Q-band Electron Nuclear Double Resonance (ENDOR) and X-band EPR of the Sulfobetaine 12 Heat-treated Cytochrome c Oxidase Complex*

(Received for publication, July 16, 1996, and in revised form, October 10, 1996)

Siegfried M. Musser Dagger §, Yang-Cheng Fann par , Ryszard J. Gurbiel **, Brian M. Hoffman and Sunney I. Chan Dagger Dagger Dagger

From the Dagger  Division of Chemistry and Chemical Engineering, Arthur Amos Noyes Laboratory of Chemical Physics, California Institute of Technology, Pasadena, California 91125 and the  Departments of Chemistry and Biochemistry, Molecular Biology, and Cellular Biology, Northwestern University, Evanston, Illinois 60208

ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
FOOTNOTES
Acknowledgments
REFERENCES


ABSTRACT

Heat treatment of the bovine cytochrome c oxidase complex in the zwitterionic detergent sulfobetaine 12 (SB-12) results in loss of subunit III and the appearance of a type II copper center as characterized by electron paramagnetic resonance (EPR) spectroscopy. Previous authors (Nilsson, T., Copeland, R. A., Smith, P. A., and Chan, S. I. (1988) Biochemistry 27, 8254-8260) have interpreted this type II copper center as a modified version of the CuA site. By using electron nuclear double resonance spectroscopy, it is found that the CuA proton and nitrogen resonances remain present in the SB-12 heat-treated enzyme and that three new nitrogen resonances appear having hyperfine coupling constants consistent with histidine ligation. These hyperfine coupling constants correlate well with those recently found for the CuB histidines from the cytochrome aa3-600 quinol oxidase from Bacillus subtilis (Fann, Y. C., Ahmed, I., Blackburn, N. J., Boswell, J. S., Verkhovskaya, M. L., Hoffman, B. M., and Wikström, M. (1995) Biochemistry 34, 10245-10255). In addition, the total EPR-detectable copper concentration per enzyme molecule approximately doubles upon SB-12 heat treatment. Finally, the observed type II copper EPR spectrum is virtually indistinguishable from the EPR spectrum of CuB of the as-isolated cytochrome bo3 complex from Escherichia coli. These data indicate that the type II copper species that appears results from a breaking of the strong antiferromagnetic coupling of the heme a3-CuB binuclear center.


INTRODUCTION

The aa3-type cytochrome c oxidase (CcO)1 complexes contain four redox-active metal ions, two hemes A (heme a and heme a3) and two copper centers (CuA and CuB). Electrons from ferrocytochrome c are input into the CcO complex at the CuA center; electron transfer to heme a or the heme a3-CuB binuclear site quickly follows electron input. The highly exergonic reduction of dioxygen to water at the heme a3-CuB binuclear center is coupled to the pumping of protons against an electrochemical gradient (1). While for a long time the CuA center was thought to consist of a single copper ion coordinated by two histidines and two cysteines, the recent literature (2, 3, 4, 5, 6, 7) provides a convincing argument in favor of the original hypothesis of Kroneck et al. (8) that the CuA site actually is a mixed-valence, binuclear copper center. The recent crystal structures of the CcO complex have confirmed a binuclear configuration with two bridging thiolates (9, 10).

The structural lability of the CuA site has been inferred from mild heat treatment, AgNO3 incubation, and p-(hydroxymercuri)benzoate (p-HMB) modification experiments which resulted in the appearance of type II copper EPR signals and diminution of the 830 nm absorption (thought to arise predominantly from the CuA site (11)) of the enzyme (12, 13, 14, 15, 16). A scheme which describes the perturbations to the CuA site under these various conditions has been postulated by Li et al. (14). The ligand rearrangements proposed by these workers must be re-evaluated, however, because they are based on a mononuclear structure for the CuA site. As type II signals certainly can arise from CuB under appropriate conditions, it should be noted that it has not been demonstrated definitively which copper ions are EPR-detectable as a result of these various modification procedures. The possibility exists that CuB becomes visible under one set of conditions, whereas the CuA site is modified under a different set of conditions. Also, perturbations to both copper sites can occur concurrently either by independent means or through allosteric interactions between the two copper sites. Electron nuclear double resonance (ENDOR) spectroscopy can be used to interpret ligand rearrangement reactions much more accurately than EPR and absorption spectroscopies. Thus, it was thought prudent to examine whether the type II signals that appear in the above modification experiments are accompanied by loss of the strongly coupled cysteinyl protons in the ENDOR spectrum associated with the CuA site. A change in the nitrogen ENDOR of the CuA center is certainly expected as well.

Proton and nitrogen ENDOR (35 GHz) are reported here for the bovine CcO complex after short, mild heat treatment in the zwitterionic detergent sulfobetaine 12 (SB-12), a procedure which results in a type II copper EPR signal that is very similar to that observed for the as-isolated Escherichia coli cytochrome bo3 complex, a structurally related ubiquinol oxidase complex. Special emphasis is placed on double integration of the first derivative 9.2-GHz EPR signals obtained. Integration of EPR signals is important for two reasons. 1) The model of a mixed-valence, binuclear CuA center makes possible a scenario in which modification of this center leads to oxidation of the cuprous copper. An increase in EPR intensity would result if the magnetic coupling between the two copper ions is broken; a decrease in EPR intensity is expected if the magnetic coupling remains. 2) Uncoupling of the heme a3-CuB binuclear center as a result of the SB-12 heat treatment procedure would make CuB EPR-detectable. The second explanation is most consistent with the data gathered in this study.


