A plant thylakoid ATP/ADP carrier

1

IDENTIFICATION, EXPRESSION AND FUNCTIONAL ANALYSES OF A THYLAKOIDATP/ADP CARRIER FROM ARABIDOPSIS*

Sophie Thuswaldner†, Jens O. Lagerstedt ‡,§,|| , Marc Rojas-Stütz ¶,|| , Karim Bouhidel**, ChristopheDer**, Nathalie Leborgne-Castel**, Arti Mishra#, Francis Marty**, Benoît Schoefs**, Iwona Adamska¶,

Bengt L. Persson‡, and Cornelia Spetea†

From the †Division of Cell Biology and #Department of Physics, Chemistry and Biology, LinköpingUniversity, SE-581 85 Linköping, Sweden, ‡Department of Chemistry and Biomedical Sciences, KalmarUniversity, SE-391 82 Kalmar, Sweden, ¶Department of Physiology and Plant Biochemistry, University

of Konstanz, D-78457 Konstanz, Germany, and **Dynamique Vacuolaire et Réponses aux Stress del’Environnement, UMR Plante-Microbe-Environnement CNRS 5184/ INRA 1088/UB, Université de

Bourgogne, BP 47870, F-21078 Dijon Cedex, FranceRunning title: A plant thylakoid ATP/ADP carrier

Address correspondence to: Cornelia Spetea, Division of Cell Biology, Linköping University, SE-581 85Linköping, Sweden, Phone: 46 13 225788; Fax 46 13 224314; E-mail cornelia.spetea@ibk.liu.se

In plants, the chloroplast thylakoid membraneis the site of light-dependent photosyntheticreactions coupled to ATP synthesis. The ability ofthe plant cell to build and alter this membranesystem is essential for efficient photosynthesis. Anucleotide translocator homologous to the bovinemitochondrial ADP/ATP carrier (AAC) waspreviously found in spinach thylakoids. Here wehave identified and functionally characterized athylakoid ATP/ADP carrier (TAAC) fromArabidopsis. (i) Sequence homology with thebovine AAC and the prediction of chloroplasttransit peptides indicated a putative carrierencoded by the At5g01500 gene, as a TAAC. (ii)Transiently expressed TAAC-green fluorescentprotein fusion construct is targeted to thechloroplast. Western blotting using a peptide-specific antibody together with immunogoldelectron microscopy revealed a major location ofTAAC in the thylakoid membrane. Previousproteomic analyses identified this protein inchloroplast envelope preparations. ( i i i )Recombinant TAAC protein specifically transportsadenine nucleotides across the cytoplasmicmembrane of Escherichia coli. The transportoccurs in a counter exchange mode, namely importof ATP and export of ADP. Studies on isolatedthylakoids from Arabidopsis confirmed theseobservations. (iv) The lack of TAAC in anArabidopsis T-DNA insertion mutant caused 30-40% reduction in the thylakoid ATP transport andmetabolism. (v) TAAC is readily expressed indark-grown Arabidopsis seedlings, and its levelremains stable throughout the greening process. Its

expression is highest in developing green tissuesand in leaves undergoing senescence or abioticstress. We propose that the TAAC protein suppliesATP for energy-dependent reactions duringthylakoid biogenesis and turnover in plants.

Chloroplasts perform oxygenic photosynthesisin algae and plants, and have evolved byendosymbiosis from cyanobacteria. They have twodistinct membrane systems, the double envelopesurrounding the organelle, and an internalmembrane system named thylakoids. Theenvelope membrane represents the interfacebetween the cytoplasm and chloroplast stroma,whereas the thylakoid membrane separates thestroma and the lumenal space. Altogether approx.800 membrane proteins have been identified byproteomics in the envelope and thylakoidmembranes of Arabidopsis thaliana (for reviews,see ref. 1,2). As expected, the main function forthe identified envelope proteins was transport ofions and metabolites, whereas photosynthesis wasattributed to most of the identified thylakoidproteins. The major protein complexes inthylakoids are photosystems (PS)1 I and II, thecytochrome b6f complex and the proton-translocating ATP synthase. These photosyntheticcomplexes contain not only proteins but alsopigments and other cofactors. Their assembly,activity and removal require a large number ofauxiliary, regulatory and transport proteins (3,4).Many biochemical reports pointed to the existenceof transport activities in the thylakoid membrane,such as calcium transport (5), copper transport (6),

http://www.jbc.org/cgi/doi/10.1074/jbc.M609130200The latest version is at JBC Papers in Press. Published on January 29, 2007 as Manuscript M609130200

Copyright 2007 by The American Society for Biochemistry and Molecular Biology, Inc.

by guest on July 12, 2018http://w

ww

.jbc.org/D

ownloaded from

A plant thylakoid ATP/ADP carrier

2

anion channels (7), cation channels (8,9), andnucleotide transport (10). Only the thylakoidcopper transporter was identified at the geneticlevel in Arabidopsis (11). No hydrophobicproteins related to the above-mentioned transportactivities were identified in the previous proteomicworks on Arabidopsis thylakoid membranes(reviewed in ref. 2). Therefore, genetic strategiesare required for identification and elucidation oftheir role in optimal function of the thylakoid.

ATP is produced during the light-dependentphotosynthetic reactions on the stromal side of thethylakoid membrane. Besides its utilization duringCO2-fixation in the stroma, ATP drives manyenergy-dependent processes in thylakoids,including protein phosphorylation, folding, importand degradation. Alternatively, ATP is transportedacross the membrane into the lumenal space, andconverted to GTP by a nucleoside diphosphatekinase, NDPK3 (10). There are also several recentindications for the presence of phosphorylatedproteins in the thylakoid lumen of green algae andplants, as provided by mass spectrometry analyses(12,13). The transport of ATP across the thylakoidmembrane proceeds via a nucleotide-bindingprotein of 36.5 kDa (10). This protein ishomologous to the bovine mitochondrialADP/ATP carrier (AAC), which is the moststudied member of the mitochondrial carrierfamily (MCF, Prosite PS50920). The structure ofthe bovine AAC was solved by X-raycrystallography at 2.2 Å resolution in the presenceof carboxyatractyloside (14). In Arabidopsis, thereare three genes encoding for mitochondrial AACs(15). All three proteins were identified in themitochondrial proteome (16), and were shown totransport adenine nucleotides when expressed inheterologous system (17). The protein(s)responsible for the thylakoid ATP transportactivity (10) have not yet been isolated oridentified at the genetic level.

