Biochimica et Biophysica Acta, 494 (1977) 191-197 © Elsevier/North-Holland Biomedical Press
BBA 37738 D E T E R M I N A T I O N OF T H E E X C H A N G E I N T E G R A L IN B I N U C L E A R I R O N - S U L F U R CLUSTERS IN P R O T E I N S OF V A R Y I N G C O M P L E X I T Y
J. C. SALERNO, TOMOKO OHNISHI, HAYWOOD BLUM and J. S. LEIGH Johnson Research Foundation, Department of Biochemistry and Biophysics, University of Pennsylvania Philadelphia, Pa. 19174 (U.S.A.) (Received January 25th, 1977) (Revised manuscript received May 4th, 1977)
SUMMARY The coupling constants J between the iron atoms in ferredoxin type ironsulfur proteins containing binuclear clusters were evaluated by two parallel methods. The temperature dependence of the EPR linewidths and integrated absorption intensities are both related to the energy of the first excited state. The values of J obtained were: center S-1 in succinate dehydrogenase, 90 c m - l ; Rieske's iron-sulfur center, 65 cm-~; adrenodoxin, 270 cm -1. The behavior of iron-sulfur center N-la in N A D H :UQ reductase was also examined; its similarity to that of center S-1 indicates that center N - l a is also a binuclear iron-sulfur center, with J 90 cm-1. Greater rhombic distortion present in the EPR spectrum of a binuclear cluster was associated with smaller values of J.
INTRODUCTION Scholes et al. [1] have described an elegant method for determining the strength of the zero field splitting in high spin heroes. A similar method was described and applied to the measurement of the exchange coupling in ferredoxin from the blue green algae Spirulina m a x i m a by Gayda et al. [2]. Both these determinations relied on the measurement of relaxation times and related the temperature dependence of these times to the energy of the excited state assuming that the relaxation occurred via an Orbach mechanism [3] (see, for example, ref. 4 for a discussion of relaxation mechanisms in the temperature range studied). A well established model for binuclear iron-sulfur clusters exist [5, 6]; two acid labile sulfur atoms bridge two antiferromagnetically coupled high spin iron atoms. In the reduced state, in which EPR signals can be observed, the formal valences of the iron atoms are + 3 and --2. The exchange coupling leads to a term in the Hamiltonian of the form --2JS1"$2 where $1 and $2 are the spin vectors of the high spin Fe ÷3 and Fe +z respectively, and J is the exchange coupling constant (J < 0). The states of the coupled system can then be described by a total spin number S -- 1/2, 3/2, 5/2, 7/2, 9/2. The energies relative to ground spin states thus
192 characterized are 0, - 3 J , - 8 J , 15J, and 24J, respectively. Only the ground (S 1/2) state is known to be EPR detectable. At low temperatures ( k T ~ 3J), only the ground state will be significantly populated. The first excited state will become accessible as the temperature is raised, however; the ratio of the population of the ground state (S ~ 1/2) to the first excited state ( S - 3/2) is given by the Boltzmann expression N ( I / 2 ) / N ( 3 / 2 ) ~ ( M e 3J/~,7-)~ where M is the ratio between the multiplicities of the two states. Since the excited state (S -- 3/2) has no observable EPR spectrum, the temperature dependence of the EPR intensity should deviate from the simple Curie Law dependence predicted by the Boltzmann distribution between the Zeeman levels S 1/2, M~ ± 1/2. We can thus use the above expression to determine J from the temperature dependence of the observed absorption area. This is analogous to the determination of J from magnetic susceptibility measurements, except that we observe the de-population of the ground state rather than the population of the excited states. For a ferredoxin with J 80 cm -1, at 77 K, the magnitude of the observed intensity would be given by I I0 (1 + Me-3J/kT) - ~ where M 2 (since the multiplicities of the two states are (2S -t 1), or 2 and 4, respectively) corresponding to a diminution of the signal by about 3 ~ . At 110 K, the intensity should be 10% smaller than that predicted by Curie Law for a simple S 1/2 system. MATERIALS AND METHODS Reconstitutively active soluble succinate dehydrogenase was prepared as described in ref. 7. N A D H : U Q reductase was prepared by the method of Hatefi et al. [8]. Rieske's iron sulfur center was studied in preparations of bcl complex obtained by the procedure of Erecinska et al. [9]. Adrenodoxin was studied in intact beef adrenal cortex mitochondria isolated as described by Omura et al. [10]. EPR measurements were performed on Varian E-4 and E-109 microwave spectrometers. Temperature control was obtained by use of a flowing helium cryostat (Air Products LTD-3-110). Temperature measurements were made with a carbon resistor installed below the sample at temperatures below 20 K; above 20 K thermocouples (a chromel vs. Au-0.07 ~ Fe) located below the sample and inside the sample tube were used. Integrations were performed on a Nicolet 1074 instrument computer. Resistor and thermocouples were calibrated by Air Products, and rechecked by immersion in liquid nitrogen, liquid helium and at 25 "C (vs. ice water in the case of the thermocouples). Use of two thermocouples, one at either end of the sample, provided insurance that the sample was in a narrow range of temperature ( ~ 1 K at 150 K); additional temperature errors were due to drift during each experiment of about 3 K maximum at 150 K: serious errors were avoided since temperature was continuously monitored. At temperatures below 35 K, no such drift was observed. RESULTS AND DISCUSSION Center S-1
Fig. 1 shows a plot of the double integrated intensity of EPR spectra of iron sulfur center S-I [11] in succinate dehydrogenase (reduced with succinate) vs. the
193
0.8-
06 ¸
04 oN[o 02
.002
• SH
,0~5
.6~
.62
I/T Fig. 1. Deviation from Curie Law of the integrated absorption 1 of center S-1 in succinate dehydrogenase and the high temperature EPR signals in NADH:ubiquinone reductase primarily from center N-la. Vertical axis, integrated absorption multiplied by temperature; horizontal axis, reciprocal temperature. A simple spin 1/2 system would follow Curie Law behavior and a horizontal straight line at 1.00 would result. The curves were drawn from the antiferromagnetic coupling model using J 60, 80 and 120 cm-1 for the lower, middle and upper curves. O, succinate dehydrogenase, succinate reduced; Z], NADH:UQ reductase, dithionite NADH reduced. EPR conditions were: frequency, 9.137 GHz; microwave power, 5 roW; modulation frequency, 100 KHz; modulation anaplitude, 10 gauss; time constant, 1 s; scan rate, 250 gauss/rain.
