doi:10.1016/j.jmb.2008.09.074
J. Mol. Biol. (2008) 384, 865–877
Available online at www.sciencedirect.com
A D-Pathway Mutation Decouples the Paracoccus
denitrificans Cytochrome c Oxidase by Altering
the Side-Chain Orientation of a Distant
Conserved Glutamate
Katharina L. Dürr 1 ⁎, Juergen Koepke 2 , Petra Hellwig 3 ,
Hannelore Müller 2 , Heike Angerer 2 , Guohong Peng 2 , Elena Olkhova 2 ,
Oliver-Matthias H. Richter 1 , Bernd Ludwig 1,4 and Hartmut Michel 2,4
1
Department of Biochemistry,
Molecular Genetics Group,
Johann Wolfgang Goethe
University, Max-von-Laue-Str. 9,
D-60438 Frankfurt/Main,
Germany
2
Department of Molecular
Membrane Biology, Max Planck
Institute of Biophysics,
Max-von-Laue-Str. 3, D-60438
Frankfurt/Main, Germany
3
Institute of Biophysics, Johann
Wolfgang Goethe University,
Max-von-Laue-Str. 1, D-60438
Frankfurt/Main, Germany
4
Cluster of Excellence Frankfurt
Macromolecular Complexes,
Johann Wolfgang Goethe
University, Max-von-Laue-Str. 3,
D-60438 Frankfurt/Main,
Germany
Asparagine 131, located near the cytoplasmic entrance of the D-pathway in
subunit I of the Paracoccus denitrificans aa3 cytochrome c oxidase, is a residue
crucial for proton pumping. When replaced by an aspartate, the mutant
enzyme is completely decoupled: while retaining full cytochrome c oxidation activity, it does not pump protons. The same phenotype is observed for
two other substitutions at this position (N131E and N131C), whereas a
conservative replacement by glutamine affects both activities of the enzyme.
The N131D variant oxidase was crystallized and its structure was solved to
2.32-Å resolution, revealing no significant overall change in the protein
structure when compared with the wild type (WT), except for an alternative
orientation of the E278 side chain in addition to its WT conformation.
Moreover, remarkable differences in the crystallographically resolved chain
of water molecules in the D-pathway are found for the variant: four water
molecules that are observed in the water chain between N131 and E278 in the
WT structure are not visible in the variant, indicating a higher mobility of
these water molecules. Electrochemically induced Fourier transform infrared
difference spectra of decoupled mutants confirm that the protonation state of
E278 is unaltered by these mutations but indicate a distinct perturbation in
the hydrogen-bonding environment of this residue. Furthermore, they
suggest that the carboxylate side chain of the N131D mutant is deprotonated.
These findings are discussed in terms of their mechanistic implications for
proton routing through the D-pathway of cytochrome c oxidase.
© 2008 Elsevier Ltd. All rights reserved.
Received 27 June 2008;
received in revised form
28 August 2008;
accepted 17 September 2008
Available online
9 October 2008
Edited by R. Huber
Keywords: Proton pumping; proton gating; water chain; electron transfer;
FTIR spectroscopy
*Corresponding author. Max-Volmer-Laboratory for Biophysical Chemistry, Institute of Chemistry, Technical University of
Berlin, Secr. PC 14, Strasse des 17. Juni 135, D-10623 Berlin, Germany. E-mail address: [email protected].
Present address: P. Hellwig, Institut de Chimie, UMR 7177, CNRS, Université Louis Pasteur, 4, rue Blaise Pascal, 67000
Strasbourg, France.
Abbreviations used: COX, cytochrome c oxidase; WT, wild type; CCCP, carbonyl cyanide m-chlorophenylhydrazone; FTIR, Fourier transform infrared; SHE, standard hydrogen electrode; P side, positively charged aqueous side of the
membrane; PLS, proton loading site; DM, n-dodecyl-β-D-maltoside.
0022-2836/$ - see front matter © 2008 Elsevier Ltd. All rights reserved.
866
Introduction
Cytochrome c oxidase (COX) is a terminal member
of the aerobic respiratory chain of mitochondria and
many bacteria, including Paracoccus denitrificans and
Rhodobacter sphaeroides. It catalyzes electron transfer
from cytochrome c to molecular oxygen (dioxygen)
and utilizes the free energy provided by this redox
reaction to generate an electrochemical proton gradient across the inner mitochondrial and bacterial
cytoplasmic membrane, respectively. During a full
dioxygen reduction cycle, eight protons are taken up
from the inside, four for water formation (“substrate
protons”) and four for translocation across the
membrane (“vectorial or pumped protons”).1 In
aa3-type COXs, which also include the mitochondrial
oxidase, four redox-active cofactors participate in
electron transfer: CuA, the primary electron acceptor
from cytochrome c; heme a; and the binuclear center
consisting of CuB and heme a3, where binding and
reduction of oxygen take place.
At least two proton-conducting pathways, termed
D-pathway and K-pathway (see Fig. 1), have been
Fig. 1. Structures of subunits I and II of the four-subunit
complex of the aa3 COX from P. denitrificans (Protein Data
Bank code 1AR1) illustrating the two proton-conducting
pathways of the enzyme: D-pathway (red) and K-pathway
(green). A putative proton exit pathway is also indicated
(black arrow). Functionally important amino acid side
chains in the pathways (including N131) are represented as
sticks. The VMD software was used for visualization.2
Structural Insights from a Decoupled COX Variant
identified by analyzing both structural data3–8 and
mutagenesis studies.9–20 These pathways connect
the inner aqueous phase and the binuclear site. The
initial concept that protons consumed in water
formation at the active site use the K-pathway and
translocated protons are transferred via the Dpathway3 could not be maintained because measurements of charge translocation21,22 and results
from flow-flash experiments23,24 indicated that the
K-pathway delivers the first one or two protons to
the binuclear center during the reductive phase of
the catalytic cycle only, with the remaining two or
three substrate protons transported via the D-pathway during the “oxidative” part of the cycle (after
the binding of molecular oxygen to the reduced binuclear center of COX). All pumped protons appear
to use the D-pathway.
Proton translocation in these pathways has been
proposed to occur via hydrogen-bonded chains, so
called “proton wires,”25–28 which include amino
acid residues and bound water molecules. 29,30
According to the Grotthuss mechanism of proton
transfer (reviewed in Refs. 26 and 31), an excess
proton can travel along such a chain of hydrogen
bonds when each non-hydrogen atom, usually an
oxygen or a nitrogen, in the chain is a donor and an
acceptor of a hydrogen bond, giving rise to a much
higher rate of proton translocation compared to a
simple diffusion process. The individual proton is
just relayed from the donor to the acceptor, followed
by a conformational reorganization.
As a consequence of the observation that the Dpathway is used for pumped protons and substrate
protons, a mechanism differentiating substrate
protons from pumped protons (“gating”) in the
D-pathway must operate, but its molecular mechanism remains unknown. Results from several
studies3,32–47 indicate that the side chain of Glu278,
located at the upper end of the D-pathway, may be
the key element for gating. From Glu278, protons
are funneled into two sites (see Fig. 1 for illustration), with one being the binuclear center and the
other a yet unidentified protonatable cluster [often
referred to as pump site X or proton loading site
(PLS)] from which pumped protons are eventually
expelled to the P side (positively charged aqueous
side) of the membrane. The direction of proton
transfer beyond Glu278 has been discussed in
terms of a reorientation of proton-conducting
water arrays controlled by the redox state-dependent electric field between heme a and the binuclear center.32,39
Several mutations that selectively abolish proton
pumping have been identified in positions N13115,48,49
and N19915,50 (P. denitrificans numbering is used
unless otherwise stated). Since they are located in a
region of the D-pathway far below E278 (Fig. 1), they
might a priori be expected to impair the uptake of
both substrate and pumped protons, for example as
reported for D124N.12–14,16 The selective loss of
proton pumping in the former variants has therefore
been attributed to impeding the branching function
of E278 by a yet unknown mechanism.