EXPERIMENTAL PROCEDURES

Enzyme Purification

The beef heart CcO complex was isolated essentially by the Hartzell and Beinert method (17). Mitochondria were lysed with 2.5 g of Triton X-114/g of protein in TEH buffer (20 mM Tris, 10 mM EDTA, 1 mM histidine, pH 8.0) supplemented with 200 mM KCl and centrifuged at 13,700 × g for 10 h. The pellet was washed three times with TEH buffer and solubilized with 2 g of potassium cholate (twice recrystallized)/g of protein in TEH buffer at a protein concentration of 40 mg/ml. The enzyme was precipitated with ammonium sulfate and the pellets were redissolved in 25 mM Tris, 0.1% n-dodecyl-beta -D-maltoside (DDM), 10 mM EDTA, pH 8.0 (dialysis buffer I). The enzyme was dialyzed (50- kDa membrane) against 50-100 volumes of dialysis buffer I and then 50-100 volumes of 25 mM Tris, 0.1% DDM, pH 8.0. After centrifugation at 32,500 × g, 4 °C for 30 min, the enzyme (150-250 µM) was aliquoted, frozen in liquid nitrogen, and stored at -80 °C.

Growth of E. coli cells and isolation of "His-tagged" bo3-type ubiquinol oxidase (UQO) complex was accomplished as described elsewhere (18).

Concentration and Activity Assays

Stock CcO concentrations were calculated from the reduced minus oxidized difference spectrum at 605 nm (Delta epsilon 605red-ox = 24 mM-1 cm-1) using sodium dithionite as the reductant. This method, however, was found to be unreliable for the various modified enzyme samples, and the pyridine hemochrome method (19) was used instead. A dual wavelength extinction coefficient of 46.4 ± 1.0 mM-1 cm-1 (mean ± S.D.) at 588-638 nm for the reduced minus oxidized pyridine hemochrome was estimated from 10 different determinations on five different batches of as-isolated enzyme. Turnover numbers were calculated from the initial rate of ferrocytochrome c (1-80 µM initial concentration) oxidation monitored optically (550 nm or 520 nm) in 100 mM sodium phosphate, 0.1% DDM, pH 7.4. The kcat and Km for a particular enzyme sample were obtained from Eadie-Hofstee plots.

Stock UQO concentrations and enzyme activity were calculated as described previously (18). The kinetic constants for the UQO complex used here were identical to those reported earlier (18).

SB-12 Heat Treatment

The SB-12 heat-treated CcO complex was prepared by a modification of the procedure of Nilsson and co-workers (12). Three parts stock enzyme was diluted with ten parts 15 mM SB-12, 100 mM sodium phosphate, 500 mM NaCl, pH 7.4, and incubated in a 40 °C water bath for 15 min. The solution was cooled on ice for 10 min and centrifuged at 32,500 × g, 4 °C for 30 min to remove large aggregates. After dialysis (50-kDa membrane) at 8 °C against 30-40 volumes of 5 mM SB-12, 10 mM Tris, pH 8.0, samples were applied to a short (2.5-3.5 × 2.5 cm) DE52 (Whatman) column equilibrated with 10 mM Tris, 0.1% DDM, pH 8.0. The column was then washed with 50 ml of 10 mM Tris, 0.1% DDM, pH 8.0, and the enzyme was eluted with 200 mM NaCl, 25 mM Tris, 0.1% DDM, pH 8.0. The fractions containing enzyme were pooled and concentrated to greater than 150 µM using Centricon-100s (Amicon). After dialysis (50-kDa membrane) at 8 °C against about 100 volumes of 25 mM Tris, 0.1% DDM, pH 8.0, for 4-10 h, the enzyme was centrifuged at 32,500 × g, 4 °C for 30 min, frozen in liquid nitrogen, and stored at -80 °C until use.

Subunit III Depletion

Subunit III was removed from the CcO complex using high detergent and salt concentrations. One part stock enzyme was diluted with six parts 5% Triton X-100, 300 mM Tris-HCl, 50 mM EDTA, pH 8.5, and incubated at room temperature for 20 h. After 10-fold dilution with distilled water, the enzyme was applied to a DE52 column equilibrated with 10 mM Tris, 0.1% DDM, pH 8.0. The column was then washed with 100 ml 10 mM Tris, 0.1% DDM, pH 8.0, and the enzyme was eluted with 200 mM NaCl, 25 mM Tris, 0.1% DDM, pH 8.0. The fractions containing enzyme were pooled and concentrated to greater than 150 µM using Centricon-100s. After dialysis (50-kDa membrane) at 8 °C against about 100 volumes of 25 mM Tris, 0.1% DDM, pH 8.0, for 4-10 h, the enzyme was centrifuged at 32,500 × g, 4 °C for 30 min, frozen in liquid nitrogen, and stored at -80 °C until use.

Electrophoresis

The absence or presence of subunit III in various enzyme samples was determined by SDS-polyacrylamide gel electrophoresis. Enzyme samples were dissociated using 250 mM Tris-HCl, 8 M urea, 3.3% beta -mercaptoethanol, 5% SDS, 15% glycerol, pH 6.2, at room temperature for about 1 h and run on a Hoefer SE 250 vertical electrophoresis unit using a 7.2%/0.19% (acrylamide/bisacrylamide) stacking gel, a 14%/0.37% running gel, and a running buffer of 20 mM Tris, 240 mM glycine, 0.1% SDS, pH 8.4. Gels were stained with Coomassie Blue.

