Upload
others
View
5
Download
0
Embed Size (px)
Citation preview
ANRV277-BI75-28 ARI 12 May 2006 15:19
Relations Between Structureand Function of theMitochondrialADP/ATP CarrierH. Nury,1,∗ C. Dahout-Gonzalez,2,∗
V. Trezeguet,3 G.J.M. Lauquin,3 G. Brandolin,2
and E. Pebay-Peyroula1
1Institut de Biologie Structurale Jean-Pierre Ebel, UMR 5075CEA-CNRS-Universite Joseph Fourier, F-38027 Grenoble cedex 1,France; email: [email protected], [email protected] de Biochimie et Biophysique des Systemes Integres, UMR 5092CEA-CNRS-Universite Joseph Fourier, Departement de Reponse et DynamiqueCellulaires, F-38054 Grenoble cedex 9, France; email: [email protected],[email protected] de Physiologie Moleculaire et Cellulaire, UMR 5095 CNRS-UniversiteBordeaux 2, Institut de Biochimie et Genetique Cellulaires, F-33077 Bordeaux cedex,France; email: [email protected], [email protected]
Annu. Rev. Biochem.2006. 75:713–41
First published online as aReview in Advance onFebruary 1, 2006
The Annual Review ofBiochemistry is online atbiochem.annualreviews.org
doi: 10.1146/annurev.biochem.75.103004.142747
Copyright c© 2006 byAnnual Reviews. All rightsreserved
0066-4154/06/0707-0713$20.00
∗These authorscontributed equally to thework.
Key Words
membrane protein, nucleotide transport, oligomeric state
AbstractImport and export of metabolites through mitochondrial membranesare vital processes that are highly controlled and regulated at thelevel of the inner membrane. Proteins of the mitochondrial carrierfamily (MCF) are embedded in this membrane, and each memberof the family achieves the selective transport of a specific metabo-lite. Among these, the ADP/ATP carrier transports ADP into themitochondrial matrix and exports ATP toward the cytosol after itssynthesis. Because of its natural abundance, the ADP/ATP carrier isthe best characterized within MCF, and a high-resolution structureof one conformation is known. The overall structure is basket shapedand formed by six transmembrane helices that are not only tilted withrespect to the membrane, but three of them are also kinked at thelevel of prolines. The functional mechanisms, nucleotide recogni-tion, and conformational changes for the transport, suggested fromthe structure, are discussed along with the large body of biochemicaland functional results.
713
Ann
u. R
ev. B
ioch
em. 2
006.
75:7
13-7
41. D
ownl
oade
d fr
om a
rjou
rnal
s.an
nual
revi
ews.
org
by U
nive
rsity
of
Cal
ifor
nia
- Ir
vine
on
09/0
5/06
. For
per
sona
l use
onl
y.
ANRV277-BI75-28 ARI 12 May 2006 15:19
Contents
INTRODUCTION. . . . . . . . . . . . . . . . . 714BRIEF HISTORY . . . . . . . . . . . . . . . . . . 715STRUCTURE ANALYSIS . . . . . . . . . . 717
The Cavity . . . . . . . . . . . . . . . . . . . . . . . 717CATR Binding . . . . . . . . . . . . . . . . . . . 722The MCF Motif . . . . . . . . . . . . . . . . . . 722Location of the ADP/ATP Carrier
Signature . . . . . . . . . . . . . . . . . . . . . 725Lipids . . . . . . . . . . . . . . . . . . . . . . . . . . . 726Deviation to Pseudo-Threefold
Symmetry . . . . . . . . . . . . . . . . . . . . . 727Conserved Residues: A Structural
or Functional Role? . . . . . . . . . . . 728NUCLEOTIDE ATTRACTION
AND BINDING . . . . . . . . . . . . . . . . . 728Specificity for Binding and
Transport . . . . . . . . . . . . . . . . . . . . . 730Biochemical Evidence for
IMS-Binding Sites . . . . . . . . . . . . . 730Structural Features . . . . . . . . . . . . . . . 731Binding ATP from the Matrix . . . . . 731
TRANSPORT MECHANISM . . . . . . 732Conformational Changes . . . . . . . . . 732What Triggers the Changes?
Kinetic Aspects . . . . . . . . . . . . . . . . 733Is the Functional Unit a Dimer? . . . 734
HYPOTHESES ANDCONCLUSIONS. . . . . . . . . . . . . . . . 736A Proposed Mechanism Involving
CDLs. . . . . . . . . . . . . . . . . . . . . . . . . 736Open Questions . . . . . . . . . . . . . . . . . . 736
INTRODUCTION
Cell compartmentalization into organelleslimited by membranes, allowing segregation
Mitochondrialcarrier family(MCF): proteins ofthe innermitochondrialmembrane, whichshuttle substrates inand out
IMS:intermembranespace
TM:transmembrane
of various cell functions in different loca-tions of the cellular space, implies that a com-plex communications network needs to be setup and regulated. In most cases, moleculesinvolved in intracellular trafficking have topass through membranes, and specific trans-port proteins generally catalyze these jour-neys. This is particularly true with charged
species because the lipid bilayer of membranesis an effective barrier to anions and cations.
Solute transports through membranes areessential steps in many metabolic pathways.Mitochondria are particularly rich in solutecarriers, and most of these constitute the so-called mitochondrial carrier family (MCF) (1),also referred to as SLC25. About 20 distincttransport functions are involved in the fluxesof the various metabolites through the in-ner mitochondrial membrane, which is theonly mitochondrial permeability barrier, theouter membrane being freely permeable tomolecules up to approximately 5000 Da. Allof the characterized carriers are not alwayspresent in all tissues, according to the occur-rence of tissue-specific metabolic pathways.
Among MCF members, the ADP/ATPcarrier is the most abundant and, along withthe phosphate carrier, appears indispensableto mitochondria. It can be considered theparadigm of mitochondrial metabolite carri-ers because several major findings of carrierstructures and mechanisms have been firstproduced through studies of the ADP/ATPcarrier. Albeit the function of all the membersof the MCF have not yet been identified, theyshare main structural and functional charac-teristics that are summarized as follows: (a) allmitochondrial carriers are encoded by nucleargenes, (b) the primary structure of most car-riers displays three repeated homologous re-gions of about 100 amino acids each, (c) the Nand C termini face the intermembrane space(IMS) and six transmembrane (TM) segmentscan be delineated, (d) a common sequence,the MCF motif, can be found in each re-peated region with slight deviations on oneor two signature sequences for some carri-ers, and (e) comparison of primary structuresindicates that mitochondrial carriers have noorthologues in prokaryotes, their emergenceseems to be the evolutionary consequence ofthe capture of an ancient aerobic procaryoticcell by the primitive eucaryotic cell.
Although high-resolution structures areimportant tools to tackle functional mecha-nisms, structures of membrane proteins are
714 Nury et al.
Ann
u. R
ev. B
ioch
em. 2
006.
75:7
13-7
41. D
ownl
oade
d fr
om a
rjou
rnal
s.an
nual
revi
ews.
org
by U
nive
rsity
of
Cal
ifor
nia
- Ir
vine
on
09/0
5/06
. For
per
sona
l use
onl
y.
ANRV277-BI75-28 ARI 12 May 2006 15:19
still far behind expectations owing to diffi-culties in getting sufficient amounts of pureand well-folded proteins and in crystallizingsuch amphipatic molecules (2). Furthermore,transporters often undergo large conforma-tional changes, and therefore solutions of sol-ubilized proteins are highly inhomogeneousin terms of conformations. Obtaining well-ordered three-dimensional crystals is morelikely to occur when a single conformation hasbeen locked either by a mutation as for lac-tose permease (3) or by the presence of an in-hibitor as discussed herein for the ADP/ATPcarrier (4). Two-dimensional crystallizationis an alternative approach to structure de-termination by X-ray crystallography, whichmight favor protein stability because the pro-tein is embedded in a flat lipidic bilayer. Eventhough the resolution obtained by electronmicroscopy is usually lower, it is an interest-ing complementary approach that might helpelucidate larger conformational changes. TheADP/ATP carrier is the only MCF carrier forwhich the three-dimensional structure of oneconformation is known. The first projectedstructure of a yeast isoform complexed toatractyloside (ATR) was obtained by electronmicroscopy (5, 6). The high-resolution struc-ture of the bovine carrier was solved to 2.2 A inthe presence of carboxyatractyloside (CATR)(4), and a second crystal form of the sameprotein gave additional information on car-diolipins (CDLs) (7). This review discussesmainly structure-function relationships, com-bining structural with biochemical and func-tional data.
BRIEF HISTORY
The concept of a mitochondrial carrier foradenine nucleotides arose in the 1960s as aresult of the convergence of two indepen-dent series of observations described by re-searchers working at deciphering the mito-chondrial phosphorylation process.
In 1955, a pioneering study by Siekevitz &Potter (8) established that two distinct pools ofadenine nucleotides were operating coopera-
HUMAN DISEASE AND THE ADP/ATPCARRIER
Disorders in the mitochondrial energy-generating system arereflected by a variety of clinical symptoms ranging from my-opathy, with lactic acidosis, to severe multisystem disease, in-volving the central nervous or cardiac system. In most cases,the origin of the disease is evidenced by a defect in one of thecomponents of the respiratory chain and is due to mutationsin either the mitochondrial or the nuclear genome. For somepatients, however, impairment of the mitochondrial functioncannot be ascribed to such defects. Almost 25% of the pa-tients suffering from such a dysfunction do not show clearimpairments of the respiratory chain. In some cases, metabo-lite carriers, including the ADP/ATP carrier, might be thecause of the pathology. Mitochondrial myopathies implicat-ing the ADP/ATP carrier have been described in human andin mouse, and these myopathies may be involved in severalcases of autosomal dominant progressive ophthalmoplegia.
High-resolutionstructure:three-dimensionalprotein structureswith atomic detailsusually obtained byX-raycrystallography
ATR: atractyloside
CATR:carboxyatractyloside
Cardiolipin (CDL):an anionicphospholipid inwhich a glycerolmoiety links twophosphatidyl groups
tively in the oxidative phosphorylation path-way, one located in the mitochondrial matrixand the other in the cytosol. Soon after, Press-man (9) reported that although the intramito-chondrial concentrations of nucleotides werelargely kept unaffected by the extramitochon-drial concentration, “an extremely dynamicinterchange of nucleotides” took place be-tween the two pools. To rationalize thesefindings, the hypothesis put forward at thattime was “the existence of a fixed quantity” ofan intramitochondrial component with suffi-cient affinity for adenine nucleotides. Curi-ously enough, there was no mention of a puta-tive adenine nucleotide transporter althoughBartley & Davies (10) had already advancedthe possibility of transporters for mitochon-drial respiratory substrates.
The second series of findings came fromstudies aimed at elucidating the poisonouseffect of a natural drug, ATR, extractedfrom the thistle Atractylis gummifera, which iswidespread in Mediterranean countries. Thisplant had been already described two thou-sand years ago by Dioscorides (∼40–90 AD)as a medicinal herb in his famous herbal
www.annualreviews.org • The Mitochondrial ADP/ATP Carrier 715
Ann
u. R
ev. B
ioch
em. 2
006.
75:7
13-7
41. D
ownl
oade
d fr
om a
rjou
rnal
s.an
nual
revi
ews.
org
by U
nive
rsity
of
Cal
ifor
nia
- Ir
vine
on
09/0
5/06
. For
per
sona
l use
onl
y.
ANRV277-BI75-28 ARI 12 May 2006 15:19
BA: bongkrekic acid
De Materia Medica. ATR is a diterpene het-eroglucoside of which two sulfate residuesmake it strongly anionic at physiological pHand thereby nonpermeable through the mito-chondrial membrane.
In 1962, Bruni et al. (11) and Vignaiset al. (12) reported that ATR inhibited specif-ically the oxidative phosphorylation of ex-tramitochondrial ADP. In the following years,these results were confirmed by several oth-ers groups (13–15), and it was clearly demon-strated that ATR did not inhibit the phos-phorylation of intramitochondrial ADP. Thenthe obvious conclusion suddenly appearedthat ATR was inhibiting the entry of ADPinto mitochondria, and consequently, the ex-istence of a transporter for ADP was postu-lated. Subsequently, detailed analyses of thekinetic properties of the adenine nucleotidetransport (16, 17) allowed the characteriza-tion of an exchange-diffusion process betweenthe intramitochondrial ADP and ATP with ei-ther ADP or ATP added to mitochondria; noother nucleotide could be a substrate for thecarrier. The finding, not fully appreciated atthat time, by Bruni et al. (11) that ATR in-hibits the binding of adenine nucleotides torat liver mitochondria, was instrumental laterin the identification of the substrate carriersites when radioactively labeled inhibitors be-came available.
In 1970, following preliminary studies byWelling et al. (18), Henderson & Lardy (19)introduced bongkrekic acid (BA), a complexfatty acid derivative, as an ADP/ATP carrierinhibitor. It is produced and secreted by thebacteria Pseudomonas cocovenenans and bindsto the matrix side of the carrier after diffu-sion through the membrane. From that timeto 1982, when the protein sequence of theADP/ATP carrier was determined, a largebody of results dealing with the properties ofthe carrier accumulated. These studies greatlybenefited from the availability of the two fam-ilies of inhibitors of exquisite specificity andaffinity. The asymmetrical binding of the twoclasses of inhibitors allowed the characteriza-tion of two stable conformational states of the
carrier, which possibly represent two extremeconfigurations close to that exhibited duringthe transport cycle by the carrier.
The other major findings are briefly sum-marized as follows: (a) When mitochon-dria are actively respiring in the presence ofphosphate and ADP, the latter is exchangedagainst intramitochondrial ATP with a 1- to 1-stoichiometry; (b) the only physiological sub-strates are ADP and ATP, surprisingly, in theirfree forms, i.e., Mg-ADP and Mg-ATP are notrecognized by the carrier; (c) the ADP/ATPexchange is electrogenic, which means onenegative charge is extruded from the matrixto the cytosol for each cycle, and this processis driven by the membrane potential; (d) thekinetic parameters of the carrier are consis-tent with the mitochondrial ATP productionand the cell nucleotide concentrations underphysiological conditions; and (e) the carriercould be purified in detergent solutions, andtransport activity could be reconstituted afterreincorporation into liposomes.
