Two Residues of a Conserved Aromatic Ladder of the Mitochondrial ADP/ATP
Carrier Are Crucial to Nucleotide Transport†
Claudine David,‡Bertrand Arnou,‡,§Jean-Fre ´de ´ric Sanchez,‡Ludovic Pelosi,|Ge ´rard Brandolin,|
Guy J.-M. Lauquin,‡and Ve ´ronique Tre ´ze ´guet*,‡
Laboratoire de Physiologie Mole ´culaire et Cellulaire, Institut de Biochimie et Ge ´ne ´tique Cellulaires, UMR 5095,
CNRS-UniVersite ´ Bordeaux 2, 1, rue Camille Saint-Sae ¨ns, F-33077 Bordeaux Cedex, France, and Laboratoire de Biochimie et
Biophysique des Syste `mes Inte ´gre ´s (LBBSI), Institut de Recherches en Technologies et Sciences du ViVant (iRTSV), UMR 5092,
CNRS-CEA-UniVersite ´ Joseph Fourier, F-38054 Grenoble Cedex 9, France
ReceiVed July 4, 2008; ReVised Manuscript ReceiVed October 24, 2008
ABSTRACT: The mitochondrial ADP/ATP carrier is the paradigm of the mitochondrial carrier family (MCF),
whose members are crucial for cross-talks between mitochondria, where cell energy is mainly produced,
and the cytosol, where cell energy is mainly consumed. These carriers share structural and functional
characteristics. Resolution of the 3D structure of the beef mitochondrial ADP/ATP carrier, in a complex
with one of its specific inhibitors, revealed interesting features and suggested the involvement of some
particular residues in substrate binding and transfer from the outside to the inside of mitochondria. To
ascertain the role of these residues, namely, Y186, Y190, F191, and Y194, they were mutated into alanine
in the yeast mitochondrial ADP/ATP carrier at equivalent positions (Y203, Y207, F208, and Y211). Two
residues, Y203 and F208, appeared to be crucial for transport activity but not for substrate binding per se,
indicating their involvement in the substrate transfer process through the carrier. Furthermore, it was
possible to show that these mutations precluded conformational changes of the matrix loop m2, whose
movements were demonstrated to participate in substrate transport by the wild-type carrier. Therefore,
these aromatic residues may be involved in substrate gliding, and they may also confer specificity toward
adenine nucleotides for the ADP/ATP carrier as compared with the MCF members.
The three-dimensional structure of the bovine mitochon-
drial adenine nucleotide carrier (BAncp),1a model member
of the mitochondrial carrier family (MCF), has been solved
by X-ray crystallography at 2.2 Å resolution (1). The
structure is that of a monomer in complex with carboxyatrac-
tyloside (CATR), a highly specific inhibitor of Ancp to which
it binds with very high affinity. Ancp has the general shape
of a basket delineated by six tilted transmembrane segments
It has been postulated on the basis of numerous biochemi-
cal data that Ancp changes its conformation to perform
nucleotide transport (ref 2 and references therein). The
conformation deciphered by X-ray analyses opens to the
intermembrane space (IMS). Bongkrekic acid (BA) and
isoBA are the second class of inhibitors that are very specific
to Ancp. They bind with very high affinity but only to the
matrix side of the carrier, in contrast to CATR and ATR
(the decarboxylated form of CATR), which bind only to the
IMS side. ATR and CATR prevent transport of ADP added
on the cytosolic side, and biochemical studies have suggested
that ATR/CATR and nucleotide binding sites partially
overlap (3). Therefore, and given that Ancp transports only
the magnesium-free forms of ADP and ATP (4, 5), analyses
of residues located in the Ancp cavity suggest the mechanism
of nucleotide binding, at least from the IMS side, and
In the following, numbering of BAnc1p amino acid
residues starts after the initiating methionine, and numbering
of ScAnc2p starts at the initiating methionine to avoid
renumbering of already characterized mutations of this carrier
(for example, R96H of op1 (6)). Basic residues at the
entrance of the cavity (K22, R79, and R279 of the bovine
Anc1p or BAnc1p) may attract nucleotides that would glide
further with their adenine ring along a “tyrosine ladder”
(Y186, Y190, and Y194 in TMS 4) to reach a second patch
of basic residues (K32, R137, and R234) (1, 7). These six
basic residues are very well conserved in the 116 known
†This work was supported by the University of Bordeaux 2, the
Centre National de la Recherche Scientifique, and the Re ´gion Aquitaine.
B.A. was supported by the French Ministe `re de l’Enseignement
Supe ´rieur et de la Recherche. J.-F.S. was awarded a grant from the
Fondation pour la Recherche Me ´dicale.
* To whom correspondence should be addressed. E-mail:
firstname.lastname@example.org. Phone: (33) 556 99 90 39. Fax:
(33) 556 99 90 63.
‡CNRS-Universite ´ Bordeaux 2.
§Present address: CEA, iBiTecS (Institut de Biologie et Technologies
de Saclay), CNRS, URA 2096, Universite ´ Paris-Sud 11, LRA 17V,
F-91191 Gif-sur-Yvette, France.
|CNRS-CEA-Universite ´ Joseph Fourier.
1Abbreviations: ANC, mitochondrial adenine nucleotide carrier
(ADP/ATP carrier) encoding gene; Ancp, adenine nucleotide carrier
(ADP/ATP carrier); ATR, atractyloside; BAnc1p, isoform 1 of bovine
Ancp; BA, bongkrekic acid; EMA, eosin-5-maleimide; IMS, inter-
membrane space; MCF, mitochondrial carrier family; MIM, mitochon-
drial inner membrane; N-ADP, naphthoyl-ADP; Sc, Saccharomyces
cereVisiae; ScAnc2p, isoform 2 of Saccharomyces cereVisiae Ancp;
SDS-PAGE, sodium dodecyl sulfate-polyacrylamide gel electro-
phoresis; TMS, transmembrane segment; YPD, rich yeast extract
peptone dextrose medium.
Biochemistry 2008, 47, 13223–1323113223
10.1021/bi8012565 CCC: $40.75
2008 American Chemical Society
Published on Web 11/17/2008
Ancp amino acid sequences (8). Their mutations at equivalent
positions of ScAnc2p into alanine or isoleucine inactivate
mitochondrial function in yeast (9-12). Next to Y190 of
the BAnc1p lies F191, which is perfectly conserved in all
known Ancp sequences. Therefore, the “tyrosine ladder” is
renamed the “aromatic ladder” and consists of Y186, Y190,
F191, and Y194 of BAnc1p. We showed in a previous study
that in ScAnc2p the double replacement of Y203 (equivalent
to Y186) and Y207 (equivalent to Y190) with alanine
strongly impaired yeast growth on nonfermentable carbon
sources while replacement with phenylalanine did not (1).
