotope; g = 4.3 (right-hand side). Other settings were: the spectra were indistinguishable from those of the microwave frequency, 9.17GHz; power, 0.5mW; native proteins, temperature, 23°K. 1972 MOSSBAUER EFFECT IN RUBREDOXIN 1067
Mossbauer spectra It may be that these two lines were due to the presence ofan impurity containing iron, or to a differ- The spectra of rubredoxins from Ch. ethylica and ent relaxation rate associated with the physical state Cl. pasteurianum were not significantly different, of the rubredoxin (freeze-dried powder compared indicating that the iron atoms in the two proteins with the frozen solutions used in the present work). are in very similar electronic states. Most of the Reduced rubredoxin. The reduced protein contains spectra shown in this paper are for Ch. ethylica iron in the Fe2+ state and the Mossbauer spectrum rubredoxin. shows the characteristic two-line quadrupole pattern Oxidized rubredoxin. The Mossbauer spectra of at all temperatures. This is shown for Ch. ethylica oxidized rubredoxin measured at various temper- rubredoxin (Fig. 6) and Cl. pasteurianum (Fig. 7). atures are shown for Ch. ethylica in Fig. 4, and for At 77°K the shift, 8, was 0.65mm/s relative to iron Cl. pasteurianum in Fig. 5. at room temperature. The variation of quadrupole At the lower temperature (1.3°K and 4.2°K) splitting l\EQ with temperature (Fig. 6) was small magnetic hyperfine splitting was observed, with an showing that the ligand field acting at the iron atom effective field at the iron nucleus of 370+3kG, in is very strong. The value is 3.10mm/s at 198°K, and good agreement with the results of Phillips et al. it increases to 3.16mm/s at 77°K, then remains (1970). The hyperfine field was not discernibly altered constant down to 1.4°K. It is concluded that the by application of external magnetic fields of up ground-state orbital is pure, and lies lower by a to 30kG. large energy (A/k 900'K) than the first-excited At high temperatures (77°K and 195°K) the elec- orbital state. At low temperatures magnetic hyperfine tron-spin-relaxation rates have become rapid and the splitting was observed by applying magnetic fields of hyperfine splitting collapses (Fig. 4a). The spectrum up to 60kG to the samples (Fig. 8). The splitting ofthe in fact was extremely difficult to observe, which could spectrum was much larger than that caused by the be due either (a) to the relaxation rate being close to external field alone, showing that the effective field the precession frequency of the nucleus in the at the nuclei was mainly caused by the internal field hyperfine field or (b) to a very small Mossbauer arising from the unpaired electrons ofthe Fe2+ ion. effect at these temperatures arising from a weak binding of the iron atom in the molecule. The former is the most likely explanation. Our spectra ofoxidized Ch. ethylica and Cl. pasteurianumrubredoxins did not show the two strong lines in the centre of the spectra at 4.2°K and 77°K observed by Phillips et al. (1970). (a)
0
0.08 e(b)
.:., . 0 ;,--~~~~~~~ 0
I-
0 0.4k (c,)s
0O .0 (A 0 0.8
0 .0 CA uz 0 .~~~~~z '@ -'- - '--'* (d) *; , .,. o%0-7 0.41 -L., x,.t - -_I 0.8
0 -8 -6 -4 -2 2 4 6 8 -8 -6 -4 -2 0 2 4 6 Velocity (mm/s) Velocity (mm/s) Fig. 4. Mdssbauer spectra of oxidized Ch. ethylica Fig. 5. Mdssbauer spectra ofoxidized Cl. pasteurianum rubredoxin (1.2mM) enriched with 57Fe isotope rubredoxin (0.6mM) enriched with "7Fe isotope Temperatures were: (a) 77°K; (b) 4.2°K; (c) 1.4°K; Temperatures were: (a) 77°K; (b) 4.2°K and a small (d) 4.2°K and a field of 30kG was applied per- field was applied perpendicular to the y-rays. (c) was pendicular to the y-rays. For details see the text. at 1.3-K. For details see the text. Vol. 129 1068 K. K. RAO AND OTHERS