To investigate the structure and char-acterize the conformational states of theADP/ATP carrier, photoactivable derivativesof ADP/ATP or of the inhibitors were de-signed and tested either with the bovinecarrier or with the yeast carrier (Table 1). Sev-eral authors found evidence that both confor-mational states expose quite different aminoacid regions, and the differences were fur-ther highlighted with the use of specific pro-teases and antibodies directed against theADP/ATP carrier. Site-directed mutagenesiswas applied mainly to the yeast ADP/ATPcarrier isoform 2 to identify residues po-tentially crucial for the binding or trans-port mechanism. Some positive and negativeresidues were thus proposed to be essential(Table 1). This approach, though promis-ing, was limited to some residues in the ab-sence of structural data. However, the searchfor second-site revertants starting from well-defined inactive mutants (Table 1) led to theproposed network of potentially interactingamino acids that can be now reexamined in thelight of the three-dimensional bovine carrier
716 Nury et al.
Ann
u. R
ev. B
ioch
em. 2
006.
75:7
13-7
41. D
ownl
oade
d fr
om a
rjou
rnal
s.an
nual
revi
ews.
org
by U
nive
rsity
of
Cal
ifor
nia
- Ir
vine
on
09/0
5/06
. For
per
sona
l use
onl
y.
ANRV277-BI75-28 ARI 12 May 2006 15:19
structure. So far, only one charge pair sus-pected in the yeast carrier is confirmed in thebovine structure (D134-R234 of the bovineprotein). However, the other pairs are inthe vicinity of the ones evidenced in thebovine structure. All in all, this does not in-validate the general use of mutants becausebovine and yeast amino acid sequences, al-though pretty similar, are different enoughto suppose the involvement of different net-works of structurally related amino acids.
It is interesting to note that in addition tothe adenine nucleotide carrier, four other nu-cleotide transporters belonging to MCF havebeen identified: the mitochondrial carriers forGDP/GTP (20), dNDP/dNTP (21), and Mg-ATP/Pi (22) and, surprisingly, also the per-oxisomal AMP/ADP/ATP carrier (23). All ofthese carriers are insensitive to ATR, and incontrast to the ADP/ATP carrier, the chargesof their nucleotide substrates are partially ortotally neutralized by Mg2+ or by H+, with apossible exception of the dNDP/dNTP trans-porter. In addition, they also usually exhibita lower specificity regarding either the basemoiety or the polyphosphate chain length,and finally, their rate capacity is approximately15- to 30-fold lower.
STRUCTURE ANALYSIS
The ADP/ATP carrier was crystallized intwo different forms (7, 24). In both crys-tal forms, proteins surrounded by lipids arepacked within layers as if they are sittingin a membrane. Three-dimensional crystalsconsist of stacks of these layers. The modelof the protein structure was refined fromthe best diffracting crystal form (4) and de-posited in the Protein Data Bank under ac-cession code 1okc. From the N to the C ter-mini, both located in the IMS, the overallarchitecture consists of six TM helices la-beled H1 to H6, connected by three loopsM1 to M3 on the matrix side and two loopsC1 and C2 on the IMS (Figures 1 and 2a).Matrix loops are partially structured and con-tain short amphipatic helices (labeled h1-2,
h3-4, and h5-6) spanning over 12 residues(Figure 1c). Odd-numbered TM helices aresharply kinked (20◦ to 35◦), and all the sixTM helices are also tilted with respect to themembrane. As a result, the backbone of theADP/ATP carrier is shaped similar to a basket,closed toward the matrix and opened widelytoward the IMS. The backbone also exhibitsa pseudo-threefold symmetry consistent withthe triplicated sequence of the gene. Most ofthe residues are well defined in the model, ex-cept a few residues at the N and C termini,which are probably disordered in the crystal.Residues located in IMS loops exhibit largerDebye-Waller factors than others, which isalso indicative of partial disorder, preventingsome of the side chains to be modeled.
The Cavity
The cavity, formed by the TM helices, ismainly hydrophilic and enters very deeply intothe protein. It is shaped conically with an en-trance diameter of about 20 A, followed by anarrow funnel over 20 A long with a diame-ter of 8 A, and it is closed at 10 A from thematrix side (Figure 3). Several patches of ba-sic residues are observed. They are formedby residues K198/R104, R187/K91/K95,K22/R79/R279, and K32/R137/R234/R235(labeled 1 to 4 in Figure 4) and locatedfrom the entrance to the bottom of thecavity. Facing the second patch at roughly10 A distance from it, Y194 is the first ofthree tyrosines, Y194/Y190/Y186; this ar-rangement forms a ladder along H4, enter-ing the cavity. The ladder ends at the levelof the third basic patch, where the conic cav-ity is constricted to 8 A by four residuesY186/K22/R79/R279. The presence of manycharged or polar residues within the cavityis balanced by well-ordered water molecules,forming extensive hydrogen-bond networksinvolving both water molecules and aminoacid side chains. One of the largest net-works connects all TM helices except H4and involves the side chains of K22, E29,K32, Q36, N76, R79, D134, R137, T232,
www.annualreviews.org • The Mitochondrial ADP/ATP Carrier 717
Ann
u. R
ev. B
ioch
em. 2
006.
75:7
13-7
41. D
ownl
oade
d fr
om a
rjou
rnal
s.an
nual
revi
ews.
org
by U
nive
rsity
of
Cal
ifor
nia
- Ir
vine
on
09/0
5/06
. For
per
sona
l use
onl
y.
ANRV277-BI75-28 ARI 12 May 2006 15:19
Tab
le1
Com
pila
tion
ofch
emic
ally
labe
led
and
mut
ated
amin
oac
ids
ofA
DP
/AT
Pca
rrie
rsan
dth
eir
cons
eque
nces
once
llan
dki
neti
cpr
oper
ties
Loc
atio
na
Bov
isof
orm
1b
Hu
isof
orm
1c
Scis
ofor
m2d
Scis
ofor
m1e
Nc
isof
orm
1fO
rgan
ism
Mut
atio
nSe
cond
site
reve
rtan
tof
Gro
wth
ongl
ycer
olP
rote
inco
nten
t
ATR
/C
ATR
bind
ing
BA
bind
ing
Nuc
leot
ide
exch
ange
Oth
erR
efer
ence
s
MM
MM
MSc
2M
atur
atio
naf
ter
tran
slat
ion
(44)
Nt
——
S1S1
A1
Sc2
N-a
cety
latio
n(4
4)
Nt
D2
D2
P14
E4
—H
uH
uA
,KE
Nog
Yesg
0h(8
8)
H1
K9
K9
I24
VA
14H
uH
uA D
,EYe
sg
Yesg
148 pm
oles
/m
gh
Yes
Yesi,j
(88)
H1
D10
D10
D25
D15
D15
Sc2
Sc2
E VR
253I
R96
AYe
skIs
olat
edfr
omth
epa
rent
mut
atio
n(8
9)(9
0)
H1
G14
G14
G29
G19
G19
Sc2
V C S1
R96
T+D
149G
R25
3IYe
sYe
sYe
s24
6nM
m(9
0)(8
9)(9
1)
H1
A17
A17
S32
S22
S22
Sc2
Sc2
N IR
253I
R96
LYe
sk(8
9)(9
0)
H1
K22
K22
K37
K27
K27
Sc2
Ncn
Bov
A AN
o57
%o
57%
p;
38%
q
73%
v
80%
r ;89
%s
97%
w
0%;4
9%t
22%
t1.
2%u
x(3
3,92
)(9
3)(9
4)
H1
E29
E29
E44
E34
E34
Sc2
Sc2
Q GR
151A
orR
293A
K37
Aor
R29
3A
Yes
Yes
49%
o46
%p
;33
%q
52%
r ;86
%q
83%
t37
%u
(33)
(92)
H1
K32
K32
K47
K37
K37
Sc2
Ncn
A AN
o43
%o
6%p
5%r
9%t
1%t
No
ATP
synt
hesi
s(9
2)(9
3)
H1
V37
V37
N52
N42
N42
Sc2
IR
252I
Yesk
Isol
ated
from
the
pare
ntm
utat
ion
(89)
M1
K42
K42
E55
E45
E45
Bov
qq(9
4)
M1
K48
K48
T61
S51
R51
Bov
qq(9
4)
M1
K51
K51
A67
K57
G57
Bov
Tri
met
hyla
tion
(63)
h12
C56
C56
C72
C62
C62
Sc2
Sc2
Sc1
Bov
S A S
Yes
Yes
53%
y0.
4%;
26%
oo+/
−CL
z
aa bb,c
c
(32)
(95)
(70)
(51,
81,9
6)
M1
W70
W70
W86
W76
W76
Sc2
YYe
sYe
sYe
sYe
sdd
(97,
98)
M1
R71
R71
R87
R77
R77
Sc2
A,D
Yes
(90)
H2
R79
R79
R95
R85
R85
Sc2
Sc2
Sc2
Sc2
Ncn
A D,T
,K,Q
Hrr
L,P
A
No
No
No
No
0o 35%
o-
WT
h
0p 28%
p;
49%
v
Yes
0r 36%
r ;74
%w
Yes
0y 6%;
14%
t,ee
13%
t
0%–7
%i,u
3%–6
%i,u
(99)
(99;
G.
Lau
quin
,un
publ
ishe
dda
ta)
(100
,101
)(9
3)
718 Nury et al.
Ann
u. R
ev. B
ioch
em. 2
006.
75:7
13-7
41. D
ownl
oade
d fr
om a
rjou
rnal
s.an
nual
revi
ews.
org
by U
nive
rsity
of
Cal
ifor
nia
- Ir
vine
on
09/0
5/06
. For
per
sona
l use
onl
y.
ANRV277-BI75-28 ARI 12 May 2006 15:19
H2
Y80
Y80
Y96
Y86
Y86
Sc2
CYe
slYe
s22
3 nMm
Yes
(91)
H2
F88
F88
F104
F94
F94
Sc2
L,V
R25
3I(9
3)
H2
A89
A89
A10
5A
95A
95Sc
2T
R25
2I(9
3)
H2
I97
L97
M11
3L
103
M10
3H
uP
Ass
ocia
ted
with
adP
EO
(102
)
C1
D10
3D
103
K11
7D
107
K10
7H
uG
Ass
ocia
ted
with
adP
EO
(103
)
H3
Y11
1Y
111
W12
5W
116
W11
6Sc
2Y
Yes
Yes
Yes
Yes
ff(9
7,98
)
H3
A11
3A
113
A12
7A
118
A11
8H
uSc
2P P
Nog
Yes
0h WT
hYe
sN
oW
Tgg
Ass
ocia
ted
with
adP
EO
(88,
104)
H3
L12
7L
127
L14
1L
132
L13
2Sc
2S
Yesl
hh21
5 nMm
(91)
H3
D13
4D
134
D14
8D
139
D13
9Sc
2Sc
2S N
R15
1AN
oYe
s46
%o
5%p
6%r
0tN
oAT
Psy
nthe
sis
(92)
(30)
H3
R13
7R
137
R15
1R
142
R14
2Sc
2Sc
2N
cn
A K A
R15
1AN
oYe
s18
%o
21%
p;
22%
q
19%
ii,v
31%
r ;90
%s
87%
w
23%
;92%
t
34%
tN
oAT
Psy
nthe
sis
(92)
(98)
(93)
H3
R13
9R
139
R15
3R
144
R14
4N
cnA
84%
v95
%w
5%t
(93)
h34
C15
9C
159
V17
5V
166
V16
6B
ovjj
Ver
yra
pid
labe
ling
with
EM
Ain
inve
rted
part
icle
s
(96,
105)
h34
K16
2K
162
K17
8K
169
K16
9B
ovSc
2Sc
2
kk,l
lM I
Yes
Yes
97%
o90
%p
;39
%q
146%
r ;61
%s
36%
;149
%t
32%
t60
%u
z(4
2,43
,94)
(99,
101)
(33)
h34
K16
5K
165
K18
1K
172
K17
2B
ovSc
2ll Im
mYe
s49
%o
48%
p;
35%
q70
%r ;
68%
s69
%t
z(4
2,43
)(9
2)
h34
h34
M2
S166
D16
7G
168
S166
D16
7G
168
S182
D18
3G
184
T17
3D
174
G17
5
S173
D17
4G
175
Sc2
2-A
zNA
DP
labe
ling
ofth
eSe
r182
-Arg
190
pept
ide
(44)
M2
L16
9L
169
V18
5L
176
I176
M2
R17
0R
170
A18
6L
177
A17
7
M2
G17
1G
171
G18
7G
178
G17
8
M2
L17
2L
172
L18
8L
179
L17
9
M2
Y17
3Y
173
Y18
9Y
180
Y18
0
M2
Q17
4Q
174
R19
0R
181
R18
1
H4
V18
0V
180
V19
6V
187
V18
7H
uM
,Fg
>Ye
sYe
s�
(106
)
H4
R18
7R
187
R20
3R
194
R19
4Sc
2N
cnL A
No
0h,o
0p Yes
0r Yes
0%t
6%t
No
ATP
synt
hesi
san
dno
seco
ndsi
tere
vert
ant
(33,
99,1
00)
(93)
H4
Y19
0Y
190
Y20
6Y
197
Y19
7Sc
2H
R25
3IN
oAT
Psy
nthe
sis
(89) (Con
tinue
d)
www.annualreviews.org • The Mitochondrial ADP/ATP Carrier 719
Ann
u. R
ev. B
ioch
em. 2
006.
75:7
13-7
41. D
ownl
oade
d fr
om a
rjou
rnal
s.an
nual
revi
ews.
org
by U
nive
rsity
of
Cal
ifor
nia
- Ir
vine
on
09/0
5/06
. For
per
sona
l use
onl
y.