To assess more precisely the role of the aromatic ladder
residues located in the C-terminal end of TMS 4 in nucleotide
gliding during transport, they were sequentially replaced with
alanine either one at a time or in pairs in ScAnc2p (Y203,
Y207, F208, and Y211, Figure 1A). ATR binding properties
were determined with isolated mitochondria, as well as
nucleotide binding, which was examined using of a non-
transportable ADP analogue, 3′-O-(1-naphthoyl)adenosine 5′-
diphosphate (N-ADP). The results suggest that the aromatic
patch residues are not involved in nucleotide binding per se
but rather in substrate gliding via conformational changes
We have then generated a cysteine-less variant of ScAnc2p
(Anc2CLp) into which a single cysteine residue was intro-
duced at position 176 (Anc2CLpV176C) located in the small
R-helix of the matrix loop m2 at a position equivalent to
C159 of BAnc1p (Figure 1A). Indeed, movements of the
matrix loop m2 during nucleotide transport have been
revealed from various experimental approaches: limited
and photolabeling of the carrier with photoactivatable
analogues of atractyloside and adenine nucleotides (18-20).
It was proposed that the N-terminal part of m2 is accessible
from the IMS and that the C-terminal part is exposed to the
mitochondrial matrix in the Ancp·CATR complex (“CATR”
conformation) (Figure 1B). Conversely, in the Ancp·BA
complex, the C-terminal part of m2 is accessible from the
IMS, and the N-terminal part is exposed to the matrix (“BA”
conformation) (Figure 1B) (14).
The single cysteine variant Anc2CLpV176Cis active.
EMA labels V176C in the absence or in the presence of
ADP or BA. In contrast, CATR prevents this labeling,
suggesting that large conformational changes of the matrix
loop m2 may take place upon ligand binding. The Y203A
and F208A mutations were introduced into Anc2CLpV176C
(Anc2CLpV176CY203Aand Anc2CLpV176CF208A), and these vari-
ants were never labeled with EMA, whatever the added
ligand, indicating that the Y203A and F208A mutations
might preclude some of the conformational changes essential
for ADP/ATP transport, thus pointing to the role of these
residues in nucleotide gliding.
MATERIALS AND METHODS
Chemicals. [3H]Atractyloside ([3H]ATR) and 3′-O-(1-
naphthoyl)adenosine 5′-diphosphate (N-ADP) were synthe-
sized as previously described (21, 22). Protein concentration
was determined using the bicinchoninic acid reagent kit from
Sigma. Nucleotides and CATR were purchased from Sigma,
and P1,P5-di(adenosine-5′)-pentaphosphate was from Cal-
biochem. Hexokinase/glucose-6-phosphate dehydrogenase
enzyme mix was obtained from Roche Diagnostics GmbH.
Strains, Media, and Transformation. Escherichia coli strain
used for plasmid propagation was XL1-Blue (recA1 endA1
gyrA96 (Nalr) thi hsdR17 (rK-mK+) supE44 relA1 lac-F′
[Tn10 (tetr) proAB+lacIqlacZ∆M15]). Bacteria were
transformed according to standard methods either with
calcium chloride or by electroporation, as already described
(23). The following Saccharomyces cereVisiae strains were
used in this study: JL1∆2∆3u-(MATR leu2-3,112 his3-11,15
ade2-1 trp1-1 ura3-1 can1-100 anc1::LEU2 ∆anc2::HIS3
∆anc3) (this work); JL1-3ANC2 (MATR leu2-3,112 his3-
which refers to the 2N1-3 strain (24). The JL1∆2∆3u-strain
was obtained after replacement by homologous recombina-
tion of the interrupted anc3 gene of JL1-3∆2 (23) with a
fragment corresponding to the deletion of ScANC3 nucle-
otides +269 (HindIII site) to +975 (ClaI site), the STOP
codon being at position +922. This fragment contained 48
FIGURE 1: (A) Modeling of the ScAnc2p 3D structure in complex
with CATR. A 3D model of Anc2p from S. cereVisiae based on
the template structure of BAnc1p (PDB entry 1okc (1)) was
constructed as indicated in Materials and Methods. Only part of
TMS 3 and 4 (H3 and H4) are represented. The carboxyatractyloside
ligand was inserted using the docking program AutoDock3. The
“tyrosine ladder” described in the beef Anc1p 3D structure
corresponds to residues Y186, Y190, and Y194 to which F191 is
added to constitute the “aromatic ladder”, corresponding to Y203,
Y207, F208, and Y211 of ScAnc2p. CATR ) carboxyatractyloside.
(B) Movements of the matrix loop m2 upon inhibitor binding (from
ref 14). Only transmembrane helices 3 and 4 and the small helix
h34 of the matrix loop m2 are represented. The stars represent
residues 162 (gray) and 176 (black). Serine 162 was mutated into
a cysteine in ref 14, and V176 was mutated into cysteine in this
study. Numbering of ScAnc2p amino acid residues starts at the
initiating methionine and thus corresponds to numbering of the gene
codons. In ref 14, S162 is numbered S161 since amino acid
numbering in this paper corresponds to that of the mature protein
whose initiating methionine has been removed (20).
Biochemistry, Vol. 47, No. 50, 2008
David et al.
bp of the 5′ noncoding region of ScANC3 followed by the
first 269 bp of the ScANC3 sequence and 324 bp of the 3′
noncoding region (from the ClaI site to the BamHI site of
the ScANC3 gene). Removal of the URA3 gene in the anc3
locus was controlled by the ability of the recombinant strain
to grow in the presence of 5-fluoroorotic acid (25).
The strains were cultivated and transformed as described
in ref 23. The composition of all the media used (YPD,
YPLact, SGal-W) is described in ref 23.
Site-Directed Mutagenesis. Site-directed mutagenesis of
ScANC2 was performed using the Transformer site-directed
mutagenesis kit (Clontech Laboratories) with the following
mutagenic primers (mutated bases are underlined and amino
acid substitution is indicated in parentheses):5985′-ATTGT-
The mutated ScANC2 genes were subcloned into a
centromeric plasmid, pRS314, under the control of ScANC2
regulatory sequences as described in ref 23. The resulting
plasmids were used to transform the JL1∆2∆3u-strain to
assess their ability to complement the Scanc2 deletion.
Isolation of Mitochondria, Kinetic Measurements, and
Immunodetection. Yeast cells used for isolation of mito-
chondria were grown in minimal synthetic medium consisting
of 0.17% yeast nitrogen base without amino acids and
ammonium sulfate, 0.5% (NH4)2SO4, 2% galactose, and all
amino acids except tryptophan (SGal-W) to allow the plasmid
to be maintained in transformants. The protocols and
materials used to perform isolation of mitochondria, ADP/
ATP transport, [3H]ATR binding measurements, and protein
immunostaining are described in ref 23. Free ADP concen-
trations were calculated using WEBMAXC STANDARD
tm). After ECL detection with a GENE GNOME system
(Syngene, Ozyme), signal intensities of the bands were
quantified with Genetools, a Syngene software. The time
course of CATR-induced release of bound 3′-O-(1-naph-
thoyl)adenosine 5′-diphosphate (N-ADP) binding was studied
by incubating freshly isolated mitochondria in 2 mL of 120
mM KCl, 10 mM MOPS, pH 6.8, and 1 mM EDTA at 16
°C (0.5 mg of mitochondrial proteins/mL). After N-ADP
addition (from 0.1 to 10 µM) the fluorescence level was set
to zero. Then the increase in fluorescence induced upon
addition of 5 µL of 2 mM CATR (∆F) was recorded until it
became stable. The apparent dissociation constant, K1/2, was
determined after plotting ∆F/∆Fmaxas a function of added
EMA Labeling of Membrane-Embedded ScAnc2p Variants.
