Evidence for a ferryl Fea3 in oxygenated cytochrome c oxidase.

Communication

THEJOURNAL OF BIOLOGICAL CHEMISTRY vol. 262 No. 4 Issue of February 5 pp. 1446-1448 1987 0 1987 by T i e American Society of Bioiogical Chemists, Inc. Printed in U S.A.

Evidence for a Ferry1 Fea, in Oxygenated Cytochrome c Oxidase*

dioxygen, a subpopulation of an EPR-silent, three-electronreduced dioxygen intermediateistrapped at the dioxygen reduction site. We (15) showed that this trapped intermediate reacts rapidly with carbon monoxide (CO) at room tempera(Received for publication, June 6, 1986) ture to produce a rhombic CUBEPR signal which is identical Stephan N. Witt and Sunney I. ChanS to the rhombic CUB EPR signal observed by Blair et al. (16) From the Arthur A. Noyes Laboratory of Chemical Physics, during reoxidation by dioxygen at low temperatures.The California Institute of Technology Pasadena, trapped intermediate exhibits optical bands in the reoxidized California 91 125 minus resting difference spectrum at 580 and 537 nm and a Soret maximum at 424-425 nm.Upon addition of CO to Evidence is reported which shows that a reactive samples displaying these optical features, both bands in the ferryl Fe,,/cupric Cua binuclear couple is present at a-p region were abolished. Based on its reactivity with CO to the dioxygen reduction site (i) in “oxygenated” cyto- produce a half-reduced dioxygen reduction site, in which Fea, chrome c oxidase; (ii) when the fully reduced enzyme is thought to be stabilized in a low-spin ferrous-dioxygen (or is reoxidized at low temperatures; and (iii) when par- possibly GO) adduct, we proposed that the EPR-silent, threetially reduced cytochrome c oxidase is reoxidized with electron-reduced dioxygen intermediate isa ferryl Fe.,/cupric dioxygen at room temperature. CuBcouple (15, 16). In this communication we report chemical and spectroscopic evidencefor anEPR-silent,three-electron-reduced When concentrated hydrogen peroxide (H202)is added to dioxygen intermediate at the dioxygen reduction site in the pulsed or reduced cytochrome c oxidase, a species is formed 428/580 nm component of oxygenated cytochrome c oxidase. with optical bands at 580-582 and 532-537 nm in the reoxi- Thisintermediate is identicaltotheintermediatethatis dized minus resting difference spectrum and with a Soret formed at low temperatures and which is trapped at room maximum at 427-428 nm (1-4). This state of the enzyme is temperature when partially reduced samples are reoxidized spectroscopically similartothe “oxygenated” state of the with dioxygen; we propose that the intermediate contains a enzyme originally prepared by Okunuki et al. ( 5 ) and studied ferryl Fe,,/cupric CuBcouple. by numerous investigators (6-11). Over the years there has MATERIALS ANDMETHODS been considerable confusion and speculation as to the redox and ligation states of Fess and CuBin this state of the enzyme. Beef heart cytochrome c oxidase was prepared by the method of Confusion over these issues resulted because the oxygenated Hartzell and Beinert (17) and wasdissolved in 50 mM phosphate state, when prepared by reducing the enzyme with excess buffer containing 0.5% Tween-20 at pH 7.4. Sample preparation were as described elsewhere (15) with the modification dithionite and reoxidizing with dioxygen, was shown by Orii procedures that the samples were reduced with NADH only; phenazine methoand King (12) to be an admixture of up to three distinct sulfate was not used as a mediator since it inhibits catalase (18). components. These species include Compound C (Compound Samples were maintained at ice temperature throughout the sample I in the nomenclature of Orii and King), which exhibits an manipulations and the recording of optical spectra. Reduced cytointense absorption at 605-607 nm in the reoxidized minus chrome c oxidase samples in EPR tubes were reoxidized with suffiresting difference spectrum (13), a species which exhibits a cient H202 (Baker) to give a final concentration of approximately 4 mM. Catalase (0.01-0.10%) (Sigma) was added within 100 s from the and p bands a t 580-582 and 532-537 nm in the difference addition of H202in order to remove the excess H202. After recording spectrum (CompoundI1 of Orii and King), anda species with optical spectra of samples from 750-450 nm, the samples were rebands at 600 and 580 nm (Compound I11of Orii and King). frozen to 77 K. Carbon monoxide was added to the EPR samples by Apparently a more homogeneous population of Compound I1 the methods described earlier (15). Integrations of the rhombic CUB is produced when large excesses of H202are used to reoxidize EPR signal were carried out following the method of Aasa and only the ViinngHrd (19) in which the area under the low-field hyperfine line is reduced cytochrome c oxidase, as the sample exhibits to the double integral of a standard sample, in this case, 580-582 and 532-537 nmbandsinthe reoxidized minus compared Cu(I1)-o-phenanthroline.The g values used in the integration of the resting difference spectrum anda 428 nm Soretmaximum (2, CUBEPR signal were determined from computer simulations of the 4). This state of the enzyme (Compound 11) has also been spectrum kZ= 2.28, g, = 2.109, and g, = 2.052) (20). recently referred to as pulsed peroxide I by Naqui et al. (14). EPR spectra were recorded on a Varian E-line Century series XInstead of adopting the conventional namesfor this species, band spectrometer equipped with a liquid nitrogen dewar to maintain namely Compound 11, oxygenated, or pulsed peroxide I, we the samples at 77 K. The optical spectra were recorded on a Beckman CIII which was interfaced to a Spex Industries SC-31 SCAMP will refer to this species simply as the 428/580 nm form of Acta data processor. the enzyme. In a recent communication (15), we reported that, upon RESULTSANDDISCUSSION reoxidation of partially reduced cytochrome c oxidase with An optical spectrumof the restingenzyme is shown in Fig. * This work was supported by Grant GM22432 from the National 1, a. Afterreoxidation of the reducedsample with excess Institute of General Medical Sciences, United States Public Health H202,the optical spectrum inFig. 1,b, was obtained following Service. This is Contribution 7421 from the Division of Chemistry the addition of catalase. Comparison of the optical spectrum and Chemical Engineering. The costs of publication of this article of the resting enzyme sample (Fig. 1, a) to the spectrum of were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with the reoxidized enzyme sample (Fig. 1, b ) reveals a shift of the (Y band from 600 to 596 nm and a corresponding shift of the 18 U.S.C. Section 1734 solely to indicate this fact. p band from 547 to 534 nm, respectively. The Soretmaximum $ To whom reprint requests should be addressed.