Instrumentation

Optical absorption spectra and kinetic measurements were obtained with Hewlett-Packard 8452 or 8453 diode array UV/Vis spectrophotometers. The X-band EPR spectra were recorded using a Varian E-109 spectrometer equipped with a Varian E-231 TE 102 rectangular cavity. The modulation frequency used was 100 kHz, and temperature was controlled with a helium cryostat (Oxford Instruments) or liquid nitrogen finger Dewar (Wilmad). The ENDOR spectra were recorded on a modified Varian E-109 EPR spectrometer equipped with an E-110 35-GHz microwave bridge using 100-kHz field modulation as described previously (20). To a first approximation, the ENDOR spectrum for a single orientation of a nucleus (N) of spin I consists of 2I transitions at frequencies given by:
&ngr;<SUB>±</SUB>(m<SUB>I</SUB>)=&cjs3539;A<SUB><UP>N</UP></SUB>/2±&ngr;<SUB><UP>N</UP></SUB>+<FR><NU>3</NU><DE>2</DE></FR> P<SUB><UP>N</UP></SUB> (2m<SUB>I</SUB>−1)&cjs3539;   −I+1≤m<SUB>I</SUB>≤I (Eq. 1)
where AN and PN are orientation-dependent hyperfine and quadrupole coupling constants and nu N is the nuclear Larmor frequency (21). All ENDOR spectra were taken with a radiofrequency source mixing with a noise source generator described elsewhere to enhance the ENDOR response (22).

Data Analysis

Double integration of first derivative X-band EPR spectra was done using the baseline correction and integration capabilities of Lab CalcTM. The EPR visible copper concentration was calculated from double integrals of standard solutions of CuSO4 in 100 mM imidazole, pH 8.4, or in 100 mM histidine, pH 8.0, under the same experimental conditions.


RESULTS

Development of a Type II Copper EPR Signal

As shown in Fig. 1A, SB-12 heat treatment of the CcO complex results in the development of a type II copper EPR signal. The appearance of this type II copper EPR signal is accompanied by a high-spin heme signal in the g = 6 (~1100 G) region of the spectrum. This high-spin heme signal is best seen at low temperature (Fig. 1B) where line-broadening is less of a problem. Curiously, the development of this type II EPR signal is critically dependent on the length of the first dialysis step (see "Experimental Procedures"). Double integration and correction for spectrometer gain and enzyme concentration as determined by the pyridine hemochrome technique reveals that the amount of EPR-visible copper approximately doubles after 40 h of dialysis (Fig. 1, inset). The integration error is estimated to be in the range of 10-20%; this error arises from the protein concentration estimate, the baseline correction routine, and the fact that there are small heme signals in the g = 2 region even at 80 K. Copper standards were used to estimate the EPR-visible copper in the native enzyme. As expected, the ratio of EPR-visible copper to enzyme complex is approximately unity for native enzyme; therefore, this was assumed to be the case, and the integration of modified enzyme was standarized to this value. There appear to be two "isosbestic points" (at 2938 and 3059 G, data not shown); these isosbestic points suggest that a single process occurs to produce the type II signal.


Fig. 1. A, the EPR spectra of the SB-12 heat-treated, 40-h SB-12-dialyzed CcO complex (top) and the native CcO complex (bottom) at 80 K (average of four scans). B, the EPR spectra of the same two samples at 10 K. Spectra are scaled to protein concentration based on the pyridine hemochrome method, and the buffer is 25 mM Tris, 0.1% DDM, pH 8.0. The inset shows the linear dependence of the integrated intensity (77 K) on dialysis time (see "Results" for details). Spectrometer settings: modulation amplitude, 10 G; time constant, 32 ms; power, 1 milliwatt; microwave frequency, 9.22 GHz; scan rate, 8.3 G/s.
[View Larger Version of this Image (14K GIF file)]


Various attempts were made to determine what step of the SB-12 heat treatment procedure is responsible for the perturbations to the EPR spectrum of the enzyme. Gel electrophoresis reveals that subunit III is lost during the SB-12 heat treatment procedure. Subunit III can be removed by high salt and detergent concentration (see "Experimental Procedures"). In and of itself, it is found that subunit III depletion is insufficient to yield the large type II copper signal shown in Fig. 1. However, a small amount of type II copper signal is sometimes apparent in the EPR spectrum of the subunit III-depleted CcO complex; this signal intensifies after 40 h of dialysis against 10 mM Tris, 0.1% DDM, pH 8.0 at 8 °C. Dialysis of the native enzyme itself against this same buffer results in the appearance of a small amount of the type II copper signal. A substantial type II copper signal develops after 40 h of dialysis of the CcO complex in SB-12 buffer without a 15-min heat treatment. Typical results of these experiments are tabulated in Table I. The catalytic activity (kcat) of the enzyme decreases as a result of subunit III depletion, SB-12 heat treatment, and/or extensive dialysis, but the Km for cytochrome c binding is virtually unaffected. As the decreased kcat (20-80% of wild-type) indicates that one or more electron transfer pathways have been perturbed, it is possible that the structural modification to the copper center that gives rise to the type II signal is largely responsible for the decreased catalytic activity. However, we cannot rule out the possibility that an electron transfer path topologically distant from the modified copper site has been perturbed. For example, a structurally modified heme a3-CuB binuclear site and a disrupted CuA electron input pathway is consistent with the EPR and turnover data. The optical spectra shown in Fig. 2 reveal that there is very little perturbation to the 830 nm absorption band except in the case of the SB-12-treated samples. In general, the 830 nm absorption intensity diminishes roughly according to time in contact with SB-12. No correlation is found, however, between the 830 nm absorption intensity and the appearance of the type II copper signal. In fact, a substantial type II copper signal is seen for the subunit III-depleted, 40-h DDM-dialyzed enzyme, yet no diminution of the 830 nm absorption intensity is apparent. As shown in Fig. 3, the type II copper signal is characterized by a gpar of about 2.19, an Apar of about 190 G, and a seven-line hyperfine pattern at gperp  = 2.06 with a splitting of about 15 G. When analyzing the ENDOR data, it is important to note that at fields at or above that corresponding to g ~2.2, EPR signals of the CuA site and the type II copper species overlap (see Fig. 3). At lower fields, only the type II copper species can contribute to the ENDOR spectrum.