Mitochondria were isolated in buffer A (0.6 M mannitol, 10
mM Tris·HCl, 0.1 mM Na2EGTA, pH 7.4), resuspended in
buffer B (0.6 M mannitol, 10 mM MOPS, 0.1 mM
Na2EGTA, pH 6.8) at 8 mg of proteins/mL (final concentra-
tion), and incubated for 15 min at 4 °C in the absence or in
the presence of 20 µM CATR or 20 µM BA. They were
then diluted twice with buffer A and incubated in the dark
for 30 min at 4 °C in the presence of 200 µM EMA. Labeling
was stopped by 20 mM DTT (10 min, 4 °C, in the dark).
After addition of 1 volume of 2× electrophoresis sample
buffer (125 mM Tris·HCl, pH 6.8, 20% glycerol, 4% SDS,
0.2% 2-mercaptoethanol, 0.001% bromophenol blue), samples
were incubated for 2 min at 70 °C and then loaded on a
20% polyacrylamide gel (SDS-PAGE). After migration,
fluorescence was detected at λex ) 532 nm with the
fluorescent image analyzer Fujifilm FLA-500.
Construction of a Three-Dimensional Model Structure for
ScAnc2p. A pairwise alignment between ScAnc2p and
bAnc1p sequences was constructed using ClustalX (26), and
manual refinement was based on multiple sequence alignment
of 116 Ancp sequences generated with ClustalX. A three-
dimensional model of ScAnc2p was built using the homology
modeling program Modeller (27) and the template structure
of bAnc1p (PDB entry: 1okc). We used the AutoDock 3.05
package (28) for ligand docking. The CATR structure was
generated using the prodrg server (http://davapc1.bioch.dundee.
ac.uk/prodrg/ (29)) and AutoDockTools program (30). Blind
docking was carried out using the Lamarckian genetic
algorithm for 512 trials and with a rmsd tolerance of 0.5 Å
for clustering and 2 Å for reclustering.
Growth Phenotypes on a Nonfermentable Carbon Source
Suggest Differing InVolVement of the Aromatic Ladder
Residues. Y186 of the BAnc1p aromatic ladder is conserved
in 100% of the Ancp amino acid sequences, Y190 in 98%,
F191 in 100%, and Y194 in 87%. In the remaining cases
they are replaced with phenylalanine, another aromatic amino
acid (8). Mutation of Y203 and Y207 of ScAnc2p (equivalent
to Y186 and Y190 of BAnc1p) (Anc2pY203FY207F) into
phenylalanine did not prevent JL1∆2∆3u-growth on non-
fermentable carbon sources (Figure 2, line 3) but conferred
cold sensitivity on lactate. In contrast, when both residues
were mutated simultaneously into alanine (Anc2pY203AY207A),
yeast growth on nonfermentable carbon sources was impaired
(Figure 2, line 4, and ref 1), and the transformants were
temperature (12 and 37 °C) sensitive on glucose (data not
shown). Thus, one or both of these residues may play a
crucial role in nucleotide transport.
The role of the aromatic ladder was further investigated
by mutating the four residues of ScAnc2p into alanine, in
pairs or one at a time, namely, Y203, Y207, F208, and Y211.
Introduction of two alanines simultaneously prevented growth
on lactate at 28 °C (Figure 2, lines 4-6) and conferred
temperature sensitivity on glucose (data not shown). Mutation
into alanine of one residue at a time clearly showed that Y207
was dispensable to ScAnc2p function (Figure 2, line 8), in
contrast to Y203 and F208. Indeed, ScAnc2pY203Aand
ScAnc2pF208Astrains exhibited a growth defect on lactate at
28 °C (Figure 2, lines 7 and 9) and 12 or 37 °C (not shown),
while the ScAnc2pY211Astrain displayed a reduced growth
phenotype at 28 °C (Figure 2, line 10) and no growth at 12
or 37 °C (not shown).
Mitochondrial ADP/ATP Carrier and Nucleotide Transport
Biochemistry, Vol. 47, No. 50, 2008 13225
All Variant Genes Lead to ScAnc2p Production in Mito-
chondria. In the cases of ScAnc2pY203Aand ScAnc2pF208A
variants, the lactate minus growth phenotype might reflect
the lack of ScAnc2p variants in mitochondria should these
proteins be unstable either before or after their insertion into
MIM. Amounts of carrier proteins in mitochondria were first
evaluated from immunostaining experiments. Ten micro-
grams of proteins from mitochondria isolated from
JL1∆2∆3u-transformed with the plasmids encoding the
variants was loaded on a 12% acrylamide gel. After
SDS-PAGE and transfer, the nitrocellulose membrane was
first immunostained with an antibody raised against the last
14 amino acids of ScAnc2p (1/10000) (31). After ECL
detection, the membrane was immunodecorated with an
antibody raised against yeast porin (1/10000) (32). The signal
intensities were quantified and plotted as ScAnc2p/porin
ratios (Figure 3). ScAnc2p was present in every transformant,
but its relative amount varied with the variant. The relative
amounts of the more active variants, ScAnc2pY203FY207Fand
ScAnc2pY207A, were similar to that of the wild type, where-
as ScAnc2pY211Aand the inactive forms were less abundant.
Importantly, however, all of the variant transporters, even
the inactive ones, were present in mitochondria in reasonable
To quantify ScAnc2p variants more precisely, we per-
formed [3H]ATR binding experiments with isolated mito-
chondria. Results are given in Table 1 and roughly correlate
with those of Figure 3, when considering the maximum
number of binding sites (BmaxATR) for two variants,
ScAnc2pF208Aand ScAnc2pY211A. However, they were quite
different for ScAnc2pY203Aand ScAnc2pY207A. The BmaxATR
values were higher than that of the wild type whereas the
relative amounts as evidenced by immunodecoration were
lower (ScAnc2pY203A) or similar (ScAnc2pY207A).
ATR binding experiments may be used to estimate the
absolute amount of properly folded ScAnc2p in mitochondria
while immunodecoration indicates only the amount of
ScAnc2p relative to porin. The greatest divergence was
observed with the active variant ScAnc2pY207A. This could
mean that, in cells expressing this variant, the amount of
mitochondrial proteins is modified.
Some of the ATR dissociation constant values were
dramatically changed. They decreased by a factor of 5-7
for the ScAnc2pF208Aand ScAnc2pY211Avariants, while they
increased by a factor of 6 for the ScAnc2pY203Avariant. This
suggests a large decrease in ATR affinity for the latter, which
could correlate with a decrease in nucleotide binding ability.