In this work, we provide computer-based andexperimental evidence that the product of theArabidopsis At5g01500 gene is a thylakoidATP/ADP carrier (TAAC). In a previousproteomic study, this protein was concluded to belocalized in the chloroplast (inner) envelope (18).Here we demonstrate a major thylakoid and aminor envelope location for the TAAC protein.

Our studies also show that TAAC is readilyexpressed in Arabidopsis dark-grown seedlings,and appears to be needed for both synthesis ofphotosynthetic complexes during greening and fortheir recycling during senescence and stress.

EXPERIMENTAL PROCEDURES

Plant Material–Arabidopsis (Arabidopsisthaliana cv. Columbia) plants were grownhydroponically (19) or on soil (for expressionstudies) at 120 mmol photons m–2 s–1 and 22°Cwith 8-h-light/16-h-dark cycles. For localizationu s i n g immunogold, Arabidopsis cv.Wassilewskaja was grown on soil. For transientexpression studies, tobacco (Nicotiana tabacumcv. Xanthi) plants were grown on MurashigeSkoog (MS) medium at 100 mmol photons m–2 s–1

and 24°C under a 16-h-light/8-h-dark photoperiod.Arabidopsis (cv. Wassilewskaja) seeds of the

T-DNA insertion line FLAG 443D03 (taacmutant, Fig. S1A) were obtained from PublicLinesa t I N R A(http://dbsgap.versailles.inra.fr/publiclines/). Forscreening of homozygous mutant plants, genomicDNA isolated from wild type (WT) and the taacmutant was analyzed by two sets of PCRreactions, using gene specific forward (F) 5'-AGAACAACGTGTCGTCGAATC -3' andr e v e r s e ( R ) 5 ' -CAACACCATTCTTTCCAAAAGG-3' primers,and the T-DNA left border (LB) 5'-TGGTTCACGTAGTGGGCCATCG-3' primer.(Fig. S1B). Homozygoty was confirmed in thenext generation.

Structural Analyses–Homology and orthologysearches using BLAST were performed in MIPS(h t tp : / /mips .gs f .de /pro j / tha l /db) , TIGR(http://www.tigr.org/tdb/tgi/ego), and Cyanobase(http://www.kazusa.or.jp/cyanobase). Prediction ofintracellular location and membrane topologyw e r e p e r f o r m e d u s i n g T a r g e t P(http://www.cbs.dtu.dk/services) , the PPDBdatabase (http://www.ppdb.tc.cornell.edu), and apackage of programs at ARAMEMNON website(http://aramemnon.botanik.uni-koeln.de) . T h eamino acid sequences were aligned with ClustalW(http://www.ebi.ac.uk/clustalw/index.html).

A structural homology model of the

by guest on July 12, 2018http://w

ww

.jbc.org/D

ownloaded from

A plant thylakoid ATP/ADP carrier

3

Arabidopsis TAAC protein was built from itsamino acid sequence by use of several localstructure and fold recognition methods at theMetaServer (http://bioinfo.pl/meta/) (20). Theestablished structure of the protein with thehighest scores (bovine AAC; PDB #1okc A) wasused as template, and the comparison was doneusing the Swiss-Model program (21). Thecalculated 3D-model was obtained by optimallysatisfying spatial restraints derived from the 3D-Jury sequence alignment. Analysis of the TAACstructural model was performed by use of theD e e p V i e w / S w i s s - P d b V i e w e r p r o g r a m(http://www.expasy.org/spdbv), and also by use ofInsight II software (version 2005) on the Octaneworkstation by Silicon Graphics. The structuralmodel shown (Fig. 1D) was produced by use ofthe latter.

Transient Expression of a TAAC-GFP FusionConstruct and Microscope Analysis–The entirecoding region of the TAAC gene was amplified byPCR using an Arabidopsis CD4-14 cDNA library(22), and subcloned into the expression vectorpDR303. In the resulting pNL1500 plasmid, theTAAC sequence was fused in-frame to the 5'-endof a green fluorescent protein (GFP) -codingsequence, and placed under the control of thetranscriptional 35S promoter from the cauliflowermosaic virus (CaMV35S). This plasmid wasintroduced into tobacco leaf protoplasts byelectroporation. After 24 h of incubation indarkness, protoplasts were imaged using aconfocal laser scanning microscope. The detailedprotocols for construct design, protoplast isolation,electroporation, and microscope analysis aredescribed in Supplemental Experimentalprocedures.

C h l o r o p l a s t M e m b r a n ePreparations–Arabidopsis chloroplasts wereprepared (23), and lysed in 50 mM Tricine/KOH(pH 7.6) and 5 mM MgCl2 (Tricine buffer) for 10min on ice. Thylakoids and envelope membraneswere purified from the lysate by centrifugation at113,000 ¥ g for 1 h on a sucrose step gradient (0.6M, 0.93 M, 1.2 M and 1.5 M sucrose). Thylakoidswere collected at the 1.2/1.5 M interface, dilutedwith Tricine buffer and centrifuged at 10,000 ¥ gfor 10 min. Envelope membranes were collected atthe 0.93/1.2 M interface, diluted with Tricine

buffer and centrifuged at 10,000 ¥ g for 10 min (todiscard thylakoid vesicles) followed by 150,000 ¥g for 30 min. The thylakoids and envelopemembranes were finally resuspended in Tricinebuffer supplemented with 0.3 M sucrose.Chlorophyll (Chl) and protein concentrations weredetermined as described (24,25). For topologystudies, thylakoid membranes (0.2 mg Chl mL-1)were treated with 1 M NaCl, 0.1% or 1% (w/v)Triton X-100 for 30 min on ice in darknessfollowed by centrifugation at 150,000 ¥ g for 30min. Both soluble and membrane fractions wereanalyzed by Western blotting.