(a)
0 0
2
4 6 I.'.'4e> . ._
,0O o:
I-I0
o <2 ._- 0 4 1". I tee)~~~.* .. 6
d 0 2
4 2 6
-6 -4 -2 0 2 4 6 0 Velocity (mmlns) Fig. 6. Mdssbauer spectra of reduced Ch. ethylica rubredoxin (1.2mM) enriched with 57Fe isotope were: Temperatures (a) 198°K; (b) 77°K; (c) 4.2°K; -4 -2 0 2 4 6 see the text. (d) 1.4°K. For details Velocity (mm/s) Fig. 8. Mdssbauer spectra of reduced Ch. ethylica rubredoxin at 4.2°K, showing the effect of large magnetic fields applied perpendicular to the y-rays
0 .i% V...A.-w. The fields were: (a) 5kG; (b) IOkG; (c) 15kG; (d) 1-1 30kG; (e) 60kG. For spectra (a), (b), (c) and (e) the 0 sample concentration was 0.7mM; for spectrum (d) it 2 was 1.2mM. For further details see the text. ._
0 4
It is seen that the lines of the spectrum in magnetic 6 fields are quite sharply defined. The fact that sharp F I lines are observed in a sample containing anisotropic -4 -2 0 2 4 6 ions, which are randomly oriented, is partly a con- Velocity (mm/s) sequence of the anisotropic susceptibility of the iron ions. There is a tendency in a randomly oriented Fig. 7. Mossbauer spectrum of reduced C1.Cl. pasteur- sample for the effective field to be directed per- 5 iasotope ianum rubredoxin (0.8mM) enriched witi 57Fe isotope pendicular to the axis of symmetry of the ion, since at 770K the probability P(Q) that the external field H will be For details see the text. oriented at an angle 0 to the axis, is given by P(0)dO = sin 9d9 so that P(Q) is largest for 0 = 90°. In addition if Xl > X this tendency will be enhanced, as the internal field H. at the nucleus will then take From the behaviour of the two quadriupole lines in up an angle with the axis where tan# = XJX. tan0 the field, the lower-energy line is seen 1to arise from (Johnson, 1967b). X,, and Xl are the susceptibilities the I. = ±3/2 excited state of 57Fe, i.e. the electric- per mol measured parallel and perpendicular re- field gradient, q has a negative sign. T'his identifies spectively to the symmetry axis and the ligand field. the ground-state orbital as d,2. The effective field, Hafr at the nuclei will then be: 1972 MOSSBAUER EFFECT IN RUBREDOXIN 1069
M The hyperfine field in Fe2+ salts may be considered Heff. = H+H, = H+ Hn,(O) to be the sum of the contributions: Hn = H, + HL + Hd For small-field magnetization M = X,H and where H. arises from the s-electrons, HL from the H Hff±.=- H+ Hn (O) orbital current and Hd from the spin dipolar moment XN of the d-electrons. HL is related to the g value by: where H,,(0) is the hyperfine field, i.e. the internal HL-= 2B
These values may assist in interpreting the structure Dunham, W. R., Palmer, G., Sands, R. H. & Bearden, of more complex iron-sulphur proteins such as A. J. (1971) Biochim. Biophys. Acta 253, 373-384 ferredoxins. Several proteins of this important class Eaton, W. A., Palmer, G., Fee, J. A., Kimura, T. & have been shown to contain iron atoms co-ordinated Lovenberg, W. (1971) Proc. Nat. Acad. Sci. U.S. 68, 3015-3020 to sulphur ligands (either inorganic sulphide or Evans, M. C. W. & Smith, R. V. (1971) J. Gen. Microbiol. cysteine sulphur from the protein). The simplest of 65, 95-98 these are the two-iron, two-sulphur (plant-type) Gibson, J. F., Hall, D. O., Thornley, J. H. M. & Whatley, ferredoxins; although X-ray crystallography has not F. R. (1966) Proc. Nat. Acad. Sci. U.S. 56, 987-996 yet been applied successfully to these proteins, Hirs, C. H. W. (1967) Methods Enzymol. 