Indeed, numerous studies have suggested overlapping of
ATR and ADP binding sites, at least partially.
Some of the Rungs of the Aromatic Ladder Are Crucial to
Nucleotide Translocation or Binding. In a previous paper
we showed that the op1 mutation of ScAnc2p (R96H),
FIGURE 2: Yeast growth phenotype associated with aromatic ladder
residue mutations. JL1∆2∆3u-strain was transformed with plasmid
pRS314 containing no gene (line 1), the wild-type ScANC2 gene
(line 2), or the various Scanc2 mutated genes (lines 3-10), under
the control of ScANC2 regulating sequences. Yeast transformants
were isolated and inoculated in liquid complete minimum medium
free of tryptophan containing 2% glucose as carbon source (SC-
W). When the cultures reached log phase, cells were diluted to
obtain 104, 103, 102, or 10 cells per 2 µL and then spotted onto
SC-W (glucose) or rich lactate-containing medium (lactate) plates
and incubated at 12, 28, or 37 °C. For the sake of clarity, only
results obtained after 3 days (glucose) or 6 days (lactate) of
incubation at 28 °C are shown.
FIGURE 3: Relative ScAnc2p content determined after immunostain-
ing of mitochondria proteins. Mitochondria were isolated from cells
grown in SGal-W. Ten micrograms of mitochondrial proteins was
subjected to SDS-PAGE and transferred onto nitrocellulose
membrane which was then immunodecorated with an antibody
raised against the last 14 amino acids of ScAnc2p (1/10000) and
then with an antibody raised against yeast porin (1/10000).
Intensities of the signals were quantified with GeneTools from
Syngene after each revelation. The ratio ScAnc2p/porin in arbitrary
units was set at 1 for the wild-type protein. Data are the means of
two different experiments.
Table 1: [3H]ATR Binding Parameters and Naphthoyl-ADP Dissociation
337 ( 27
445 ( 236
392 ( 21
637 ( 85
264 ( 9
211 ( 12
327 ( 26
226 ( 116
2002 ( 245
177 ( 35
68 ( 12
45 ( 13
1.7 ( 0.4
2.6 ( 0.2
1.8 ( 0.2
aFrom ref 33.bVarious [3H]ATR concentrations were incubated with
isolated mitochondria (1 mg). After incubation on ice (45 min),
mitochondria were pelleted, washed, and lyzed with 5% Triton X-100
and 0.5 M NaCl. Radioactivity associated with the pellet was counted to
calculate the amount of [3H]ATR bound to mitochondria. Nonspecific
binding was measured in the presence of 500 µM CATR. BmaxATR)
number of maximum ATR binding sites.
N-ADP were added to isolated mitochondria (0.5 mg/mL). The
fluorescence level was set to zero. Addition of CATR released N-ADP
from mitochondria inducing an increase in fluorescence (∆F). The K1/2
value was determined by plotting ∆F/∆F max as a function of added
N-ADP.dnd, not determined.
cVarious concentrations of
Biochemistry, Vol. 47, No. 50, 2008
David et al.
though impairing yeast growth on a nonfermentable carbon
source, did not modify the maximum rate of ADP/ATP
exchange but rather the KMfor external ADP (33), whose
value was increased about 500-fold as compared to the wild-
type protein. It was therefore of crucial importance to
determine whether Y203A and F208A mutations had similar
effects on ScAnc2p activity. We measured ADP/ATP
exchange kinetic parameters with mitochondria isolated from
the strains expressing the corresponding variants. Results are
given in Table 2. Interestingly, ScAnc2pY203Abehaved
similarly to the op1 mutant: the VmaxADPvalue was similar
to that of the WT, but the KMADPvalue was increased almost
500-fold. In contrast, ScAnc2pY207A, which grew on lactate,
had a KMADPvalue similar to that of WT and a slightly
decreased VMaxADPvalue. The ScAnc2pF208Avariant could not
perform ADP/ATP exchange. The ScAnc2pY211Avariant had
a decreased VmaxADPvalue (around 2-fold) and an increased
KMADPvalue (12.5-fold) but not as much as ScAnc2pY203A,
accounting for the reduced but still significant growth
observed on lactate (Figure 2, line 10).
The Aromatic Ladder Is Not Crucial to ADP Binding per
se. 3′-O-(1-Naphthoyl)adenosine 5′-diphosphate (N-ADP) is
a nontransportable ADP analogue whose binding properties
to Ancp can be examined with isolated mitochondria (22).
Its fluorescence is quenched upon binding to Ancp, so
specific binding to the mitochondrial Ancp can be measured
by the fluorescence enhancement observed upon dissociation
after addition of CATR, a very specific Ancp inhibitor. The
K1/2value of N-ADP binding to the beef Ancp is 3 µM (22),
which is close to the value determined for the yeast ScAnc2p
(K1/2) 1.7 µM, Table 1). As can be seen in Table 1, the
K1/2 values are in the same range for ScAnc2pY203Aand
ScAnc2pF208A, which are inactive for transport (2.6 and 1.8
µM, respectively). This suggests that the absence of nucle-
otide transport catalyzed by these variants does not result
from their inability to bind nucleotides but rather from their
inability to catalyze the substrate translocation step.
The m2 Loop of ScAnc2pY203Aand of ScAnc2pF208AIs
Locked in a “CATR”-like Conformation. CATR and BA are
known to induce conformational changes similar to those
involved in nucleotide translocation during which the matrix
loop m2 plays an important role. C159 of BAnc1p in the
middle of loop m2 was differentially labeled by the non-
permeant sulfhydryl-alkylating reagent eosin-5-maleimide
(EMA) in the presence or absence of nucleotides, CATR or
BA (15). This labeling inhibited ADP/ATP transport. C159
corresponds to V176 of the yeast ScAnc2p, which nonethe-
less contains four cysteinyl residues: C73 in matrix loop m1
(equivalent to C56 of BAnc1p), C244 in TMS 5, C271 in
matrix loop m3 (equivalent to C256 of BAnc1p), and C288
in the C-terminal end of m3. None of them was labeled with
EMA in the wild-type ScAnc2p embedded in mitochondrial
membrane (ref 14 and data not shown). We first engineered
a mutant of ScAnc2p in which the four endogenous cysteinyl
residues were mutated into alanine (Anc2CLp). We then
mutated V176 in order to introduce a cysteine in the matrix
loop m2, at a position equivalent to that of C159 of BAnc1p
(Anc2CLpV176C) (Figure 1A). Such a mutation does not
preclude yeast growth on a nonfermentable carbon source,
and its doubling time in rich lactate-containing medium is
approximately 4 h 30 min, which is comparable to that of
the wild-type cells (3 h 30 min (23)). Furthermore,
Anc2CLpV176Cexchanged nucleotides with a significant
efficacy since the KMADPvalue was in the same range though
the Vmaxvalue was half that of the wild type (Table 2). Last,
we controlled that the Anc2CLpV176Cprotein was stable over
the EMA labeling duration by Western blot analyses of
mitochondria at the end of the labeling experiments (data
EMA labeling was performed with freshly isolated and
with frozen-thawed mitochondria in the absence or in the
presence of CATR or BA. As can be seen in Figure 4A (lanes
1-3), Anc2CLpV176Cwas never labeled with EMA in freshly
isolated mitochondria. However, it was labeled in thawed
Table 2: ADP/ATP Exchange Parameters
90 ( 3
94 ( 8
94 ( 6
81 ( 6
78 ( 2
54 ( 1
47 ( 1
0.68 ( 0.08
368 ( 66
9 ( 2
328 ( 61
0.9 ( 0.1
8.5 ( 0.8
1 ( 0.2
3 h 30 min
7 h 30 min
17 h 30 min
4 h 30 min
aFrom ref 33.
was determined by measuring at various time intervals the absorbance at
600 nm of yeast cells cultivated in YPlact.dNM, not measurable.
bCorresponding to Anc2CLpV176C.