Protein Analysis–SDS/PAGE and Westernblotting were performed as described (26). Ananti-TAAC antibody was produced in rabbitagainst a peptide corresponding to the last 15residues at C-terminus of the Arabidopsis protein,and purified by affinity chromatography(Innovagen, Lund, Sweden). Where indicated,antibodies against the 110 kDa protein of thetranslocon of the chloroplast inner envelopemembrane (TIC110, gift from Prof. J. Soll), alight-harvesting Chl a/b-binding protein of PSII(Lhcb2, Agrisera, Umeå, Sweden) and an anti-Xpress antibody (Invitrogen, Carlsbad, CA) werealso used.

Immunogold Electron Microscopy–Arabidopsis rosette leaves were thinly diced andfixed overnight. After blocking in 1% (w/v)bovine serum albumin (BSA), the grids wereincubated with the anti-TAAC antibody followedby incubation with goat anti-rabbit serumconjugated with colloidal gold particles, stainingand electron microscopy, as earlier described (27).Detailed protocols are given in SupplementalExperimental procedures.

Cloning, Heterologous Expression andPurification of a Recombinant TAAC Protein–AllDNA manipulations, including PCR, restrictiondigestion, agarose gel electrophoresis, ligation,and transformation into Escherichia coli DH5a,were performed by standard procedures (28). TheArabidopsis TAAC gene was PCR amplified witholigonucleotides designed to exclude the first 177basepairs of the gene (corresponding to thepredicted transit peptide), cloned and recombinedas a fusion construct with an N-terminal

by guest on July 12, 2018http://w

ww

.jbc.org/D

ownloaded from

A plant thylakoid ATP/ADP carrier

4

hexahistidine tag followed by Xpress epitope anda C-terminal FLAG epitope (DYKDDDDK), intoa pTrcHisB plasmid for expression in E. coli. Sucha set-up has previously been employed to purifyanother membrane-embedded protein (29). Thedetailed protocols are given in SupplementalExperimental procedures.

Uptake of Radioactive ATP and ADP into E.col i Cells–For uptake experiments, cellstransformed with the TAAC-expressing plasmid(or control expression plasmid) were induced withisopropyl 1-thio-b,D-galactopyranoside (IPTG) for6 h in Terrific Broth (TB) medium supplementedwith 10 mM malate and 10 mM pyruvate (17).Uptake experiments were carried out according to(17,30), with a few modifications. Cells (30 mL,100 mg mL-1) were incubated in 50 mM phosphatebuffer (pH 7.0) containing 50 mM [a-32P]ATP(500 mCi/mmol; 1 m Ci=37 kBq, AmershamBiosciences, Little Chalfont, UK) or [a-32P]ADP,prepared according to ref. (30), at 30°C for theindicated time periods. Where indicated, theuptake experiments were carried out in thepresence of 2.5 mM various nonlabelednucleotides. For determination of the transportaffinity (Km) and the maximal rate (Vmax), uptakeof a range of concentrations (0-250 mM) of [a-32P]ATP or [a-32P]ADP was carried out in E. colicells, preincubated or not with 100 m M m -chlorocarbonyl-cyanide phenylhydrazone (CCCP)for uncoupling. The cells were thereafter quicklyfiltrated through a 0.45 mm filter (Pall, New York)under vacuum, and washed three times with 1 mLice-cold phosphate buffer. The radioactivityretained on the filters was quantified in 3.5 mLwater using scintillation spectrometry. Backexchange experiments were carried out essentiallyas previously described (17,30). The detailedprotocol is provided in SupplementalExperimental procedures.

Uptake of Radioactive ATP into Thylakoidsand Assay of NDPK Activity–Arab idops i sthylakoids (30 m L, 0.3 m g Chl m L -1) wereincubated in Tricine buffer containing 10 mM[a-32P]ATP in darkness at 22°C for the indicatedperiods of time. In some experiments, the uptakewas carried out using 0-50 mM [a-32P]ATP or inthe presence of 1 mM nonlabeled nucleotides. The

thylakoids were washed twice, recovered by rapidcentrifugation (45 s, 13,000 ¥ g ), and theirradioactivity counted in 1 mL water. Controlexperiments in the presence of 10 mM nonlabeledATP indicated 10% nonspecific binding. Thisvalue was subtracted from all measured activities.Back exchange experiments were performedessentially as described for E. coli. Details aregiven in Supplemental Experimental procedures.NDPK activity was assayed as the amount of[g-32P]GTP produced from 1 mM [g -32P]ATP(12.5 Ci/mmol) and 1 mM GDP, as described (10).

RNA Extraction and Northern BlotAnalysis–Total RNA was extracted from frozenArabidopsis tissues using TRIzol“ and Plant RNAPurification Reagent (Invitrogen, Carlsbad, CA) orRNeasy Kit (Qiagen, Germantown, CA),according to the manufacturer’s instructions.Equal amounts (5-10 µg) of RNA were separatedo n a 1 . 2 % ( w / v ) denaturatingformaldehyde/agarose gel, and transferred to apositively charged nylon membrane (Roche,Indianapolis, IN) (28). The TAAC probe wasobtained by PCR amplification of a 430 bpfragment using the At5g01500 cDNA as template,and labeled by adding the DIG-labeleddeoxynucleotide mixture (Roche, Indianapolis,IN). The following PCR primers were used:forward 5'-GGCACCGCTTGACCGAATAA andr e v e r s e 5 ’ -AAGCGAAGGACCTAGACCGTTGTA.Hybridization was carried out according to theinstructions in the DIG-Nonradioactive NucleicAcid Labeling and Detection System (Roche,Indianapolis, IN).

RESULTS

Structural Analyses of the TAACProtein–Similarity search with the amino acidsequence of the bovine AAC against theArabidopsis protein database combined withprediction of transit peptides revealed six plastidMCF members, which is in line with previouslypublished data (31). Among them, the product ofthe At5g01500 gene (MIPS code for TAAC) with aputative carrier function, corresponded in size tothe previously reported nucleotide thylakoid

by guest on July 12, 2018http://w

ww

.jbc.org/D

ownloaded from

A plant thylakoid ATP/ADP carrier

5

transporter (10). The coding region of theAt5g01500 gene obtained by screening of anArabidopsis cDNA library (22) confirmed theexon assignment in the MIPS database. In thesame database, five expressed sequence tags(ESTs) were found for Arabidopsis TAAC,indicating that the gene is transcriptionally active.