1446

A Ferry1 Fe,, in Oxygenated Cytochrome c Oxidase

1447

activation of all of the molecules may not be achieved. By reoxidizing fully reduced cytochrome c oxidase with excess Hz02, we ensure that all of the enzyme molecules are activated, and indeed, the difference spectrum in Fig. 1, c, does not contain the substantial component at 606 nm which is found in the spectra reported by Chance and co-workers (2) and Wrigglesworth (4). Therefore, theLY maximum of 596 nm reported herein isprobably closer to the actualvalue for this state of the enzyme. Significantly, the optical difference spectrum of the 428/580 nm componentof the oxygenated enzyme between 700 and 475 nm (Fig. 1,c ) is similar to the difference spectrum recently reported by this laboratoryfor the trapped a a , , , , , I three-electron-reduced dioxygen intermediate (15). 650 600 550 500 270 290 310 330 350 Insights into the nature of the 428/580 nm component of A (nrn) MAGNETIC FLUX DENSITY ( m T ) the oxygenated enzyme were derived from additional optical and parallel E P R experiments designed to monitor the reacW tivity of this species toward CO. The optical spectrum obV z a tained within approximately 20 s after the addition of CO to m a the 428/580 nm species (Fig. 1, d ) shows two interesting 2m features: 1) the 01 and p bands became red-shifted relative to a the spectrum of the 428/580 nm species by 6 and 12 nm, respectively (Fig. 1, b ) ; and 2) the ratio of the intensities of the a and p band maxima increased from 1.28 to 1.36 upon addition of CO to the 428/580 nm species. The difference spectrum in Fig. 1, e, shows that, upon addition of CO to the a a reoxidized species, the 580/537 nm spectral features (Fig. 1, c) are abolished andtheband at 604 nm becomes more X (nrn) MAGNETIC FLUX DENSITY CmTl intense. FIG. 1. Absorption and EPR spectra of the 428/580 nm An E P R spectrum a t 77 K after thereoxidation of the fully component of oxygenated cytochrome c oxidase before and reduced sample with excess H2O2 (Fig. 1, f ) consists of the after the addition of CO. After the addition of CO to the oxygenated sample, the samplewas flushed with dioxygen and then incubated at typical E P R spectrum of CuA (gz = 2.18, g, = 2.03, and g, = 277 K for 8 h to obtain a resting spectrum ( a ) . The reduced sample 1.99), with a slightyield of a radical EPR signal at g = 2. The was reoxidized with a 24-fold molar excessof Hz02, followed in 105 s addition of CO to this sample resulted in a large yield (50 f by the addition of 0.1% catalase ( b and c), and then frozen (360 s from the addition of H202) to 77 K to record an EPR spectrum ( f ) . 10%) of the rhombic CuB EPR signal (gll = 2.28, A,, = 0.0106 The resultant reoxidized sample was incubated under 1 atm of CO cm”) (Fig. 1,g ) . This signal has been reported toform under for 20 s, at approximately 283 K, and then refrozen to 77 K. After a variety of experimental conditions (15, 16, 20, 21). After recording an EPR spectrum (g), the sample was thawed again to recording the optical spectrum (Fig. 1, d and e ) the sample obtain the optical spectra in d and e and subsequently refrozen (620 was immediately refrozento 77 K and another EPR spectrum s from the addition of H202) to obtain a final EPR spectrum (h).The latter EPR spectrum verifiesthat the species exhibiting therhombic was recorded (Fig. 1, h ) to verify that the intermediate exhibCUB EPR signal (Fe:t-CO Cug) was present during the optical iting the rhombic CuB EPR signal was present during the measurement. The difference spectra (c and e ) are the 428/580 nm optical measurements. The EPR spectra in Fig. 1, g and h, species minus resting and the 428/580 nm species plus CO minus are a superposition of the EPR signals from CuA and CuB. A