867
Structural Insights from a Decoupled COX Variant
The best characterized example of such a decoupled variant is N131D.15,48,51–54 The corresponding mutation of the homologous residue N139 in
R. sphaeroides has been suggested to decouple the
proton pump by increasing the apparent pKa value
of E286 (E278 in P. denitrificans) in the PR state from
9.4 to 11,52 as concluded from pH level-dependent
measurements of the rate for the PR → F transition,
reflecting the protonation state of E286. Continuum
electrostatic calculations yield still higher values for
this variant.55,56 More recently, it has been reported
that the R. sphaeroides decoupling mutation N207D
(N199D in P. denitrificans) is correlated with an even
higher shift of the E286 apparent pK.50,56 Since the
restoration of the original charge distribution by the
additional mutation D132N (D124N in P. denitrificans) at the entrance of the D-pathway was shown to
restore proton pumping in the N139D (N131D in
P. denitrificans) variant,57 researchers have suggested
that introducing a charged side chain influences the
glutamate side chain via long-range electrostatic
interactions. However, the relatively long distance of
about 20 Å between the two residues renders an
electrostatic effect rather unlikely. Furthermore, continuum electrostatic calculations suggest that the
effect of the introduced negative charge in the N131D
variant on E278 is very small in terms of changing its
pKa.55 Olsson and Warshel were able to calculate the
effect of the N139D (N131D in P. denitrificans) mutation on proton pumping.54 They concluded from
changes in the free energy surface for proton
translocation of the variant that the electrostatic
potential of water molecules below E286 (E278 in
P. denitrificans) is altered, thereby stopping the flow
of protons to the P side and trapping the proton in
the binuclear center. More recently, in a replacement
of N139 by the neutral amino acid Thr, 49 the
apparent pKa of E286 was shown to be shifted in
the opposite direction, again supporting the view
that the observed pKa changes have no electrostatic
origin but rather reflect a possible E286 side-chain
reorientation caused by the mutations.
Despite all experimental and computational efforts
to elucidate the mechanism of decoupling in these
enzyme variants, it is presently still unclear how these
mutations lead to a selective loss of proton pumping
and whether this phenomenon requires a negatively
charged group in the center of the D-pathway.
In this study, we systematically replaced the N131
residue with different amino acids in order to
evaluate its role in proton pumping. Interestingly,
replacement by strongly acidic residues, such as
aspartate and glutamate, leads to selective loss of
proton pumping. In addition, an exchange by the
weakly acidic cysteine results in a decoupled enzyme. Redox Fourier transform infrared (FTIR) spectroscopy was used to monitor the effects of mutations introduced at position 131 on the protonation
state of residue E278. The crystal structure of the
N131D variant determined at 2.32 Å resolution
surprisingly reveals an additional E278 conformation along with changes in the positions and occupancies of the crystallographically resolved water
molecules in the D-pathway, which, taken together,
appear responsible for the decoupled phenotype of
this variant.
Results
Cytochrome c oxidation activity and proton
pumping
The wild type (WT) and all mutant enzymes,
histidine tagged at the C-terminus of subunit I, were
purified using a Ni–nitrilotriacetic acid affinity
column. As judged from SDS-gel electrophoresis
(data not given), the isolated COX complexes showed
a normal four-subunit composition, with all mutated
I subunits migrating identically with their WT
counterpart. UV/VIS spectroscopic features were
indistinguishable from those obtained for the WT,
with heme/protein ratios varying between 14 and
17 nmol heme a/mg.
We first characterized the mutant enzymes with
respect to their steady-state cytochrome c oxidation
activity (listed in Table 1). The activity of the N131D
variant is, within experimental error, identical to the
WT, as reported earlier.15,48 Two further replacements of N131 by Glu and Cys affected the turnover
number only marginally (Table 1), thus also displaying cytochrome c oxidation activities again
comparable with the WT enzyme. In contrast, a
substitution of the Asn by the longer glutamine side
chain diminished the steady-state activity by almost
50%. When hydrophobic amino acid residues are
introduced in this position, the turnover number of
the enzyme decreases drastically, resulting in residual cytochrome c oxidation activity of 11% and 3%
for the N131A and N131V variants, respectively. A
similarly substantial loss in turnover rate was also
observed for the N131Y variant, which retains 6% of
the WT activity.
Proton pump stoichiometries were determined with
the use of a stopped-flow spectrometric approach
after reconstitution of the purified COXs into
Table 1. Catalytic activities and proton pump stoichiometries of purified P. denitrificans aa3 COX variants in
comparison with the WT activity
Oxidase
type
Cytochrome c
oxidation activity (%)
Proton translocation
(H+/e−)
WT
N131D
N131E
N131C
N131Q
N131A
N131V
N131Y
100.0 ± 12.3
114.0 ± 16.7
77.9 ± 14.9
86.4 ± 11.4
52.2 ± 4.1
10.7 ± 1.0
2.6 ± 0.2
6.2 ± 0.9
1.04 ± 0.15
0.09 ± 0.02
0.12 ± 0.02
0.12 ± 0.02
0.40 ± 0.09
0.04 ± 0.01
n.d.
n.d.
One hundred percent activity corresponds to a turnover number
of 458 electrons s− 1. Data are expressed as the mean ± SE of at least
three independent experiments. n.d. indicates not determined.
868
Fig. 2. Changes in phenol red absorbance followed at
555.6 nm (isosbestic point of cytochrome c) in the presence
and in the absence of the protonic uncoupler CCCP for the
WT and several mutant enzymes. Proton pump stoichiometries obtained by an exponential fitting procedure are
listed in Table 1.
proteoliposomes. Figure 2 illustrates the transient
changes in phenol red absorbance recorded after
mixing the proteoliposomes with reduced cytochrome c for WT and several mutant COXs. A
rapid exponential decrease of the phenol red
absorbance for the WT enzyme indicates transient
acidification of the extravesicular medium. In the
presence of the protonophore carbonyl cyanide mchlorophenylhydrazone (CCCP), a net alkalization
is the result of the consumption of protons in water
formation. With the WT enzyme pumping one
Structural Insights from a Decoupled COX Variant
proton per transferred electron, the phenol red
absorbance changes in the absence and in the
presence of CCCP should be identical in absolute
numbers. In fact, after exponential fitting of the
absorbance changes, we obtained a H+/e− ratio of
1.04 for the WT enzyme. In contrast, none of our
mutant enzymes reached this proton pump activity
(see Table 1). For example, the N131Q variant
resulted in reduced proton pumping with a stoichiometry of about 0.4 H+/e−.
Three mutants, N131D, N131C, and N131E, although retaining full cytochrome c oxidation activity,
display a proton pump stoichiometry of approximately 0.1 H+/e− and may thus be safely classified
as “ideally decoupled” mutants. In contrast, the
pump stoichiometry of 0.04 H+ /e− experimentally
determined for the N131A variant is difficult to
assess: while this mutation may indeed lead to a
defect in pumping, it cannot be excluded that its
substantially reduced cytochrome c oxidation rate
results in experimental difficulties in the system used
here. Since respiratory control ratios of the COX
vesicles determined separately were between 4.1 and
7.5 for the variants and approximately 10.0 for the
WT enzyme (data not shown), none of the mutations
produces a leak for protons across the membrane.
FTIR spectroscopic characterization of
uncoupled variants
Figure 3A presents the oxidized-minus-reduced
FTIR difference spectra of WT, N131D, N131C, and
N199D mutant enzymes for a potential step from
− 0.292 to 0.708 V at pH 7.0. Overall, the difference
spectra of the various mutant enzymes are similar in
shape, indicating that no major structural change
occurs upon mutation. The signature previously
assigned to the reorganization of the protonated
Fig. 3. (A) Oxidized-minus-reduced FTIR difference spectra of WT and N131D, N131C, and N199D mutant COXs from
P. denitrificans for a potential step from −0.5 to 0.5 V versus Ag/AgCl (corresponding to a potential step from −0.292 to
0.708 V SHE′). (B) Oxidized-minus-reduced FTIR difference spectra of WT (dashed dotted line), N131D (continuous line),
and N131C (dashed line) COXs from P. denitrificans for a potential step from −0.5 to 0.5 V versus Ag/AgCl (corresponding
to a potential step from −0.292 to 0.708 V SHE′). (C) Oxidized-minus-reduced FTIR difference spectra of WT (dasheddotted line) and N131C variant (continuous line) COXs from P. denitrificans, with samples equilibrated at pH 9.2, for a
potential step from − 0.5 to 0.5 V versus Ag/AgCl (corresponding to a potential step from − 0.292 to 0.708 V SHE′).
Structural Insights from a Decoupled COX Variant
form of Glu278 58,59 is present for all variants
analyzed; a small upward shift of 4 cm− 1 indicates
a minor variation in the hydrogen-bonding environment for the reduced form. In the oxidized form,
shifts of about 1–2 cm− 1 occur for all variants, but
this observation may not be significant since the
resolution of the spectra is 4 cm− 1. Some variations
can be found between 1580 and 1500 cm− 1 for the
N131D mutant enzyme, which is shown in direct
comparison with the WT enzyme and the N131C
variant enzyme in Fig. 3B. Interestingly, a broad
signal at 1586 cm− 1 is perturbed together with
signals around 1420–1390 cm− 1. These spectral
ranges are characteristic for the signals of the
respective asymmetric and symmetric modes of
deprotonated acidic side chains, respectively. Since
the modes are absent in the N131C variant enzyme,
they may be assigned to originate from the
introduced aspartic acid in the N131D variant. We
suggest that its carboxylate is deprotonated and
reorganizes upon redox reaction.