reciprocal o f temperature. The intensities have been corrected for Curie Law dependence arising from the B o l t z m a n n d i s t r i b u t i o n between the Ms - - + 1/2 sublevels o f the S 1/2 state by m u l t i p l y i n g by T. The theoretical lines were o b t a i n e d by considering the p o p u l a t i o n o f the first excited state at the values o f J shown for t e m p e r a t u r e between 50 K a n d 350 K. Neglecting higher excited states (S ~ 5/2) results in only small errors below 200 K ; this is the region in which useful spectra m a y be obtained. The best fit to the experimental points is o b t a i n e d by using a value o f J o f a b o u t 90 c m - 1 A p p l i c a t i o n o f a m e t h o d similar to that o f G a y d a et al. [2] using the linewidth o f the gy c o m p o n e n t as a measure o f the relaxation time resulted in a t e m p e r a t u r e d e p e n d e n c e as shown in Fig. 2. Similar results were o b t a i n e d using gz and gx. The relation 6 = A ( e + 3 S / k T - - 1 ) - - ~ ~ A e - 3 s / k r (for e - 3 s / k r ~ 1), for the O r b a c h r e l a x a t i o n process predicts a straight line o f slope - - 1 . 3 J for the plot o f log 6 agains l / T ; at high t e m p e r a t u r e s this becomes the d o m i n a n t c o n t r i b u t i o n to the linewidth and the plot becomes linear. A s s u m i n g that the overall linewidth is given by H 2 .... HL 2 + HG 2 where HL is the width o f the individual L o r e n t z i a n packets and HG the width o f the G a u s s i a n envelope due to i n h o m o g e n e o u s b r o a d e n i n g , the L o r e n t z i a n linewidth can be extracted from the overall linewidth by a simple calculation: HL - - ( H 2 - HG2) 1/2. While it is true t h a t m e a s u r e m e n t o f the a m p l i t u d e o f a derivative signal is directly related to linewidth, the d o u b l e integrated intensity o f the signal is not so related. Deviations from Curie Law b e h a v i o r o f this p a r a m e t e r c o r r e s p o n d to the p o p u l a t i o n o f the first excited state at the expense o f the g r o u n d state K r a m e r ' s d o u b l e t from which the E P R signal arises. This is an " e q u i l i b r i u m " measurement, while the line-
194 ~00
50 Adrenodox~n "E
~ 20
~~
\
Io-
~o
i '~ "~'~"'~RieskesCenfer ~"o. "
.............
~
~
N-I~
I
,~ - I ,
0
I0
10ST
, 20
30
Fig. 2. Temperature dependence of the ]inewidth of various binuclear ~ron sulfur clusters. Vertical axis, log; horizontal axis, reciprocal temperature. EPR conditions as in Fig. 1, except modulation amplitude, 5 gauss, when the overall ]Jnewidth was below 20 gauss. The ]inewidths are in arbitrary units scaled to allow the curves to be shown without overlap; on the logarithmic scale used this does not alter the slope o[the lines from which J,is calculated. Open symbols indicated measured lJnewJdth; closed symbols indicate calculated Lorentzian ]inew~dth: (~-F ~&), adrenodoxin; ( l + ~), center S-I in succinat¢ dehydrogenas¢; ( l + ~), N-]a in N A D H ubiquJnon¢ reductase; (~4 ~), Rieske's Center.