ANRV277-BI75-28 ARI 12 May 2006 15:19
Tab
le1
(Con
tin
ued
)
Loc
atio
na
Bov
isof
orm
1b
Hu
isof
orm
1c
Scis
ofor
m2d
Scis
ofor
m1e
Nc
isof
orm
1fO
rgan
ism
Mut
atio
nSe
cond
site
reve
rtan
tof
Gro
wth
ongl
ycer
olP
rote
inco
nten
t
ATR
/C
ATR
bind
ing
BA
bind
ing
Nuc
leot
ide
exch
ange
Oth
erR
efer
ence
s
H4
G19
2G
192
G20
8G
199
G19
9Sc
2S
R25
3I(8
9)
H5
A21
6A
216
G23
3G
224
G22
4Sc
2S
R25
3I(8
9)
H5
Q21
7Q
217
W23
4W
225
W22
5Sc
2Sc
2Sc
2
Y F L
R25
3IYe
sYe
sYe
s
Yes
Yes
Yes
nn(9
7,98
)(8
9,97
)(8
9,97
)
H5
V22
6V
226
C24
3A
234
A23
4Sc
2Sc
2S A
Yes
Yes
93%
yYe
s95
%oo
(32)
H5
P22
9P
229
P24
6P
237
P23
7Sc
2G
Yes
No
ATP
synt
hesi
s(9
9)
H5
D23
1D
231
D24
8D
239
D23
9Sc
2S
No
22%
o0p
0r0
No
ATP
synt
hesi
s(9
2)
H5
R23
4R
234
R25
1R
242
R24
2Sc
2N
cnI A
No
32%
o24
%p
;38
%v
Yes
24%
r ;56
%w
Yes
1%;7
%t
23%
tN
oAT
Psy
nthe
sis
and
nose
cond
site
reve
rtan
t
(33,
99,1
00)
(93)
H5
R23
5R
235
R25
2R
243
R24
3Sc
2N
cnI A
No
23%
o24
%p
;48
%v
Yes
24%
r ;84
%w
Yes
2%t
15%
tN
oAT
Psy
nthe
sis
and
4se
cond
site
reve
rtan
ts
(33,
99,1
00)
(93)
H5
R23
6R
236
R25
3R
244
R24
4Sc
2N
cnI A
No
29%
o24
%p
;26
%v
Yes
24%
r ;91
%w
Yes
1%;3
%t
127%
tN
oAT
Psy
nthe
sis
and
11se
cond
site
reve
rtan
ts
(33,
99,1
00)
(93)
h56
C25
6C
256
C27
0C
261
A26
1Sc
2Sc
2Sc
1
S A A
Yes
Yes
46%
y55
%y,
oo(3
2)(9
5)(7
0)
H6
A27
3A
273
C28
7C
278
A27
8Sc
2Sc
2Sc
1
Y A A
R96
AYe
sYe
sYe
s(3
2)(9
5)(7
0)
H6
R27
9R
279
R29
3R
284
R28
4Sc
2N
cnA A
No
74%
o78
%p
;3%
v94
%r ;
43%
w57
%y,
oo;7
3%to
125%
t
21%
t
11%
u(3
3,99
,100
)(9
3)
H6
G28
3G
283
G29
7A
288
G28
8Sc
2S
Yes
Yes
307 nM
mYe
sl
(91)
H6
A28
4A
284
A29
8A
289
A28
9Sc
2S
R25
3I(8
9)
H6
L28
7L
287
I301
I292
L29
2Sc
2T
R25
3I(8
9)
H6
V28
8V
288
S302
S293
S293
Hu
Sc2
M MN
og
Yesp
pA
ssoc
iate
dw
ithad
PE
O(8
8)
H6
Y29
0Y
290
Y30
4Y
295
Y29
5Sc
2H
R96
Aor
R25
2I(8
9)(9
0)
H6
I293
I293
L30
7L
298
L29
8Sc
2P
R25
2I(8
9)
H6
F296
Y29
6I3
10I3
01L
301
Sc2
2-A
zN
AD
Pla
belin
gof
the
Ile3
10–L
ys31
7pe
ptid
espr
even
ted
byC
ATR
bind
ing
(44)
720 Nury et al.
Ann
u. R
ev. B
ioch
em. 2
006.
75:7
13-7
41. D
ownl
oade
d fr
om a
rjou
rnal
s.an
nual
revi
ews.
org
by U
nive
rsity
of
Cal
ifor
nia
- Ir
vine
on
09/0
5/06
. For
per
sona
l use
onl
y.
ANRV277-BI75-28 ARI 12 May 2006 15:19
a Loc
atio
nof
the
amin
oac
ids
dedu
ced
from
the
thre
e-di
men
sion
alst
ruct
ure
ofth
ebo
vine
AD
P/A
TP
carr
ier
(isof
orm
1)an
dfr
omth
eam
ino
acid
sequ
ence
alig
nmen
t.A
min
oac
ids
are
indi
cate
dus
ing
the
one-
lett
erco
dean
dar
efo
llow
edw
ithth
eir
num
beri
ngin
the
sequ
ence
.Tho
seco
nser
ved
inal
lkno
wn
AD
P/A
TP
carr
ier
sequ
ence
sar
ein
red.
Abb
revi
atio
ns:N
t:no
n-or
dere
dN
-ter
min
alre
gion
;Ct:
non-
orde
red
C-t
erm
inal
regi
on;H
1to
H6:
TM
helix
1to
6;M
1an
dM
3:m
atri
xlo
op1
and
3;h1
2:sm
allh
elic
ein
M1;
h34:
smal
l
helic
ein
M2;
h56:
smal
lhel
ice
inM
3;C
1:cy
toso
liclo
op1,
asde
fined
inFi
gure
2.b
Bov
:bov
ine.
c Hu:
hum
an.
dSc
:Sac
char
omyc
esce
revi
siae
with
Scis
ofor
m2
orSc
2,w
hich
isis
ofor
m2
ofSc
AD
P/A
TP
carr
ier.
e Sc
isof
orm
1or
Sc1,
whi
chis
isof
orm
1of
ScA
DP
/AT
Pca
rrie
r.f N
c:N
euro
spor
acr
assa
.g H
eter
olog
ous
expr
essi
onin
yeas
t.h
Max
imum
num
ber
of[3
H]A
TR
-bin
ding
site
son
isol
ated
mito
chon
dria
inpm
oles
/mg
mito
chon
dria
lpro
tein
s.T
his
valu
eis
450
pmol
es/m
gfo
rth
ew
ild-t
ype
Scis
ofor
m2
and
174
pmol
es/m
gfo
rth
ew
ild-t
ype
Hu
isof
orm
1pr
oduc
edin
yeas
t.i E
xcha
nge
activ
itym
easu
red
with
isol
ated
mito
chon
dria
.j A
DP
/AT
Pex
chan
gepa
ram
eter
sw
ithis
olat
edm
itoch
ondr
iaar
eV
AD
PM
=32
.6nm
oles
/mn/
mg;
KM
=3.
7μ
Mfo
rth
ew
ild-t
ype
Hu
isof
orm
1an
dV
AD
PM
=95
.5nm
oles
/mn/
mg;
KM
=7.
2μ
Mfo
rth
eK
10A
mut
ant.
k Doe
sno
tsup
port
grow
thon
glyc
erol
at37
◦ C.
l Con
fers
invi
voB
Are
sist
ance
.m
Cor
resp
onds
toth
ein
hibi
tor
conc
entr
atio
nyi
eldi
ngha
lfth
em
axim
umva
lue
ofth
eA
DP
/AT
Pex
chan
gew
ithis
olat
edm
itoch
ondr
ia(I
CBA 50
).T
his
valu
eis
145
nMfo
rth
ew
ild-t
ype
Scis
ofor
m2.
nA
DP
/AT
Pca
rrie
rex
pres
sed
inE
.col
iinc
lusi
onbo
dy.
oA
DP
/AT
Pca
rrie
rco
nten
tdet
erm
ined
byco
mpe
titiv
een
zym
e-lin
ked
imm
unos
orbe
ntas
say.
p[3
H]C
ATR
bind
ing
tom
itoch
ondr
iaex
pres
sed
asa
perc
enta
geof
bind
ing
obta
ined
with
mito
chon
dria
isol
ated
from
yeas
tcel
lsex
pres
sing
the
epis
omic
wild
-typ
eSc
isof
orm
2en
codi
ngge
ne.
qP
erce
ntag
eof
inhi
bitio
nof
AD
P/A
TP
exch
ange
byC
ATR
(10
μM
)in
reco
nstit
uted
vesi
cles
with
isol
ated
Scis
ofor
m2.
Wild
-typ
eSc
isof
orm
2A
DP
/AT
Pex
chan
gein
hibi
tion
is7%
.r [
3 H]B
Abi
ndin
gto
mito
chon
dria
expr
esse
das
perc
enta
geof
bind
ing
obta
ined
with
mito
chon
dria
isol
ated
from
yeas
tcel
lsex
pres
sing
the
epis
omic
wild
-typ
eSc
isof
orm
2en
codi
ngge
ne.
s Per
cent
age
ofin
hibi
tion
ofA
DP
/AT
Pex
chan
geby
BA
(10
μM
)in
reco
nstit
uted
vesi
cles
with
isol
ated
Scis
ofor
m2.
Wild
-typ
eSc
isof
orm
2A
DP
/AT
Pex
chan
gein
hibi
tion
is83
%.
t Per
cent
age
ofw
ild-t
ype
activ
ity(A
DP
/AD
Pex
chan
ge)i
na
reco
nstit
uted
syst
em.
uP
erce
ntag
eof
wild
-typ
eAT
Psy
nthe
sis
activ
ityw
ithis
olat
edm
itoch
ondr
ia.
v Per
cent
age
ofin
hibi
tion
ofA
DP
/AD
Pex
chan
geby
CAT
R(1
0μ
M)i
nre
cons
titut
edve
sicl
esw
ithis
olat
edSc
isof
orm
2.W
ild-t
ype
Scis
ofor
m2
AD
P/A
DP
exch
ange
inhi
bitio
nis
11%
.w
Per
cent
age
ofin
hibi
tion
ofA
DP
/AD
Pex
chan
geB
A(1
0μ
M)i
nre
cons
titut
edve
sicl
esw
ithan
isol
ated
Nc
AD
P/A
TP
carr
ier.
Wild
-typ
eca
rrie
rA
DP
/AD
Pex
chan
gein
hibi
tion
is97
%.
x Onl
yly
sine
prot
ecte
dby
CAT
R,n
otby
ATR
orB
A,a
gain
stSc
hiff
base
form
atio
nw
ithpy
rido
xalp
hosp
hate
incu
bate
dw
ithin
tact
mito
chon
dria
.y R
econ
stitu
tion/
lipos
omes
.z C
ardi
olip
inre
quir
emen
tdur
ing
reco
nstit
utio
nex
peri
men
tsfo
rA
DP
/AT
Pex
chan
geac
tivity
.aa
Seve
nty-
five
perc
entd
ecre
ase
indi
mer
form
atio
naf
ter
impo
rtof
Scis
ofor
m1
into
the
mito
chon
dria
linn
erm
embr
ane.
bbV
ery
rapi
dN
-eth
ylm
alei
mid
ela
belin
gon
subm
itoch
ondr
ialp
artic
les
inhi
bite
dby
CAT
R.
ccIn
term
olec
ular
disu
lfide
brid
geca
taly
zed
byco
pper
-o-p
hena
nthr
olin
ean
dbi
func
tiona
lmal
eim
ides
.dd
Ver
yw
eak
fluor
esce
nce
chan
geup
onAT
Pbi
ndin
g:�
F/F
=0.
5%in
stea
dof
5%fo
rw
ild-t
ype
Scis
ofor
m2.
eeT
hem
axim
umra
teof
AD
P/A
TP
exch
ange
with
isol
ated
mito
chon
dria
issi
mila
rto
that
ofw
ild-t
ype
Scis
ofor
m2
and
the
KM
cons
tant
is50
0tim
eshi
gher
(V.P
ostis
,per
sona
lcom
mun
icat
ion)
.ff
No
fluor
esce
nce
chan
geup
onAT
Pan
dB
Abi
ndin
g.gg
AD
P/A
TP
exch
ange
expe
rim
ents
wer
epe
rfor
med
with
isol
ated
mito
chon
dria
.Kin
etic
para
met
ers
for
the
mut
anta
resi
mila
rto
that
for
the
wild
-typ
eSc
isof
orm
2,w
hich
are
VA
DP
M=
69.6
nmol
es/m
n/m
g;K
M=
0.67
μM
(fre
eA
DP
).hh
KAT
Rd
=32
30nM
asco
mpa
red
to22
0nM
for
the
wild
-typ
eSc
isof
orm
2.ii
Nin
etee
npe
rcen
tinh
ibiti
onof
activ
ityw
ith10
μM
CAT
Rin
stea
dof
55%
for
the
wild
-typ
epr
otei
n.jj [
3 H]7
-azi
do-4
-iso
prop
ylac
rido
nela
belin
gof
Cys
159
ofB
ovA
DP
/AT
Pca
rrie
rin
mito
chon
dria
.kk
Schi
ffba
sefo
rmat
ion
with
pyri
doxa
lpho
spha
tein
mito
chon
dria
,not
insu
bmito
chon
dria
lpar
ticle
s,pr
even
ted
byC
ATR
and
BA
.ll
May
bela
bele
dw
ith2-
azid
oA
DP
(Bov
AD
P/A
TP
carr
ier)
,or
2-az
ido
and
8-az
ido
ATP
(yea
stSc
isof
orm
2).
mm
Inco
mbi
natio
nw
ithK
178M
mut
atio
n.nn
�F/
Fup
onAT
Por
BA
bind
ing
issi
mila
rto
that
ofth
ew
ild-t
ype
prot
ein.
ooP
erce
ntag
eof
wild
-typ
eac
tivity
(AD
P/A
TP
exch
ange
)with
adde
dca
rdio
lipin
s.pp
C.D
eM
arco
sL
ousa
,per
sona
lcom
mun
icat
ion.
qqSc
hiff
base
form
atio
nw
ithpy
rido
xalp
hosp
hate
inSM
Pin
the
pres
ence
ofB
Abu
tpre
vent
edby
CAT
R.
rrT
hear
gini
neto
hist
idin
ech
ange
corr
espo
nds
toth
eop
1m
utat
ion
first
desc
ribe
dby
Slon
imsk
iand
colla
bora
tors
(107
)and
late
rid
entifi
edby
Kol
arov
and
colla
bora
tors
(108
).
www.annualreviews.org • The Mitochondrial ADP/ATP Carrier 721
Ann
u. R
ev. B
ioch
em. 2
006.