The encoded Arabidopsis TAAC proteincontains 415 amino acids and an N-terminalchloroplast transit peptide (amino acids 1-59), asindicated by TargetP (score 0.764, RC3) (Fig. 1A).Using advanced bioinformatics analysis for intra-chloroplast protein sorting (32), a low cysteinecontent (0.07%) and a negative GRAVY index (-0.079) were determined (data extracted fromPPDB), which are in favor of a thylakoid ratherthan envelope location for this putative protein.Nevertheless, it was annotated as an envelopeprotein due to its high hydrophobicity, highisoelectric point and putative carrier function (18).Notably, two proteomic studies disagree on theenvelope localization of the At5g01500 geneproduct (18,33).

The mature form shows 28% identity (43%similarity) to the bovine AAC protein (bAAC).The TAAC termini are 50 (N) and 30 (C) residues,respectively, longer than the correspondingregions in bAAC, and are rich in both positivelyand negatively charged residues (Fig. 1A). Thepresence of a five glycine-repeat in the middle ofthe N-terminus indicates that this region mayundergo large conformational changes, whichcould regulate TAAC activity. A ring of four well-conserved positively charged residues (K22, R79,Y186 and R279 in bAAC) was proposed to act asa selectivity filter for adenine nucleotides in allAACs (14,34). The alignment (Fig. 1A) shows thatthese residues are also conserved in TAAC (K130,R186, Y282, K369), suggesting a possiblesubstrate preference for adenine nucleotides. As itwas found in all MCF members, TAAC containsthree repeated homologous regions, with highlyconserved consensus sequences (Fig. 1B). Theunique AAC signature (RRRMMM) in the thirdrepeat region is only partially conserved in TAAC(RRqMql, where the lower case letters indicatenonconserved residues), as in the case of two otherAACs (34). The relevance of this signature forbinding and/or transport of adenine nucleotides is

not clearly understood. Nevertheless, the detailedanalysis of the bovine AAC X-ray structureindicated that only the first two arginines (alsoconserved in TAAC) interact withcarboxyatractyloside (34).

Various software tested at ARAMEMNONwebsite indicated zero to six transmembranedomains (TMDs) for the TAAC protein. Thereason for this large variability is the presence ofmany charged residues as well as conservedprolines in the putative TMDs of this type oftransport protein, unusual features for whichvarious software are more or less sensitive, assuggested previously (32). Six TMDs have beenrevealed in the crystal structure of the bovine AAC(14). Therefore, we have used the alignment withthe bovine carrier (Fig. 1A) as basis for thepresence of six TMDs in the TAAC structure (Fig.1B). In the folding model of the TAAC protein(Fig. 1C), the TMDs are connected by three loopson the stromal side and two loops on the lumenalside. The stromal loops contain amphipatic helices(h12, h34, h56, Fig. 1B). For the mitochondrialAAC, the N- and C-termini are in theintermembrane space, and the adenine nucleotide-binding site is on the matrix side of the membrane,where ATP is synthesized. Taking into accountthat, in chloroplasts, ATP is synthesized on thestromal side of the thylakoid membrane, and istransported from the stroma into the thylakoidlumen (10), the N- and C-termini of the TAACprotein are expected to be located on the lumenalside of membrane (Fig. 1C).

The TAAC amino acid sequence was analyzedfor structural homologues at the MetaServer(bioinfo.pl/meta/) (20). As expected, substantialstructural similarities to the bovine AAC (PDB#1okc A) were identified, and a highly significant3D-jury homology score (140.2) was provided.The presented structural homology model ofTAAC (Fig. 1D) is limited to amino acid residues110 to 384, excluding the large N- and C-termini,which have no correspondence in the bovine AAC.Like the bovine protein, the overall structure isbasket-shaped, with a closure on the stromal sideand opening on the lumenal side (Fig. 1D, sideview). The backbone also exhibits pseudo-threefold symmetry due to the presence of threerepeated homologous domains. The TMDs form a

by guest on July 12, 2018http://w

ww

.jbc.org/D

ownloaded from

A plant thylakoid ATP/ADP carrier

6

cavity that enters deeply into the protein, and thefour indicated residues making up the selectivefilter (K130, R186, Y282, K369), surround thebottleneck of the cavity (Fig. 1D, lumenal view).Taken together, the structural analyses indicatethat the product of the At5g01500 gene possessesa chloroplast transit peptide as well as thecharacteristic sequence features of a MCF andAAC member.

In Vivo and in Vitro Localization Studies ofthe TAAC Protein–The targeting of a TAAC-GFPconstruct was investigated in tobacco leafprotoplasts (Fig. 2A). As shown in the mergeimage (Fig. 2D), GFP fluorescence (Fig. 2B) andChl red autofluorescence (Fig. 2C) colocalized.These data show that TAAC is efficiently targetedto the chloroplast. However, no information aboutthe intra-chloroplast localization could beobtained. For this purpose, Arabidopsis chloroplastmembrane subfractions, obtained by sucrosedensity gradient centrifugation, were analyzed byWestern blotting using the anti-TAAC antibodyand its preimmune serum. As shown in Fig. 3A(upper panel), a single band with Mr of 36.5 kDacrossreacted with the anti-TAAC antibody, andcorresponded to a chloroplast protein mainlylocated in the thylakoid membrane. The data donot, however, refute earlier findings (18),suggesting that this protein is also present in theenvelope membrane. No crossreaction wasdetected when the corresponding rabbitpreimmune serum was used, further supporting thespecificity of the anti-TAAC antibody (Fig. 3A,upper panel). The high purity of the analyzedthylakoid and envelope preparations wasconfirmed by Western blotting with antibodiesagainst corresponding marker proteins, such as thethylakoid Lhcb2 protein, and the inner envelopeTIC110 protein (Fig. 3A, lower panels).

Treatment with 1 M NaCl or 0.1% (w/v)Triton X-100 rendered the 36.5 kDa protein bandin the membrane fraction, as in the case of controlthylakoids (Fig. 3B). Only treatment with 1%(w/v) Triton X-100 solubilized the TAAC protein,indicating that this is a membrane-spanningprotein, and supporting the TMDs prediction data(Fig. 1).