Spectra at pH 9.2 have been obtained and are
presented in Fig. 3C for the WT and the N131C
variant in order to probe a possible effect of the
mutations on the pH level dependence of the
redox-dependent behavior. Furthermore, in this
study, major spectral features remained essentially
unchanged in both samples. We note that the broad
positive mode at 1439 cm− 1 arises from the B = O
vibration of the borate buffer. For both samples,
WT and N131C, the presence of difference signals
at 1746 and 1734 cm− 1 in the spectra at pH 9.2
reveals that Glu278 is protonated in the reduced
and oxidized forms. Again, a small shift of the
signals by approximately 4 cm− 1 is seen, obviously
arising from a small variation in the hydrogenbonding environment.
X-ray structure of the N131D variant at 2.32 Å
resolution
When comparing the atomic coordinates of the
N131D variant structure (resolved to 2.32 Å) to the
ones of the recently refined WT structure (Koepke
et al., unpublished results) truncated to 2.29 Å
resolution (see Materials and Methods), no significant difference in the overall protein structure is
detectable (RMSD = 0.498 Å). The backbone and
side-chain conformations seem to be virtually
identical with the WT structure. However, two features distinguish the variant structure from WT:
(i) Two electron densities are seen for the side
chain of the highly conserved residue E278,
indicating two side-chain orientations: One
(shown in orange in Fig. 4) is identical to the
E278 orientation observed in the WT structure
(in magenta in Fig. 4), and is also found in other
COX structures, such as those from bovine heart
mitochondria8,60 and R. sphaeroides.6 In this
downward orientation, the E278 side chain is
within hydrogen-bonding distance to a water
molecule (Wat551), which participates in a
869
Fig. 4. Superposition of the P. denitrificans N131D
variant (carbon atoms in yellow) with the WT (magenta)
structure. Hydrogen bonds between water molecules of
the WT accounting for a water chain along the D-pathway
are represented by green dashed lines, with their respective lengths given in Å. These hydrogen bonds were
restrained during the final refinement cycles to 2.8 Å. The
electron density of the variant structure around water
molecules accounting for the D-pathway is given in green
wire-frame representation at the level of 0.2σ. A difference
Fourier density between variant and WT structures was
calculated and is presented as blue (positive density) and
red (negative) wire frames at the level of ±2σ. The electron
density for the water molecules 465, 899, 962, 968, and 969
is missing in the variant structure, indicated by a negative
difference density (red) and in addition by magenta labels
of these water positions. The missing water molecules
reflect a higher flexibility of the D-pathway water chain in
the variant structure. The side chain of Glu278 was refined
in two orientations. Thus, one orientation of this residue is
marked by carbon atoms shown in orange.
hydrogen-bonding network important for proton transfer in the D-pathway. This conformation of E278 might be required for taking up
protons from the D-pathway, thus being often
referred to as the “input” conformation of this
residue. In the other conformational state
revealed by the present N131D variant structure, the E278 side chain points upward (in
870
Structural Insights from a Decoupled COX Variant
yellow in Fig. 4). A similar configuration of the
corresponding residue was observed in the Xray structure of the variant E286Q6 and in WT
COX from R. sphaeroides at pH 10.0 (as cited in
Refs. 61 and 62). According to previous theoretical studies on COX,29,33,39,47,63 such a conformational state of E278 is considered as a “proton
release” configuration from which a “loaded”
proton is transferred either to the binuclear
center or to a “proton loading/pump site” near
the heme a3 propionates. The routing of protons
further on from this “output” configuration
depends on the relative energy barriers, which
are presumably modulated by the redox state of
the oxygen intermediates at the active site, as
well as on the presence of a transient pathway
for proton transfer, which may exist for very
short periods only because of rapid fluctuations
of water molecules.29
(ii) Distinct changes in the positions of water
molecules are observed in the N131D structure,
in the vicinity of the D-pathway: In the cavity
between residue 131 and E278, several water
molecules can be resolved from the electron
density of the variant (Fig. 4, green wire frame)
and WT (data not shown) structures, forming a
continuous hydrogen-bonded water chain connecting the two residues. These water molecules
are represented by spheres in Fig. 4, shown in
magenta for the WT structure and in red for the
N131D variant structure. Notably, five water
molecules are missing in the variant structure,
which becomes obvious when the difference
Fourier electron density between the variant
and the WT electron density is considered (red
and blue wire frames in Fig. 4, representing
negative and positive difference densities at a
level of ±2σ, respectively). Distinct negative
difference densities are observed below E278
(missing Wat969 and Wat899), close to residue
131 (missing Wat962 and Wat968), and near the
entrance of the D-pathway (missing Wat465).
For comparison, in the WT structure, the
electron densities for these water molecules
near N131 are approximately 0.6σ and more
than 1σ for those close to E278. To exclude the
possibility that the observed differences arise
from the different resolutions of the two relevant structures, we calculated both applying the
same reflections (see Materials and Methods).
The remaining minor difference in resolution is
due to the differences in unit cell dimensions of
the data sets.
Discussion
In this study, further replacements in the N131
position of the Paracoccus COX were analyzed and
characterized by FTIR spectroscopy, and the N131D
variant was crystallized and its structure was determined at 2.32 Å. An interesting correlation is found
between its defects in proton pumping and structural alterations in the mutant enzyme.
E278 conformation is altered in the N131D
decoupled variant
The central finding in this work is an additional
density for the E278 side chain observed in the
N131D crystal structure. Considering the generally
accepted role of this conserved glutamate as a
potential redox-controlled switch directing protons
either to the binuclear center or to a PLS around the
heme a3 propionates, the structural alteration found
for the N131D variant appears to be of particular
importance for the observed selective loss of proton
pumping. Whereas the WT (downward) configuration is also found in X-ray COX structures from other
species,6,8 the simultaneous existence of both sidechain orientations of this glutamate as observed here
in the N131D mutant enzyme has never been seen
directly before. On the other hand, it has already
been suggested in the first description of a COX
structure3 and is in agreement with results from
redox-FTIR spectroscopy.58 In addition, simulation
calculations have provided evidence for the existence of (several) additional glutamate side-chain
conformations.33,39,40,47,63 Our own full-scale molecular dynamics simulations29 of COX in the fully
oxidized state have indicated that the side chain of
E278 moves into the upward orientation (see Refs. 40
and 55) if it is deprotonated. Interestingly, this
orientation was also found in the R. sphaeroides WT
structure at pH 10.0 (unpublished results by Gennis
& Brzezinski, cited in Refs. 61 and 62). These findings
suggest that the alternative “up” orientation may be
populated only transiently in specific steps of the
regular reaction cycle and thus cannot be revealed in
static X-ray structures of the WT enzyme.
The additionally observed E278 “up” (Gluup)
conformation of the N131D variant structure reflects
a higher occupancy of this transient state that is
apparently caused by the mutation. According to
results from a recent molecular dynamics study of
the corresponding E242 (bovine heart mitochondria
COX numbering) side-chain rotational isomerization,63 this transient conformational “up” state is
supposed to be extremely vulnerable to leakage of
protons from the PLS back to E242. In the bovine WT
COX, a rapid (∼ 1 ps), highly exergonic downward
flip of the deprotonated E242 side chain (Glu− up) to
the other conformation (Glu−down) is assumed to be
essential for “valve” function, thus ensuring that
protons stored at the PLS are only ejected to the P
side.64 The additional “up” conformation revealed
in the Paracoccus N131D variant structure seems to
indicate that this crucial fast transition is impeded as
a consequence of the mutation, which would
compromise the E278 valve function, thus explaining its decoupled phenotype.