Table I.

EPR-visible copper resulting from various treatments of the CcO complex


Subunit III present?a Km kcat EPR-visible copper (80 K)

µM s-1
Native Yes 47 230 1.00
Subunit III-depletedb No 56 165 1.11
40-h DDM-dialyzedc Yes 52 182 1.25
40-h DDM-dialyzed,c subunit III-depletedb No 49 152 1.49
40-h SB-12-dialyzed No 49 128 1.81
SB-12 heat-treated, 40-h SB-12 dialyzed No 60 47 2.25

a  An entry of "no" implies that greater than 90% of subunit III is missing as determined by SDS-polyacrylamide gel electrophoresis.
b  The phrase "subunit III-depleted" refers to the high detergent, high salt treatment described under "Experimental Procedures."
c  The phrase "DDM-dialyzed" denotes dialysis against 10 mM Tris, 0.1% DDM, pH 8.0, at 8 °C followed by the short DE52 column treatment described for the SB-12 heat-treated enzyme.


Fig. 2. UV-visible absorption spectra of the six samples listed in Table I. Native, ----; subunit III-depleted, - ·- ·-; 40-h DDM-dialyzed, - - - - -; subunit III-depleted, 40-h DDM-dialyzed, ---···; 40-h SB-12-dialyzed, - - -; SB-12 heat-treated, 40-h SB-12-dialyzed, ·····. The protein concentration in each case is 2.5 µM in 25 mM Tris, 0.1% DDM, pH 8.0.
[View Larger Version of this Image (25K GIF file)]



Fig. 3. The EPR spectra at 80 K of the six samples listed in Table I. From top to bottom: native; subunit III-depleted; 40-h DDM-dialyzed; subunit III-depleted, 40-h DDM-dialyzed; 40-h SB-12-dialyzed; SB-12 heat-treated, 40-h SB-12 dialyzed. Spectra are scaled to protein concentration based on the pyridine hemochrome method. The inset shows the seven-line hyperfine pattern for the SB-12 heat-treated, 40-h SB-12 dialyzed enzyme. The bottom panel shows the difference spectrum between the SB-12 heat-treated, 40-h SB-12 dialyzed and the native samples. Note that at lower fields (<g ~2.2), only the type II copper species can contribute to the ENDOR spectrum; at higher fields, both the native CuA center and the type II species can contribute. Spectrometer settings are the same as in Fig. 1 except for a modulation amplitude of 5 G for the inset spectrum (average of two scans).
[View Larger Version of this Image (24K GIF file)]


Fig. 4 shows the EPR spectrum of the as-isolated UQO complex. The type II copper signal in this spectrum is characterized by a gpar of 2.19, an Apar of about 190 G, and a seven-line hyperfine pattern at gperp  = 2.04 with a splitting of about 15 G. At liquid helium temperatures, a substantial high-spin heme signal is present (data not shown) indicating that the heme o3-CuB binuclear center is magnetically uncoupled.


Fig. 4. The EPR spectrum at 80 K of the as-isolated cytochrome bo3 complex. Spectrometer settings are the same as in Fig. 1 except for a time constant of 64 ms (average of 16 scans). Compare with Fig. 3.
[View Larger Version of this Image (19K GIF file)]


No attempt was made to reverse the structural changes that give rise to the type II copper EPR signal, and no diminution of this EPR signal was ever observed once it appeared. However, extensive efforts in the past to reverse the type II modification were unsuccessful.2

1H ENDOR

The 35-GHz continuous-wave 1H ENDOR spectrum of the SB-12 heat-treated, 40.5-h SB-12-dialyzed CcO complex obtained in the gpar region of the type II copper species at fields low enough that the CuA site cannot contribute shows a doublet centered at the proton Larmor frequency (nu H approx 46.3 MHz) and split by a hyperfine coupling constant Apar approx 10 MHz (Fig. 5A) as interpreted with the relation: nu ± (1H) = |nu (1H) ± A(1H)/2|. This proton, which is not seen for the native enzyme, is not associated with the hemes because the hyperfine-split doublet vanishes when the spectrum is taken outside the type II copper EPR envelope. In addition, the proton ENDOR signals associated with the CuA center are still present and remain unperturbed at all fields that fall within the CuA EPR envelope indicating this site remains intact under the experimental conditions (data not shown). The proton ENDOR spectrum is very similar to that reported by Fann et al. (23) for the CuB site of the cytochrome aa3-600 ubiquinol oxidase complex isolated from Bacillus subtilis (Fig. 5C). As is the case with the Bacillus enzyme, this strongly coupled proton (Apar approx 10 MHz) is solvent-exchangeable (Fig. 5B) and is consistent with a bound hydroxide anion.