The TAAC protein was immunolocalized on

cryosections containing mesophyll cells ofArabidopsis rosette leaves, using serial dilutions ofthe TAAC antibody. Specific labeling wasobtained at 1:50 dilution of the antibody. Theanalysis of several sections showed that theimmunogold particles were mainly found inchloroplasts (Fig. 4). Particles were alsooccasionally detected on peroxisomes (data notshown). The nucleus, the cell wall, the vacuole,the cytosol, and organelles including mitochondriawere not labeled (Fig. 4A). In chloroplasts, thedistribution of the immunogold particles, whichreflects the distribution of the TAAC protein, wasnonuniform. The particles were mostly associatedwith the stroma-exposed regions of the thylakoidmembrane and of the grana stacks (Fig. 4B and C),as also demonstrated for the chloroplast ATPsynthase (35), and rarely associated with thestroma and envelope (Fig. 4A) or inside the granastacks (Fig. 4D). Control sections, without primaryantibodies, were not labeled (Fig. 4E). Takentogether, these studies demonstrate a chloroplastlocation for the TAAC protein, mainly in thethylakoid membrane.

Expression and Functional Characterizationof the TAAC Protein in E. coli Cells–Sequenceanalyses data indicated adenine nucleotides as themost likely substrates for transport by TAAC (Fig.1). To study its substrate specificity and kineticsfor transport, we expressed a recombinant His6-Xpress-TAAC-FLAG protein in E. coli cells. Fig.5A shows a Western blot with Xpress antibody ofprotein extracts obtained by cell fractionation, andat different steps of the purification process. Asingle cross-reacting band was detected in theIPTG-induced E. coli cells containing the TAAC-expressing plasmid (pos.), but not in the IPTG-induced cells lacking the coding sequence in theplasmid (neg.). The majority of the expressedTAAC protein was found in the membranefraction (M), a finding not least important for thefunctional characterization in intact bacterial cells(see below). The TAAC protein was purified fromthe detergent-solubilized membrane protein extractby chromatography on a Ni-column (M1) followedby a FLAG-column (M2a). The TAAC proteinwas 3,500 and 7,000-fold enriched in M1 andM2a, respectively, compared to the crudemembrane extract. The identity of the purified

by guest on July 12, 2018http://w

ww

.jbc.org/D

ownloaded from

A plant thylakoid ATP/ADP carrier

7

protein was confirmed by Western blotting withthe anti-TAAC antibody (M2b) as well as by massspectrometry analysis (data not shown).Coomassie staining revealed the TAAC protein asthe main protein band in M2a sample,emphasizing the importance of this proteinpurification system as a tool for crystallization andfurther characterization.

Fig. 5B shows a time dependency for theuptake of [a-32P]ATP and [a-32P]ADP into intactE. coli cells expressing the recombinant TAACprotein. In the background, non-time dependentbinding of nucleotides was observed in IPTG-treated cells lacking the TAAC insert, representinga maximum of 19 and 23% of the assayed uptakeof ATP and ADP, respectively. Furthermore,radioactive nucleotides were detected in washedand disrupted expressing cells (see Fig. 4D below)but not in the control cells (data not shown),indicating that the assayed transport activity isdependent on the presence of the TAAC protein.The ratio of ATP to ADP uptake was about 2.2,suggesting a preferred import of ATP versus ADPinto E. coli cells. This is opposite to the preferredimport of ADP versus ATP by heterologouslyexpressed Arabidopsis mitochondrial AACs (17).

For determination of the Km and Vmax values ofthe TAAC-catalyzed transport, the E. coli cellswere incubated with 0-250 mM radioactiveadenine nucleotides for 1 min (Fig. S2). Theapparent Km values for ATP and ADP (Table S1),were about 45 mM, i.e. intermediate between thosereported for Arabidopsis (10-15 m M) andmammalian (100-150 mM) AACs, when expressedand assayed in E. coli cells (17). The calculatedVmax value for ATP and ADP uptake were 0.71 and0.53 nmol mg-1 protein h-1 (Table S1). Both Vmax

values are in the same range as those determinedfor the Arabidopsis mitochondrial AACsexpressed in E. coli, i.e. 0.18-4.41 nmol mg-1

protein h-1 (17).The presence of CCCP increased the apparent

Km value of TAAC for ATP to 170 mM, and thecorresponding Vmax value to 1.79 nmol mg-1

protein h-1 (Fig. S2 and Table S1). Thecorresponding values for ADP were not affectedby the presence of CCCP. The fact that TAAC hasa lower affinity for binding ATP in non-energized

E. coli membranes could be due to a proteinconformational change such that the substratebecomes loosely bound and is faster released onthe other side of the membrane. The Km valueswere only slightly lowered by CCCP in the case ofthe mitochondrial AACs expressed in E. coli (17).The mechanistic reason for this discrepancy withrespect to CCCP is not known. Nevertheless, theTAAC transport activity shows other differencesas well. For example, export of ADP by TAAC isinhibited by CCCP (see below) at variance withthe same activity performed by the mitochondrialAACs (17).

To investigate the nucleotide transportspecificity of TAAC, we performed uptake of [a-32P]ATP into E. coli cells expressing therecombinant protein, in the presence of excess ofvarious nonlabeled nucleotides (Fig. 5C). Incomparison with the control uptake, ATPcompeted with the highest efficiency (83%),closely followed by ADP (60%). The guaninenucleotides did not affect the ATP uptake,suggesting that although they could bind to thecarrier protein (10), they are not substrates fortransport. ADP-glucose (Fig. 5C) and phosphate(data not shown) did not compete with the ATPuptake, indicating that TAAC is not involved inneither nucleotide-sugar nor phosphate transport.The competition data strongly support thatArabidopsis TAAC specifically transports adeninenucleotides, as reported for mammalian AACs,and in contrast to maize AAC, which can alsotransport guanine nucleotides (36 and referencestherein).

Most MCF members catalyze strict soluteexchange reactions. TAAC-expressing E. coli cellswere preloaded with [a-32P]ATP followed by backexchange with nonlabeled external substrates, andthe released nucleotides were separated by thinlayer chromatography (TLC). As shown in Fig.5D, disrupted preloaded cells contained bothradioactive ATP and ADP, indicating that anuptake of ATP had taken place, and that part of itwas converted to ADP in the cytosol. Nosignificant amount of radioactive nucleotides wasreleased after the addition of buffer. A preferentialexport of [a-32P]ADP occurred during the backexchange with externally added adenine

by guest on July 12, 2018http://w

ww

.jbc.org/D

ownloaded from

A plant thylakoid ATP/ADP carrier

8

nucleotides, and was effectively inhibited byCCCP (Fig. 5D).