was 146 pM. The samples consequence of this superpositionis that thepositive comporesting, respectively. Sample concentration were contained in 3.4-mm, inner diameter, quartz EPR tubes. Conditions for obtaining EPR spectra were: temperature, 77 K; microwave nent of the CuAEPR spectrum at g = 2.08 (Fig. 1,f ) is reduced in intensity because that region of the spectrum is dominated power, 4 milliwatts; modulation amplitude, 1 mT; and gain, 1.25 X by the intense broad negative feature centered at g = 2.07 of 10‘ ( f - h ) . the rhombic CUB EPR signal (20). After warming the sample adecreasein the of the reoxidized samples was typically 428 nm. The Soret to 277 K to obtain an optical spectrum, maximum of 428 nm in the optical spectrum of the oxygenated concentration of the species giving rise to the rhombic CuB sample and thetwo bands a t 582 ( L e = 5 mM” cm”) and 534 EPR signal is observed. Thus, the overall intensity of the nm in the reoxidized minus resting difference spectrum (Fig. rhombic CuB E P R signal is diminished, as evidenced by the 1, c ) agree well with published results (1-4), although the N decrease in intensity of the hyperfine lines, while the g = 2.08 maximum of 596 nm differs substantially from the value of component of CuA has become more prominent (Fig. 1, h ) . 601 and 600 nm as reportedby Wrigglesworth (4) and Kumar The increased intensity in the N band upon the addition of et al. (2), respectively. This discrepancy in the 01 region for CO with the simultaneous formation of the rhombic CuBEPR samples prepared in this laboratory and in the laboratory of signal is consistent with a two-electron reduction of a ferryl Chance and co-workers and others is most likely related to Fea3/cupric CUB couple to alow-spinferrous-dioxygen (or the incomplete formation of the 428/580 nm component of possibly CO) Fe,,/cupric CuB couple (Equation 4). the oxygenated enzyme. Incomplete formation of the 428/580 In Equations 1-3 we propose a reaction scheme to account nm species can resultwhen excessH202is added to the restingfor ferryl Fea3 inthe preparation of oxygenated cytochrome c enzyme (4), since all of the molecules are probably not acti- oxidase. Whenthe H202 concentration is approximately vated, andwhen excess H,02 is added to the “pulsed” enzyme equivalent to the enzyme concentration, a peroxidic adduct, in the presence of catalase in the solution (2). In this latter Compound C, is produced at the dioxygen reduction site (4); case, due to the presenceof catalase, the effective concentra- however, if large excesses of H202are added to Compound C, tion of H 2 0 , is lower than expected; thus,the complete or are added initially to the pulsed or reduced enzyme, the OPTICAL

A Ferry1 Fe,, in Oxygen,zted Cytochrome c Oxidase

1448

extra H202can act as a direct one-electron donor to Compound C to produce the ferryl Fe,,/cupric CuBcouple and superoxide. A small yield of a radical EPR signal is, in fact, observed in the EPR spectrum of the 428/580 nm species sample (Fig. 1,f). In other experiments we have observed 510 times theyield of the radical E P R signal reported here. [FeilCu:Fe~~Cu~] + 2H202+ 4H+ -+

+

(1)

+ 2H+

(2)

[Fe'.'Cu~Fe~$u~] 4H2O [Fei:'Cu!]

+ H,O,

+ [Fe~~-O-O-Cu~]

temperatures (183-210 K) decayed to the EPR-silent intermediate, proposed to be a ferryl Fe,,/cupric CUB couple. In the present study, we show that intermolecular one-electron reduction of Compound C also produces the ferryl Fe,/cupric CUBcouple. The present results strengthen our earlier proposal (16) that thedioxygen bond is cleaved in cytochrome c oxidase at the three-electronlevel of reduction in an entropically promoted bond breaking step. Acknowledgment-We helpful discussions.

would like to thank Joel E.