Alternatively, the altered side-chain orientation of
E278 may affect its proton funneling properties: The
rotational mobility of the E278 carboxylate in the
variant may be reduced and, consequently, the E278
Structural Insights from a Decoupled COX Variant
side chain is preferentially “locked” in one position,
unable to supply protons to the pump site at all but
nevertheless allowing an ample number of protons
to reach the binuclear site for oxygen reduction. With
the aforementioned assumption that the altered E278
orientation now delivers all protons taken up from
the D-pathway to the binuclear site instead of
discriminating between pumped and substrate protons in mind, we may even explain the increased
cytochrome c oxidation activity reported for the
N139D variant from R. sphaeroides.48
Observed effects on E278 in decoupled variants
are probably not of electrostatic nature
Considering the distance of at least 18 Å, it is far
from obvious how the E278 conformational change
observed for the P. denitrificans N131D variant in this
work or the increased E286 pKa reported for the R.
sphaeroides N139D variant may be linked to the site
of mutation. An electrostatic effect of the ionized
aspartate side chain introduced by the mutation in
position N139 was proposed by Namslauer et al.65
Similarly, an electrostatically derived pKa shift of
E286 was also assumed to be responsible for the
observed decoupling of the variant N207D in R.
sphaeroides.50 These pKa increases were computationally confirmed for both ionic variants in a recent
continuum electrostatic calculation.56 Moreover, the
coupled phenotype in the charge-shift double
variant N139D/D132N was attributed to a restoration of the E286 pKa,57 thus corroborating the idea
that the effects of the decoupled variants are of
electrostatic nature.
In our work, the deprotonated state of N131D
was confirmed by redox-FTIR spectroscopy, in
support of this explanation. On the other hand,
the spectra of decoupled mutants did not reveal a
different protonation state for the E278 side chain
reflecting possible pKa alterations. However, it
must be emphasized that the shifted pKa values
observed for the decoupled R. sphaeroides mutants
are “apparent” ones that are characteristic only of
the transient PR state49,50,65—such pKa alterations
are therefore most likely not revealed by the redoxFTIR method applied here. According to our
calculations,55 the charge–charge interaction energies between the two acidic residues E278 and
N131D are quite low, mainly due to the long
distance of 18 Å. Very recently, an electrostatic
influence was considered more likely for charged
residues introduced closer to E278 (e.g., at a
distance of 7 Å, such as S197D in R. sphaeroides20).
Furthermore, our findings show that not only
acidic replacement of N131 by Asp or Glu but
also the presumably neutral side-chain substitution
by Cys results in decoupling of the COX redoxdriven proton pump. According to our electrostatic
calculations, which predict a pKa between 10.7 and
22.2 depending on the dielectric constant applied,55
deprotonation of the N131C side chain, causing a
change in the pKa of E278 by electrostatic interaction, is rather unlikely. Therefore, the decoupled
871
phenotype of at least this particular variant must
be explained by another mechanism, similarly to
the recently characterized N139T variant from R.
sphaeroides, in which the loss of proton pumping
also cannot be attributed to electrostatic effects due
to the neutral Thr replacement.49
Distorted water structure in the D-pathway is
characteristic of decoupled variants
Molecular dynamics simulations indicate that a
continuous chain of water molecules, connecting
D124 up to E278, is not present in the D-pathway.29,30
Accordingly, in several X-ray structures, 6 the
crystallographically resolved water molecules
form a continuous hydrogen-bonded proton wire
only between E278 and N131, which is connected
to D124 via a single water molecule (see also Fig.
4). It was proposed that the side chains of N131
and another conserved asparagine (N113) in its
direct vicinity are required to maintain the tight
connectivity between these two water chains.53
However, one has to keep in mind that asparagines in the given structure cannot be part of a
proton-connecting hydrogen-bonded network.
With the side-chain carbonyl oxygen only being
a hydrogen-bond acceptor, they interrupt the
“proton wire” and block proton transfer. Therefore, they may offer a potential gating function,
controlling the access of solvent protons further up
the D-pathway. A reorientation of these asparagine side chains is necessary to allow the formation of an effective proton-transferring hydrogenbonded chain from D124 up to E278. These
conclusions are in agreement with the results of
free energy profile calculations for excess protons
in the D-pathway.53
With the aforementioned considerations, it is
interesting to note the reduced number of observable water molecules in the D-pathway of the
N131D variant structure presented here when
compared with the WT. We interpret this as an
indication of a higher mobility of these water
molecules since large cavities in protein structures
without any crystallographically defined water
molecules are often filled with disordered water
molecules.66,67 Changes are most prominent in the
immediate vicinity of the mutated N131D side
chain, at the entrance of the D-pathway near D124,
and around E278. Such changes in the water
structure around E286 were also postulated for
the decoupled N139D and N207D variants from R.
sphaeroides,68 since the redox-FTIR spectra of these
two variants exhibited a slight shift of the E286
signature, indicating an alteration of the chain of
water molecules running from the site of mutation
up to E286. A very similar effect is seen in the
corresponding FTIR spectrum for the N131C
variant presented in this work (Fig. 3B and C),
suggesting a common mechanistic explanation for
the decoupling observed in all such variants (i.e., a
perturbation of the position or orientation of water
molecules within the D-pathway).
872
Correct hydrogen bonding may be crucial to
stabilize the proton-conducting water chain
The asparagine side chain can form hydrogen
bonds as a donor and as an acceptor. This property
may be crucial to the abovementioned functions of
the three asparagines. In contrast, the mutated
N131D side chain in its deprotonated form, as
evidenced by our FTIR experiments, can only be a
hydrogen acceptor, unable to provide hydrogen
bonds to Asn113 and Wat972 (see Fig. 4). As a result
of the mutation introduced here, the polarity of the
hydrogen-bonded chain is changed and the hydrogen-bond pattern is disturbed. These effects may be
responsible for the observed perturbation of the
water structure in the middle of the D-pathway and
be further propagated up to E278 (see below). The
notion that the decoupled phenotype of the N131D
mutant is rather a consequence of the perturbed
hydrogen-bonding capability of the mutated aspartate side chain and not due to its ionic state is
supported by the newly identified N131C (this work)
and N139T49 (R. sphaeroides) variants, in which uncharged side chains are also less suited for hydrogen
bonding (as donors of only a single hydrogen bond
instead of two hydrogen bonds), resulting in a
decoupled phenotype as well.
However, any conclusive explanation as to how a
proton can be transferred from Asp124/Wat622 via
Asn/Asp131 to Wat972 (see Fig. 4) on the opposite
side of the asparagine cluster is still missing. A
protonated Asp side chain could transfer a proton
via a side-chain rotation, as also indicated by the
results of a theoretical study53 that predicted an
accelerated proton translocation by a neutral Asp
side chain compared with a negatively charged one.
However, one has to keep in mind that assuming a
pKa of 5 for Asp131, it would be protonated in a
rapid dynamic equilibrium for 0.5% of the time at
pH 7.0. Whenever it is protonated, the side chain
may rotate and deliver a proton to Wat972 and thus
lead to an acceleration of proton transfer from the
inner side to E278 compared with the WT. Actually,
the question on whether the N131D mutation leads
to a deceleration or an acceleration of proton transfer
is far from being settled. Indeed, results of measurements of the transmembrane charge separation for
the F-to-O transition with the homologous N139D
variant from R. sphaeroides suggest that the mutation
leads to an acceleration of the rate (τ) of the slowest
proton uptake phase via the D-pathway from 1.5 to
0.6 ms.51 In addition, the fact that the decoupled
turnover of the homologous N139D variant in R.
sphaeroides is more than 400% of the coupled turnover of the WT enzyme48 is more easily reconciled
with an enhanced proton transfer through the Dpathway than with a diminished one (but see below
and Ref. 69 for an alternative explanation for the
increased N139D turnover rates).
Provided that the changes in the water mobility and
the hydrogen-bonding pattern actually affect proton
translocation in the D-pathway, still the question
arises as to why, selectively, proton pumping but not
Structural Insights from a Decoupled COX Variant
water formation at the active site is impaired. On the
basis of the two variant structure-derived key
observations, the perturbed E278 side-chain position
(s) and the changes in the proton-wire water structure,
we envisage two possible explanations:
(i) Precise timing of proton delivery may be more
demanding for proton pumping than for water
formation (“kinetic gating”)
A changed proton conduction due to the altered
water structure and hydrogen bonding in the Dpathway may not deliver protons within a distinct
time window required for proton pumping. Assuming that a less precise timing is necessary for substrate protons consumed at the binuclear center for
water formation, a reduction in proton transfer rate
would selectively affect proton pumping and not
the steady-state cytochrome c oxidation activity.