In Planta Functional Analyses of the TAACProtein–The nucleotide uptake and exchange dataobtained from functional expression of therecombinant TAAC in E. coli were compared withthose in Arabidopsis thylakoids. All experimentsusing thylakoids were carried out in darkness,when ATP binding and hydrolysis by the catalyticportion (CF1) of the ATP synthase is extremelylow (37,38, and references therein). Under ourexperimental conditions, the uptake of [a-32P]ATPinto thylakoids had a fast linear phase (up to 1min), and reached a plateau level of 150 pmol ofATP/mg Chl within about 5 min (Fig. S3A). Theestimated apparent Km for ATP was about 0.5 mM,(Fig. S3B), which is significantly lower than theKm value of recombinant TAAC in E. coli (47 mM,Table S1). Another difference from TAAC-mediated transport in E. coli is that, in addition toATP and ADP, GDP also competed with theuptake of [a-32P]ATP in thylakoids (Fig. S3C).These data imply that, in addition to TAAC, thereare other sites for ATP binding and transport inthylakoids, including the previously characterizedbinding to the ATP synthase (37,39,40).Nevertheless, our data also indicate similarkinetics to the ones reported for nucleotide uptake(by AACs) in isolated mitochondria (17,41).

To investigate the internal affinities ofthylakoids for adenine nucleotides, we carried outback exchange experiments in thylakoidspreloaded with [a-32P]ATP (Fig. S3D). Preloadedthylakoids contained radioactive ATP and ADP ina ratio of about 4:1, and showed a preferentialexport of [a -32P]ADP, supporting the dataobtained for TAAC in E. coli cells (Fig. 5D).Since incubation with ADP is known to abolishthe ATPase activity of CF1 (38), its interference inthe observed exchange. Is not very likely. The factthat even GDP induced export of ADP (Fig. S3D)has significance for the previously characterizedNDPK activity in the thylakoid lumen (10). SinceGDP competed with ATP uptake and inducedADP export, there must exist another transportsystem than TAAC for GDP in thylakoids.

To determine the physiological influence ofTAAC on the availability of adenine nucleotides

inside the chloroplast thylakoid lumen, a FLAGline with a T-DNA insertion in the TAAC gene(taac mutant, Fig. S1A) was analyzed andcompared with WT. The homozygoty of this linewas confirmed by PCR (Fig. S1B), and the lack ofTAAC was verified by Western blotting ofisolated thylakoids (Fig. 6A). As a control, theamount of Lhcb2 was analyzed, and shown to beat comparable level in the WT and the t aacmutant. Uptake of [a-32P]ATP into dark-incubatedthylakoids isolated from the taac mutant wasreduced to 37% (at 1 min) and 68% (at 5 min) ofthe WT levels (Fig. 6B). Interconversion of ATPto GTP in the thylakoids (via lumenal NDPK3)was reduced to 60-70% of the WT levels (Fig.6C). Additional phenotypic effects of T A A Cdisruption included about 40% reduction in thethylakoid content as compared to WT (expressedas mg Chl/g leaf, Fig. S4), resulting in pale greenleaves of the taac mutant.

TAAC Expression Pattern in Arabidopsis–Toget information about the in planta role of TAAC,we addressed several questions: whetherexpression of the TAAC gene in Arabidopsis isrestricted to photosynthetic tissues, and whether itoccurs at specific stages of development and/orunder certain stress conditions. As shown byNorthern blot experiments (Fig. 7A) a significantamount of T A A C transcript was detected inseedlings grown in darkness, and its level did notchange significantly during a period of up to 12 hof greening. The level of the TAAC protein in thecorresponding protein extracts followed thetranscript pattern, as revealed by Western blottingwith the anti-TAAC antibody (Fig. 7A). Forcomparison, the transcription of maizemitochondrial AACs readily occurs in darkness,and decreases as soon as the tissue becomesphotosynthetically active (36) and referencestherein).

We also investigated expression duringvarious leaf developmental stages (Fig. 7B). Ourdata revealed that the amount of the T A A Ctranscript was high in developing leaves,decreased during maturation, and increased againduring senescence to levels exceeding thosepresent in developing leaves. The TAAC level inthe corresponding protein extracts generallyfollowed the level of corresponding transcripts.

by guest on July 12, 2018http://w

ww

.jbc.org/D

ownloaded from

A plant thylakoid ATP/ADP carrier

9

Next, we assayed expression in leaves exposed tovarious abiotic stress conditions (Fig. 7C). Whilethe amounts of the TAAC transcript were increasedduring wounding, light stress, oxidative stress, saltstress and desiccation, a drastic reduction wasdetected under heat shock conditions. Similarly,the exposure of leaves to low temperature led to aslight decrease in the TAAC transcript level. Asbefore, the protein pattern resembled the one ofthe transcript. Finally, the TAAC expression wasinvestigated in various Arabidopsis organs (Fig.7D). The highest levels of TAAC transcript andprotein were detected in developingphotosynthetic organs such as leaves (see also Fig.7B), flower buds and green siliques, followed byfully developed organs such as roots, flowers,cauline and rosette leaves, and least in the stem.For comparison, the mitochondrial AACs show asimilar expression pattern in all tissues (36),pointing to a more general role in plantmetabolism. High TAAC levels in Arabidopsisdark-grown seedlings, in developing and senescentleaves may have relevance for initiation ofbiogenesis and turnover of the photosyntheticapparatus in the thylakoid membrane.

DISCUSSION

The old dogma was that mitochondria andplastids have different types of nucleotidetransporters, namely mitochondrial AACs thatexport ATP to the cytosol, and plastidic NTTs thatperform ATP uptake into the stroma (42). The datapresented in this work demonstrate that thechloroplast thylakoid membrane contains an MCFmember, namely an ATP/ADP carrier, hereinnamed TAAC. We identified this protein inArabidopsis thylakoids as the product of theAt5g01500 gene. Detection of the TAAC proteinin low amounts in the chloroplast envelope and innon-photosynthetic plastids points to an apparentdual localization. We show that the recombinantTAAC protein is specific for adenine nucleotidetransport, and imports ATP in exchange for ADPacross the E. coli cytoplasmic membrane. Thesame exchange pattern occurs across Arabidopsisthylakoids. Finally, we show reduction of ATPtransport and metabolism within the thylakoidmembrane in a TAAC knockout mutant.