FIGURE 4: (A) EMA labeling of the membrane-bound Anc2CLpV176C
mitochondria were incubated (8 mg of proteins/mL) without or with
20 µM CATR or 20 µM BA for 15 min on ice. EMA labeling
(200 µM) was performed for 30 min on ice in the dark. The reaction
was stopped with 20 mM DTT. Samples were subjected to
SDS-PAGE, and fluorescence was visualized at λex) 532 nm.
The position of Anc2CLpV176C, indicated by an arrow, was
controlled by immunoblotting after transfer of the gel onto a
nitrocellulose membrane. The EMA-labeled bands, which migrated
in the gels immediately ahead of the Anc2p band, have been
tentatively identified to the phosphate carrier and the mitochondrial
porin (VDAC1) according to refs 44 and 14. (B, C) Y203A (B)
and F208A (C) prevent EMA labeling of the membrane-bound
Anc2CLpV176CY203A(B) and Anc2CLpV176CY203A(C). Frozen-thawed
mitochondria (8 mg of proteins/mL) containing Anc2CLpV176C(B
and C), Anc2CLpV176CY203A(B), or Anc2CLpV176CF208A(C) were
incubated without or with 20 µM CATR or 20 µM BA for 15 min
on ice. EMA labeling was performed as indicated in panel A.
Mitochondrial ADP/ATP Carrier and Nucleotide Transport
Biochemistry, Vol. 47, No. 50, 2008 13227
mitochondria in the absence of ligand, and this labeling was
inhibited by preincubation of mitochondria with CATR but
not with BA (Figure 4A, lanes 4-6). Therefore, V176C
EMA labeling appears to be a good reporter of carrier
conformational changes induced by inhibitor binding, though
at this stage we cannot exclude that CATR prevents EMA
labeling only by steric hindrance. This point is discussed
below. The difference between freshly isolated or thawed
mitochondria can be accounted for by rupture of the outer
and/or the inner membrane that would give EMA access to
Since the affinity of ScAnc2pY203Afor ATR was dramati-
cally reduced as compared to the wild-type protein, Y203
of Anc2CLpV176Cwas mutated into alanine, and EMA
labeling was examined for this new variant. As expected,
the Y203A mutation precluded yeast growth on a nonfer-
mentable carbon source. However, the new variant was stable
when embedded in mitochondria, and its ADP/ATP exchange
capacity was similar to that of ScAnc2pY203A(data not
shown). As can be seen in Figure 4B (lanes 4-6), this variant
was not labeled whatever the incubation conditions as if
“BA” and “CATR” conformations were similar with respect
to V176C EMA labeling. Similar results were obtained with
freshly isolated mitochondria (data not shown). The Y203A
variant might be locked in a “CATR”-like conformation, and
this locking cannot be reversed by BA binding, probably
because the matrix loop m2 cannot swing in a manner
appropriate for nucleotide translocation. In addition, these
results allow us to exclude that the absence of EMA labeling
of V176C in the presence of CATR is due to steric hindrance.
Considering such a dramatic effect of the F208A mutation
on ADP/ATP transport, it was also of interest to introduce
this mutation in Anc2CLpV176C. As expected, the new variant
Anc2CLpV176CF208Adid not restore yeast growth in a lactate-
containing medium though the protein was present in
significant amount in mitochondria (data not shown).
Anc2CLpV176CF208Abehaved similarly to Anc2pF208Aregarding
nucleotide transport. Interestingly, this completely inactive
mutant was never labeled with EMA (Figure 4C), similarly
to Anc2CLpV176CY203A. The F208A mutation might also lock
ScAnc2p in a “CATR”-like conformation, which however
would differ from that adopted by ScAnc2pY203Asince F208A
increases ATR affinity, contrary to Y203A. Furthermore,
F208A completely inactivated the nucleotide transport
whereas some transport was still detected with Y203A.
This paper assigns a crucial role to a structural amino acid
motif initially revealed in the 3D structure of BAnc1p in
complex with CATR and formerly named the “tyrosine
ladder” (1) (Figure 1A). It was first proposed that adenine
nucleotides from the IMS would be attracted and “preori-
ented” through the combined effect of a patch of positively
charged amino acids and of the “tyrosine ladder”. In this
ladder, Y186 constricts the funnel-shaped cavity of BAnc1p
in complex with CATR (7). This would restrict access of
the nucleotide to the other side of the inhibited carrier,
consisting of a second patch of basic and acidic residues
belonging to the third signature sequence of the mitochondrial
235RRMMM239, which is the end of the Ancp signature
sequence (7). Right beside Y190 is F191, which is perfectly
conserved in all known Ancp amino acid sequences (8). The
orientation of its aromatic ring is almost orthogonal to that
of the other tyrosyl residues (Figure 1). Thus, we hypoth-
esized involvement of the aromatic ladder in nucleotide
progress into the funnel. This was assessed by mutating the
corresponding residues of yeast ScAnc2p, Y203, Y207, F208,
and Y211, into alanine, in pairs or individually. Two residues,
Y203 and F208, which are perfectly conserved among Ancp
sequences, obviously play a crucial role in nucleotide
exchange since their mutation impaired the transport function.
Estimates of the variant carrier amounts in the MIM ruled
out impaired Ancp variant biogenesis being responsible for
the observed phenotype. However, ATR binding properties
were dramatically modified for ScAnc2pY203A, in line with
the observation that Y186 of BAnc1p, which is equivalent
to yeast Y203, is stacked on the diterpene moiety of CATR
(1). CATR blind docking on the 3D model of yeast Anc2p
indicates that the lowest binding energy cluster corresponds
to a CATR conformer located within the Ancp cavity which
is open to the intermembrane space. Upon careful examina-
tion, it appears that the critical residues of ScAnc2p involved
in the CATR binding, as exhibited in the crystal structure
of the bAncp1·CATR complex, are also those involved in
the lowest energy conformation of the CATR docking
experiment. Y203, on one hand, and L142, V145, and I200,
on the other, are involved respectively in stacking with the
diterpene moiety of CATR and van der Waals contacts with
the isovaleric residue, while N104, K108, and R204 are
hydrogen-bonded to the glucose moiety of the CATR sulfate
groups. In addition, R96 is involved in a salt bridge with a
carboxylic group of CATR. These results indicate, first, that
the analysis of our ScAnc2p mutagenesis results with the
help of the bAnc1p crystal structure is perfectly relevant and,
second, that the 3D model of ScAnc2p could be safely used
for interpreting our results with the yeast carrier.