Fig. 5. Comparison of 1H ENDOR spectra of the SB-12 heat-treated, 40-h SB-12 dialyzed CcO complex (35.04 GHz, gpar approx 2.20) and the cytochrome aa3-600 ubiquinol oxidase complex (35 GHz, gpar approx 2.23). The spectra from both enzymes show a hyperfine-split doublet centered at the proton Larmor frequency that vanishes upon exchange into D2O buffer (shown here for the modified CcO complex, demonstrated previously for the cytochrome aa3-600 complex (23)). Experimental conditions: modulation amplitude, 0.6 G; scan rate, 1 MHz/s; temperature, 2 K (average of 50 scans). The cytochrome aa3-600 ubiquinol oxidase spectrum is reproduced from Fann et al. (23).
[View Larger Version of this Image (16K GIF file)]


14N ENDOR

Fig. 6, top, shows the ENDOR spectrum of the SB-12 heat-treated, 40.5-h SB-12-dialyzed CcO complex obtained in the "gpar " region of the type II copper species where the CuA center does not resonate. Also shown in Fig. 6, bottom, is the ENDOR spectrum obtained near g ~2.04, which is near gperp for both the type II copper species and the CuA site; thus, each copper center can contribute to the ENDOR spectrum at this higher field. The observed resonance frequencies are those expected for 14N ligands coordinated to copper where each ligand yields an ENDOR pattern as described by Equation 1. In the gpar spectrum (Fig. 6, top), only resonances arising from the type II copper species are expected; ENDOR peaks with frequencies greater than 10 MHz can be assigned to nu + branches of three 14N ligands, denoted as N1, N2, and N3, with Apar hyperfine couplings of 38, 24, and 17 MHz, respectively. No quadrupole splittings were resolved. The resonance at 7 MHz (marked with an asterisk) is associated with heme nitrogens as the signal persists outside the type II copper EPR envelope. The 14N ENDOR peaks from the three distinct type II copper nitrogens can be unambiguously tracked in the field-dependent ENDOR spectra taken as the field increases from gpar into the CuA region (2.2 g < 1.94), where the 14N ENDOR pattern becomes more complicated due to the overlap of the 14N signals from the native CuA center and the type II copper species (Fig. 6, bottom). Although the Aperp values for N1, N2, and N3 are not precisely measurable, they fall within the range of those from 14N ligands to CuB of the cytochrome aa3-600 ubiquinol oxidase complex (23). The 14N ENDOR spectra from this ubiquinol oxidase complex and those reported here for the SB-12 heat-treated CcO complex are compared in Fig. 6; hyperfine coupling constants are summarized in Table II.


Fig. 6. Comparison of 14N ENDOR spectra of the SB-12 heat-treated, 40-h SB-12 dialyzed CcO complex (35.04 GHz, gpar approx 2.22 ("low field"), gperp  = 2.04 ("high field")), the native CcO complex (35.1 GHz, gperp  = 2.04), and the cytochrome aa3-600 ubiquinol oxidase complex (35.2 GHz, gpar approx 2.23 ("low field"), gperp = 2.05 ("high field")). Experimental conditions are the same as for Fig. 5 except: modulation amplitude, 2.5 G (CcO)/1.25 G (aa3-600); scan rate, 2 MHz/s. The cytochrome aa3-600 spectra are reproduced from Fann et al. (23).
[View Larger Version of this Image (21K GIF file)]


Table II.

14N hyperfine coupling constants (in MHz) for the type II copper center in the SB-12 heat-treated CcO complex and for CuB of the B. subtilis cytochrome aa3-600 ubiquinol oxidase complex

Uncertainty in A: ±3 MHz. Values for the cytochrome aa3-600 ubiquinol oxidase complex were originally reported by Fann et al. (23).
N1 N2 N3

SB-12 heat-treated CcO complex Apar 38 24 17
Cytochrome aa3-600 complex Apar 37 25 17
Aperp 42 31 20


DISCUSSION

The appearance of a type II copper species as a consequence of SB-12 heat treatment of the CcO complex has been reported previously by Nilsson et al. (12). In addition, p-HMB treatment or incubation with AgNO3 has been found to cause the appearance of this type II species (13, 15, 16). In terms of EPR characterization, the type II copper species we report here appears to be identical to the one described by these authors. This type II species has been interpreted to result from modification of the CuA site of the CcO complex. In the scheme of Li et al. (14), CuA is assumed to be mononuclear with two nitrogen and two cysteine ligands. Mild heat treatment in DDM results in loss of one of the cysteine ligands thereby yielding a type I (blue) copper center. Heat treatment with SB-12 or treatment with p-HMB or AgNO3 causes deligation of the second cysteine and the ligation of at least one more histidine to produce a type II copper center. The multifrequency EPR work of Kroneck and co-workers (2, 3, 4, 5, 6, 7, 8, 9, 10), biophysical data on overexpressed subunit II fragments, and the recent crystal structures of the CcO complex indicate, however, that the CuA center is actually a mixed-valence, binuclear copper center with two bridging thiolates (Fig. 7). The CuA modification scenario of Li et al. (14) obviously conflicts with these more recent data.