75:7
13-7
41. D
ownl
oade
d fr
om a
rjou
rnal
s.an
nual
revi
ews.
org
by U
nive
rsity
of
Cal
ifor
nia
- Ir
vine
on
09/0
5/06
. For
per
sona
l use
onl
y.
ANRV277-BI75-28 ARI 12 May 2006 15:19
Figure 1Overall structure of the bovine ADP/ATP carrier. The ribbon diagram, colored blue to red from the Nterminus (N) to the C terminus (C), depicts transmembrane helices (H1–H6), loops facing the IMS (C1and C2), and loops facing the matrix (M1–M3). Matrix loops are partially structured in short helices(h1–2, h3–4, and h5–6). Three cardiolipins (CDLs), CDL800, CDL801, and CDL802, are bound to thestructure and represented as ball and sticks in gray. The inhibitor, CATR, complexed with the protein isdepicted in yellow. Panels a, b, c are viewed from the IMS, the side, and the matrix, respectively. Thecolor code for the ribbon diagram is the same as for Figures 2, 3, 6, 7, and 11.
R234, R235, N276, R279 and about 20 watermolecules (Figure 5). Almost all the residuesdescribed in this paragraph are conservedamong ADP/ATP carriers (see the supple-mentary material). Follow the SupplementalMaterial link from the Annual Reviews homepage at http://www.annualreviews.org.
CATR Binding
The inhibitor CATR, cocrystallized with thecarrier, is located within the cavity with itsditerpene moiety oriented toward the bottomand its sulfate groups toward the IMS (seefigure 5 in Reference 4). The inhibitor bindsto the carrier through numerous interactionsthat involve all the chemical groups exceptthe primary alcohol located on the sugar ring.This structural observation is consistent withthe inhibitory properties of CATR, which arereduced once the molecule is truncated ormodified, except for the primary alcohol thatcan be modified without changing the inhibi-
tion capacity (25). The two carboxylates andthe hydroxyl group on the diterpene moiety,as well as both sulfates on the glucose ring, arelinked to the carrier through electrostatic in-teractions or hydrogen bonds, some of whichalso involve water molecules. The isovalericchain and the diterpene ring interact via vander Waals contacts.
The MCF Motif
MCF members are characterized by triplica-tion of the following motif: PxD/ExxK/RxK/R-(20 to 30 residues)-D/EGxxxxaK/RG,where the letter a represents an aromaticresidue (Figure 2b). Several authors refinedthis motif by restricting some residues la-beled as x (26). The PxD/ExxK/R sequencecan be directly related to the basket shape ofthe carrier. Indeed, the prolines located in thesecond half of odd-numbered helices, closeto the matrix, induce sharp kinks responsiblefor the closed form toward the matrix. It is
722 Nury et al.
Ann
u. R
ev. B
ioch
em. 2
006.
75:7
13-7
41. D
ownl
oade
d fr
om a
rjou
rnal
s.an
nual
revi
ews.
org
by U
nive
rsity
of
Cal
ifor
nia
- Ir
vine
on
09/0
5/06
. For
per
sona
l use
onl
y.
ANRV277-BI75-28 ARI 12 May 2006 15:19
C1 C2 C
H2
H3
H4
H5
H6
h5-6h
3-4h
1-2
H1
Intermembranespace
Matrix37
53 64
99 108
142
156
199 209 290
238
253 264167
73 273176
4
30 40 50 60 70
P I E RV K L L LQVQHASKQISAEKQYKGIIDCVVRIPKE Q FLSF WRepeat 1
P L D FA R T R LAADVGKGAAQREFTGLGNCITKIFKS D LRGL YRepeat 2140 150 160 170
P F D TV R R R MMM QSGRKGADIMYTGTVDCWRKIAKD E PKAF FRepeat 3
230 2 40 250 260 270
a
b
N
PxD/ExxK/RxK/R (20 to 30 residues)
RG
G
G
Q
K
G
G
G
D/EGxxxxaK/RG
Figure 2Overall topology and MCF motifs of the bovine ADP/ATP carrier. (a) Schematic diagram of thesecondary structure. Regions containing MCF motif residues are colored in gray, and the RRRMMMmotif is in back. Kinks in H1, H3, and H5 are induced by the prolines. (b) Alignment of the three MCFmotifs. On the top, the consensus MCF sequence boxed in gray. The ADP/ATP carrier signature presentin the third repeat is boxed in black.
Figure 3Surface representation of the cavity. Thelongitudinal section through the cavity shows thewide cavity present in the bovine ADP/ATP carrierand accessible from the IMS. R234, R235, andR236, the three arginines of the ADP/ATP carriersignature, located on the C-terminal end of H5 areshown, as well as E264, which forms a salt bridgewith R236 (yellow). From Pebay-Peyroula et al. (4).
known that, in the absence of other specificinteractions, proline residues can adopt abroad range of stable conformations (27).Therefore, when present in helices, surround-ing interactions might induce a given kink an-gle. This is probably partly achieved by theacidic and basic residues that follow the pro-lines and form salt bridges that strengthenthe closed conformation of the helix bun-dle (Figure 6). If the surrounding interac-tions are modified, the kink angles may bechanged. Therefore, the prolines were pro-posed to act as hinges that could allow theopening toward the matrix, which could betriggered by the disruption of the salt bridges(28). However, it was also proposed that, dur-ing evolution, prolines could have mutatedinto serines and that subsequent packing de-fects were compensated by other mutationslocking the kink in the structure (29). Thiscould explain why the proline of the sec-ond motif is not conserved among ADP/ATP
www.annualreviews.org • The Mitochondrial ADP/ATP Carrier 723
Ann
u. R
ev. B
ioch
em. 2
006.
75:7
13-7
41. D
ownl
oade
d fr
om a
rjou
rnal
s.an
nual
revi
ews.
org
by U
nive
rsity
of
Cal
ifor
nia
- Ir
vine
on
09/0
5/06
. For
per
sona
l use
onl
y.
ANRV277-BI75-28 ARI 12 May 2006 15:19
-180°
-180° 180°
-120° -60° 60°0°
0°
120° 180°H105
D291
Q84 Q217
T220
Y190
F191 N115
Y186
R187
G123 S126T83
K91
N87
G14A17
A284
A216A197
S21 G283A18
R79
D134N76
N276
Y228K32
E29
R137
V25
K22 G280
R279V277
G224S227
R235 F230
G182
I183V130
R234
L127
D134
D231
L287Y80
S212
Y194 K95
K198 Y111
R104
1
Y
Y2
Y
3
4
Selectivity filter
Bottom of the cavity
Top of the cavity
IMS
Figure 4Distribution of residues within the cavity. The figure represents a two-dimensional projection of theresidues present at the surface of the cavity. Each circle represents an atom of a residue located within thecavity, with a size proportional to its solvent accessibility. Residues are colored as follows: basic, K, R(blue); acidic, D, E (red); aromatic, F, Y, W ( gray); hydrophobic, A, V, P, M, I, L, G ( yellow); and polar, S,T, H, C, N, Q ( green). The positive patches are labeled 1 to 4, and the tyrosine ladder is marked by Y.
carriers and is often found to be a serine. Thesecond basic residue of the first part is re-placed with a leucine, L34, in the first mo-tif, although in the second and third, it isbasic, R139 and R236. R139 participates inthe interaction between M2 and M3 throughE152, M238, and S241. The interaction ofR236 with E264 is discussed below. The sec-ond part of the motif spans from the end ofthe short amphipatic helices to the N terminiof the even-numbered helices. Each glycinetherefore delineates a helix extremity and al-lows flexibility of the loop that links both he-lices. The acidic residues of the two first mo-tifs, replaced with a glutamine Q64 in the firstmotif, participate with the interactions be-tween M1/M2 and M2/M3, respectively. Theresidue of the third motif, E264, forms a saltbridge with R236, belonging to the ADP/ATPcarrier signature, but also contacts K271,the basic residue of the third MCF motif(Figure 7). These interactions clamp togetherthe C terminus of H5 with the C terminus
of h5–6 and the N terminus of H6. The ba-sic residue at the end of the first MCF mo-tif, R71, interacts with a CDL through awater molecule and participates with the in-teraction of loop M1 with M2 (through D143and G145), and in the second motif, theresidue is replaced with a glutamine, Q174.
The sequence alignments of severalMCF carriers highlight further conserva-tions within the range of the second partof the MCF motif. Indeed, for ADP/ATPcarriers, the short amphipatic helices andthe second half of the three matrix loopsare conserved in length and in sequences(alignments shown in supplementary ma-terial). Follow the Supplemental Materiallink from the Annual Reviews home pageat http://www.annualreviews.org. On thecontrary, the first half of the matrix loops, pre-ceding the short amphipatic helices, are rathervariable in length and in composition withinMCF members or within the three motifs ofa single carrier.
724 Nury et al.
Ann
u. R
ev. B
ioch
em. 2
006.
75:7
13-7
41. D
ownl
oade
d fr
om a
rjou
rnal
s.an
nual
revi
ews.
org
by U
nive
rsity
of
Cal
ifor
nia
- Ir
vine
on
09/0
5/06
. For
per
sona
l use
onl
y.
ANRV277-BI75-28 ARI 12 May 2006 15:19
Figure 5Schematicrepresentation of alarge hydrogen-bondnetwork. Thenetwork connects allthe TM helices,except H4. Itimplicates side chainsof polar, acidic, andbasic residues thatare highly conservedwithin ADP/ATPcarriers, as well asmain-chaincarbonyls (labeledCO) and watermolecules.Hydrogen bonds arededuced from atomicdistances and arerepresented as dottedlines.
Location of the ADP/ATP CarrierSignature
All ADP/ATP carriers that belong to theMCF are characterized by a unique signature,RRRMMM. These residues in the bovine car-rier are located at the C terminus of H5. Thefirst and the third arginines, R234 and R236,are part of the MCF motif (Figure 2a). Thestructure reveals that the three arginines spanthe thinnest part of the protein; the side chainsof R234 and R235 are accessible from the cav-ity open toward the IMS, and R236 points to-ward the matrix but is shielded from the sur-face by a salt bridge involving E264. R234 andR235 are part of a large hydrogen-bond net-work involving water molecules at the bottomof the cavity (Figure 5). R234 is also involvedin a salt bridge described above, with the MCFmotif acidic residue D134, and R235 inter-acts with the MCF motif residue D231. In-terestingly, the salt bridge between R234 andD134 was predicted from revertant studies(30). Both arginines, R234 and R235, inter-act with CATR. In addition to the salt bridgewith E264, R236 is implicated in a network
Figure 6The kinked conformation of odd-numbered helices. H1, H3, and H5,represented as ribbons, are kinked after prolines P27, P132, and P229,which are the first residues in each MCF motif. Acidic and basic residuesalso belonging to the MCF motif form salt bridges (dotted lines) that tie thethree helices together.
www.annualreviews.org • The Mitochondrial ADP/ATP Carrier 725
Ann
u. R
ev. B
ioch
em. 2
006.
75:7
13-7
41. D
ownl
oade
d fr
om a
rjou
rnal
s.an
nual
revi
ews.
org
by U
nive
rsity
of
Cal
ifor
nia
- Ir
vine
on
09/0
5/06
. For
per
sona
l use
onl
y.
ANRV277-BI75-28 ARI 12 May 2006 15:19
Figure 7MCF motif of thethird repeat. Theprotein betweenthe C terminus ofH5 and the Nterminus of H6 isrepresented asribbons. Sidechains of MCFmotif residues andCDL801 areshown inball-and-stickform. A salt bridgebetween R236(ADP/ATP carriersignature) andE264 (MCF motif)is highlighted.F270 is sandwichedbetween P229 andCDL801.
of hydrogen bonds including K271 and sev-eral water molecules, which link M3 to thefirst part of M1 (several interactions implicateresidues L35, Q36, Q38, H39, I44, Q49, andY50).
Lipids
The primitive crystal form that diffracted to2.2-A resolution showed the presence of twoor three CDLs tightly bound to the carrier(Figure 1b,c). Though it could not be identi-fied as a CDL from the experimental elec-tron density maps, the third lipid was des-ignated as a CDL because its position wasrelated to the two others by the pseudo-threefold symmetry. The presence of threeCDLs was confirmed by the second crystalform in which the three CDLs were nicelyidentified. CDLs are known to bind stronglyto the carrier and to remain present after ex-traction and purification from the mitochon-drial membrane (31). CDLs are located inthe inner leaflet of the membrane and par-tially nested in small grooves formed by matrix
loops. CDL801 clamps the N terminus of h1-2 to the C terminus of h5-6 and H6. CDL800and CDL802 interact similarly. CDL acylchains interact with the protein through hy-drophobic or aromatic residues, whereas thephosphate groups and their glycerol linkersare involved in hydrogen bonds with main-chain nitrogens or carbonyls of symmetry-related residues located at the beginning ofthe short matrix helices and at the beginningof even-numbered helices. CDL801 interactsmainly with I53, I54, F270, G272, W274,and S275; CDL800 interacts with W70, G72,L74, L156, and G157; and CDL802 interactswith Y173, G175, V177, G252, T253, andV254 (7). Except for aromatic residues, mostof the other residues have short side chainsand interact with the lipids via main-chainatoms. Impaired activity of the carrier bear-ing a single mutation of either C56 or K162could be retrieved by adding CDLs to the pro-teoliposomes (32, 33). Both mutations occurin the short helices, h1-2 and h3-4, respec-tively, and probably destabilize the structure.CDLs could compensate for this.
726 Nury et al.
Ann
u. R
ev. B
ioch
em. 2
006.
75:7
13-7
41. D
ownl
oade
d fr
om a
rjou
rnal
s.an
nual
revi
ews.
org
by U
nive
rsity
of
Cal
ifor
nia
- Ir
vine
on
09/0
5/06
. For
per
sona
l use
onl
y.