width determination is a "kinetic" measurement depending on the rate of excitation rather than a significant population of the excited state. From the temperature dependence of the Lorentzian linewidth we can deduce a value of J of about 95 cm -1, in good agreement with our estimate based on absorption areas. Rieske's iron sul[ur center The integrated absorption signal of Rieske's Center [12] is difficult to determine with good accuracy at high temperatures, since the linewidth becomes quite broad and causes lack of resolution in the highly anisotropic lineshape. The temperature dependence of the estimated linewidth (gy shown; gz similar) of Rieske's iron sulfur center in cytochrome bc~ complex may still be used as a measure of relaxation times as shown in Fig. 2. From the slope of the line a value of J of around 65 c m - ~ may be calculated, considerably smaller than that reported for any other binuclear iron sulfur cluster. Center N - l a High temperature g = 1.94 signals in N A D H dithionite reduced complex I [13], particulate N A D H : U Q ubiquinone reductase, have a temperature dependence strikingly similar to that of center S-I in succinate dehydrogenase. In this complex preparation the contribution to the signal from center N-1 b is much smaller than that of N - l a [14]. The temperature dependence suggests that center N-la, at least, is also
195 a binuclear iron sulfur cluster with a value of J of about 90 c m - l . The J value of center N - l b is probably similar, but is difficult to evaluate precisely because of its large temperature independent linewidth in this preparation. Integration of the absorption indicated an apparent non-Curie Law fall in intensity as the temperature was raised. However, the large molecular weight of the enzyme (10 6) places a limit on the signal to noise of about one order of magnitude worse than that attained with succinate dehydrogenase, making accurate double integration rather difficult in this case at temperatures above 77 K. Adrenodoxin
We were unable to observe significant departure from Curie Law behavior of the integrated absorption intensity of adrenodoxin [10, 15] in beef adrenal cortex preparations. Above 200 K the signal to noise ratio worsens due to Curie Law effects, decreased temperature stability and the onset of rapid broadening, making evaluation of the integrated intensity impractical. EPR spectra were recorded up to 230 K, however, which were of sufficient quality for linewidth measurements. The linewidth of the EPR spectrum was almost entirely temperature independent below 150 K. As the temperature was raised, further extremely rapid broadening of the spectrum occurred. Evaluation of J yielded a value of about 270 c m - 1, much larger than any previously reported J value for a binuclear iron-sulfur cluster. Such a value of J could explain the failure of susceptibility experiments to observe departure from Curie Law behavior below 200 K, as postulated by Kimura et al. [16]. It is interesting to note that the binuclear iron sulfur cluster with the smallest value of J, Rieske's iron sulfur center, has the greatest rhombic distortion. Those binuclear clusters with intermediate rhombic distortion (spinach ferredoxin, S p i r u l i n a m a x i m a ferredoxin, center S-l and center N-la) have intermediate values of J between 80 and 100 cm -1. Adrenodoxin, with a strikingly axial EPR spectrum, has by far the largest value of J. Fig. 3 shows the variation of J with A g ~- gy g×, which we take
500-
J (cm') 2OO
I00-
/t~'l~ ~ 4 . ~ o _ _ o _ _ _ ~ _ -
-~
50-
o 062 0.04 o,6e o.0~ o.io Ag Fig. 3. Variation of J with the rhombic distortion of the EPR spectrum as measured by .Ig (gy gx). Points labeled • were from our measurements; ( , for measurements of J by other workers ; ~A,from estimates based on published data (see footnotes p. 196). In order of decreasing J, the points represent putaredoxin, adrenodoxin, Azotobacter Vinelandii protein I, centers N-la and S-l, Azotobacter Vinelandii protein II, S. Maxirnus ferredoxin, spinach ferredoxin, and Rieske's Center.
196 as a measure of the rhombic distortion of the EPR spectra of the binuclear clusters in several iron sulfur proteins. Estimates of J have been made in some cases* by comparing the published temperature dependence of the EPR spectra to that of iron sulfur cluster of known J. While no great precision** is claimed in these cases, good enough estimates can be made to show that J decreases monotonically as the rhombic distortion increases. Using the assumptions of the model of Gibson et al., the value of~]g ,~ g~ --g~ is proportional to (E, -- E~)/E~E2, where E~ ~ (E~ E_z) and E2 ~ ( E ~ - - E ~ z ) are the energies of the dxz and dyz ferrous iron orbitals relative to the dz z orbital. Increasing rhombic distortion is thus brought about primarily by inequivalence of the dxz and dyz iron orbitals, which can in turn be caused by deviation from parallel alignment of the lines joining the cysteinyl sulfur ligands of each iron atom by rotation about the iron iron axis. This would be determined by the alignment of the polypeptide chains. Other mechanisms causing rhombic distortion include the effects of steric hindrance and electrostatic forces on the position of the bridging sulfur ligands, and differential electrostatic effects of nearby charges on the dxz and dyz orbitals. The results presented in Fig. 3 demonstrate that maximal orbital overlap occurs for the axial case, which is reasonable in view of the fact that in this configuration either the d orbitals themselves or tetrahedrally directed hybrid bonding orbitals of both iron atoms could be directed at the same point, while rhombic distortion of the type proposed above would lead to rotation of the angular frame of each set of iron orbitals relative to the other, decreasing the orbital overlap for both direct terms and terms involving the bridging sulfur orbitals. In addition, Ag is inversely related to the ligand field strength, which in turn is inversely related to the interatomic distance. Since the value