The loss of ring stacking loosens the Ancp·ATR complex
(Table 1), which is consistent with a dramatically increased
KDATRvalue (Table 1). This stacking cannot be compensated
for by the nearby aromatic residues, which point toward the
cavity (Y207 and Y211). In the case of ScAnc2pF208A, a
completely inactive variant, and of ScAnc2pY211A, a less
active variant, the KDATRvalues are 5-7-fold lower than that
of the wild type, thus indicating the formation of much more
stable ScAnc2p·ATR complexes than with the wild-type
ScAnc2p. F208 and Y211 are in the largest part of the cavity
and oriented toward the IMS. It seems unlikely that they
restrict CATR access to the cavity. Instead, they may play a
role in ScAnc2p molecular “plasticity”. Their mutation into
alanine might lock the carrier in the “CATR” conformation,
as reflected by the high affinity for ATR and by the absence
of EMA labeling of V176C when associated with F208A.
Aromatic residues can contribute to protein conformation
through various interactions and notably through cation-π
ones. Analyses of the 3D structure of BAnc1p·CATR (PDB
entry: 1okc) with the CAPTURE program (34) identified
three probable R/Y interactions that were energetically
significant: R59/Y50, R104/Y111, and R234/Y131. None of
them involves a residue of the aromatic ladder, though we
cannot rule out this pattern as being different in the
noninhibited conformation. Similar results were obtained
when the presence of putative cation-π interactions was
Biochemistry, Vol. 47, No. 50, 2008
David et al.
sought within the ScAnc2p 3D model constructed by
homology modeling with the BAnc1p·CATR crystal structure.
Nucleotide transport by ScAnc2pY203Aor ScAnc2pF208A
was either dramatically impaired (Anc2pY203A) or abolished
(Anc2pF208A) (Table 2). The KMADPvalue for ScAnc2pY203A
was almost 500-fold higher than that of the WT. However,
these variants can still bind a nucleotide analogue (N-ADP)
with K1/2values similar to that of ScAnc2p (Table 1), which
is puzzling at first glance. Thus the nucleotide binding step
is not the main concern for these mutants but rather the
nucleotide transfer from one side of the carrier to the other.
This step very likely involves conformational changes.
Therefore, we have introduced into ScAnc2p a cysteinyl
residue in position 176 in the middle of the loop m2, which
might be considered as an intrinsic conformational probe.
C176 is labeled in the presence of BA but not in the pres-
ence of CATR, contrary to what is observed for
Anc2CLpV176CY203Aand Anc2CLpV176CF208A, in which C176
is never labeled, whatever the carrier inhibitor. ScAnc2pY203A
and ScAnc2pF208Amight be locked in conformations similar
to the “CATR” conformation and might be unable to swing
to the “BA” conformation upon BA binding. Yet this
swinging is one of the requisites of nucleotide transport, as
has been shown by numerous approaches (for a review, see
for example ref 8).
Therefore, we propose that Y203 and F208, two aromatic
residues of TM4 from the ScAnc2p carrier, play an essential
role in nucleotide attraction and its progress through the
carrier leading to a productive ADP/ATP exchange.
Our results are substantiated by two recently published
papers reporting on the ADP binding on BAnc1p using a
molecular dynamics approach on time scales of 0.3 (35) and
0.53 µs (36). The time scales explored are probably too short
to give an insight into the conformational changes involved
in the substrate translocation step. In addition, a more realistic
environment mimicking the ionic composition of the IMS
water phase in equilibrium with both the carrier protein and
the lipid bilayer could probably contribute to a partial
shielding of the electrostatic forces at work. Nevertheless,
some interesting results are worth underscoring. In these
studies the ADP·Ancp complex forms rapidly because of a
strong electrostatic attraction between the nucleotide and the
protein. The phosphate groups are attracted first, and their
binding to positive conserved residues would thus trigger
the conformational changes inherent to transport. In one of
the simulations, adenine moiety forms successively π-stack-
ing interactions with Y194 and Y190 (Y211 and Y207 of
ScAnc2p, respectively) either individually or together (36).
In both studies the pentose ring interacts with Y186 (Y203
of ScAnc2p) at the end of the simulation. Therefore, Y186
plays a key role in ADP transport. Interestingly, one of the
simulations is abortive for ADP transport since the nucleotide
remains rapidly stuck between Y194 and F191 (Y211 and
F208 of ScAnc2p, respectively) (36). This is puzzling
considering the essential role of F208 evidenced by our
studies but points out to the absolute requirement of an
experimental approach to corroborate molecular dynamics
Y203 and F208 are 100% conserved among all known
Ancp sequences. A phenylalanine residue at a position
equivalent to F208 is present in 71% of the MCF members
of the yeast S. cereVisiae (Figure 5). This residue might
therefore also play a role in substrate transport in many
carriers, regardless of substrate specificity. Y203 is much
less conserved among MCF (26%) (Figure 5), and its role
may be restricted to substrates bearing an adenine moiety.
Indeed, it is conserved in Sal1p and Leu5p whose substrates
are ATP-Mg/Piand CoA, respectively (37, 38). In contrast,
Ggc1p, which transports guanine nucleotides (39), and
Rim2p, which transports pyrimidine nucleotides (40), contain
a glycine at this position. However, the peroxisomal yeast
Ant1p, which transports adenine nucleotide (41), presents a
very different sequence, PxxxYAxFQ, in this region. Its
peroxisomal location may account for this divergent evolu-
tion as compared to the mitochondrial carriers.
Along this line, it is interesting to consider the mitochon-
drial peripheral benzodiazepine receptor (TSPO/PBR), which
contains a CRAC sequence (cholesterol recognition amino
acid consensus), L/V-X(1-5)-Y-X(1-5)-R/K, in the cyto-
solic C-terminal region. NMR experiments combined with
molecular modeling of a synthetic peptide corresponding to
FIGURE 5: Sequence alignment of 34 members of S. cereVisiae MCF
(identified or putative). (1) The aromatic ladder sequence corre-
sponds to that defined by comparing all known Ancp sequences
(8). The aromatic residues are in bold and italicized whatever their
position. (2) The locus tags are defined in the Saccharomyces
Genome Database (http://www.yeastgenome.org/). Sequences were
aligned with ClustalX (26). (3) Conservation of ScAnc2p Y203, Y207,
F208, and Y211 is indicated as percent of the aromatic residue at
equivalent positions in the 116 Ancp sequences analyzed (8).