ANRV277-BI75-28 ARI 12 May 2006 15:19
Deviation to Pseudo-ThreefoldSymmetry
The threefold symmetry results from evolu-tion. The existence of a structural motif re-peated three times in each MCF carrier isprobably related to stability but also to a com-mon transport mechanism that allows a smallmolecule to cross over the membrane. Diver-gence from a strict threefold repeat was in-duced by the necessity of selective transport.Therefore, internal symmetry and also de-viations from it are interesting features thathave to be analyzed. The three CDLs alsofollow the threefold symmetry. Maximum dif-ferences are at both extremities of each motif(close to IMS loops C1 and C2) and in the firstpart of the matrix loops (preceding the shorthelices). The three repeats have a very similar
skeleton, and root mean square deviations be-tween backbone atoms of the three motifs areabout 2 A [see figure 3 in (4)]. In particular,the first repeat has an additional turn in thefirst part of its matrix loop (one or two addi-tional residues compared to the first parts ofM3 and M2, respectively) (Figure 8), whichallows the interaction between Q43 and theside chain of D143 and main-chain atoms ofV144 and Q150, as well as between K42 andD247. As a result, the loop M1 protrudes to-ward the center of the protein surface on thematrix side (Figure 1c).
The backbone of H2 is bent and deviatestherefore from a straight α-helix in contrastto H4 and H6. Negative charges are rathersymmetrically distributed with the exceptionof E264, an extra negative charge, which is at
Figure 8Deviation from the pseudo-threefold symmetry. The superposition of repeat 1 (blue), 2 (pale green), and 3(pale yellow) highlights the similarity of helical parts. Repeats 2 and 3 were rotated to be superimposed onrepeat 1. The figure shows also a significant difference at the beginning of matrix loops (between theodd-numbered and the short helices), which is slightly longer for M1. M1 folds back toward the center ofthe protein, as seen in Figure 1c, and thus allows interactions with M2 and M3. Small ribbon portions ofM2 and M3 are depicted in green and yellow and represent these interactions.
www.annualreviews.org • The Mitochondrial ADP/ATP Carrier 727
Ann
u. R
ev. B
ioch
em. 2
006.
75:7
13-7
41. D
ownl
oade
d fr
om a
rjou
rnal
s.an
nual
revi
ews.
org
by U
nive
rsity
of
Cal
ifor
nia
- Ir
vine
on
09/0
5/06
. For
per
sona
l use
onl
y.
ANRV277-BI75-28 ARI 12 May 2006 15:19
the end of h5-6. Positively charged residuesare numerous and asymmetrically distributedwithin the cavity, in particular at the entrance.Many charged residues are also present withinthe matrix loops. A global analysis of the dis-tribution shows that all loops have an excessof positively charged residues with a ratio ofpositive to negative residues of 6/3, 5/3, and7/4 for M1, M2, and M3, respectively. Aro-matic residues are mainly located on the ex-ternal surface of the protein, probably at theboundary between hydrophobic lipid chainsand their hydrophilic head groups, as pre-viously described for various proteins (34).However, the third MCF repeat contains lessaromatic residues, especially in the IMS re-gion (only three compared to eight for repeat1 and seven for repeat 2). The tyrosine ladderof the second repeat, located on H4 and com-posed of Y186, Y190, Y194, and also F191,has side chains oriented toward the cavity, incontrast to most aromatic residues of this IMSregion in repeat 1 or 3. Aromatic residues arenot uniformly distribued within the cavity butare grouped along H4. Interestingly, H4 isalso the only TM helix not to be implicated inthe large hydrogen-bond network describedabove.
The two first MCF repeats also deviatefrom the strict consensus sequence, and thethird MCF motif is the only one that strictlyobeys the motif definition. Noticeable inter-actions are observed between K271 and E264as well as between E264 and R236. In addi-tion, the third motif contains many chargedamino acids, one of which, K267, interactswith a neighboring molecule in the centeredcrystal form both directly and also throughCDL801 (7).
Conserved Residues: A Structural orFunctional Role?
In order to utilize the bovine ADP/ATPstructure for a general analysis of mitochon-drial transport, it is of interest to com-pare the sequences within the ADP/ATPcarriers (see supplementary material; fol-
low the Supplemental Material link fromthe Annual Reviews home page at http://www.annualreviews.org) and also within theMCF. In a first step, we relate the sequencesimilarities or differences within ADP/ATPcarriers to structural elements observed forthe bovine carrier. As for many membraneproteins, the parts protruding from the mem-brane are less conserved than the hydropho-bic membrane-inserted parts. The most con-served region of the protein is the cavity,which has roughly twice as many strictly con-served residues (60%) (Figure 9) as the wholeprotein (33%) (Figure 10). The hydrophobicregion facing CDL800 is also more conserved.
Analysis of conserved residues by typeshows that aromatic residues are conservedto an unusual degree (92% of the phenylala-nines and 86% of the tyrosines are conservedas aromatic residues within the ADP/ATP car-rier subfamily) together with charged residues(80% of aspartic acids and 76% of argininesare conserved as acid or basic residues, respec-tively). The cavity of ADP/ATP carriers con-tains an unusually high amount of positivelycharged residues compared to correspondingresidues in other MCF carriers.
Glycine, the smallest amino acid, is knownto allow large conformational flexibility for itshigh entropic cost of insertion in secondarystructures. It is often found at the extrem-ity of α-helices for soluble protein. Glycinesare also involved in dimerization motifs GxxGfor monotopic membrane proteins and fa-vor helix-helix interactions in polytopic mem-brane proteins (35). In the bovine ADP/ATPcarrier structure, 70% of the glycines are con-served, and the glycine motifs present in TMhelices probably play a role in intramolecularTM-TM interactions.
NUCLEOTIDE ATTRACTIONAND BINDING
The specificity of nucleotide recognition andlocalization of their binding sites were deci-phered using nucleotide derivatives and pho-tolabeling approaches. The results obtained
728 Nury et al.
Ann
u. R
ev. B
ioch
em. 2
006.
75:7
13-7
41. D
ownl
oade
d fr
om a
rjou
rnal
s.an
nual
revi
ews.
org
by U
nive
rsity
of
Cal
ifor
nia
- Ir
vine
on
09/0
5/06
. For
per
sona
l use
onl
y.
ANRV277-BI75-28 ARI 12 May 2006 15:19
-180°
-180° 180°
-120° -60° 60°0°
0°
120° 180°
Selectivityfilter
Bottom of the cavity
Top of the cavity
IMS
Figure 9Conserved residues in the cavity. The representation is the same as in Figure 4, except that residuesaccessible within the cavity are colored according to their conservation among ADP/ATP carriers: nosimilarity (gray), medium or high similarities ( yellow or orange), respectively, and identical (red ).
Figure 10Conserved residues on external surfaces. Orientations in panels10a and 10b are the same as in 1a and 1c,respectively. Residues are colored according to conservation among ADP/ATP carriers from white to red(0% to 100% of similarity).
www.annualreviews.org • The Mitochondrial ADP/ATP Carrier 729
Ann
u. R
ev. B
ioch
em. 2
006.
75:7
13-7
41. D
ownl
oade
d fr
om a
rjou
rnal
s.an
nual
revi
ews.
org
by U
nive
rsity
of
Cal
ifor
nia
- Ir
vine
on
09/0
5/06
. For
per
sona
l use
onl
y.
ANRV277-BI75-28 ARI 12 May 2006 15:19
can now be combined with the bovine struc-ture, which revealed a cavity accessible tonucleotides from the IMS and highlightedthe precise location of functionally importantresidues.
Specificity for Binding and Transport
The binding and transport properties of eachMCF carrier are rather specific and werestudied particularly for ADP/ATP carriers.None of the naturally occurring pyrimidineribonucleotides 5′ diphosphate and triphos-phate binds to the carrier. Using syntheticanalogues of purine nucleotides, it becamepossible to characterize the two-step sequenceof the transport process, namely the recog-nition of a nucleotide to a specific site, fol-lowed by the vectorial process of transport.Because several adenine nucleotide analoguescan bind to the carrier with a high affinitywithout being transported, binding requires alower specificity than transport (36). A centralissue regarding the transport process involvesthe anti or syn conformation of nucleotides,which rely on the orientation of the planarpurine base with respect to the ribose ring.To be transported by the ADP/ATP carrier,a nucleotide must have a nonfixed anti con-formation, with an additional amino groupon C6 and an unsubstituted C2 atom. Forexample, 8-Br ADP and the derived 8-azidoADP, blocked in the syn conformation, bindto the carrier but are not transported. 2-azidoADP adopts the anti conformation but sub-stitution at position 2 prevents transport. Incontrast, analogues, such as formycin or tube-ricidin diphosphate and triphosphate and 1-Noxide-ADP or -ATP, fulfill the required struc-tural criteria and are transported. The pres-ence of bulky substituents at positions 2′ or3′ of the ribose moiety is tolerated for bind-ing (37) or even for transport in mitochondria(38).
Fluorescein derivatives are structurally re-lated to adenine nucleotides and thereforeare recognized by the ADP/ATP carrier, al-though not transported. For example, eosin
Y binds to the carrier from the matrix sideand inhibits the ADP/ATP transport in bovineheart inside-out submitochondrial particles(39). These effects were interpreted on the ba-sis of common structural features between theA/B and D rings of eosin Y, with the adeninering and the ribose moiety of the anti form ofADP, respectively. In contrast, the absence ofa negative potential at position N1 of guaninemight explain why GDP is not recognized bythe carrier.
The demonstration that the free forms ofADP and ATP are the actual substrates forthe ADP/ATP carrier came from experimentscarried out with the isolated carrier either in adetergent solution or after incorporation intothe membrane of liposomes. For example, thefact that ADP- or ATP-induced conforma-tional transitions of the isolated carrier (fol-lowed by tryptophanyl fluorescence changes)were abolished by Mg2+ ions afforded directevidence that Mg nucleotides are not rec-ognized by the carrier (40). Similarly, it wasclearly demonstrated that Mg2+ ions inhib-ited the ADP/ATP exchange in reconstitutedproteoliposomes (41). Finally, although AMPnucleotides are recognized by the carrier, onlyADP or ATP is transported.
Biochemical Evidence forIMS-Binding Sites
Localization of two specific nucleotide-binding regions of the ADP/ATP carrier wasachieved with photoaffinity radiolabeled nu-cleotides carrying a reactive azido group onthe adenine ring. Two segments of the peptidechain of the bovine carrier, spanning residuesF153 to M200 and Y250 to M281, were cova-lently labeled with 2-azido-ADP (42). Map-ping of the yeast carrier with 2-azido-ADPled to the labeling of a segment delimited byresidues G172 and M210 (43). A more pre-cise assignment of the binding region wasachieved with 2-azido-3′-O-naphthoyl-ADPand restricted to the S183-R191 segment,which is located in the second matrix loopbetween h3-4 and H4 (44). Although not
730 Nury et al.
Ann
u. R
ev. B
ioch
em. 2
006.
75:7
13-7
41. D
ownl
oade
d fr
om a
rjou
rnal
s.an
nual
revi
ews.
org
by U
nive
rsity
of
Cal
ifor
nia
- Ir
vine
on
09/0
5/06
. For
per
sona
l use
onl
y.
ANRV277-BI75-28 ARI 12 May 2006 15:19
directly accessible from the IMS in theCATR-ADP/ATP carrier structure, theseresidues are not far from the cavity, and it ispossible that binding the ADP, even labeled,already induces a conformational change. Anadditional segment spanning residues I311-K318, corresponding to the C-terminal end ofthe carrier, was also labeled (44). It was shownfrom the characterization of deletion mutantsthat this nucleotide-binding region plays acritical role for ADP/ATP transport in yeast(G. Brandolin, unpublished data). Whetherthe two nucleotide-binding segments belongto the same carrier monomer or to adjacentmonomers is still an open question. The otherpeptide segments implicated in nucleotidebinding also remain to be elucidated.
Structural Features
From the IMS, the patch of positive chargeslocated at the entrance of the cavity attractsADP3− in spite of the opposing membrane po-tential and possibly provides a first nucleotide-binding site. Furthermore, positive patches inthe middle and at the bottom of the cavityattract nucleotides to the bottom. The cav-ity narrows at a bottleneck located at 20 Afrom the entrance. The four residues sur-rounding the bottleneck are conserved onlywithin the ADP/ATP carrier subfamily (ex-cept for the carrier involved in Grave’s diseasefor which the transported molecule has notyet been clearly identified). These are threebasic residues, K22, R79, and R279 (the twoarginines related by the internal pseudosym-metry), and Y186, which belongs to the tyro-sine ladder. Mutations of the residues, corre-sponding to K22 and R79 in yeast, severelyimpair the transport properties (Table 1).Coming from the IMS, a nucleotide couldglide with its adenine ring along H4 usingthe tyrosine ladder, whereas the phosphateswould follow the basic patches. At the levelof the constriction, the four residues couldrestrict the entrance to the bottom of thecavity to adenine nucleotides. The cavity islimited to 8 A. It is therefore interesting to
note that the residues of the putative selec-tivity filter, K22, R79, Y186, and R279, cor-respond to R190, K241, Y338, and K434 inthe human Mg-ATP/Pi carrier, which importsMg-ATP and exports Pi. Although the lysinesand arginines are switched, the same typesof residues are found at the same locations.In addition, the Mg-ATP/Pi carrier also ex-hibits two out of the three tyrosines formingthe ladder, and the positive patches are con-served, except for the second one (R187 andK91 do not correspond to positive residues).Regarding the cavity, the Mg-ATP/Pi car-rier and the carrier implicated in Grave’s dis-ease are the most similar to the ADP/ATPcarrier in terms of charged and aromaticresidues.
The phosphate carrier is another MCFmember in which several residues, includingH32, T79, Y83, K90, Y94, and K98, were ex-pected to be important in the transport path-way (45). Interestingly, H32 and T79, residuesthat are unique to phosphate carriers, corre-spond to K22 and N76, respectively, of thebovine ADP/ATP carrier. K22 is describedherein as belonging to a putative selectivityfilter, and N76 is located close to it.
Binding ATP from the Matrix
From the structure obtained in the presence ofCATR, it is difficult to deduce ATP-bindingsites. However, it highlights a salt bridgebetween R236 and E264 (Figure 3). R236could be involved in ATP binding, thus mod-ifying the salt bridge and inducing a con-formational change. Binding of eosin Y, anon-thiol-reactive molecule, was shown to bedisplaced by ADP or ATP, and therefore eosinY was supposed to bind in close vicinity to thenucleotide site. In addition, binding of eosinY also prevented the thiol-reactive eosin-5-maleimide molecule to bind on C159 (39).The authors therefore suggested that ATPbinds in the vicinity of C159. However, in theabsence of further structural data, it is diffi-cult to discuss the ATP-binding site from thematrix.
www.annualreviews.org • The Mitochondrial ADP/ATP Carrier 731
Ann
u. R
ev. B
ioch
em. 2
006.