of the exchange integral should be a sensitive function of interatomic distance as well as bond angle, this effect could also contribute to the observed relationship between g and J. This simplified viewpoint ignores contributions to g from sources such as anisotropy in the g tensor of the ferric iron atoms, and is limited by the crystal field approach of the Gibson model ignoring the effects of covalency. However, the empirical relationship presented is strong enough to suggest that iron sulfur clusters * Estimates of J were made as follows: putaredoxin, from the temperature dependence of the Mossbauer spectra of M u n k et al. [17], J 225 325 c m - l . Azotobacter vinelandii proteins 1 and I1, from the temperature dependence of the EPR spectra of Shetna et al. [18], about 130 and 90 c m - ~ respectively. Spirulina m a x i m a ferredoxin, measured by Gayda et al. [2] by a similar method as J 83 cm -~. Spinach ferredoxin, measured by Palmer et al. [19] by magnetic susceptibility as J 80-100 cm -1, the temperature dependence of the EPR spectra [18] suggests that 80 cm -1 is close to the correct figure using our measurements as a basis for the estimate. While the absolute value of o/ our measurements may be in error by as much as 15/o, the differences between the values of J for different centers are easily detected, and we can confidently state, for example, that the magnitude of the exchange integral in Azotobacter protein II lies between those o r S . m a x i m a ferredoxin and center S-I. *" Error estimates were made as follows: for our own measurements, the standard error was small for all measurements except that of Rieske center. Accordingly, the error shown reflects the sum of the squares of the estimated uncertainities from such sources as temperature and linewidth measurement. F o r the crude estimates of J described above, since J was estimated by comparison with one or more centers measured directly by us or other workers, the error was assumed to be the m a x i m u m deviation consistent with the observation that, for example, the temperature dependence of the EPR spectra of Azotobaeter protein II was intermediate between spinach ferredoxin and succinate dehydrogenase center S-1.
197 which deviate significantly from this behavior are either n o t binuclear or have an additional relaxation mechanism such as cross relaxation with a faster relaxing species. ACKNOWLEDGEMENTS The authors express their gratitude to Dr. T. E. K i n g for his generous gift of succinate dehydrogenase a n d N A D H : u b i q u i n o n e reductase, and to Dr. Maria Erecinska for her generosity in providing bcl complex. We also t h a n k Ms. M. Mosley and Mr. K. Ray for help in preparing the manuscript. This work was supported by United States Public Health G r a n t s G M 12202, USPHS-HL-15061-0551 and N a t i o n a l Science F o u n d a t i o n G r a n t BMS 7513459. REFERENCES 1 Schole~, C. P., lsaacson, R. A. and Feher, G. (1971) Biochim. Biophys. Acta 244, 206-210 2 Gayda, J. P., Gibson, J. E., Cammack, R., Hall, D. P. and Mullinger, R. (1976) Biochim. Biophys. Acta 434, 154 163 3 Orbach, R. (1961) Proc. R. Soc. Lon. A. 264, 458-484 4 Abragam, A and Bleaney, B. (1970) Electron Paramagnetic Resonance of Transition lons, p. 560, Claredon Press, Oxford 5 Gibson, J. F., Hall, D. O., Thornley, J. F. and Whatley, F. (1966) Proc. Natl. Acad. Sci. U.S. 56, 987-990 6 Sands, R. H. and Dunham, W. R. (1975) Q. Rev. Biophys. 7, p. 443-507 7 Ohnishi, T., Salerno, J. C., Winter, D. B., Lira, J., Yu, C. A., Yu, L. and King, T. E. (1976) J. Biol. Chem. 251,2094-2104 8 Hatefi, Y., Haavik, A. G. and Grittiths, D. E. (1962) J. Biol. Chem. 237, 1676-1680 9 Erecinska, M., Oshino, R., Oshino, N. and Chance, B. (1973) Arch. Biochem. Biophys. 157, 431 4~,5 10 Omura, 1., Sato, R., Cooper, D. Y., Rosenthal, O. and Estabrook, R. W. (1967) in Methods in Enzymology, Vol. 10, pp. 362-367, Academic Press, N.Y. I1 Beinert, H. and Sands, R. H. (1960) Biochem. Biophys. Res. Commun. 3, 41-46 12 Rieske, J. S., McLennan, D. H. and Coleman, R. (1964) Biochem. Biophys. Res. Commun. 15, 338 344 13 Orme-Johnson, H. R., Orme-Johnson, W. H., Hansen, R. E., Beinert, H. and Hatefi, Y. (1971) Biochem. Biophys. Res. Commun. 44, 446-452 14 Ohnishi, T., Leigh, J. S., Ragan, C. I. and Racker, E. (1974) Biochem. Biophys. Res. Commun. 56, 775 782 15 Watari, H., and Kimura, T. (1966) Biochem. Biophys. Res. Commun. 24, 106-111 16 Kimura, T., Tasaki, A. and Watari, H. (1970) J. Biol. Chem. 245, 4450 17 Munck, E., Debrunner, P. G., Tsibris, J. C. M. and Gunsalus, 1. C. (1972) Biochemistry 11,855 18 Shetna, Y. I., Dervertanian, D. V. and Beinert, H. (1968) Biochem. Biophys. Res. Commun. 31, 862-868 19 Palmer, G., Dunham, R. W., Fee, J. A., Sands, R. H., lizuka, T. and Yonetani, T. (1971) Biochim. Biophys. Acta 253, 373-384
BBA 37738 D E T E R M I N A T I O N OF T H E E X C H A N G E I N T E G R A L IN B I N U C L E A R I R O N - S U L F U R CLUSTERS IN P R O T E I N S OF V A R Y I N G C O M P L E X I T Y
J. C. SALERNO, TOMOKO OHNISHI, HAYWOOD BLUM and J. S. LEIGH Johnson Research Foundation, Department of Biochemistry and Biophysics, University of Pennsylvania Philadelphia, Pa. 19174 (U.S.A.) (Received January 25th, 1977) (Revised manuscript received May 4th, 1977)
SUMMARY The coupling constants J between the iron atoms in ferredoxin type ironsulfur proteins containing binuclear clusters were evaluated by two parallel methods. The temperature dependence of the EPR linewidths and integrated absorption intensities are both related to the energy of the first excited state. The values of J obtained were: center S-1 in succinate dehydrogenase, 90 c m - l ; Rieske's iron-sulfur center, 65 cm-~; adrenodoxin, 270 cm -1. The behavior of iron-sulfur center N-la in N A D H :UQ reductase was also examined; its similarity to that of center S-1 indicates that center N - l a is also a binuclear iron-sulfur center, with J 90 cm-1. Greater rhombic distortion present in the EPR spectrum of a binuclear cluster was associated with smaller values of J.