Fig. 7. Left, schematic of the binuclear nature of the CuA site as revealed by x-ray crystallography (9, 10). The fourth ligand for the lower copper is the main chain carbonyl of a glutamic acid residue. Right, possible perturbation to the CuA site that results in the appearance of a type I EPR signal (lower copper atom) and loss in 830 nm absorption intensity. See "Discussion" for details. Solid lines denote strong bonds; dashed lines denote weak ligation.
[View Larger Version of this Image (10K GIF file)]


The explanation for the appearance of the type II signal upon SB-12 heat treatment must be re-evaluated in light of the binuclear nature of the CuA center and the data presented here. There are two possibilities. The first is that the SB 12 heat treatment procedure causes disruption of the CuA site; the approximate doubling of the copper EPR intensity arises from full oxidation of the two copper ions. The second possibility is that the CuA site remains in a native configuration and the magnetic coupling of the heme a3-CuB binuclear center is broken making CuB EPR-visible. In this scenario, the extra EPR-visible copper, all of which appears to be in a type II configuration, is CuB. The strongest evidence in support of the first explanation is the diminution in the 830 nm absorption intensity, which is believed to arise predominantly from the CuA center (11), upon SB-12 heat treatment of the enzyme. It has been found here, however, that the type II EPR signal can be created without any apparent loss in the 830 nm absorption intensity (Fig. 2). In addition, 1H and 14N ENDOR spectroscopy reveals that the CuA resonances are present when a stoichiometric amount of the type II copper species is EPR-detectable. Instead, three additional 14N ENDOR resonances are detected (Fig. 6). The seven-line hyperfine pattern in the gperp region of the EPR spectrum (Fig. 3) is consistent with three approximately equivalent nitrogen ligands. The excellent agreement between the hyperfine coupling constants of the type II center described here and those for the CuB center of the cytochrome aa3-600 ubiquinol oxidase complex of B. subtilis (Table II), which does not contain a CuA center, allows confident assignment of the three new nitrogen ENDOR resonances as CuB histidines. Finally, the EPR spectrum of the as-isolated UQO complex contains the identical type II copper species that results upon SB-12 heat treatment of the CcO complex. As the UQO complex does not contain a CuA center but is otherwise structurally similar to the CcO complex (1), and, in this particular sample, the heme o3-CuB binuclear center is magnetically uncoupled, these data provide strong confirmation that the type II signals that appear as a result of the various CcO treatments discussed here do indeed arise from CuB. The changes in the resonance Raman spectrum of heme a3 when the type II copper signal is apparent in the CcO EPR spectrum are now more easily understood (12, 24). The perturbations that uncouple the heme a3-CuB binuclear center and make CuB EPR-detectable also disrupt the heme a3 pocket.

Given that the type II EPR signal seen arises predominantly from CuB, it is important to question to what extent the structural integrity of the CuB redox site is preserved. Cline and co-workers (25) report an EPR spectrum of CuB obtained by full reduction of the enzyme, flushing with O2, and quick freezing in liquid N2. This spectrum is characterized by a gpar of about 2.28 and an Apar of about 100 G. Photolysis and freezing of the fully reduced CO-bound CcO complex in the presence of oxygen has been found to yield a type II copper signal with a gpar of 2.26-2.28 and an Apar of 102-137 G (26, 27). High pH causes the appearance of a type II species with a gpar of 2.30 and and Apar of 136 G (28). Addition of cyanide to the Thermus thermophilus cytochrome ba3 complex produces a type II copper species with a gpar of 2.28 and an Apar of about 140 G (29, 30). These spectra contrast with the type II signal obtained by SB-12 heat treatment which has a larger Apar (~190 G) and a smaller gpar (~2.19). As harsher treatments tend to result in larger Apar values, the indication is that the structural integrity of the CuB site has been perturbed in the experiments described here. A small perturbation to the heme a3-CuB site clearly must have occurred to break the magnetic coupling in this binuclear center. Both the SB-12 heat-treated CcO and the cytochrome aa3-600 complexes contain a proton with an Apar of about 10 MHz (Fig. 5). In the case of the cytochrome aa3-600 complex, this proton is solvent-exchangeable and has been tentatively assigned as arising from a bound hydroxide anion (23). A hydroxide anion could certainly be the fourth ligand for the type II copper site reported here.

The finding that SB-12 heat treatment makes CuB EPR-detectable raises the issue of which copper ions give rise to the type II signals seen upon AgNO3 or p-HMB treatment (13, 15, 16). Since ENDOR experiments have not been attempted on such enzyme samples, a definitive conclusion cannot be made here. It is certainly possible that these treatments affect the CuA site exclusively since they are thiol-specific reagents. Gelles and Chan (13) observed an approximate 40% increase in copper EPR intensity upon p-HMB treatment of the enzyme. This observation is consistent with a scenario where the type II copper signals arise from CuB due to perturbations similar to those caused by SB-12 heat treatment. On the other hand, disruption of the CuA center and oxidation of the cuprous copper would also result in a similar increase in the integrated EPR intensity.

While the ENDOR data reveal the presence of native CuA resonances in the presence of stoichiometric amounts of type II copper, the possibility that a substoichiometric population of CuA sites has been perturbed cannot be ruled out. The creation of a small amount of a type II copper species from severe disruption of the CuA ligation structure would be undetectable in the EPR and ENDOR spectra due to the strong signals arising from the CuB center. The diminution in the 830 nm absorption intensity and the slightly larger than stoichiometric increase in the integrated intensity of the copper EPR signals are consistent with this possibility.