Morgan for

REFERENCES [Feiy=O Cu#(OH)l

L [Fei: -L + CO +

Cu&OH)]

+ CO,

(L = 0, or CO)

(4)

1. Lemberg, R., and Mansley, G. E. (1966) Biochim. Biophys. Acta 118,19-35 2. Kumar. N., Naaui. 59. _ .A.. and Chance. B. (1984) J. Biol. Chem. 2~~. 11668-11671 3. Chance, B., Kumar. C.. Powers. L.. and Chine. C. (1983) ", Y. Biophys. J. 44,353-363 4. Wridesworth, J. M. (1984) Biochem. J. 217. 715-719 5. Okunuki, K., Hagihora, B., Sekuzu, I., and Horio, J. (1958) I

I

,

It is evident from these studies that the spectral features of the 428/580 nm component of oxygenated cytochrome c oxidase in the a,p region are similar to (i) the spectral features of the trapped, EPR-silent, three-electron-reduced dioxygen intermediate that is formed when partially reduced samples are reoxidized with dioxygen (15); and (ii) to the spectral features of the intermediate formed upon a one-electron reversal of the dioxygen reduction reaction reported by Wikstrom (22). In addition, we have demonstrated here that the 428/580 nm component of the oxygenated enzyme also reacts rapidly with CO at room temperature to produce large yields of the rhombic CuBE P R signal previously obtained (i) at low temperatures when theEPR-silent,three-electron-reduced dioxygen intermediate reacts with CO in samples containing 40% ethylene glycol (16); and (ii) when CO is added to samples at 277-298 K containing the trapped three-electron-reduced intermediate formed by reoxidizing partially reduced samples in the absence of ethylene glycol (15). Based on these chemical and spectroscopic similarities, we conclude that the same reactive, EPR-silent, ferryl Fes3/cupric CUB binuclear couple is formed at thedioxygen reduction site, whether the reduced enzyme is reoxidized by dioxygen or excess HZ02and when the pulsed enzyme is activated by excess H202. In our previous work we showed that a ferryl Fe,, is formed via intramolecular electron transferof a third electron to the two-electron-reduced dioxygen intermediate (Compound C) from either CuAor Fe, (16). In that study, the transferof the third electron to the peroxidic adduct at low temperature (180 K) first produced an EPR-detectable intermediate,suggested to be a ferrous Fe,,/cupric CuBhydroperoxide, which at higher

6. 7. 8. 9. 10. 11.

12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22.

.

I

I

Proceedings of the International Symposium on Enzyme Chemistry, Tokyo and Kyoto, 1958,p. 264, Maruzen, Tokyo Orii, Y., and Okunuki, K. (1963a) J. Biochem. (Tokyo) 5 3 , 489499 Lemberg, R., and Gilmour, M. V. (1967) Biochim. Biophys. Acta 143,500-517 Williams, G. R., Lemberg, R., and Cutler, M. E. (1968) Can. J. Biochem. 46, 1371-1379 Lemberg, R., and Cutler, M. E. (1970) Biochim. Biophys. Acta 197,l-10 Muijsers, A. O., Tiesjema, R. H., and Van Gelder, B. F. (1971) Biochim. Biophys. Acta 234,481-492 Tiesjema, R. H., Muijsers, A. O., and Van Gelder, B. F. (1972) Biochim. Biophys. Acta 256,32-42 Orii, Y., and King, T. E. (1976) J. Biol. Chem. 2 5 1 , 7487-7493 Nicholls, P. (1978) Biochem. J. 175,1147-1150 Naqui, A., Chance, B., and Cadenas, E. (1986) Annu. Reu. Biochem. 55, 137-166 Witt, S. N., Blair, D. F., and Chan, S. I. (1986) J. Biol. Chem. 261,8104-8107 Blair, D. F., Witt, S. N., and Chan, S. I. (1985) J . Am. Chem. SOC.107, 7389-7399 Hartzell, C.R., and Beinert, H. (1974) Biochim. Biophys. Acta 368,318-338 Dawson, R. M. C., Elliott, D. C., Elliott, W. H., and Jones, K. M. (eds) (1969) in Data for Biochemical Research, 2nd Ed., p. 441, Oxford University Press, New York, Oxford Aasa, R., and VLinngHrd, T. (1975) J. Magn. Reson. 19,308-315 Reinhammer, B., Malkin, R., Jensen, P. Karlsson, B., Andrbasson, L.-E., Aasa, R.,VanngHrd, T., and Malmstrom, B. G. (1980) J. Biol. Chem. 2 5 5 , 5000-5003 Wilson, M. T., Jensen, P., Aasa, R., Malmstrom, B. G., and VanngHrd, T. (1982) Biochem. J. 2 0 3 , 483-492 Wikstrom, M. (1981) Proc. Natl. Acad. Sci. U. S. A. 78,40514054