INTRODUCTION Scholes et al. [1] have described an elegant method for determining the strength of the zero field splitting in high spin heroes. A similar method was described and applied to the measurement of the exchange coupling in ferredoxin from the blue green algae Spirulina m a x i m a by Gayda et al. [2]. Both these determinations relied on the measurement of relaxation times and related the temperature dependence of these times to the energy of the excited state assuming that the relaxation occurred via an Orbach mechanism [3] (see, for example, ref. 4 for a discussion of relaxation mechanisms in the temperature range studied). A well established model for binuclear iron-sulfur clusters exist [5, 6]; two acid labile sulfur atoms bridge two antiferromagnetically coupled high spin iron atoms. In the reduced state, in which EPR signals can be observed, the formal valences of the iron atoms are + 3 and --2. The exchange coupling leads to a term in the Hamiltonian of the form --2JS1"$2 where $1 and $2 are the spin vectors of the high spin Fe ÷3 and Fe +z respectively, and J is the exchange coupling constant (J < 0). The states of the coupled system can then be described by a total spin number S -- 1/2, 3/2, 5/2, 7/2, 9/2. The energies relative to ground spin states thus
192 characterized are 0, - 3 J , - 8 J , 15J, and 24J, respectively. Only the ground (S 1/2) state is known to be EPR detectable. At low temperatures ( k T ~ 3J), only the ground state will be significantly populated. The first excited state will become accessible as the temperature is raised, however; the ratio of the population of the ground state (S ~ 1/2) to the first excited state ( S - 3/2) is given by the Boltzmann expression N ( I / 2 ) / N ( 3 / 2 ) ~ ( M e 3J/~,7-)~ where M is the ratio between the multiplicities of the two states. Since the excited state (S -- 3/2) has no observable EPR spectrum, the temperature dependence of the EPR intensity should deviate from the simple Curie Law dependence predicted by the Boltzmann distribution between the Zeeman levels S 1/2, M~ ± 1/2. We can thus use the above expression to determine J from the temperature dependence of the observed absorption area. This is analogous to the determination of J from magnetic susceptibility measurements, except that we observe the de-population of the ground state rather than the population of the excited states. For a ferredoxin with J 80 cm -1, at 77 K, the magnitude of the observed intensity would be given by I I0 (1 + Me-3J/kT) - ~ where M 2 (since the multiplicities of the two states are (2S -t 1), or 2 and 4, respectively) corresponding to a diminution of the signal by about 3 ~ . At 110 K, the intensity should be 10% smaller than that predicted by Curie Law for a simple S 1/2 system. MATERIALS AND METHODS Reconstitutively active soluble succinate dehydrogenase was prepared as described in ref. 7. N A D H : U Q reductase was prepared by the method of Hatefi et al. [8]. Rieske's iron sulfur center was studied in preparations of bcl complex obtained by the procedure of Erecinska et al. [9]. Adrenodoxin was studied in intact beef adrenal cortex mitochondria isolated as described by Omura et al. [10]. EPR measurements were performed on Varian E-4 and E-109 microwave spectrometers. Temperature control was obtained by use of a flowing helium cryostat (Air Products LTD-3-110). Temperature measurements were made with a carbon resistor installed below the sample at temperatures below 20 K; above 20 K thermocouples (a chromel vs. Au-0.07 ~ Fe) located below the sample and inside the sample tube were used. Integrations were performed on a Nicolet 1074 instrument computer. Resistor and thermocouples were calibrated by Air Products, and rechecked by immersion in liquid nitrogen, liquid helium and at 25 "C (vs. ice water in the case of the thermocouples). Use of two thermocouples, one at either end of the sample, provided insurance that the sample was in a narrow range of temperature ( ~ 1 K at 150 K); additional temperature errors were due to drift during each experiment of about 3 K maximum at 150 K: serious errors were avoided since temperature was continuously monitored. At temperatures below 35 K, no such drift was observed. RESULTS AND DISCUSSION Center S-1
Fig. 1 shows a plot of the double integrated intensity of EPR spectra of iron sulfur center S-I [11] in succinate dehydrogenase (reduced with succinate) vs. the
193
0.8-
06 ¸
04 oN[o 02
.002
• SH
,0~5
.6~
.62
I/T Fig. 1. Deviation from Curie Law of the integrated absorption 1 of center S-1 in succinate dehydrogenase and the high temperature EPR signals in NADH:ubiquinone reductase primarily from center N-la. Vertical axis, integrated absorption multiplied by temperature; horizontal axis, reciprocal temperature. A simple spin 1/2 system would follow Curie Law behavior and a horizontal straight line at 1.00 would result. The curves were drawn from the antiferromagnetic coupling model using J 60, 80 and 120 cm-1 for the lower, middle and upper curves. O, succinate dehydrogenase, succinate reduced; Z], NADH:UQ reductase, dithionite NADH reduced. EPR conditions were: frequency, 9.137 GHz; microwave power, 5 roW; modulation frequency, 100 KHz; modulation anaplitude, 10 gauss; time constant, 1 s; scan rate, 250 gauss/rain.