This is in line with the “kinetic gating mechanism”
proposed by Popovic and Stuchebrukhov.70 They
suggested that E278 may be connected to the PLS
and the active site by two separate proton-transporting water networks offering different proton
conduction rates. It is assumed that the “faster” one
delivers pumped protons to the PLS, whereas the
chemical protons are provided by the “slower” one,
thus ensuring that a proton is preloaded into the
pump site before the driving redox event (protonation of the reduced oxygen intermediates at the
active site by substrate protons) occurs. Since both
networks are supplied with protons via the Dpathway, a moderately decreased proton translocation rate, possibly caused by the D-pathway water
rearrangement, might become rate limiting only for
the fast proton-conducting network, thereby decoupling the proton pump. Moreover, this explanation
is also in agreement with a suggestion recently
provided by Wikström and Verkhovsky.69 They
assumed that a slowed proton translocation caused
by the mutated N131D side chain may delay
reprotonation of E278, which decouples the pump
by lowering the barrier for proton transfer from the
“pump site” back to E278 (see also below). In their
model, this would result in a bypass of the slowest
step of the proton pumping cycle and explain even
the enhanced steady-state cytochrome c oxidation
activity of the N139D variant (R. sphaeroides).
Furthermore, hydrophobic replacements of the
N131 side chain, which cannot provide any
stabilizing hydrogen bonds at all, are expected to
cause even more pronounced perturbations of the
water structure. Consequently, the proton translocation rate would be slowed substantially and the
delivery of substrate protons via the mutated Dpathway eventually becomes rate limiting even for
oxygen chemistry at the binuclear site, thus also
affecting steady-state turnover. This is consistent
with the experimentally observed drastically
reduced cytochrome c oxidation activity of the
hydrophobic variants N131A and N131V or the
N131Y variant (Table 1), with its limited ability to
form hydrogen bonds.
873
Structural Insights from a Decoupled COX Variant
(ii) A lowered rate of proton conductance in the Dpathway impairs the E278 switch (by shifting
the conformational equilibrium of the E278
side-chain orientation)
It was proposed in a recent molecular dynamics
simulation of the E242 (corresponding to E278) sidechain rotational isomerization63 that a lowered rate
of proton conductance in the D-pathway may raise
the kinetic barrier for the Glu−down-to-GluHdown
transition (reprotonation of the anionic Glu−down via
the D-pathway). Consequently, the Glu−down state
accumulates, which in turn increases the occupancy
of the Glu−up state. This seems reasonable since the
additional E278 conformation revealed by the
N131D variant structure actually supports such an
increased population of the Glu−up state. As already
outlined above, this state is assumed to be vulnerable to leakage of protons from the pump site back
to E278, thus explaining the selective loss of proton
pumping of the decoupled variants.
Both the kinetic (i) and conformational (ii) arguments
may also be linked to each other and thus contribute to
various extents to the observed phenotype.
(iii) Caused by local environment changes in the
variant enzyme exhibiting an alternative E278
orientation, the pKa of E278 may be shifted,
thus rendering the side chain unsuitable to
deliver protons to the pump site.
Finally, with the N131C mutant enzyme phenotype largely identical with the corresponding N131D
mutant COX, former explanations invoking electrostatic effects appear less substantiated.
Materials and Methods
Site-directed mutagenesis
Site-directed mutagenesis was performed according to
the QuikChange site-directed mutagenesis protocol (Stratagene). All mutations introduced into the ctaDII gene encoding subunit I15,16 were confirmed by DNA sequencing.
Mutated ctaDII genes were subcloned into pKH48, a derivative of pUP3915 carrying a C-terminal six-histidine affinity tag.
Bacterial growth, protein purification, and
crystallization
Conclusions
The X-ray structure of a decoupled variant
(N131D) of the P. denitrificans COX provides
evidence for the simultaneous population of an
alternative conformational state of the E278 side
chain considered to be of prime importance for the
routing of “chemical” and “pumped” protons,
acting as a kinetic valve in the WT COX to prevent
proton leakage from the P side. In addition, several
distinct differences are seen for the arrangement of
water molecules in the proton-conveying D-pathway of the variant.
These structural findings suggest that the distorted water structure in the D-pathway and/or
an impeded branching function of the E278 side
chain in the variant may be responsible for the
observed phenotype, explained by one, singly, or
by a combination of the following mechanistic
scenarios:
(i) A shifted conformational equilibrium of the
two E278 side-chain orientations impairs
either (a) its function as a switch separating
pumped and substrate protons delivered by
the D-pathway or (b) its recently proposed
“valve” function, resulting in leakage of
pumped protons from the PLS back to E278,
which selectively inhibits proton pumping
without affecting the routing of substrate
protons to the binuclear site.
(ii) An impeded proton transfer rate through the
D-pathway, due to changes in water positions
observed in the N131D variant structure, is not
compatible with the required time regime for
proton pumping in the variants.
pKH48 plasmids carrying mutated subunit I genes were
conjugated into the P. denitrificans recipient strain AO1,
lacking functional aa3- and cbb3-type COXs.16 Bacterial
cells were grown at 32 °C to an optical density at 600 nm
N 3 in succinate medium16,71 supplemented with 25 μg/ml
of streptomycin. Membranes isolated from these cells
were solubilized in n-dodecyl-β-D-maltoside (DM) and
applied to a Ni–nitrilotriacetic acid affinity column
(QIAGEN), which allows purification of the four-subunit
COX complex in a single-step procedure.72 Isolated COX
complexes were characterized for their subunit composition with the use of SDS-gel electrophoresis, by recording
their visible redox difference spectra and by determining
their heme a/protein ratios. For crystallization purposes,
WT and N131D mutant COXs were expressed in P.
denitrificans strain AO1 using construct pUP3916 lacking
a histidine tag. Accordingly, the two-subunit COX was
purified via Fv antibody fragment 7E2C50S73 and crystallized as described previously.3,4
Determination of enzymatic activities of isolated
enzymes
Enzymatic activity was determined as previously
described74 with 20 μM reduced horse heart cytochrome
c (Sigma) using a reaction buffer containing 20 mM Tris–
HCl, pH 7.5, 20 mM KCl, 1 mM ethylenediaminetetraacetic acid, and 0.2 g/l of DM. The assay was performed at
550 nm in a Hitachi U-3000 spectrophotometer.
Reconstitution of purified COX into liposomes and
determination of proton pumping
Asolectin (phospatidylcholine, type II-S, Sigma) and
cholate were purified as described elsewhere.75 A total of
40 mg/ml of purified asolectin was resuspended in
100 mM Hepes–KOH, pH 7.3, 10 mM KCl, and 2% cholate
and stirred at 4 °C until all particles were dissolved.
874
Subsequently, the solution was sonicated at intervals of
30 s (Branson Sonifier II 250, 30% duty cycle) until the
solution appeared translucent (both steps were carried out
under nitrogen atmosphere). After a short centrifugation
(6000g, 15 min, 4 °C) to remove particulate material, purified COX previously eluted from a Q-Sepharose column
by 100 mM Hepes–KOH, 500 mM KCl, and 0.15 g/l of DM
was added to a final concentration of 4 μM. After a
sequence of dialysis steps,76 the mixture was centrifuged
again (6000g, 15 min, 4 °C) and the supernatant was used
for proton pump measurements.
Stopped-flow measurements recording the absorbance
changes of the pH-sensitive dye phenol red photometrically were carried out essentially as described in Ref.
77 using a stopped-flow spectrometer (Hi-Tech Scientific
SF-61).
A suspension containing 0.4 μM COX in proteoliposomes, 60 μM phenol red, and 10 μM valinomycin was
prepared in the last dialysis buffer, and reduced horse
heart cytochrome c at a final concentration of 200 μM
was equilibrated in the same buffer. After adjusting the
pH levels of both solutions to 7.3, solutions were filled
into 2.5- and 0.25-ml syringes of the stopped-flow
spectrometer, respectively. The 10:1 ratio of syringes
was chosen to avoid mixing artefacts.76,78 The absorbance change of phenol red was monitored at 25 °C and
555.6 nm, which was determined to be the isosbestic
point for cytochrome c under the experimental conditions. For determination of the decoupled rates, the
procedure was repeated after adding CCCP (final
concentration = 10 μM) to the solution containing the
proteoliposomes.
Electrochemistry
The ultra-thin-layer spectroelectrochemical cell for the
UV/VIS and IR was used as previously described.79
Sufficient transmission in the range of 1800 to 1000 cm− 1,
even in the region of strong water absorbance around
1645 cm− 1, was achieved with the cell path length set to
6–8 μm. The gold-grid working electrode was chemically
modified with a 2 mM cysteamine solution as reported
before.58 In order to accelerate the redox reaction, we
added 16 mediators as reported in Ref. 58 (except nmethylphenazoniummethosulfate and n-ethylphenazoniumsulfate) to a final concentration of 45 μM each. At
this concentration, and with the cell path length below
10 μm, no spectral contribution from the mediators in the
VIS and IR ranges could be detected in control experiments with samples lacking the protein, except for
the PO modes of the phosphate buffer between 1200 and
1000 cm− 1. Potentials quoted with the data refer to the
Ag/AgCl/3 M KCl reference electrode. The potential
step from − 0.5 to + 0.5 V applied here therefore
corresponds to a potential step from −0.292 to 0.708 V
SHE′ (standard hydrogen electrode). Electrochemistry
samples were equilibrated in 100 mM phosphate buffer,
pH 7.0, or borate buffer, pH 9.2, 100 mM KCl, and 0.15%
DM each.