75:7
13-7
41. D
ownl
oade
d fr
om a
rjou
rnal
s.an
nual
revi
ews.
org
by U
nive
rsity
of
Cal
ifor
nia
- Ir
vine
on
09/0
5/06
. For
per
sona
l use
onl
y.
ANRV277-BI75-28 ARI 12 May 2006 15:19
TRANSPORT MECHANISM
The binding of nucleotides to the carrier hasto induce conformational changes that trig-ger a one-to-one ADP/ATP exchange. Thestructure in the presence of CATR and bio-chemical data offer a basis for the discussionon transport mechanism.
Conformational Changes
The very high selectivity of the ADP/ATPcarrier for adenine nucleotides and the recog-nition of fully charged species instead ofpartially neutralized Mg2+-nucleotide com-plexes considerably enhance the binding en-ergy of nucleotide-carrier interaction, whichin turn is used to drive the ADP and ATPtranslocation. Energy supply has to sustain theconformational rearrangements of the carrierrequired to transport large and charged nu-cleotides. The general concept of an inducedtransition fit mechanism that would provideenergy for transport through the formationof a transient carrier-substrate complex hasrecently been discussed (46). Indeed, normalmode calculations based on the structure ofthe bovine ADP/ATP carrier show that low-energy movements are not sufficient to inducestructural conformations that would allow thetransport of the nucleotides (P. Amara, per-sonal communication). However, conforma-tion fluctuations in the unloaded carrier couldbe necessary to expose transiently the confor-mations to which ADP and ATP bind. The ad-ditional energy needed for the transport pro-cess could come from the relaxation of bothADP3− and ATP4− in their binding sites.
The conformation changes undergone bythe carrier are a central issue with regard tothe molecular mechanism of the ADP/ATPtransport. This point was addressed by study-ing the modification of the carrier topog-raphy in the presence of CATR and BA.Conformers of the carrier were differenti-ated on the basis of their enzymatic, immuno-chemical, and chemical reactivities (for a re-view, see Reference 47). Experiments with the
bovine heart ADP/ATP carrier have shownthat the CATR-carrier complex is more io-dinated than the BA-carrier complex or thedenatured carrier (48). The N-terminal re-gion of the membrane-bound bovine carrier,which is exposed to the IMS in intact mito-chondria, is particularly sensitive to confor-mational changes; its reactivity to antibod-ies is much higher in the presence of CATRthan of BA (49). In addition, systematic single-cysteine mutants on the yeast carrier followedby thiol accessibility exploration showed thatthe accessibility of residues 98 to 106 (equiv-alent to 81–89 in the bovine carrier) is drasti-cally modified in the presence of BA comparedto CATR (50). The modifications were inter-preted by the authors as a 180◦ twist of H2.Residues 81 to 89 of the bovine carrier arelocated in the vicinity of the N-terminal endup to residue 12; therefore structural modifi-cations of H2 could be correlated to that ofthe N terminus. The conformational changesundergone by the ADP/ATP carrier also af-fect peptide segments located on the matrixside. Thus, the extent of C56 alkylation byN-ethylmaleimide in the presence of ADP orATP is enhanced by BA and counteracted byCATR (51). In addition, the K42-Q43, K146-G147, and K244-G245 bonds, which are notaccessible in the native membrane of invertedmitochondrial particles, become unmaskedand accessible for cleavage by specific pro-teases in the presence of BA (52). Additionalconformation-dependent thiol-labeling andcross-linking experiments carried out on thebovine carrier with maleimide reagents wereinterpreted in terms of the participation of thematrix-exposed loops to the transport pro-cess. In particular, a critical gating role wasproposed for loop M1, which is positivelycharged and could attract a nucleotide andconvey it to a binding domain located on loopM2 (53). We have recently demonstrated byproteolytic and chemical labeling approachesthat loop M2 of the yeast ADP/ATP carrierundergoes conformation-dependent swing-ing, probably related to the transport process(53a).
732 Nury et al.
Ann
u. R
ev. B
ioch
em. 2
006.
75:7
13-7
41. D
ownl
oade
d fr
om a
rjou
rnal
s.an
nual
revi
ews.
org
by U
nive
rsity
of
Cal
ifor
nia
- Ir
vine
on
09/0
5/06
. For
per
sona
l use
onl
y.
ANRV277-BI75-28 ARI 12 May 2006 15:19
The bovine ADP/ATP carrier structuredetermined in the presence of CATR showsseveral interactions between matrix loops. Be-cause of its difference in length and asym-metrical position, M1 plays a central role(Figures 1c and 8). The first part of M1 (be-tween H1 and h1-2) interacts with the C-terminal ends of H3 and H5 as well as withthe first part of M3 (notably through a saltbridge between K42 and D247). The interac-tions between M1 and M2 involve the secondpart of M1 (between h1-2 and H2) and thefirst part of M2 (between H3 and h3-4). Simi-lar interactions, obtained by circular permuta-tion, are observed between M2 and M3. Theinteraction between M1 and M3 implicatesthe C-terminal ends of H1 and h5-6. Manypolar or electrostatic interactions among thematrix loops are observed. It is therefore dif-ficult to conceive that one single loop couldmove during transport without affecting theconformation of the other loops. Conversely,if dimerization of the carrier is necessary forits function, the dimerization interface breaksthe threefold symmetry. Combining both as-pects, it is conceivable that even if all threematrix loops move during transport, the am-plitude of these movements might be differentfor each of them.
The structure also suggests that modifi-cations of the kink angles in odd-numberedhelices would induce large conformationalchanges, reorienting the short amphipaticloops (helped by the presence of conservedglycines of the MCF motif) and changingthe accessibility of the cysteines as demon-strated by cross-linking experiments. Suchmodifications may be general for all MCFmembers.
What Triggers the Changes?Kinetic Aspects
The kinetic properties of the ADP/ATP car-rier have been studied in detail in isolated mi-tochondria and after functional reconstitutionin proteoliposomes. Most of the mitochon-drial carriers were shown to catalyze strict so-
lute exchange reactions (54, 55); this is alsothe case of the ADP/ATP carrier. In isolatedmitochondria, stoichiometry of the exchangewas assessed in experiments in which nonme-tabolized transportable analogues were usedinstead of ADP. This avoided the difficultmeasurements linked to dephosphorylationand transphosphorylation of the external nu-cleotide once it has been transported intothe matrix space. Using AOPCP, a methyleneanalogue of ADP, the stoichiometry of the ex-change with intramitochondrial nucleotides is1 (for a review, see Reference 56). A 1-to-1 stoichiometry of exchange was also deter-mined for the ADP/ATP transport system re-constituted in proteoliposomes from the iso-lated bovine heart carrier (41). In contrast, auniport function of the ADP/ATP carrier wasdeduced from investigation of the reconsti-tuted ADP/ATP transport in black lipid mem-branes (57). In this approach, measurementsof electrical currents associated with the func-tioning of the carrier upon photolysis of cagedADP/ATP were interpreted as reflecting nettransport of nucleotides.
In isolated rat heart and rat liver mi-tochondria, the ADP/ATP transport wasdemonstrated to proceed according to asequential mechanism in which both theexternal substrate and the internal substratebind to the carrier before the transloca-tion occurs (58, 59). This intermolecularmechanism implies the existence of positiveinteractions between distinct binding sitesexposed to the outer and inner face of themembrane-embedded carrier. Consistentwith these findings, the occurrence of distinctspecific nucleotide-binding sites on theADP/ATP carrier was deduced from thebinding of ADP and ATP derivatives tothe carrier either in the membrane-boundstate or when isolated in detergent solution(60–62). For example, in the mitochondrialmembrane, two specific nucleotide-bindingsites are located on the same face of themembrane, either the IMS face of intactmitochondria or the matrix face of inside-outparticles. Negative interactions between
www.annualreviews.org • The Mitochondrial ADP/ATP Carrier 733
Ann
u. R
ev. B
ioch
em. 2
006.
75:7
13-7
41. D
ownl
oade
d fr
om a
rjou
rnal
s.an
nual
revi
ews.
org
by U
nive
rsity
of
Cal
ifor
nia
- Ir
vine
on
09/0
5/06
. For
per
sona
l use
onl
y.
ANRV277-BI75-28 ARI 12 May 2006 15:19
adjacent sites were evidenced for the bindingof transported nucleotides but not for that ofnontransportable nucleotides, such asnaphthoyl-ADP, thus illustrating that thebinding step triggers conformational changesresponsible for negative cooperativity, whichis set up prior to the transport process.
The structure shows that the negativelycharged ADP attracted to the bottom of thecavity could interact with the basic residuesof the MCF motifs and thus interfere withthe salt bridges that strengthen the closedform. The binding of ADP itself could beone element that triggers the conformationalchanges, which result in nucleotide transport.An additional element could be the bindingof ATP from the matrix. Combining the hy-pothesis of a selectivity filter with the con-formational changes triggered by substratebinding would explain the selectivity of eachMCF member, despite a common mechanismbased on a similar structure induced by theMCF motifs for all the carriers of this fam-ily. The slow turnover of ADP/ATP carriers(about 1000 min−1) (63) is compatible with thelarge structural changes hypothetized fromthe kinked helices and is necessary for thetransport of nucleotides but not for ion trans-porters (64).
Is the Functional Unit a Dimer?
Experimental evidence for oligomeriza-tion. Ever since the discovery of the ADP/ATP carrier in the 1970s, several bio-chemical and biophysical experiments in-dicated that MCF members are dimeric.Mainly three types of experiments were per-formed. The first type led to a stoichiome-try, the second gave a particle mass, and thethird was based on distances between givenresidues of neighboring molecules. Klingen-berg and coworkers first proposed the func-tional unit to be a dimer (65, 66) on thebasis of inhibitor stoichiometry studies. Mea-suring CATR amounts by 35S radioactivityand protein concentration with modified Bi-uret or Lowry protocols, a 0.55 molar ratio
of CATR-to-protein was determined. Analyt-ical ultracentrifugation was carried out withthe ADP/ATP carrier (67) and the uncouplingprotein (UCP) (68). Both proteins were sol-ubilized in Triton X-100. The total CATR-ADP/ATP carrier micelle mass was ∼180 kDaand dropped to 65 kDa, which corresponds toa dimer, after subtracting the contributions ofdetergent and lipids. The mass of UCP mi-celles was also compatible with a dimeric orga-nization. However, similar experiments on theyeast ADP/ATP carrier solubilized in dodecylmaltoside (M. LeMaire, personal communi-cation) or the bovine carrier (C. Ebel, per-sonal communication) evidenced a monomer.Small-angle neutron scattering of the bovineADP/ATP carrier solubilized in LAPAO ledto a particle mass compatible with a dimerin the presence of both inhibitors CATRor BA (69). However, the bovine ADP/ATPcarrier is usually solubilized with a largeamount of detergent and lipids, and it is dif-ficult to correct for these contributions to thescattering.
Native electrophoresis has been used forseveral mitochondrial carriers [ADP/ATP(70), 2-oxoglutarate (71), citrate (72), tricar-boxylate (73)], and this led the authors to con-clude that a dimeric organization exists. Fromin vivo and in vitro assembly of ADP/ATP and2-oxoglutarate carriers, it was proposed thatnewly imported carriers assemble rapidly withthe few preexisting monomers in the innermembrane, with the vast majority of the car-rier pool being dimeric. Using double-taggedproteins, the phosphate carrier is the only onefor which functional observations were relatedto the existence of a dimer (74).
Covalent dimers of the yeast ADP/ATPcarrier, wherein the C terminus of the firstmonomer is fused to the N terminus of thesecond one, were found to be functional (75–77). The transport activity or inhibitor bind-ing of chimeric proteins is similar to the nativeones. From these experiments, the authorssuggest a proximity of H1 from one monomerwith H6 from the second one. However, thesegment bridging H6 from one monomer to
734 Nury et al.
Ann
u. R
ev. B
ioch
em. 2
006.
75:7
13-7
41. D
ownl
oade
d fr
om a
rjou
rnal
s.an
nual
revi
ews.
org
by U
nive
rsity
of
Cal
ifor
nia
- Ir
vine
on
09/0
5/06
. For
per
sona
l use
onl
y.
ANRV277-BI75-28 ARI 12 May 2006 15:19
H1 from the other could, if unstructured,span over more than 100 A. A disulfide bridgebetween C28 of two yeast phosphate carri-ers inhibits transport (78), and thus, it wasproposed that C28 is at the interface be-tween monomers. An interface involving H3and H4 also was proposed after homologymodeling of the yeast citrate carrier structure(79).
The oligomeric state and the conforma-tional changes of the ADP/ATP carrier wereextensively investigated for years using thiollabeling or cross-linking with naturally occur-ring cysteines or single-cysteine mutants. Al-most all the loops of the protein were exploredin the presence of CATR or BA. Such exper-iments on the first IMS loop showed that theconformation of the first part of the loop isdifferent with both inhibitors (80). The au-thors suggest that C1 could act as a swinginggate during ADP uptake from the IMS. Thegate would be related to a partial unwindingof the C terminus of H2. These suggestionsare in line with the proposed dimer in whichC2 loops from both monomers interact onthe IMS side, thus allowing flexibility for C1loops. In addition, the secondary structure ofH2 is bent toward its C-terminal end, possiblyindicating a certain flexibility of the helix. Onthe basis of cysteine interactions (direct or in-duced through linkers), it was shown that C56from both monomers located in loop M1 canbe cross-linked in the presence of BA (53, 81).
Altogether, the experimental results do notconverge clearly toward a single dimeric or-ganization among all MCF carriers or evenwithin a specific type of carriers. The discrep-ancies among the ADP/ATP carriers couldresult from a rather loose oligomeric organi-zation in which protein-protein interactionscould occur for the transport process but inwhich carriers are otherwise only in closevicinity. The oligomeric state could also de-pend on experimental conditions, such as na-tive membrane versus detergent micelles orprotein and detergent concentration. It couldalso be that other MCF carriers have evolveddifferently and that their dimeric interfaces
are varied. Another possibility is the existenceof higher oligomeric interactions.