11, 197-199 evidence from Mossbauer spectroscopy, electron- Johnson, C. E. (1967a) in Magnetic Resonance in Bio- nuclear double resonance and proton magnetic logical Systems (Ehrenberg, A., Malmstr6m, B. G. & resonance (see Rao et al., 1971; Cammack et al., Vanngard, T., eds.), pp. 405-406, Pergamon Press 1971; Dunham et a!., 1971; Munck et al., 1972; Poe Ltd., Oxford Johnson, C. E. (1967b) Proc. Phys. Soc. London 92, et a!., 1971) confirms the model of Gibson et al. 748-757 (1966) for these proteins, in which the two iron Johnson, C. E. (1971) J. Appl. Physics. 42, 1325-1331 atoms are co-ordinated to sulphur ligands, probably Lang, G. (1970) Quart. Rev. Biophys. 3, 1-60 with tetrahedral symmetry (Eaton et al., 1971). The Lode, T. & Coon, M. J. (1971) J. Biol. Chem. 246, iron atoms, which are antiferromagnetically coupled 791-802 together, are both high-spin Fe3+ in the oxidized Lovenberg, W. (1966) in Protides of Biological Fluids ferredoxin, and one of them is reduced to high-spin (Peeters, H., ed.), p. 165, Elsevier Publishing Co., Fe2+ in the reduced state. X-ray crystallography has Amsterdam been applied to the four-iron, four-sulphur high- Lovenberg, W. & Sobel, B. E. (1965) Proc. Nat. Acad. Sci. U.S. 54, 193-199 potential iron protein from Chromatium (Carter et Lovenberg, W. & Williams, W. M. (1969) Biochemistry al., 1971) and the eight-iron, eight-sulphurferredoxin 8, 141-148 from Micrococcus aerogenes (Sieker et al., 1972), Marshall, W. & Johnson, C. E. (1962) J. Phys. Radium and in both cases the iron atoms appear to be tetra- 23, 733-737 hedrally co-ordinated to sulphur ligands. Therefore Meyer, T. E., Sharp, J. J. & Bartsch, R. J. (1971) the iron atom in rubredoxin can be considered as a Biochim. Biophys. Acta 234, 266-269 model for those in the more complex iron-sulphur Munck, E., Debrunner, P. G., Tsibris, J. C. M. & proteins. Gunsalus, I. C. (1972) Biochemistry 11, 855-863 Newman, D. J. & Postgate, J. R. (1968) Eur. J. Biochem. 7, 45-50 We are grateful to Mr. L. Becker, Miss V. Minder and Peisach, J., Blumberg, W. E., Lode, E. T. & Coon, M. J. Miss J. Zantovska for skilled technical assistance. This (1971) J. Biol. Chem. 246, 5877-5881 work was supported by grants from the Science Research Peterson, J. A. & Coon, M. J. (1968) J. Biol. Chem. 243, Council. 329-334 Peterson, J. A., Basu, D. & Coon, M. J. (1966) J. Biol. References Chem. 241, 5162-5164 Phillips, W. D., Poe, M., Weiher, J. F., McDonald, C. C. Atherton, N. M., Garbett, K., Gillard, R. D., Mason, R., & Lovenberg, W. (1970) Nature (London) 227, 574-577 Mayhew, S. J., Peel, J. L. & Stangroom, J. E. (1966) Poe, M., Phillips, W. D., Glickson, J. D., McDonald, C. C. Nature (London) 212, 590-593 & San Pietro, A. (1971) Proc. Nat. Acad. Sci. U.S. Bachmayer, H., Piette, L. H., Yasunobu, K. T. & 68, 68-71 Whiteley, H. R. (1967) Proc. Nat. Acad. Sci. U.S. 57, Rabinowitz, J. C. & Pricer, W. E. (1962) J. Biol. Chem. 122-127 237, 2898-2902 Bachmayer, H., Yasunobu, K. T., Peel, J. L. & Mayhew, Rao, K. K., Cammack, R., Hall, D. 0. & Johnson, C. E. S. (1968) J. Biol. Chem. 243, 1022-1030 (1971) Biochem. J. 122, 257-265 Boyer, R. F., Lode, E. T. & Coon, M. J. (1971) Biochem. Sieker, L. C., Adman, E. & Jensen, L. H. (1972) Nature Biophys. Res. Commun. 44, 925-930 (London) 235, 40-42 Cammack, R., Rao, K. K., Hall, D. 0. & Johnson, C. E. Tsibris, J. C. M. & Woody, R. W. (1970) Coordin. Chem. (1971) Biochem. J. 125, 849-856 Rev. 5, 417-458 Carter, C. W., Freer, S. T., Xuong, Ng. H., Alden, R. A. & Watenpaugh, K. D., Sieker, L. C., Herriott, J. R. & Kraut, J. (1971) Cold Spring Harbour Symp. Quant. Jensen, L. H. (1971) Cold Spring Harbor Symp. Quant. Biol. 36, 381-386 Biol. 36, 359-368
1972