reciprocal o f temperature. The intensities have been corrected for Curie Law dependence arising from the B o l t z m a n n d i s t r i b u t i o n between the Ms - - + 1/2 sublevels o f the S 1/2 state by m u l t i p l y i n g by T. The theoretical lines were o b t a i n e d by considering the p o p u l a t i o n o f the first excited state at the values o f J shown for t e m p e r a t u r e between 50 K a n d 350 K. Neglecting higher excited states (S ~ 5/2) results in only small errors below 200 K ; this is the region in which useful spectra m a y be obtained. The best fit to the experimental points is o b t a i n e d by using a value o f J o f a b o u t 90 c m - 1 A p p l i c a t i o n o f a m e t h o d similar to that o f G a y d a et al. [2] using the linewidth o f the gy c o m p o n e n t as a measure o f the relaxation time resulted in a t e m p e r a t u r e d e p e n d e n c e as shown in Fig. 2. Similar results were o b t a i n e d using gz and gx. The relation 6 = A ( e + 3 S / k T - - 1 ) - - ~ ~ A e - 3 s / k r (for e - 3 s / k r ~ 1), for the O r b a c h r e l a x a t i o n process predicts a straight line o f slope - - 1 . 3 J for the plot o f log 6 agains l / T ; at high t e m p e r a t u r e s this becomes the d o m i n a n t c o n t r i b u t i o n to the linewidth and the plot becomes linear. A s s u m i n g that the overall linewidth is given by H 2 .... HL 2 + HG 2 where HL is the width o f the individual L o r e n t z i a n packets and HG the width o f the G a u s s i a n envelope due to i n h o m o g e n e o u s b r o a d e n i n g , the L o r e n t z i a n linewidth can be extracted from the overall linewidth by a simple calculation: HL - - ( H 2 - HG2) 1/2. While it is true t h a t m e a s u r e m e n t o f the a m p l i t u d e o f a derivative signal is directly related to linewidth, the d o u b l e integrated intensity o f the signal is not so related. Deviations from Curie Law b e h a v i o r o f this p a r a m e t e r c o r r e s p o n d to the p o p u l a t i o n o f the first excited state at the expense o f the g r o u n d state K r a m e r ' s d o u b l e t from which the E P R signal arises. This is an " e q u i l i b r i u m " measurement, while the line-
194 ~00
50 Adrenodox~n "E
~ 20
~~
\
Io-
~o
i '~ "~'~"'~RieskesCenfer ~"o. "
.............
~
~
N-I~
I
,~ - I ,
0
I0
10ST
, 20
30
Fig. 2. Temperature dependence of the ]inewidth of various binuclear ~ron sulfur clusters. Vertical axis, log; horizontal axis, reciprocal temperature. EPR conditions as in Fig. 1, except modulation amplitude, 5 gauss, when the overall ]Jnewidth was below 20 gauss. The ]inewidths are in arbitrary units scaled to allow the curves to be shown without overlap; on the logarithmic scale used this does not alter the slope o[the lines from which J,is calculated. Open symbols indicated measured lJnewJdth; closed symbols indicate calculated Lorentzian ]inew~dth: (~-F ~&), adrenodoxin; ( l + ~), center S-I in succinat¢ dehydrogenas¢; ( l + ~), N-]a in N A D H ubiquJnon¢ reductase; (~4 ~), Rieske's Center.