Consequences for the mechanism. Onthe basis of inhibitor and substrate-bindingstudies, a single-binding-center gated pore(SBGP) mechanism has been advocated byKlingenberg (82). According to this model,a single binding site for substrates and in-hibitors, located within the core of the pro-tein, is assumed to be alternately accessible toeach side of the membrane during the differ-ent steps of transport. The entrance and theexit of the substrate would be mediated by twogates, each gate facing one side of the mem-brane. A key feature of this model is that inorder to prevent leakage, binding of the sub-strate controls the closing of the entrance gateand the opening of the exit gate on the op-posite side of the membrane. Because of thepossible dimeric organization of the carrier,it has been proposed that the SBGP mecha-nism operates with a single translocation path-way formed at the interface of two monomers(83) or formed by the merging of the 12 TMhelices (84). Both arrangements are most un-likely because the bovine ADP/ATP carrierstructure shows that a monomer presents adeep cavity, probably belonging to a translo-cation pathway. Therefore the transport func-tion of a dimeric carrier would involve twotransport pathways, which is consistent witha sequential type of mechanism because thecarrier would be loaded simultaneously withan internal and an external nucleotide.
Gating of the channels is a compulsoryfunction that can be generalized to most trans-port systems, including channels with eitherone or two gates opened simultaneously. It isconceivable that the switch of a carrier modeto a channel mode results from concertedopening of the gates, induced, for example,by site-directed mutagenesis or by chemicalmodifications of the carriers. Conversion ofcarrier function to channel-like function hasbeen illustrated for a number of mitochon-drial carriers, including the ADP/ATP carrier,as reviewed in Reference 85.
www.annualreviews.org • The Mitochondrial ADP/ATP Carrier 735
Ann
u. R
ev. B
ioch
em. 2
006.
75:7
13-7
41. D
ownl
oade
d fr
om a
rjou
rnal
s.an
nual
revi
ews.
org
by U
nive
rsity
of
Cal
ifor
nia
- Ir
vine
on
09/0
5/06
. For
per
sona
l use
onl
y.
ANRV277-BI75-28 ARI 12 May 2006 15:19
Figure 11Protein-proteininteraction mediatedby CDLs. The twomonomers seen inthe crystal packinginteract directly nextto the matrix sideand to the IMS. Theinteraction alsoinvolves cardiolipins(gray). Van der Waalssurfaces of proteinsand lipids are shownsuperposed on theribbons for theprotein and on theballs and sticks forthe lipids.
HYPOTHESES ANDCONCLUSIONS
A Proposed Mechanism InvolvingCDLs
A second crystal form of the bovine ADP/ATPcarrier revealed possible monomer-monomerinteractions favored by CDLs (Figure 11)(7). The importance of CDLs was alreadyknown for ADP/ATP carrier activity, andthree CDLs were shown to be tightly asso-ciated to one monomer (31). Although CDLsare not essential for the growth of yeast cellson fermentable and nonfermentable carbonsources (86), the activity of ADP/ATP trans-port in the absence of CDL has been shownto be lowered to 20% of the level seen inmitochondria of wild-type cells (87). Retriev-ing the activity of different mutants in pro-teoliposomes in the presence of CDLs in-dicated that these lipids are important for
the function and the stability of the carrier(32). On the matrix side, numerous interac-tions between the N-terminal end of h1-2 ofone carrier with the C-terminal end of h5–6from the second carrier involve K267 and K51and also the phosphates of CDL801. On theIMS, the interaction is looser and implicatesloops C2 from both monomers. Within themembrane, hydrophobic contacts are madethrough the lipid alkyl chains. Prolines of theMCF motif, P27, P132, and P229, have aro-matic environments similar to W70, Y173,and F270, each of which stacks on one ofthe three CDLs. We postulate that the crosstalk between monomers during the transportprocess could implicate the CDLs and thataromatic residues relay the signal from theproline kink modification to the lipid and actas a driving belt. Indeed, P229 (located in thethird MCF repeat) interacts with CDL801(Figure 7) close to the putative dimer in-terface (Figure 11). It is also interesting tonote that residues Y228 and F230, which areclose to P229, prolongate the tyrosine lad-der present in the cavity. In this model, H2is at the opposite side of the dimer interface.This location is consistent with the hypoth-esis that this helix twists during the trans-port. In addition, several residues belongingto the basic patches of the cavity belong toH2. From the entrance, the first patch com-prises R104, belonging to C1, the second K91and K95, and the third R79, known to be cru-cial for the binding and the transport of nu-cleotides. Therefore, binding of ADP to thecavity from the IMS could induce this twist(50).
Open Questions
Elucidating the transport mechanism of mi-tochondrial ADP/ATP carriers necessitatesa combination of various approaches. High-resolution structural approaches such as X-raycrystallography are very powerful and high-light the quasi-atomic structure of the car-rier locked into a precise conformational state.This structure provides a good starting point
736 Nury et al.
Ann
u. R
ev. B
ioch
em. 2
006.
75:7
13-7
41. D
ownl
oade
d fr
om a
rjou
rnal
s.an
nual
revi
ews.
org
by U
nive
rsity
of
Cal
ifor
nia
- Ir
vine
on
09/0
5/06
. For
per
sona
l use
onl
y.
ANRV277-BI75-28 ARI 12 May 2006 15:19
for exploring various hypotheses by guidingthe choice of mutants and following theirtransport properties. Table 1 summarizes thelarge body of mutations that have alreadybeen addressed. Most of the residues shownto have important functions are highlighted inthe structural discussion. Nevertheless, manyothers still must be studied. However, mem-brane carriers are more difficult to handle thansoluble enzymes for which enzymatic assaysof modified proteins can be set up more eas-ily. In addition, the oligomeric state of mem-brane proteins in their native lipidic environ-ment is still difficult to approach because all
the experiments in which the carrier is ex-tracted from the membrane (solubilized in de-tergent or even reconstituted in liposomes)are potentially biased and not entirely satisfac-tory. Electron or atomic-force microscopiesmight offer alternative approaches to followthe oligomeric state under different condi-tions and environments.
Finally, the transport of charged nu-cleotides, which is triggered by in vivo con-ditions, nucleotide concentration, membranepotential, and possibly interactions with otherproteins, needs to be explored along with theirstructure.
SUMMARY POINTS
1. The ADP/ATP carrier structure highlights a bundle of six tilted—with half of themkinked—helices forming a cavity that is wide open toward the IMS.
2. MCF members may share a common transport mechanism, which is based on a com-mon scaffold and could rely on the kink and tilt modifications of the TM helices.
3. Substrate specificity may be related to the geometry and the chemical properties ofthe residues in the cavity, illustrated for instance by the distribution of patches of basicresidues as well as by a ladder of aromatic residues.
4. The functional properties of many mutants were explored and are compiled inTable 1.
5 The sequential transport mechanism might be induced by the simultaneous bindingof ADP and ATP on both sides of the membrane.
6 Many published results, such as cross-linking experiments, protein/inhibitor stoi-chiometries, chimeric dimers, analytical ultracentrifugation or neutron scattering,indicate that the ADP/ATP carrier is a dimer.
FUTURE ISSUES TO BE RESOLVED
1. The oligomeric state of the ADP/ATP carrier must be ascertained along with a de-scription of the mechanism for cross talk between monomers.
2. Structural data on other conformations of the carrier should be obtained.
3. Descriptions are needed of the transport process at a molecular level, the commonfeatures for the whole MCF, and the differences that built specificity.
4. The putative partners of the ADP/ATP carrier, or supercomplexes to which it couldbelong, should be characterized.
www.annualreviews.org • The Mitochondrial ADP/ATP Carrier 737
Ann
u. R
ev. B
ioch
em. 2
006.
75:7
13-7
41. D
ownl
oade
d fr
om a
rjou
rnal
s.an
nual
revi
ews.
org
by U
nive
rsity
of
Cal
ifor
nia
- Ir
vine
on
09/0
5/06
. For
per
sona
l use
onl
y.
ANRV277-BI75-28 ARI 12 May 2006 15:19
LITERATURE CITED
1. Walker JE, Runswick MJ. 1993. J. Bioenerg. Biomembr. 25:435–462. White SH. 2004. Protein Sci. 13:1948–493. Abramson J, Smirnova I, Kasho V, Verner G, Kaback HR, Iwata S. 2003. Science 301:610–
154. Pebay-Peyroula E, Dahout-Gonzalez C, Kahn R, Trezeguet V, Lauquin GJM, Brandolin
G. 2003. Nature 426:39–445. Kunji ER, Harding M. 2003. J. Biol. Chem. 278:36985–886. Kunji ER. 2004. FEBS Lett. 564:239–447. Nury H, Dahout-Gonzalez C, Trezeguet V, Lauquin GJM, Brandolin G, Pebay-Peyroula
E. 2005. FEBS Lett. 579:6031–368. Siekevitz P, Potter VR. 1955. J. Biol. Chem. 215:237–559. Pressman BC. 1958. J. Biol. Chem. 232:967–78
10. Bartley W, Davies RE. 1954. Biochem. J. 57:37–4911. Bruni A, Contessa AR, Luciani S, 1962. Biochim. Biophys. Acta 60:301–1112. Vignais PV, Vignais PM, Stanislas E. 1962. Biochim. Biophys. Acta 60:284–30013. Kemp A Jr, Slater EC. 1964. Biochim. Biophys. Acta 92:178–8014. Chappell JB, Crofts AR. 1965. Biochem. J. 95:707–1615. Pfaff E, Klingenberg M, Heldt HW. 1965. Biochim. Biophys. Acta 104:312–1516. Duee ED, Vignais PV. 1965. Biochim. Biophys. Acta 107:184–8817. Heldt HW, Jacobs H, Klingenberg M. 1965. Biochem. Biophys. Res. Commun. 18:174–
7918. Welling W, Cohen JA, Berends W. 1960. Biochem. Pharmacol. 3:122–3519. Henderson PJ, Lardy HA. 1970. J. Biol. Chem. 245:1319–2620. Vozza A, Blanco E, Palmieri L, Palmieri F. 2004. J. Biol. Chem. 279:20850–5721. Dolce V, Fiermonte G, Runswick MJ, Palmieri F, Walker JE. 2001. Proc. Natl. Acad. Sci.
USA 98:2284–8822. Fiermonte G, De Leonardis F, Todisco S, Palmieri L, Lasorsa FM, Palmieri F. 2004.
J. Biol. Chem. 279:30722–3023. Palmieri L, Rottensteiner H, Girzalsky W, Scarcia P, Palmieri F, Erdmann R. 2001.
EMBO J. 20:5049–5924. Dahout-Gonzalez C, Brandolin G, Pebay-Peyroula E. 2003. Acta Crystallogr. D 59:2353–
5525. Vignais PV. 1976. Biochim. Biophys. Acta 456:1–3826. Jezek P, Jezek J. 2003. FEBS Lett. 534:15–2527. Williams KA, Deber CM. 1991. Biochemistry 30:8919–2328. Pebay-Peyroula E, Brandolin G. 2004. Curr. Opin. Struct. Biol. 14:420–2529. Yohannan S, Faham S, Yang D, Whitelegge JP, Bowie JU. 2004. Proc. Natl. Acad. Sci.
USA 101:959–6330. Nelson DR, Felix CM, Swanson JM. 1998. J. Mol. Biol. 277:285–30831. Beyer K, Klingenberg M. 1985. Biochemistry 24:3821–2632. Hoffmann B, Stockl A, Schlame M, Beyer K, Klingenberg M. 1994. J. Biol. Chem.
269:1940–4433. Klingenberg M, Nelson DR. 1994. Biochim. Biophys. Acta 1187:241–4434. Pebay-Peyroula E, Garavito RM, Rosenbusch JP, Zulauf M, Timmins PA. 1995. Structure
3:1051–5935. Curran AR, Engelman DM. 2003. Curr. Opin. Struct. Biol. 13:412–17
738 Nury et al.
Ann
u. R
ev. B
ioch
em. 2
006.
75:7
13-7
41. D
ownl
oade
d fr
om a
rjou
rnal
s.an
nual
revi
ews.
org
by U
nive
rsity
of
Cal
ifor
nia
- Ir
vine
on
09/0
5/06
. For
per
sona
l use
onl
y.
ANRV277-BI75-28 ARI 12 May 2006 15:19
36. Vignais PV, Block MR, Boulay F, Brandolin G, Lauquin GJM. 1985. In Structure andProperties of Cell Membranes, ed. G Benga, 2:139–79. Boca Raton, FL: CRC Press
37. Lauquin GJ, Brandolin G, Lunardi J, Vignais PV. 1978. Biochim. Biophys. Acta 501:10–1938. Klingenberg M, Mayer I, Dahms AS. 1984. Biochemistry 23:2442–4939. Majima E, Yamaguchi N, Chuman H, Shinohara Y, Ishida M, et al. 1998. Biochemistry
37:424–3240. Brandolin G, Dupont Y, Vignais PV. 1981. Biochem. Biophys. Res. Commun. 98:28–3541. Brandolin G, Doussiere J, Gulik A, Gulik-Krzywicki T, Lauquin GJM, Vignais PV. 1980.
Biochim. Biophys. Acta 592:592–61442. Dalbon P, Brandolin G, Boulay F, Hoppe J, Vignais PV. 1988. Biochemistry 27:5141–4943. Mayinger P, Winkler E, Klingenberg M. 1989. FEBS Lett. 244:421–2644. Dianoux AC, Noel F, Fiore C, Trezeguet V, Kieffer S, et al. 2000. Biochemistry 39:11477–
8745. Wohlrab H. 2004. Biochem. Biophys. Res. Commun. 320:685–8846. Klingenberg M. 2005. Biochemistry 44:8563–7047. Brandolin G, Le Saux A, Trezeguet V, Lauquin GJM, Vignais PV. 1993. J. Bioenerg.