On the other hand, less drastic perturbations to the CuA site could occur. Li et al. (14) found that heat treatment in DDM results in substoichiometric populations of both type I and type II copper species. Zickermann et al. (31) have recently found that mutation of the weakly coordinating methionine of the binuclear CuA center to isoleucine results in a type I copper EPR spectrum (g| = 2.18, Apar  = 61 G) and loss of 830 nm absorption intensity. This type I species could arise in the native enzyme upon dissociation of the weak Met-Cu bond and lengthening of one of the Cu-S bonds as shown in Fig. 7. As a result of this conformational rearrangement, one of the copper atoms would have a type I-like structure (one His, one Cys, one carbonyl, and a long Cu-S bond). The other copper atom would have a high reduction potential due to the two strongly coordinating cysteines and, therefore, would become completely reduced with the electron initially shared by the two atoms. As only the type I-like copper would be EPR-visible, the EPR integration would remain constant. The loss in 830 nm absorption intensity reported here for some samples may result from a CuA rearrangement of this type. The associated type I EPR signals are likely to be difficult to detect in the presence of strong type II signals and need not be identical to those resulting from the methionine mutation structure (31). The well-known interaction potentials in the CcO complex (1) indicate the presence of conformational interactions between redox centers. Therefore, the disruption of the heme a3-CuB binuclear site that produces a type II signal could potentially, under some conditions, perturb the methionine of the CuA center resulting in loss of 830 nm absorption intensity. In fact, during enzyme turnover, the CuA reduction potential could be modulated through allosteric interactions between the heme a3-CuB binuclear site and the CuA methionine.

To summarize, the CuB center undergoes a small structural modification and becomes EPR-detectable as a type II copper species upon SB-12 heat treatment. The thiol-specific reagents AgNO3 and p-HMB, both of which cause the appearance of a type II copper species, are not expected to affect the CcO complex in the same manner. However, it is feasible that the type II species resulting from treatment with these reagents also is a perturbed heme a3-CuB center and not a drastically modified CuA site. It remains possible that SB-12 heat treatment perturbs a small population of CuA sites since this procedure results in a decrease in the 830 nm absorption intensity of the enzyme, perhaps through allosteric interactions with the methionine of the CuA site. It has been demonstrated, however, that, under some conditions, CuB of the fully oxidized CcO complex is EPR-detectable in the form of a type II copper center.


FOOTNOTES

*   This work was supported by National Institutes of Health Grants HL13531 (to B. M. H.) and GM22432 (to S. I. C.) from the United States Public Health Service. The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
§   Recipient of a National Research Service Predoctoral Award. Present address: Division of Biological Sciences, Section of Plant Biology, University of California, Davis, CA 95616.
par    Present address: Center for Molecular Design, Institute for Biomedical Computing, Washington University, Box 1099, St. Louis, MO 63130.
**   On leave from the Institute of Molecular Biology, Jagiellonian University, Krakow, Poland.
Dagger Dagger    To whom correspondence should be addressed. Tel.: 818-395-6508; Fax: 818-578-0471; E-mail: ChanS{at}starbase1.caltech.edu.
1    The abbreviations used are: CcO, cytochrome c oxidase, specifically the mitochondrial aa3-type enzyme; p-HMB, p-(hydroxymercuri)benzoate; ENDOR, electron nuclear double resonance; SB-12, sulfobetaine 12; DDM, n-dodecyl-beta -D-maltoside; UQO, ubiquinol oxidase, specifically the cytochrome bo3 complex from E. coli.
2    P. M. Li and S. I. Chan, unpublished observations.

Acknowledgments

Special thanks to Mårten Wikström for sharing manuscripts before publication and to Michael Stowell for help in the isolation of the cytochrome bo3 complex.