width determination is a "kinetic" measurement depending on the rate of excitation rather than a significant population of the excited state. From the temperature dependence of the Lorentzian linewidth we can deduce a value of J of about 95 cm -1, in good agreement with our estimate based on absorption areas. Rieske's iron sul[ur center The integrated absorption signal of Rieske's Center [12] is difficult to determine with good accuracy at high temperatures, since the linewidth becomes quite broad and causes lack of resolution in the highly anisotropic lineshape. The temperature dependence of the estimated linewidth (gy shown; gz similar) of Rieske's iron sulfur center in cytochrome bc~ complex may still be used as a measure of relaxation times as shown in Fig. 2. From the slope of the line a value of J of around 65 c m - ~ may be calculated, considerably smaller than that reported for any other binuclear iron sulfur cluster. Center N - l a High temperature g = 1.94 signals in N A D H dithionite reduced complex I [13], particulate N A D H : U Q ubiquinone reductase, have a temperature dependence strikingly similar to that of center S-I in succinate dehydrogenase. In this complex preparation the contribution to the signal from center N-1 b is much smaller than that of N - l a [14]. The temperature dependence suggests that center N-la, at least, is also
195 a binuclear iron sulfur cluster with a value of J of about 90 c m - l . The J value of center N - l b is probably similar, but is difficult to evaluate precisely because of its large temperature independent linewidth in this preparation. Integration of the absorption indicated an apparent non-Curie Law fall in intensity as the temperature was raised. However, the large molecular weight of the enzyme (10 6) places a limit on the signal to noise of about one order of magnitude worse than that attained with succinate dehydrogenase, making accurate double integration rather difficult in this case at temperatures above 77 K. Adrenodoxin
We were unable to observe significant departure from Curie Law behavior of the integrated absorption intensity of adrenodoxin [10, 15] in beef adrenal cortex preparations. Above 200 K the signal to noise ratio worsens due to Curie Law effects, decreased temperature stability and the onset of rapid broadening, making evaluation of the integrated intensity impractical. EPR spectra were recorded up to 230 K, however, which were of sufficient quality for linewidth measurements. The linewidth of the EPR spectrum was almost entirely temperature independent below 150 K. As the temperature was raised, further extremely rapid broadening of the spectrum occurred. Evaluation of J yielded a value of about 270 c m - 1, much larger than any previously reported J value for a binuclear iron-sulfur cluster. Such a value of J could explain the failure of susceptibility experiments to observe departure from Curie Law behavior below 200 K, as postulated by Kimura et al. [16]. It is interesting to note that the binuclear iron sulfur cluster with the smallest value of J, Rieske's iron sulfur center, has the greatest rhombic distortion. Those binuclear clusters with intermediate rhombic distortion (spinach ferredoxin, S p i r u l i n a m a x i m a ferredoxin, center S-l and center N-la) have intermediate values of J between 80 and 100 cm -1. Adrenodoxin, with a strikingly axial EPR spectrum, has by far the largest value of J. Fig. 3 shows the variation of J with A g ~- gy g×, which we take
500-
J (cm') 2OO
I00-
/t~'l~ ~ 4 . ~ o _ _ o _ _ _ ~ _ -
-~
50-
o 062 0.04 o,6e o.0~ o.io Ag Fig. 3. Variation of J with the rhombic distortion of the EPR spectrum as measured by .Ig (gy gx). Points labeled • were from our measurements; ( , for measurements of J by other workers ; ~A,from estimates based on published data (see footnotes p. 196). In order of decreasing J, the points represent putaredoxin, adrenodoxin, Azotobacter Vinelandii protein I, centers N-la and S-l, Azotobacter Vinelandii protein II, S. Maxirnus ferredoxin, spinach ferredoxin, and Rieske's Center.
196 as a measure of the rhombic distortion of the EPR spectra of the binuclear clusters in several iron sulfur proteins. Estimates of J have been made in some cases* by comparing the published temperature dependence of the EPR spectra to that of iron sulfur cluster of known J. While no great precision** is claimed in these cases, good enough estimates can be made to show that J decreases monotonically as the rhombic distortion increases. Using the assumptions of the model of Gibson et al., the value of~]g ,~ g~ --g~ is proportional to (E, -- E~)/E~E2, where E~ ~ (E~ E_z) and E2 ~ ( E ~ - - E ~ z ) are the energies of the dxz and dyz ferrous iron orbitals relative to the dz z orbital. Increasing rhombic distortion is thus brought about primarily by inequivalence of the dxz and dyz iron orbitals, which can in turn be caused by deviation from parallel alignment of the lines joining the cysteinyl sulfur ligands of each iron atom by rotation about the iron iron axis. This would be determined by the alignment of the polypeptide chains. Other mechanisms causing rhombic distortion include the effects of steric hindrance and electrostatic forces on the position of the bridging sulfur ligands, and differential electrostatic effects of nearby charges on the dxz and dyz orbitals. The results presented in Fig. 3 demonstrate that maximal orbital overlap occurs for