Mitochondrial ADP/ATP Carrier and Nucleotide Transport
Biochemistry, Vol. 47, No. 50, 2008 13229
the putative TM5 and C-terminal region of rat TSPO/PBR
revealed a central patch of aromatic residues in the vicinity
of the CRAC sequence (42). A girdle of three aromatic
builds up “a strong interfacial anchoring motif”, two residues
of which are essential for cholesterol binding (Y152 and Y153).
The aromatic rings of these amino acids might constitute the
gate of a groove in TM5 delineated by residues Y152, T148,
and L144 on one side and Y153, M149, and A145 on the other
side and where cholesterol, a planar molecule, might bind.
and of the distances between cholesterol and aromatic residue
rings needs further investigation.
Also of interest is the bacterial nucleoside transporter Tsx
whose crystal structure was obtained in the presence of
nucleosides (43). Although its structure is more like that of
a porin, Tsx presents well-defined binding sites for nucleo-
sides. The Tsx pore is lined with eight aromatic residues, of
which six are paired. The three pairs of aromatic residues
are highly conserved among Tsx homologues. Two nucleo-
sides are situated along the center of the channel consisting
of the eight residues. The base moieties are stacked between
aromatic residues, which also create van der Waals contacts
with the sugar moieties. Therefore, involvement of a
structural aromatic residue motif in substrate recognition and
translocation is not restricted to the mitochondrial carrier
family but might also involve proteins, whose substrates
contain unsaturated chemical rings likely involved in stacking
interactions with protein aromatic amino acids.
We thank Corinne Blancard for technical assistance. This
paper is dedicated to the memory of Jean-Fre ´de ´ric Sanchez
who passed away in May 2008.
1. Pebay-Peyroula, E., Dahout-Gonzalez, C., Kahn, R., Tre ´ze ´guet,
V., Lauquin, G. J.-M., and Brandolin, G. (2003) Structure of
mitochondrial ADP/ATP carrier in complex with carboxyatracty-
loside. Nature 426, 39–44.
2. Fiore, C., Tre ´ze ´guet, V., Le Saux, A., Roux, P., Schwimmer, C.,
Dianoux, A.-C., Noe ¨l, F., Lauquin, G. J.-M., Brandolin, G., and
Vignais, P. V. (1998) The mitochondrial ADP/ATP carrier:
structural, physiological and pathological aspects. Biochimie 80,
3. Majima, E., Yamaguchi, N., Chuman, H., Shinohara, Y., Ishida,
M., Goto, S., and Terada, H. (1998) Binding of the fluorescein
derivative eosin Y to the mitochondrial ADP/ATP carrier: char-
acterization of the adenine nucleotide binding site. Biochemistry
4. Pfaff, E., and Klingenberg, M. (1968) Adenine nucleotide trans-
location of mitochondria. 1. Specificity and control. Eur. J. Bio-
chem. 6, 66–79.
5. Verdouw, H., and Bertina, R. M. (1973) Affinities of ATP for the
dinitrophenol-induced ATPase. Biochim. Biophys. Acta 325, 385–
6. Kolarov, J., Kolarova, N., and Nelson, N. (1990) A third ADP/
ATP translocator gene in yeast. J. Biol. Chem. 265, 12711–12716.
7. Pebay-Peyroula, E., and Brandolin, G. (2004) Nucleotide exchange
in mitochondria: insight at a molecular level. Curr. Opin. Struct.
Biol. 14, 420–425.
8. Nury, H., Dahout-Gonzalez, C., Tre ´ze ´guet, V., Lauquin, G. J.-M.,
Brandolin, G., and Pebay-Peyroula, E. (2006) Relations between
structure and function of the mitochondrial ADP/ATP carrier. Annu.
ReV. Biochem. 75, 713–741.
9. Nelson, D. R., Lawson, J. E., Klingenberg, M., and Douglas, M. G.
(1993) Site-directed mutagenesis of the yeast mitochondrial ADP/
ATP translocator. Six arginines and one lysine are essential. J.
Mol. Biol. 230, 1159–1170.
10. Mu ¨ller, V., Basset, G., Nelson, D. R., and Klingenberg, M. (1996)
Probing the role of positive residues in the ADP/ATP carrier from
yeast. The effect of six arginine mutations on oxidative phospho-
rylation and AAC expression. Biochemistry 35, 16132–16143.
11. Mu ¨ller, V., Heidkamper, D., Nelson, D. R., and Klingenberg, M.
(1997) Mutagenesis of some positive and negative residues
occurring in repeat triad residues in the ADP/ATP carrier from
yeast. Biochemistry 36, 16008–16018.
12. Heidkamper, D., Mu ¨ller, V., Nelson, D. R., and Klingenberg, M.
(1996) Probing the role of positive residues in the ADP/ATP carrier
from yeast. The effect of six arginine mutations on transport and
the four ATP versus ADP exchange modes. Biochemistry 35,
13. Marty, I., Brandolin, G., Gagnon, J., Brasseur, R., and Vignais,
P. V. (1992) Topography of the membrane-bound ADP/ATP carrier
assessed by enzymatic proteolysis. Biochemistry 31, 4058–4065.
14. Dahout-Gonzalez, C., Ramus, C., Dassa, E. P., Dianoux, A.-C.,
and Brandolin, G. (2005) Conformation-dependent swinging of the
matrix loop m2 of the mitochondrial Saccharomyces cereVisiae
ADP/ATP carrier. Biochemistry 44, 16310–16320.
15. Majima, E., Koike, H., Hong, Y. M., Shinohara, Y., and Terada,
H. (1993) Characterization of cysteine residues of mitochondrial
ADP/ATP carrier with the SH-reagents eosin 5-maleimide and
N-ethylmaleimide. J. Biol. Chem. 268, 22181–22187.
16. Majima, E., Shinohara, Y., Yamaguchi, N., Hong, Y. M., and
Terada, H. (1994) Importance of loops of mitochondrial ADP/ATP
carrier for its transport activity deduced from reactivities of its
cysteine residues with the sulfhydryl reagent eosin-5-maleimide.
Biochemistry 33, 9530–9536.
17. Majima, E., Ikawa, K., Takeda, M., Hashimoto, M., Shinohara,
Y., and Terada, H. (1995) Translocation of loops regulates transport
activity of mitochondrial ADP/ATP carrier deduced from formation
of a specific intermolecular disulfide bridge catalyzed by copper-
o-phenanthroline. J. Biol. Chem. 270, 29548–29554.
18. Boulay, F., Lauquin, G. J.-M., Tsugita, A., and Vignais, P. V.
(1983) Photolabeling approach to the study of the topography of
the atractyloside binding site in mitochondrial adenosine 5′-
diphosphate/adenosine 5′-triphosphate carrier protein. Biochemistry
19. Dalbon, P., Brandolin, G., Boulay, F., Hoppe, J., and Vignais, P. V.
(1988) Mapping of the nucleotide-binding sites in the ADP/ATP
carrier of beef heart mitochondria by photolabeling with 2-azi-
do[alpha-32P]adenosine diphosphate. Biochemistry 27, 5141–5149.
