(Received for publication, July 16, 1996, and in revised form, October 10, 1996)
From the 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
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.
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.
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--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 AssaysStock CcO concentrations
were calculated from the reduced minus oxidized difference spectrum at
605 nm (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 TreatmentThe 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 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.
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%
-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.
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:
![]() |
(Eq. 1) |
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.
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.
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
g of about 2.19, an
A
of about 190 G, and a seven-line hyperfine
pattern at g
= 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.
|
Fig. 4 shows the EPR spectrum of the as-isolated UQO
complex. The type II copper signal in this spectrum is characterized by
a g of 2.19, an A
of about 190 G, and a seven-line hyperfine pattern at
g
= 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.
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 ENDORThe 35-GHz continuous-wave
1H ENDOR spectrum of the SB-12 heat-treated, 40.5-h
SB-12-dialyzed CcO complex obtained in the g 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 (
H
46.3 MHz) and split by a
hyperfine coupling constant A
10 MHz (Fig.
5A) as interpreted with the relation:
± (1H) = |
(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 (A
10 MHz) is
solvent-exchangeable (Fig. 5B) and is consistent with a
bound hydroxide anion.
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
"g" 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 g
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 g
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
+ branches of three
14N ligands, denoted as N1, N2, and N3, with
A
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 g
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
A
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.
|
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.
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 g 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 g of about 2.28 and an
A
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 g
of
2.26-2.28 and an A
of 102-137 G (26, 27).
High pH causes the appearance of a type II species with a
g
of 2.30 and and A
of 136 G (28). Addition of cyanide to the Thermus
thermophilus cytochrome ba3 complex
produces a type II copper species with a g
of 2.28 and an A
of about 140 G (29, 30). These spectra contrast with the type II signal obtained by SB-12 heat treatment which
has a larger A
(~190 G) and a smaller
g
(~2.19). As harsher treatments tend to
result in larger A
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
A
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, A = 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.
Special thanks to Mårten Wikström for sharing manuscripts before publication and to Michael Stowell for help in the isolation of the cytochrome bo3 complex.