the axial case, which is reasonable in view of the fact that in this configuration either the d orbitals themselves or tetrahedrally directed hybrid bonding orbitals of both iron atoms could be directed at the same point, while rhombic distortion of the type proposed above would lead to rotation of the angular frame of each set of iron orbitals relative to the other, decreasing the orbital overlap for both direct terms and terms involving the bridging sulfur orbitals. In addition, Ag is inversely related to the ligand field strength, which in turn is inversely related to the interatomic distance. Since the value of the exchange integral should be a sensitive function of interatomic distance as well as bond angle, this effect could also contribute to the observed relationship between g and J. This simplified viewpoint ignores contributions to g from sources such as anisotropy in the g tensor of the ferric iron atoms, and is limited by the crystal field approach of the Gibson model ignoring the effects of covalency. However, the empirical relationship presented is strong enough to suggest that iron sulfur clusters * Estimates of J were made as follows: putaredoxin, from the temperature dependence of the Mossbauer spectra of M u n k et al. [17], J 225 325 c m - l . Azotobacter vinelandii proteins 1 and I1, from the temperature dependence of the EPR spectra of Shetna et al. [18], about 130 and 90 c m - ~ respectively. Spirulina m a x i m a ferredoxin, measured by Gayda et al. [2] by a similar method as J 83 cm -~. Spinach ferredoxin, measured by Palmer et al. [19] by magnetic susceptibility as J 80-100 cm -1, the temperature dependence of the EPR spectra [18] suggests that 80 cm -1 is close to the correct figure using our measurements as a basis for the estimate. While the absolute value of o/ our measurements may be in error by as much as 15/o, the differences between the values of J for different centers are easily detected, and we can confidently state, for example, that the magnitude of the exchange integral in Azotobacter protein II lies between those o r S . m a x i m a ferredoxin and center S-I. *" Error estimates were made as follows: for our own measurements, the standard error was small for all measurements except that of Rieske center. Accordingly, the error shown reflects the sum of the squares of the estimated uncertainities from such sources as temperature and linewidth measurement. F o r the crude estimates of J described above, since J was estimated by comparison with one or more centers measured directly by us or other workers, the error was assumed to be the m a x i m u m deviation consistent with the observation that, for example, the temperature dependence of the EPR spectra of Azotobaeter protein II was intermediate between spinach ferredoxin and succinate dehydrogenase center S-1.
197 which deviate significantly from this behavior are either n o t binuclear or have an additional relaxation mechanism such as cross relaxation with a faster relaxing species. ACKNOWLEDGEMENTS The authors express their gratitude to Dr. T. E. K i n g for his generous gift of succinate dehydrogenase a n d N A D H : u b i q u i n o n e reductase, and to Dr. Maria Erecinska for her generosity in providing bcl complex. We also t h a n k Ms. M. Mosley and Mr. K. Ray for help in preparing the manuscript. This work was supported by United States Public Health G r a n t s G M 12202, USPHS-HL-15061-0551 and N a t i o n a l Science F o u n d a t i o n G r a n t BMS 7513459. REFERENCES 1 Schole~, C. P., lsaacson, R. A. and Feher, G. (1971) Biochim. Biophys. Acta 244, 206-210 2 Gayda, J. P., Gibson, J. E., Cammack, R., Hall, D. P. and Mullinger, R. (1976) Biochim. Biophys. Acta 434, 154 163 3 Orbach, R. (1961) Proc. R. Soc. Lon. A. 264, 458-484 4 Abragam, A and Bleaney, B. (1970) Electron Paramagnetic Resonance of Transition lons, p. 560, Claredon Press, Oxford 5 Gibson, J. F., Hall, D. O., Thornley, J. F. and Whatley, F. (1966) Proc. Natl. Acad. Sci. U.S. 56, 987-990 6 Sands, R. H. and Dunham, W. R. (1975) Q. Rev. Biophys. 7, p. 443-507 7 Ohnishi, T., Salerno, J. C., Winter, D. B., Lira, J., Yu, C. A., Yu, L. and King, T. E. (1976) J. Biol. Chem. 251,2094-2104 8 Hatefi, Y., Haavik, A. G. and Grittiths, D. E. (1962) J. Biol. Chem. 237, 1676-1680 9 Erecinska, M., Oshino, R., Oshino, N. and Chance, B. (1973) Arch. Biochem. Biophys. 157, 431 4~,5 10 Omura, 1., Sato, R., Cooper, D. Y., Rosenthal, O. and Estabrook, R. W. (1967) in Methods in Enzymology, Vol. 10, pp. 362-367, Academic Press, N.Y. I1 Beinert, H. and Sands, R. H. (1960) Biochem. Biophys. Res. Commun. 3, 41-46 12 Rieske, J. S., McLennan, D. H. and Coleman, R. (1964) Biochem. Biophys. Res. Commun. 15, 338 344 13 Orme-Johnson, H. R., Orme-Johnson, W. H., Hansen, R. E., Beinert, H. and Hatefi, Y. (1971) Biochem. Biophys. Res. Commun. 44, 446-452 14 Ohnishi, T., Leigh, J. S., Ragan, C. I. and Racker, E. (1974) Biochem. Biophys. Res. Commun. 56, 775 782 15 Watari, H., and Kimura, T. (1966) Biochem. Biophys. Res. Commun. 24, 106-111 16 Kimura, T., Tasaki, A. and Watari, H. (1970) J. Biol. Chem. 245, 4450 17 Munck, E., Debrunner, P. G., Tsibris, J. C. M. and Gunsalus, 1. C. (1972) Biochemistry 11,855 18 Shetna, Y. I., Dervertanian, D. V. and Beinert, H. (1968) Biochem. Biophys. Res. Commun. 31, 862-868 19 Palmer, G., Dunham, R. W., Fee, J. A., Sands, R. H., lizuka, T. and Yonetani, T. (1971) Biochim. Biophys. Acta 253, 373-384