Biomembr. 25:459–7248. Brdiczka D, Schumacher D. 1976. Biochem. Biophys. Res. Commun. 73:823–3249. Brandolin G, Boulay F, Dalbon P, Vignais PV. 1989. Biochemistry 28:1093–10050. Kihira Y, Iwahashi A, Majima E, Terada H, Shinohara Y. 2004. Biochemistry 43:15204–951. Boulay F, Vignais PV. 1984. Biochemistry 23:4807–1252. Marty I, Brandolin G, Gagnon J, Brasseur R, Vignais PV. 1992. Biochemistry 31:4058–6553. Hashimoto M, Majima E, Goto S, Shinohara Y, Terada H. 1999. Biochemistry 38:1050–56
53a. Dahout-Gonzalez C, Ramus C, Dassa EP, Dianoux AC, Brandolin G. 2005. Biochemistry44:16310–20
54. Kaplan RS. 2001. J. Membr. Biol. 179:165–8355. Palmieri F. 2004. Pflugers Arch. 447:689–70956. Vignais PV, Lunardi J. 1985. Annu. Rev. Biochem. 54:977–101457. Gropp T, Brustovetsky N, Klingenberg M, Muller V, Fendler K, Bamberg E. 1999.
Biophys. J. 77:714–2658. Duyckaerts C, Sluse-Goffart CM, Fux JP, Sluse FE, Liebecq C. 1980. Eur. J. Biochem.
106:1–659. Barbour RL, Chan SH. 1981. J. Biol. Chem. 256:1940–4860. Brandolin G, Dupont Y, Vignais PV. 1982. Biochemistry 21:6348–5361. Dupont Y, Brandolin G, Vignais PV. 1982. Biochemistry 21:6343–4762. Block MR, Vignais PV. 1984. Biochim. Biophys. Acta 767:369–7663. Klingenberg M. 1985. In Enzymes of Biological Membranes, ed. AM Martonosi, 4:511–53.
New-York: Plenum64. Hunte C, Screpanti E, Venturi M, Rimon A, Padan E, Michel H. 2005. Nature 435:1197–
20265. Riccio P, Aquila H, Klingenberg M. 1975. FEBS Lett. 56:133–3866. Klingenberg M, Riccio P, Aquila H. 1978. Biochim. Biophys. Acta 503:193–21067. Hackenberg H, Klingenberg M. 1980. Biochemistry 19:548–5568. Lin CS, Hackenberg H, Klingenberg EM. 1980. FEBS Lett. 113:304–669. Block MR, Zaccai G, Lauquin GJ, Vignais PV. 1982. Biochem. Biophys. Res. Commun.
109:471–7770. Dyall SD, Agius SC, De Marcos Lousa C, Trezeguet V, Tokatlidis K. 2003. J. Biol. Chem.
278:26757–64
www.annualreviews.org • The Mitochondrial ADP/ATP Carrier 739
Ann
u. R
ev. B
ioch
em. 2
006.
75:7
13-7
41. D
ownl
oade
d fr
om a
rjou
rnal
s.an
nual
revi
ews.
org
by U
nive
rsity
of
Cal
ifor
nia
- Ir
vine
on
09/0
5/06
. For
per
sona
l use
onl
y.
ANRV277-BI75-28 ARI 12 May 2006 15:19
71. Palmisano A, Zara V, Honlinger A, Vozza A, Dekker PJ, et al. 1998. Biochem. J. 333(Part1):151–58
72. Kotaria R, Mayor JA, Walters DE, Kaplan RS. 1999. J. Bioenerg. Biomembr. 31:543–4973. Capobianco L, Impagnatiello T, Ferramosca A, Zara V. 2004. J. Biochem. Mol. Biol.
37:515–2174. Schroers A, Burkovski A, Wohlrab H, Kramer R. 1998. J. Biol. Chem. 273:14269–7675. Hatanaka T, Hashimoto M, Majima E, Shinohara Y, Terada H. 1999. Biochem. Biophys.
Res. Commun. 262:726–3076. Trezeguet V, Le Saux A, David C, Gourdet C, Fiore C, et al. 2000. Biochim. Biophys. Acta
1457:81–9377. Huang SG, Odoy S, Klingenberg M. 2001. Arch. Biochem. Biophys. 394:67–7578. Phelps A, Wohlrab H. 2004. Biochemistry 43:6200–779. Ma C, Kotaria R, Mayor JA, Remani S, Walters DE, Kaplan RS. 2005. J. Biol. Chem.
280:2331–4080. Kihira Y, Majima E, Shinohara Y, Terada H. 2005. Biochemistry 44:184–9281. Majima E, Ikawa K, Takeda M, Hashimoto M, Shinohara Y, Terada H. 1995. J. Biol.
Chem. 270:29548–5482. Klingenberg M. 1991. In A Study of Enzymes. Vol II : Mechanism of Enzyme Action, ed. SA
Kuby, pp. 367–90. Boca Raton, FL: CRC Press83. Klingenberg M. 1992. Biochem. Soc. Trans. 20:547–5084. Terada H, Majima E. 1997. Prog. Colloid Polym. Sci. 106:192–9785. Ledesma A, de Lacoba MG, Arechaga I, Rial E. 2002. J. Bioenerg. Biomembr. 34:473–8686. Jiang F, Rizavi HS, Greenberg ML. 1997. Mol. Microbiol. 26:481–9187. Jiang F, Ryan MT, Schlame M, Zhao M, Gu Z, et al. 2000. J. Biol. Chem. 275:22387–9488. De Marcos Lousa C, Trezeguet V, Dianoux AC, Brandolin G, Lauquin GJM. 2002.
Biochemistry 41:14412–2089. Nelson DR, Douglas MG. 1993. J. Mol. Biol. 230:1171–8290. Nelson DR. 1996. Biochim. Biophys. Acta 1275:133–3791. Zeman I, Schwimmer C, Postis V, Brandolin G, David C, et al. 2003. J. Bioenerg. Biomembr.
35:243–5692. Muller V, Heidkamper D, Nelson DR, Klingenberg M. 1997. Biochemistry 36:16008–1893. Heimpel S, Basset G, Odoy S, Klingenberg M. 2001. J. Biol. Chem. 276:11499–50694. Bogner W, Aquila H, Klingenberg M. 1986. Eur. J. Biochem. 161:611–2095. Hatanaka T, Kihira Y, Shinohara Y, Majima E, Terada H. 2001. Biochem. Biophys. Res.
Commun. 286:936–4296. Majima E, Koike H, Hong YM, Shinohara Y, Terada H. 1993. J. Biol. Chem. 268:22181–
8797. Le Saux A, Roux P, Trezeguet V, Fiore C, Schwimmer C, et al. 1996. Biochemistry
35:16116–2498. Roux P, Le Saux A, Trezeguet V, Fiore C, Schwimmer C, et al. 1996. Biochemistry
35:16125–3199. Nelson DR, Lawson JE, Klingenberg M, Douglas MG. 1993. J. Mol. Biol. 230:1159–70
100. Heidkamper D, Muller V, Nelson DR, Klingenberg M. 1996. Biochemistry 35:16144–52101. Muller V, Basset G, Nelson DR, Klingenberg M. 1996. Biochemistry 35:16132–43102. Napoli L, Bordoni A, Zeviani M, Hadjigeorgiou GM, Sciacco M, et al. 2001. Neurology
57:2295–98103. Komaki H, Fukazawa T, Houzen H, Yoshida K, Nonaka I, Goto Y. 2002. Ann. Neurol.
51:645–48
740 Nury et al.
Ann
u. R
ev. B
ioch
em. 2
006.
75:7
13-7
41. D
ownl
oade
d fr
om a
rjou
rnal
s.an
nual
revi
ews.
org
by U
nive
rsity
of
Cal
ifor
nia
- Ir
vine
on
09/0
5/06
. For
per
sona
l use
onl
y.
ANRV277-BI75-28 ARI 12 May 2006 15:19
104. Kaukonen J, Juselius JK, Tiranti V, Kyttala A, Zeviani M, et al. 2000. Science 289:782–85105. Oettmeier W, Masson K, Kalinna S. 1995. Eur. J. Biochem. 227:730–33106. De Marcos Lousa C, Trezeguet V, David C, Postis V, Arnou B, et al. 2005. Biochemistry
44:4342–48107. Kovac L, Lachowicz TM, Slonimski PP. 1967. Science 158:1564–67108. Kolarov J, Kolarova N, Nelson N. 1990. J. Biol. Chem. 265:12711–16
RELATED REVIEWS
Neupert W. 1997. Annu. Rev. Biochem. 66:863–917Popot J-L, Engleman DM. 2000. Annu. Rev. Biochem. 69:881–922
www.annualreviews.org • The Mitochondrial ADP/ATP Carrier 741
Ann
u. R
ev. B
ioch
em. 2
006.
75:7
13-7
41. D
ownl
oade
d fr
om a
rjou
rnal
s.an
nual
revi
ews.
org
by U
nive
rsity
of
Cal
ifor
nia
- Ir
vine
on
09/0
5/06
. For
per
sona
l use
onl
y.
Contents ARI 22 April 2006 9:16
Annual Reviewof Biochemistry
Volume 75, 2006Contents
Wanderings of a DNA Enzymologist: From DNA Polymerase to ViralLatencyI. Robert Lehman � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 1
Signaling Pathways in Skeletal Muscle RemodelingRhonda Bassel-Duby and Eric N. Olson � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �19
Biosynthesis and Assembly of Capsular Polysaccharides inEscherichia coliChris Whitfield � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �39
Energy Converting NADH:Quinone Oxidoreductase (Complex I)Ulrich Brandt � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �69
Tyrphostins and Other Tyrosine Kinase InhibitorsAlexander Levitzki and Eyal Mishani � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �93
Break-Induced Replication and Recombinational Telomere Elongationin YeastMichael J. McEachern and James E. Haber � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 111
LKB1-Dependent Signaling PathwaysDario R. Alessi, Kei Sakamoto, and Jose R. Bayascas � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 137
Energy Transduction: Proton Transfer Through the RespiratoryComplexesJonathan P. Hosler, Shelagh Ferguson-Miller, and Denise A. Mills � � � � � � � � � � � � � � � � � � � � � � 165
The Death-Associated Protein Kinases: Structure, Function, andBeyondShani Bialik and Adi Kimchi � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 189
Mechanisms for Chromosome and Plasmid SegregationSantanu Kumar Ghosh, Sujata Hajra, Andrew Paek,
and Makkuni Jayaram � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 211
Chromatin Modifications by Methylation and Ubiquitination:Implications in the Regulation of Gene ExpressionAli Shilatifard � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 243
v
Ann
u. R
ev. B
ioch
em. 2
006.
75:7
13-7
41. D
ownl
oade
d fr
om a
rjou
rnal
s.an
nual
revi
ews.
org
by U
nive
rsity
of
Cal
ifor
nia
- Ir
vine
on
09/0
5/06
. For
per
sona
l use
onl
y.
Contents ARI 22 April 2006 9:16
Structure and Mechanism of the Hsp90 Molecular ChaperoneMachineryLaurence H. Pearl and Chrisostomos Prodromou � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 271
Biochemistry of Mammalian Peroxisomes RevisitedRonald J.A. Wanders and Hans R. Waterham � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 295
Protein Misfolding, Functional Amyloid, and Human DiseaseFabrizio Chiti and Christopher M. Dobson � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 333
Obesity-Related Derangements in Metabolic RegulationDeborah M. Muoio and Christopher B. Newgard � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 367
Cold-Adapted EnzymesKhawar Sohail Siddiqui and Ricardo Cavicchioli � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 403
The Biochemistry of SirtuinsAnthony A. Sauve, Cynthia Wolberger, Vern L. Schramm, and Jef D. Boeke � � � � � � � � � � � 435
Dynamic Filaments of the Bacterial CytoskeletonKatharine A. Michie and Jan Lowe � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 467
The Structure and Function of Telomerase Reverse TranscriptaseChantal Autexier and Neal F. Lue � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 493
Relating Protein Motion to CatalysisSharon Hammes-Schiffer and Stephen J. Benkovic � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 519
Animal Cytokinesis: From Parts List to MechanismsUlrike S. Eggert, Timothy J. Mitchison, and Christine M. Field � � � � � � � � � � � � � � � � � � � � � � � � 543
Mechanisms of Site-Specific RecombinationNigel D.F. Grindley, Katrine L. Whiteson, and Phoebe A. Rice � � � � � � � � � � � � � � � � � � � � � � � � � � 567
Axonal Transport and Alzheimer’s DiseaseGorazd B. Stokin and Lawrence S.B. Goldstein � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 607
Asparagine Synthetase ChemotherapyNigel G.J. Richards and Michael S. Kilberg � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 629
Domains, Motifs, and Scaffolds: The Role of Modular Interactions inthe Evolution and Wiring of Cell Signaling CircuitsRoby P. Bhattacharyya, Attila Remenyi, Brian J. Yeh, and Wendell A. Lim � � � � � � � � � � � � � 655
Ribonucleotide ReductasesPar Nordlund and Peter Reichard � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 681
Introduction to the Membrane Protein Reviews: The Interplay ofStructure, Dynamics, and Environment in Membrane ProteinFunctionJonathan N. Sachs and Donald M. Engelman � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 707
vi Contents
Ann
u. R
ev. B
ioch
em. 2
006.
75:7
13-7
41. D
ownl
oade
d fr
om a
rjou
rnal
s.an
nual
revi
ews.
org
by U
nive
rsity
of
Cal
ifor
nia
- Ir
vine
on
09/0
5/06
. For
per
sona
l use
onl
y.
Contents ARI 22 April 2006 9:16
Relations Between Structure and Function of the MitochondrialADP/ATP CarrierH. Nury, C. Dahout-Gonzalez, V. Trezeguet, G.J.M. Lauquin,G. Brandolin, and E. Pebay-Peyroula � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 713
G Protein–Coupled Receptor RhodopsinKrzysztof Palczewski � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 743
Transmembrane Traffic in the Cytochrome b6 f ComplexWilliam A. Cramer, Huamin Zhang, Jiusheng Yan, Genji Kurisu,
and Janet L. Smith � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 769
INDEXES
Subject Index � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 791
Author Index � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 825
ERRATA
An online log of corrections to Annual Review of Biochemistry chapters (if any, 1977 tothe present) may be found at http://biochem.annualreviews.org/errata.shtml
Contents vii
Ann
u. R
ev. B
ioch
em. 2
006.
75:7
13-7
41. D
ownl
oade
d fr
om a
rjou
rnal
s.an
nual
revi
ews.
org
by U
nive
rsity
of
Cal
ifor
nia
- Ir
vine
on
09/0
5/06
. For
per
sona
l use
onl
y.