FTIR spectroscopy
FTIR and UV/VIS difference spectra as a function of the
applied potential were obtained simultaneously from the
same sample with a setup combining an IR beam from an
interferometer (modified IFS 25, Bruker, Germany) for the
range of 4000–1000 cm− 1 and a dispersive spectrometer
Structural Insights from a Decoupled COX Variant
for that of 400–900 nm. Electrochemically induced difference spectra were recorded and processed as described in
Ref. 58.
X-ray data collection, processing, and refinement
The crystals were cryoprotected with 25% (v/v)
glycerol and flash frozen in gaseous cryo-stream cooled
by liquid nitrogen. X-ray diffraction data were collected on
synchrotron beamline X10SA of the Swiss Light Source in
Villigen, Switzerland. The collected data set was indexed
and processed with the program XDS80 to a resolution of
2.32 Å. Data collection statistics are listed in Table 2.
Recently, the structure of the cryo-cooled recombinant WT
of the two-subunit COX from P. denitrificans has been
refined (Koepke et al., unpublished results). Both data sets
were loaded into a single mtz file using the CCP4 program
CAD to truncate the native data set to exactly the same
reflections used for the variant structure.81 Writing the
data with MTZ2VARIOUS to an ASCII reflection file by
ignoring the MISS command, controlling the missing
reflection flag, leads to a file containing only entries that
are common to both data sets. Subsequently, the ASCII
file was reread with F2MTZ back into an mtz file for
further use in the CCP4 suite of programs.81 Refinement
in Refmac582 could be started directly from the WT
structure, with the side chain of residue 131 changed
according to the mutation to an aspartate. The structure
factors were calculated from the measured intensities
employing TRUNCATE of the CCP4 suite of programs.
Between each refinement round, 2Fo − Fc and Fo − Fc
electron density maps were inspected using the graphics
program XtalView.83 Water molecules were added to the
protein model at peaks N 3σ in the Fo − Fc difference
density map, when the geometry was suitable for
hydrogen bounding. In the membrane region, detergent
molecules have been fitted to difference density peaks
Table 2. Data collection statistics
Data set
1AR1
Species
Unit cell (Å)
a
93.53
b
151.04
c
156.70
Space group
Resolution (Å)a
2.7
7.4
Rsym (%)a
a
2.05
Multiplicity
Completeness 98.4
(%)a
I/σ(I)a
BWilson
23.5 (20.1)
Rfree (R) (%)
RMSD
Bond (Å)
Angle (°)
Solvent
63
molecules
Water
54
9
LDAOb
c
LMT
WT
N131D
P. denitrificans
83.40
84.54
150.47
151.40
157.19
157.66
P212121
2.29 (2.4–2.29)
2.32 (2.4–2.32)
11.4 (35.9)
7.7 (39.0)
4.9 (4.8)
10.6 (8.7)
99.1 (98.0)
98.8 (97.0)
17.0 (3.5)
31.42
27.7 (21.4)
19.1 (3.9)
46.81
24.2 (20.5)
0.028
2.704
579
0.027
2.488
396
555
10
14
375
9
12
a
Numbers in parentheses are values for the highest-resolution
shell.
b
Lauryl dimethylamine oxide.
c
DM.
Structural Insights from a Decoupled COX Variant
instead. Since DM has been exchanged against lauryl
dimethylamine oxide during purification, both detergent
molecules were found.
Accession code
Coordinates and structure factors of the N131D variant
structure have been deposited in the Research Collaboratory for Structural Bioinformatics Protein Data Bank with
accession code 3EHB.
Acknowledgements
This work was funded by Deutsche Forschungsgemeinschaft (SFB 472 and CEF EXC 115) and
supported by Max-Planck-Gesellschaft. We thank
E. Bamberg (Max Planck Institute for Biophysics,
Frankfurt, Germany) for kindly providing stoppedflow equipment and C. Bamann for assisting with
the device. We also thank the staff of beamline
X10SA at the Swiss Light Source for helping with
data collection.
References
1. Wikström, M. K. (1977). Proton pump coupled to
cytochrome c oxidase in mitochondria. Nature, 266,
271–273.
2. Humphrey, W., Dalke, A. & Schulten, K. (1996).
VMD: visual molecular dynamics. J. Mol. Graphics,
14, 33–38.
3. Iwata, S., Ostermeier, C., Ludwig, B. & Michel, H.
(1995). Structure at 2.8 Å resolution of cytochrome c
oxidase from Paracoccus denitrificans. Nature, 376,
660–669.
4. Ostermeier, C., Harrenga, A., Ermler, U. & Michel, H.
(1997). Structure at 2.7 Å resolution of the Paracoccus
denitrificans two-subunit cytochrome c oxidase complexed with an antibody Fv fragment. Proc. Natl Acad.
Sci. USA, 94, 10547–10553.
5. Tsukihara, T., Aoyama, H., Yamashita, E., Tomizaki,
T., Yamaguchi, H., Shinzawa-Itoh, K. et al. (1996). The
whole structure of the 13-subunit oxidized cytochrome c oxidase at 2.8 Å. Science, 272, 1136–1144.
6. Svensson-Ek, M., Abramson, J., Larsson, G., Tornroth,
S., Brzezinski, P. & Iwata, S. (2002). The X-ray crystal
structures of wild-type and EQ(I-286) mutant cytochrome c oxidases from Rhodobacter sphaeroides. J. Mol.
Biol. 321, 329–339.
7. Abramson, J., Riistama, S., Larsson, G., Jasaitis, A.,
Svensson-Ek, M., Laakkonen, L. et al. (2000). The
structure of the ubiquinol oxidase from Escherichia coli
and its ubiquinone binding site. Nat. Struct. Biol. 7,
910–917.
8. Yoshikawa, S., Shinzawa-Itoh, K., Nakashima, R.,
Yaono, R., Yamashita, E., Inoue, N. et al. (1998).
Redox-coupled crystal structural changes in bovine
heart cytochrome c oxidase. Science, 280, 1723–1729.
9. Mitchell, D. M., Fetter, J. R., Mills, D. A., Ädelroth, P.,
Pressler, M. A., Kim, Y. et al. (1996). Site-directed
mutagenesis of residues lining a putative proton
transfer pathway in cytochrome c oxidase from
Rhodobacter sphaeroides. Biochemistry, 35, 13089–13093.
875
10. Hosler, J. P., Ferguson-Miller, S., Calhoun, M. W.,
Thomas, J. W., Hill, J., Lemieux, L. et al. (1993).
Insight into the active-site structure and function of
cytochrome oxidase by analysis of site-directed mutants of bacterial cytochrome aa3 and cytochrome bo.
J. Bioenerg. Biomembr. 25, 121–136.
11. Thomas, J. W., Lemieux, L. J., Alben, J. O. & Gennis,
R. B. (1993). Site-directed mutagenesis of highly conserved residues in helix VIII of subunit I of the
cytochrome bo ubiquinol oxidase from Escherichia
coli: an amphipathic transmembrane helix that may
be important in conveying protons to the binuclear
center. Biochemistry, 32, 11173–11180.
12. Thomas, J. W., Puustinen, A., Alben, J. O., Gennis,
R. B. & Wikström, M. (1993). Substitution of asparagine for aspartate-135 in subunit I of the cytochrome bo ubiquinol oxidase of Escherichia coli eliminates proton-pumping activity. Biochemistry, 32,
10923–10928.
13. Fetter, J. R., Qian, J., Shapleigh, J., Thomas, J. W.,
García-Horsman, A., Schmidt, E. et al. (1995). Possible
proton relay pathways in cytochrome c oxidase. Proc.
Natl Acad. Sci. USA, 92, 1604–1608.
14. Verkhovskaya, M. L., García-Horsman, A., Puustinen,
A., Rigaud, J. L., Morgan, J. E., Verkhovsky, M. I. &
Wikström, M. (1997). Glutamic acid 286 in subunit I of
cytochrome bo3 is involved in proton translocation.
Proc. Natl. Acad. Sci. USA, 94, 10128–10131.