REFERENCES

  1. Musser, S. M., Stowell, M. H. B., and Chan, S. I. (1995) Adv. Enzymol. Related Areas Mol. Biol. 71, 79-208 [Medline] [Order article via Infotrieve]
  2. Scott, R. A., Zumft, W. G., Coyle, C. L., and Dooley, D. M. (1989) Proc. Natl. Acad. Sci. U. S. A. 86, 4082-4086 [Abstract]
  3. Antholine, W. E., Kastrau, D. H. W., Steffens, G. C. M., Büse, G., Zumft, W. G., and Kroneck, P. M. H. (1992) Eur. J. Biochem. 209, 875-881 [Abstract]
  4. Fee, J. A., Sanders, D., Slutter, C. E., Doan, P. E., Aasa, R., Karpefors, M., and Vänngård, T. (1995) Biochem. Biophys. Res. Commun. 212, 77-83 [CrossRef][Medline] [Order article via Infotrieve]
  5. Lappalainen, P., Aasa, R., Malmström, B. G., and Saraste, M. (1993) J. Biol. Chem. 268, 26416-26421 [Abstract/Free Full Text]
  6. van der Oost, J., Lappalainen, P., Musacchio, A., Warne, A., Lemieux, L., Rumbley, J., Gennis, R. B., Aasa, R., Pascher, T., Malmström, B. G., and Saraste, M. (1992) EMBO J. 11, 3209-3217 [Abstract]
  7. Kelly, M., Lappalainen, P., Talbo, G., Haltia, T., van der Oost, J., and Saraste, M. (1993) J. Biol. Chem. 268, 16781-16787 [Abstract/Free Full Text]
  8. Kroneck, P. M. H., Antholine, W. A., Reister, J., and Zumft, W. G. (1988) FEBS Lett. 242, 70-74 [CrossRef][Medline] [Order article via Infotrieve]
  9. Tsukihara, T., Aoyama, H., Yamashita, E., Tomizaki, T., Yamaguchi, H., Shinzawa-Itoh, K., Nakashima, R., Yaono, R., and Yoshikawa, S. (1995) Science 269, 1069-1074 [Medline] [Order article via Infotrieve]
  10. Iwata, S., Ostermeier, C., Ludwig, B., and Michel, H. (1995) Nature 376, 660-669 [CrossRef][Medline] [Order article via Infotrieve]
  11. Beinert, H., Shaw, R. W., Hansen, R. E., and Hartzell, C. R. (1980) Biochim. Biophys. Acta 591, 458-470 [Medline] [Order article via Infotrieve]
  12. Nilsson, T., Copeland, R. A., Smith, P. A., and Chan, S. I. (1988) Biochemistry 27, 8254-8260 [Medline] [Order article via Infotrieve]
  13. Gelles, J., and Chan, S. I. (1985) Biochemistry 24, 3963-3972 [Medline] [Order article via Infotrieve]
  14. Li, P. M., Morgan, J. E., Nilsson, T., Ma, M., and Chan, S. I. (1988) Biochemistry 27, 7538-7546 [Medline] [Order article via Infotrieve]
  15. Chan, S. I., Bocian, D. F., Brudvig, G. W., Morse, R. H., and Stevens, T. H. (1978) in Frontiers of Biological Energetics (Dutton, P. L., Leigh, J. S., Jr., and Scarpa, A., eds), Vol. 2, pp. 883-888, Academic Press, New York
  16. Chan, S. I., Bocian, D. F., Brudvig, G. W., Morse, R. H., and Stevens, T. H. (1979) in Cytochrome Oxidase (King, T. E., Orii, Y., Chance, B., and Okunuki, K., eds), pp. 177-188, Elsevier, New York
  17. Hartzell, C. R., and Beinert, H. (1974) Biochim. Biophys. Acta 368, 318-338 [Medline] [Order article via Infotrieve]
  18. Musser, S. M., Stowell, M. H. B., Lee, H. K., Rumbley, J. N., Gennis, R. B., and Chan, S. I. (1997) Biochemistry, in press
  19. Berry, E. A., and Trumpower, B. L. (1987) Anal. Biochem. 161, 1-15 [Medline] [Order article via Infotrieve]
  20. Werst, M. M., Davoust, C. E., and Hoffman, B. M. (1991) J. Am. Chem. Soc. 113, 1533-1538
  21. Hoffman, B. M., Derose, V. J., Doan, P. E., Gurbiel, R. J., Houseman, A. L. P., and Telser, J. (1993) in Biological Magnetic Resonance: EMR of Paramagnetic Molecules (Berliner, L. J., and Reuban, J., eds), Vol. 13, pp. 151-218, Plenum Press, New York
  22. Hoffman, B. M., Derose, V. J., Ong, J. L., and Davoust, C. E. (1994) J. Magn. Reson. 110, 52-57 [CrossRef]
  23. Fann, Y. C., Ahmed, I., Blackburn, N. J., Boswell, J. S., Verkhovskaya, M. L., Hoffman, B. M., and Wikström, M. (1995) Biochemistry 34, 10245-10255 [Medline] [Order article via Infotrieve]
  24. Larsen, R. W., Ondrias, M. R., Copeland, R. A., Li, P. M., and Chan, S. I. (1989) Biochemistry 28, 6418-6422 [Medline] [Order article via Infotrieve]
  25. Cline, J., Reinhammar, B., Jensen, P., Venters, R., and Hoffman, B. M. (1983) J. Biol. Chem. 258, 5124-5128 [Abstract/Free Full Text]
  26. Reinhammar, B., Malkin, R., Jensen, P., Karlsson, B., Andréasson, L.-E., Aasa, R., Vänngård, T., and Malmström, B. G. (1980) J. Biol. Chem. 255, 5000-5003 [Abstract/Free Full Text]
  27. Blair, D. F., Witt, S. N., and Chan, S. I. (1985) J. Am. Chem. Soc. 107, 7389-7399
  28. Carithers, R. P., and Palmer, G. (1981) J. Biol. Chem. 256, 7967-7976 [Abstract/Free Full Text]
  29. Surerus, K. K., Oertling, W. A., Fan, C., Gurbiel, R. J., Einarsdóttir, Ó., Antholine, W. E., Dyer, R. B., Hoffman, B. M., Woodruff, W. H., and Fee, J. A. (1992) Proc. Natl. Acad. Sci. U. S. A. 89, 3195-3199 [Abstract]
  30. Oertling, W. A., Surerus, K. K., Einarsdóttir, Ó., Fee, J. A., Dyer, R. B., and Woodruff, W. H. (1994) Biochemistry 33, 3128-3141 [Medline] [Order article via Infotrieve]
  31. Zickermann, V., Verkhovsky, M., Morgan, J., Wikström, M., Anemüller, S., Bill, E., Steffens, G. C. M., and Ludwig, B. (1995) Eur. J. Biochem. 234, 686-693 [Abstract]

©1997 by The American Society for Biochemistry and Molecular Biology, Inc.