Among photosynthetic organisms, ESTs fromthe green alga Chlamydomonas reinhardtii, maize,rice, wheat, potato and soybean show 60-80%identity to the Arabidopsis TAAC gene sequence,as revealed by orthologues search in TIGRdatabases. No homologues in the genome databaseof cyanobacteria (Cyanobase) were found,supporting a previous bioinformatic report, whichsuggested that MCFs are a later addition duringthe development of the eukaryotic cell (43).

Arabidopsis TAAC possesses thecharacteristic sequence features of MCF and AACmembers (Fig. 1). Measurements of radioactiveadenine nucleotide transport across E. colimembrane demonstrated that this protein is indeedan AAC, sharing similarities with but alsodifferences from the mitochondrial homologues.The most striking difference is the ATP/ADPcounter exchange catalyzed by TAAC (Fig. 5),which is opposite to ADP/ATP exchange by AACacross E. coli membrane (17). The pattern ofadenine nucleotide exchange for both TAAC andAAC in organello is similar to the one obtainedfor the corresponding recombinant proteinsexpressed in E. coli, but apparently opposite toeach other (Fig. 8). The explanation resides in thefact that, structurally, the thylakoid membrane andits lumenal space represent intra-chloroplastcompartments without correspondent inmitochondria. Nevertheless, the translocation ofadenine nucleotides by TAAC and AAC proceedsin an analogous manner across thylakoid andmitochondrial inner membrane, namely from thesite of ATP synthesis (stromal or matrix side) tothe other side of the membrane (thylakoid lumenor mitochondrial intermembrane space) (Fig. 8).

There is a lot of information available for theadenine nucleotide transport system inmitochondria and about the activities ofrecombinant AACs (17,41), which could becompared with the uptake data for thylakoidmembrane and recombinant TAAC (this work). Inthe past, nucleotide binding to the thylakoidmembrane has been mainly attributed to the ATPsynthase. It was shown to readily occur indarkness and to increase at least three fold duringphotophosphorylation (37,39,40). Remarkably,uncouplers and inhibitors of photophosphorylationwere found effective in suppressing the binding in

by guest on July 12, 2018http://w

ww

.jbc.org/D

ownloaded from

A plant thylakoid ATP/ADP carrier

10

the light but not the one observed in darkness (37).Using radioactive azido-ATP labeling ofthylakoids, it was later demonstrated specificbinding to a nucleotide carrier of 36.5 kDa (10),identified as a TAAC in this work. Therefore, wesuggest that the TAAC-catalyzed uptake of ATPinto thylakoids in darkness is a component of thepreviously reported tight ATP binding (37). ATPuptake in mitochondria (17,41) and ATP bindingto thylakoids in darkness (37) are both largelyunaffected by uncouplers and inhibitors ofoxidative/photo phosphorylation, indicating that aproton motive force across membrane is notabsolutely required. In light of the novel conceptof induced transition fit proposed to function intransport catalysis (44), it is the interaction of thenucleotide substrate with the (T)AAC providingthe catalytic energy for transport. Nevertheless,external energy may maximize the transition fit,hence the transport efficiency.

Data available in public microarrayscollections, such as Genevestigator database(https://www.genevestigator.ethz.ch/at/) (45),indicated the highest levels of TAAC transcript inseeds followed by shoot apex, and much lowersimilar levels in other tissues (flowers, roots androsette leaves). The same microarray dataindicated high transcript levels in young leavesand petiole and low levels in senescent leaves. Ourdata show a different pattern, with the highestexpression being detected in dark-grownseedlings, developing green tissues (leaves, flowerbuds, siliques) and in senescent leaves (Fig. 7).Expression in roots and flowers indicates presenceof the TAAC protein even in non-green plastids.Developing plastids, etioplasts and to a lowerextent also other types of plastids have, in additionto envelope, internal membranes that couldconvert to thylakoid membranes upon illumination(3). At least in developing plastids, the internalmembranes are ocassionally continuous with theenvelope (3).

The TAAC protein is found readily expressedin dark-grown seedlings, and should be very activeduring biogenesis of the photosynthetic apparatus,which starts only few minutes after light exposure(46). The HCF136 protein is also produced indark-grown seedlings and is kept at a stable levelduring light-induced greening (47). This protein

was demonstrated to be located in the thylakoidmembrane, and must be present when PSIIcomplexes are made in the stroma regions.Similarly to the HCF136 protein, TAAC has to bepresent prior to, or at least concomitantly with theaccumulation of photosynthetic complexes. It maysupply various biosynthetic processes with ATP,which was shown to be produced in increasingamounts during embryonic photosynthesis (48).

In leaf chloroplasts, TAAC was detectedmainly in the thylakoid but also in the envelopefraction to a lower extent (Fig. 3). Because aprevious proteomic work (18) has identifiedTAAC in purified envelope preparations, onecannot exclude its presence also in the envelope.The localization of another MCF member in bothmitochondria and chloroplast envelope hasrecently been reported (49). Furthermore, there areseveral examples of proteins, involved in Chlsynthesis, thylakoid biogenesis and chloroplastmorphology, targeted to both envelope andthylakoid membranes (50-52), which support ourfindings. What does a dual localization of TAACmean? The level of TAAC transcript and proteinpeaked in young green tissues that undergo rapidplastid development, and declined as the plantmatured (Fig. 7). Furthermore, the thylakoidmembrane was the predominant location underconditions in which mature tissues were used as asource of chloroplasts (Figs. 3 and 4). Theseobservations imply that TAAC is needed to bepresent in the internal membranes of developingplastids more than in the thylakoids of maturechloroplasts. Its location depends on thedevelopmental stage of plastids, and TAAC mayhave multiple functions. In premature chloroplasts,it may be initially located in the envelope andparticipate in the development of internalmembranes into thylakoid membranes. In contrast,in mature chloroplasts, TAAC is mainly located inthe thylakoid membrane, and may have a role inits turnover.