15. Pfitzner, U., Hoffmeier, K., Harrenga, A., Kannt, A.,
Michel, H., Bamberg, E. et al. (2000). Tracing the Dpathway in reconstituted site-directed mutants of
cytochrome c oxidase from Paracoccus denitrificans.
Biochemistry, 39, 6756–6762.
16. Pfitzner, U., Odenwald, A., Ostermann, T., Weingard,
L., Ludwig, B. & Richter, O. M. H. (1998). Cytochrome
c oxidase (heme aa3) from Paracoccus denitrificans:
analysis of mutations in putative proton channels of
subunit I. J. Bioenerg. Biomembr. 30, 89–97.
17. Gennis, R. B. (1992). Site-directed mutagenesis studies
on subunit I of the aa3-type cytochrome c oxidase of
Rhodobacter sphaeroides: a brief review of progress to
date. Biochim. Biophys. Acta, 1101, 184–187.
18. Gennis, R. B. (1998). Multiple proton-conducting pathways in cytochrome oxidase and a proposed role
for the active-site tyrosine. BBA/Bioenergetics, 1365,
241–248.
19. Brown, S., Moody, A. J., Mitchell, R. & Rich, P. R.
(1993). Binuclear centre structure of terminal protonmotive oxidases. FEBS Lett. 316, 216–223.
20. Namslauer, A., Lepp, H., Brändén, M., Jasaitis, A.,
Verkhovsky, M. I. & Brzezinski, P. (2007). Plasticity of
proton pathway structure and water coordination in
cytochrome c oxidase. J. Biol. Chem. 282, 15148–15158.
21. Konstantinov, A. A., Siletsky, S., Mitchell, D., Kaulen,
A. & Gennis, R. B. (1997). The roles of the two proton
input channels in cytochrome c oxidase from Rhodobacter sphaeroides probed by the effects of site-directed
mutations on time-resolved electrogenic intraprotein proton transfer. Proc. Natl Acad. Sci. USA,
94, 9085–9090.
22. Ruitenberg, M., Kannt, A., Bamberg, E., Ludwig, B.,
Michel, H. & Fendler, K. (2000). Single-electron
reduction of the oxidized state is coupled to proton
uptake via the K pathway in Paracoccus denitrificans
cytochrome c oxidase. Proc. Natl Acad. Sci. USA, 97,
4632–4636.
23. Ädelroth, P., Ek, M. S., Mitchell, D. M., Gennis, R. B. &
Brzezinski, P. (1997). Glutamate 286 in cytochrome aa
from Rhodobacter sphaeroides is involved in proton
876
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
uptake during the reaction of the fully-reduced enzyme with dioxygen. Biochemistry, 36, 13824–13829.
Ädelroth, P., Gennis, R. B. & Brzezinski, P. (1998). Role
of the pathway through K(I-362) in proton transfer in
cytochrome c oxidase from R. sphaeroides. Biochemistry,
37, 2470–2476.
Stuchebrukhov, A. (2003). Electron transfer reactions
coupled to proton translocation. Cytochrome oxidase,
proton pumps, and biological energy transduction.
J. Theor. Comput. Chem. 2, 91–118.
Nagle, J. F. & Morowitz, H. J. (1978). Molecular mechanisms for proton transport in membranes. Proc.
Natl Acad. Sci. USA, 75, 298–302.
Pomès, R. & Roux, B. (1998). Free energy profiles for
H+ conduction along hydrogen-bonded chains of
water molecules. Biophys. J. 75, 33–40.
Roux, B. (1999). Proton wires are different. Biophys. J.
77, 2331–2332.
Olkhova, E., Hutter, M. C., Lill, M. A., Helms, V. &
Michel, H. (2004). Dynamic water networks in cytochrome c oxidase from Paracoccus denitrificans investigated by molecular dynamics simulations. Biophys. J.
86, 1873–1889.
Cukier, R. I. (2004). Quantum molecular dynamics
simulation of proton transfer in cytochrome c oxidase.
Biochim. Biophys. Acta, 1656, 189–202.
Agmon, N. (1995). The Grotthuss mechanism. Chem.
Phys. Lett. 244, 456–462.
Wikström, M., Verkhovsky, M. I. & Hummer, G.
(2003). Water-gated mechanism of proton translocation by cytochrome c oxidase. Biochim. Biophys. Acta,
1604, 61–65.
Hofacker, I. & Schulten, K. (1998). Oxygen and proton pathways in cytochrome c oxidase. Proteins, 30,
100–107.
Puustinen, A. & Wikström, M. (1999). Proton exit from
the heme-copper oxidase of Escherichia coli. Proc. Natl
Acad. Sci. USA, 96, 35–37.
Michel, H., Behr, J., Harrenga, A. & Kannt, A. (1998).
Cytochrome c oxidase: structure and spectroscopy.
Annu. Rev. Biophys. Biomol. Struct. 27, 329–356.
Jünemann, S., Meunier, B., Fisher, N. & Rich, P. R.
(1999). Effects of mutation of the conserved
glutamic acid-286 in subunit I of cytochrome c
oxidase from Rhodobacter sphaeroides. Biochemistry,
38, 5248–5255.
Behr, J., Michel, H., Mäntele, W. & Hellwig, P. (2000).
Functional properties of the heme propionates in
cytochrome c oxidase from Paracoccus denitrificans.
Evidence from FTIR difference spectroscopy and sitedirected mutagenesis. Biochemistry, 39, 1356–1363.
Riistama, S., Hummer, G., Puustinen, A., Dyer, R. B.,
Woodruff, W. H. & Wikström, M. (1997). Bound water
in the proton translocation mechanism of the haemcopper oxidases. FEBS Lett. 414, 275–280.
Pomès, R., Hummer, G. & Wikström, M. (1998). Structure and dynamics of a proton shuttle in cytochrome c
oxidase. Biochim. Biophys. Acta, 1365, 255–260.
Zheng, X., Medvedev, D. M., Swanson, J. &
Stuchebrukhov, A. A. (2003). Computer simulation
of water in cytochrome c oxidase. Biochim. Biophys.
Acta, 1557, 99–107.
Michel, H. (1998). The mechanism of proton pumping
by cytochrome c oxidase. Proc. Natl Acad. Sci. USA, 95,
12819–12824.
Mills, D. A. & Ferguson-Miller, S. (2002). Influence of
structure, pH and membrane potential on proton
movement in cytochrome oxidase. Biochim. Biophys.
Acta, 1555, 96–100.
Structural Insights from a Decoupled COX Variant
43. Mills, D. A., Florens, L., Hiser, C., Qian, J. & FergusonMiller, S. (2000). Where is ‘outside’ in cytochrome c
oxidase and how and when do protons get there?
Biochim. Biophys. Acta, 1458, 180–187.
44. Brzezinski, P. & Larsson, G. (2003). Redox-driven
proton pumping by heme-copper oxidases. Biochim.
Biophys. Acta, 1605, 1–13.
45. Busenlehner, L. S., Brändén, G., Namslauer, I., Brzezinski, P. & Armstrong, R. N. (2008). Structural elements
involved in proton translocation by cytochrome c
oxidase as revealed by backbone amide hydrogendeuterium exchange of the E286H mutant. Biochemistry,
47, 73–83.
46. Gorbikova, E. A., Belevich, N. P., Wikström, M. &
Verkhovsky, M. I. (2007). Time-resolved ATR–FTIR
spectroscopy of the oxygen reaction in the D124N
mutant of cytochrome c oxidase from Paracoccus
denitrificans. Biochemistry, 46, 13141–13148.
47. Tuukkanen, A., Kaila, V. R., Laakkonen, L., Hummer,
G. & Wikström, M. (2007). Dynamics of the glutamic
acid 242 side chain in cytochrome c oxidase. Biochim.
Biophys. Acta, 1767, 1102–1106.
48. Pawate, A. S., Morgan, J., Namslauer, A., Mills, D.,
Brzezinski, P., Ferguson-Miller, S. & Gennis, R. B.
(2002). A mutation in subunit I of cytochrome oxidase
from Rhodobacter sphaeroides results in an increase in
steady-state activity but completely eliminates proton
pumping. Biochemistry, 41, 13417–13423.
49. Lepp, H., Salomonsson, L., Zhu, J. P., Gennis, R. B. &
Brzezinski, P. (2008). Impaired proton pumping in
cytochrome c oxidase upon structural alteration of the
D pathway. Biochim. Biophys. Acta, 1777, 897–903.
50. Han, D., Namslauer, A., Pawate, A., Morgan, J. E.,
Nagy, S., Vakkasoglu, A. S. et al. (2006). Replacing
Asn207 by aspartate at the neck of the D channel in the
aa(3)-type cytochrome c oxidase from Rhodobacter
sphaeroides results in decoupling the proton pump.