20. Dianoux, A.-C., Noe ¨l, F., Fiore, C., Tre ´ze ´guet, V., Kieffer, S.,
Jaquinod, M., Lauquin, G. J.-M., and Brandolin, G. (2000) Two
distinct regions of the yeast mitochondrial ADP/ATP carrier are
photolabeled by a new ADP analogue: 2-azido-3′-O-naphthoyl-
[beta-32P]ADP. Identification of the binding segments by mass
spectrometry. Biochemistry 39, 11477–11487.
21. Brandolin, G., Meyer, C., Defaye, G., Vignais, P. M., and Vignais,
P. V. (1974) Partial purification of an atractyloside-binding protein
from mitochondria. FEBS Lett. 46, 149–153.
22. Block, M. R., Lauquin, G. J.-M., and Vignais, P. V. (1982)
Interaction of 3′-O-(1-naphthoyl)adenosine 5′-diphosphate, a fluo-
rescent adenosine 5′-diphosphate analogue, with the adenosine 5′-
diphosphate/adenosine 5′-triphosphate carrier protein in the mito-
chondrial membrane. Biochemistry 21, 5451–5457.
23. De Marcos Lousa, C., Tre ´ze ´guet, V., Dianoux, A.-C., Brandolin,
G., and Lauquin, G. J.-M. (2002) The human mitochondrial ADP/
ATP carriers: kinetic properties and biogenesis of wild type and
mutant proteins in the yeast S. cereVisiae. Biochemistry 41, 14412–
24. Brandolin, G., Le Saux, A., Tre ´ze ´guet, V., Vignais, P. V., and
Lauquin, G. J.-M. (1993) Biochemical characterisation of the
isolated Anc2 adenine nucleotide carrier from Saccharomyces
cereVisiae mitochondria. Biochem. Biophys. Res. Commun. 192,
25. Boeke, J. D., Trueheart, J., Natsoulis, G., and Fink, G. R. (1987)
5-Fluoroorotic acid as a selective agent in yeast molecular genetics.
Methods Enzymol. 154, 164–175.
26. Chenna, R., Sugawara, H., Koike, T., Lopez, R., Gibson, T. J.,
Higgins, D. G., and Thompson, J. D. (2003) Multiple sequence
alignment with the Clustal series of programs. Nucleic Acids Res.
27. Sali, A., and Blundell, T. L. (1993) Comparative modelling by
satisfaction of spatial restraints. J. Mol. Biol. 234, 779–815.
Biochemistry, Vol. 47, No. 50, 2008
David et al.
28. Morris, G. M., Goodsell, D. S., Halliday, R. S., Huey, R., Hart, Download full-text
W. E., Belew, R. K., and Olson, A. J. (1998) Automated docking
using a Lamarckian genetic algorithm and empirical binding
free energy function. J. Comput. Chem. 19, 1639–1662.
29. Schuettelkopf, A. W., and van Aalten, D. M. F. (2004) PRODRGsa
tool for high-throughput crystallography of protein-ligand com-
plexes. Acta Crystallogr. D60, 1355–1363.
30. Sanner, M. F. (1999) Python: A programming language for
software integration and development. J. Mol. Graphics Mod.
31. Brandolin, G., Boulay, F., Dalbon, P., and Vignais, P. V. (1989)
Orientation of the N-terminal region of the membrane-bound
ADP/ATP carrier protein explored by antipeptide antibodies and
an arginine-specific endoprotease. Evidence that the accessibility
of the N-terminal residues depends on the conformational state
of the carrier. Biochemistry 28, 1093–1100.
32. Michejda, J., Guo, X. J., and Lauquin, G. J.-M. (1990) The
respiration of cells and mitochondria of porin deficient yeast
mutants is coupled. Biochem. Biophys. Res. Commun. 171, 354–
33. Postis, V., De Marcos Lousa, C., Arnou, B., Lauquin, G. J.-M.,
and Tre ´ze ´guet, V. (2005) Subunits of the yeast mitochondrial ADP/
ATP carrier: cooperation within the dimer. Biochemistry 44, 14732–
34. Gallivan, J. P., and Dougherty, D. A. (1999) Cation-π interactions
in structural biology. Proc. Natl. Acad. Sci. U.S.A. 96, 9459–9464.
35. Wang, Y., and Tajkhorshid, E. (2008) Electrostatic funneling
of substrate in mitochondrial inner membrane carriers. Proc.
Natl. Acad. Sci. U.S.A. 105, 9598–9603.
36. Dehez, F., Pebay-Peyroula, E., and Chipot, C. (2008) Binding
of ADP in the mitochondrial ADP/ATP carrier is driven by an
electrostatic funnel. J. Am. Chem. Soc. 130, 12725–12733.
37. Cavero, S., Traba, J., Del Arco, A., and Satru ´stegui, J. (2005) The
calcium-dependent ATP-Mg/Pi mitochondrial carrier is a target of
glucose-induced calcium signalling in Saccharomyces cereVisiae.
Biochem. J. 392, 537–544.
38. Prohl, C., Pelzer, W., Diekert, K., Kmita, H., Bedekovics, T.,
Kispal, G., and Lill, R. (2001) The yeast mitochondrial carrier
Leu5p and its human homologue Graves’ disease protein are
required for accumulation of coenzyme A in the matrix. Mol. Cell.
Biol. 21, 1089–1097.
39. Vozza, A., Blanco, E., Palmieri, L., and Palmieri, F. (2004)
Identification of the mitochondrial GTP/GDP transporter in Sac-
charomyces cereVisiae. J. Biol. Chem. 279, 20850–20857.
of a mitochondrial transporter for pyrimidine nucleotides in Saccha-
romyces cereVisiae: bacterial expression, reconstitution and functional
characterization. Biochem. J. 393, 441–446.
41. Palmieri, L., Rottensteiner, H., Girzalsky, W., Scarcia, P., Palmieri,
F., and Erdmann, R. (2001) Identification and functional reconstitu-
tion of the yeast peroxisomal adenine nucleotide transporter. EMBO
J. 20, 5049–5059.
42. Jamin, N., Neumann, J. M., Ostuni, M. A., Vu, T. K., Yao, Z. X.,
Murail, S., Robert, J. C., Giatzakis, C., Papadopoulos, V., and
Lacape `re, J.-J. (2005) Characterization of the cholesterol recogni-
tion amino acid consensus sequence of the peripheral-type ben-
zodiazepine receptor. Mol. Endocrinol. 19, 588–594.
43. Ye, J., and van den Berg, B. (2004) Crystal structure of the bacterial
nucleoside transporter Tsx. EMBO J. 3187–3195.
44. Hatanaka, T., Kihira, Y., Shinohara, Y., Majima, E., and Terada,
H. (2001) Characterization of loops of the yeast mitochondrial
ADP/ATP carrier facing the cytosol by site-directed mutagenesis.
Biochem. Biophys. Res. Commun. 286, 936–942.
Mitochondrial ADP/ATP Carrier and Nucleotide Transport
Biochemistry, Vol. 47, No. 50, 2008 13231