Biochemistry, 45, 14064–14074.
51. Siletsky, S. A., Pawate, A. S., Weiss, K., Gennis, R. B. &
Konstantinov, A. A. (2004). Transmembrane charge
separation during the ferryl-oxo → oxidized transition
in a nonpumping mutant of cytochrome c oxidase.
J. Biol. Chem. 279, 52558–52565.
52. Namslauer, A., Pawate, A. S., Gennis, R. B. &
Brzezinski, P. (2003). Redox-coupled proton translocation in biological systems: proton shuttling in
cytochrome c oxidase. Proc. Natl Acad. Sci. USA, 100,
15543–15547.
53. Xu, J. & Voth, G. A. (2006). Free energy profiles for
H+cytochrome c oxidase: a study of the wild type and
N98D mutant enzymes. Biochim. Biophys. Acta, 1757,
852–859.
54. Olsson, M. H. & Warshel, A. (2006). Monte Carlo
simulations of proton pumps: on the working principles of the biological valve that controls proton
pumping in cytochrome c oxidase. Proc. Natl Acad. Sci.
USA, 103, 6500–6505.
55. Olkhova, E., Helms, V. & Michel, H. (2005). Titration
behavior of residues at the entrance of the D-pathway
of cytochrome c oxidase from Paracoccus denitrificans
investigated by continuum electrostatic calculations.
Biophys. J. 89, 2324–2331.
56. Fadda, E., Yu, C. H. & Pomes, R. (2008). Electrostatic
control of proton pumping in cytochrome c oxidase.
Biochim. Biophys. Acta, 1777, 277–284.
57. Brändén, G., Pawate, A. S., Gennis, R. B. & Brzezinski,
P. (2006). Controlled uncoupling and recoupling of
proton pumping in cytochrome c oxidase. Proc. Natl
Acad. Sci. USA, 103, 317–322.
Structural Insights from a Decoupled COX Variant
58. Hellwig, P., Behr, J., Ostermeier, C., Richter, O. M. H.,
Pfitzner, U., Odenwald, A. et al. (1998). Involvement
of glutamic acid 278 in the redox reaction of the
cytochrome c oxidase from Paracoccus denitrificans
investigated by FTIR spectroscopy. Biochemistry, 37,
7390–7399.
59. Lübben, M., Prutsch, A., Mamat, B. & Gerwert, K.
(1999). Electron transfer induces side-chain conformational changes of glutamate-286 from cytochrome bo3.
Biochemistry, 38, 2048–2056.
60. Tsukihara, T., Shimokata, K., Katayama, Y., Shimada,
H., Muramoto, K., Aoyama, H. et al. (2003). The lowspin heme of cytochrome c oxidase as the driving
element of the proton-pumping process. Proc. Natl
Acad. Sci. USA, 100, 15304–15309.
61. Popovic, D. M. & Stuchebrukhov, A. A. (2006). Two
conformational states of Glu242 and pKas in bovine
cytochrome c oxidase. Photochem. Photobiol. Sci. 5,
611–620.
62. Namslauer, A. & Brzezinski, P. (2004). Structural
elements involved in electron-coupled proton transfer
in cytochrome c oxidase. FEBS Lett. 567, 103–110.
63. Kaila, V. R., Verkhovsky, M. I., Hummer, G. &
Wikström, M. (2008). Glutamic acid 242 is a valve in
the proton pump of cytochrome c oxidase. Proc. Natl
Acad. Sci. USA, 105, 6255–6259.
64. Kaila, V. R., Verkhovsky, M., Hummer, G. & Wikström,
M. (2008). Prevention of leak in the proton pump of
cytochrome c oxidase. Biochim. Biophys. Acta, 1777,
890–892.
65. Namslauer, A., Aagaard, A., Katsonouri, A. &
Brzezinski, P. (2003). Intramolecular proton-transfer
reactions in a membrane-bound proton pump: the
effect of pH on the peroxy to ferryl transition in cytochrome c oxidase. Biochemistry, 42, 1488–1498.
66. Ernst, J. A., Clubb, R. T., Zhou, H. X., Gronenborn,
A.M. & Clore, G. M. (1995). Demonstration of
positionally disordered water within a protein hydrophobic cavity by NMR. Science, 267, 1813–1817.
67. Yu, B., Blaber, M., Gronenborn, A. M., Clore, G. M. &
Caspar, D. L. (1999). Disordered water within a
hydrophobic protein cavity visualized by x-ray
crystallography. Proc. Natl Acad. Sci. USA, 96, 103–108.
68. Vakkasoglu, A. S., Morgan, J. E., Han, D., Pawate, A. S.
& Gennis, R. B. (2006). Mutations which decouple
the proton pump of the cytochrome c oxidase from
Rhodobacter sphaeroides perturb the environment of
glutamate 286. FEBS Lett. 580, 4613–4617.
69. Wikström, M. & Verkhovsky, M. I. (2007). Mechanism
and energetics of proton translocation by the respiratory heme-copper oxidases. Biochim. Biophys. Acta,
1767, 1200–1214.
70. Popovic, D. M. & Stuchebrukhov, A. A. (2004).
Electrostatic study of the proton pumping mechanism
in bovine heart cytochrome c oxidase. J. Am. Chem.
Soc. 126, 1858–1871.
877
71. Ludwig, B. (1986). Cytochrome c oxidase from Paracoccus denitrificans. Methods Enzymol. 126, 153–159.
72. Lucioli, S., Hoffmeier, K., Carrozzo, R., Tessa, A.,
Ludwig, B. & Santorelli, F. M. (2006). Introducing a
novel human mtDNA mutation into the Paracoccus
denitrificans COX I gene explains functional deficits in
a patient. Neurogenetics, 7, 51–57.
73. Kleymann, G., Ostermeier, C., Ludwig, B., Skerra, A.
& Michel, H. (1995). Engineered Fv fragments as a tool
for the one-step purification of integral multisubunit
membrane protein complexes. Biotechnology (NY), 13,
155–160.
74. Witt, H., Zickermann, V. & Ludwig, B. (1995). Sitedirected mutagenesis of cytochrome c oxidase reveals
two acidic residues involved in the binding of cytochrome c. Biochim. Biophys. Acta, 1230, 74–76.
75. Darley-Usmar, V. M., Capaldi, R. A., Takamiya, S.,
Millett, F., Wilson, M. T., Malatesta, F. & Sarti, P. (1987).
Mitochondria—A Practical Approach. pp. 143–152, IRL
Press/Oxford, Washington, DC.
76. Kannt, A., Soulimane, T., Buse, G., Becker, A.,
Bamberg, E. & Michel, H. (1998). Electrical current
generation and proton pumping catalyzed by the ba3type cytochrome c oxidase from Thermus thermophilus.
FEBS Lett. 434, 17–22.
77. Richter, O. M. H., Dürr, K. L., Kannt, A., Ludwig, B.,
Scandurra, F. M., Giuffre, A. et al. (2005). Probing the
access of protons to the K pathway in the Paracoccus
denitrificans cytochrome c oxidase. FEBS J. 272,
404–412.
78. Sarti, P., Jones, M. G., Antonini, G., Malatesta, F.,
Colosimo, A., Wilson, M. T. & Brunori, M. (1985).
Kinetics of redox-linked proton pumping activity of
native and subunit III-depleted cytochrome c oxidase:
a stopped-flow investigation. Proc. Natl Acad. Sci.
USA, 82, 4876–4880.
79. Moss, D., Nabedryk, E., Breton, J. & Mäntele, W. (1990).
Redox-linked conformational changes in proteins
detected by a combination of infrared spectroscopy
and protein electrochemistry. Evaluation of the technique with cytochrome c. Eur. J. Biochem. 187, 565–572.
80. Kabsch, W. (1993). Automatic processing of rotation
diffraction data from crystals of initially unknown
symmetry and cell constants. J. Appl. Crystallogr. 26,
795–800.
81. Dodson, E. J., Winn, M. & Ralph, A. (1997). Collaborative Computational Project, Number 4: providing
programs for protein crystallography. Methods Enzymol.
277, 620–633.
82. Murshudov, G. N., Vagin, A. A., Lebedev, A., Wilson,
K. S. & Dodson, E. J. (1999). Efficient anisotropic
refinement of macromolecular structures using FFT.
Acta Crystallogr., Sect. D: Biol. Crystallogr. 55, 247–255.
83. McRee, D. E. (1999). XtalView/Xfit—a versatile program for manipulating atomic coordinates and electron density. J. Struct. Biol. 125, 156–165.