We also detect increasing TAAC transcriptlevels under senescence and various abiotic stresstreatments of leaves, in contrast to theGenevestigator database reporting down regulationand no change, respectively. We suggest a role forTAAC in supplying ATP into the thylakoid lumenfor mobilization of N-resources (during

by guest on July 12, 2018http://w

ww

.jbc.org/D

ownloaded from

A plant thylakoid ATP/ADP carrier

11

senescence), for refolding of misfolded thylakoidproteins (during import), and for their degradation(during stress). Alternatively, it is converted toGTP in the thylakoid lumen (10), which v iabinding and hydrolysis by the extrinsic PsbOsubunit of PSII complex stimulates the light-induced degradation of the reaction center D1protein in plants (26,53).

To our knowledge, the product of theAt5g01500 gene is the first chloroplast AACidentified and characterized at tissue, cellular andmolecular levels. The TAAC protein represents alink between ATP synthesis on the stromal side ofthe thylakoid membrane and nucleotide-dependentreactions in the lumenal space (10,12,13). Failureto detect a more substantial reduction of ATPtransport and metabolism in the taac mutant,suggests that the At5g01500 gene may encode atransporter that normally contributes only 30-40%to the thylakoid transport. As an alternative,TAAC may quantitatively be a major player, andother yet unidentified transporters partiallycompensate for its disruption in the taac mutant.

While further analyses are required to

elucidate the physiological role of this carrier, itshould be noted that other compounds resultingfrom nucleotide metabolism such as phosphatecannot be exported by TAAC, because it is not asubstrate of this carrier, implying the presence ofother transporter(s) responsible for its removal.The identification of these proteins should be thefocus of further studies aimed at understanding thelumenal network of nucleotide metabolism.

Acknowledgments – The authors would like tothank Prof. G. von Heijne (Stockholm University)for helpful suggestions and discussions of theTAAC prediction data, Prof. J. Soll (MunichUniversity) for providing the anti-TIC110antibody and to Kent Lundh (Kalmar University)for help with the chromatographic techniques.Providing the Ecker size selected ArabidopsiscDNA library (CD4-14) by the ArabidopsisBiological Resource Center at Ohio StateUniversity is acknowledged.

Supplementary information accompanies the paperon www.jbc.org.

REFERENCES

1. Ephritikhine, G., Ferro, M., and Rolland, N. (2004) Plant Physiol. Biochem. 42, 943-9622. van Wijk, K. J. (2004) Plant Physiol. Biochem. 42, 963-9773. Vothknecht, U. C., and Westhoff, P. (2001) Biochim. Biophys. Acta 1541, 91-1014. van Wijk, K. J. (2001). In: Aro, E. M., and Andersson, B. (eds). Regulation of Photosynthesis,

Kluwer Academic, Dordrecht5. Ettinger, W. F., Clear, A. M., Fanning, K. J., and Peck, M. L. (1999) Plant Physiol. 119, 1379-13866. Shingles, R., Wimmers, L. E., and McCarty, R. E. (2004) Plant Physiol. 135, 145-1517. Vambutas, V., Tamir, H., and Beattie, D. S. (1994) Arch Biochem Biophys 312, 401-4068. Hinnah, S. C., and Wagner, R. (1998) Eur. J. Biochem. 253, 606-6139. Pottosin, I. I., and Schonknecht, G. (1996) J. Membr. Biol. 152, 223-23310. Spetea, C., Hundal, T., Lundin, B., Heddad, M., Adamska, I., and Andersson, B. (2004) Proc. Natl.

Acad. Sci. U. S. A. 101, 1409-141411. Abdel-Ghany, S. E., Muller-Moule, P., Niyogi, K. K., Pilon, M., and Shikanai, T. (2005) Plant Cell

17, 1233-125112. Wagner, V., Gessner, G., Heiland, I., Kaminski, M., Hawat, S., Scheffler, K., and Mittag, M. (2006)

Eukaryot. Cell 5, 457-46813. Rinalducci, S., Larsen, M. R., Mohammed, S., and Zolla, L. (2006) J. Proteome Res. 5, 973-98214. Pebay-Peyroula, E., Dahout-Gonzalez, C., Kahn, R., Trezeguet, V., Lauquin, G. J., and Brandolin, G.

(2003) Nature 426, 39-4415. Millar, A. H., and Heazlewood, J. L. (2003) Plant Physiol. 131, 443-45316. Heazlewood, J. L., Tonti-Filippini, J. S., Gout, A. M., Day, D. A., Whelan, J., and Millar, A. H.

(2004) Plant Cell 16, 241-256

by guest on July 12, 2018http://w

ww

.jbc.org/D

ownloaded from

A plant thylakoid ATP/ADP carrier

12

17. Haferkamp, I., Hackstein, J. H. P., Voncken, F. G. J., Schmit, G., and Tjaden, J. (2002) Eur. J.Biochem. 269, 3172-3181

18. Ferro, M., Salvi, D., Brugiere, S., Miras, S., Kowalski, S., Louwagie, M., Garin, J., Joyard, J., andRolland, N. (2003) Mol. Cell. Proteomics 2, 325-345

19. Norén, H., Svensson, P., and Andersson, B. (2004) Physiol. Plantarum 121, 343-34820. Ginalski, K., Elofsson, A., Fischer, D., and Rychlewski, L. (2003) Bioinformatics 19, 1015-101821. Schwede, T., Kopp, J., Guex, N., and Peitsch, M. C. (2003) Nucleic Acids Res. 31, 3381-338522. Kieber, J. J., Rothenberg, M., Roman, G., Feldmann, K. A., and Ecker, J. R. (1993) Cell 72, 427-44123. Eshaghi, S., Andersson, B., and Barber, J. (1999) FEBS Lett. 446, 23-2624. Bradford, M. M. (1976) Anal. Biochem. 72, 248-25425. Porra, R. J., Thompson, W. A., and Kriedemann, P. E. (1989) Biochim. Biophys. Acta 975, 384-39426. Spetea, C., Hundal, T., Lohmann, F., and Andersson, B. (1999) Proc. Natl. Acad. Sci. U. S. A. 96,

6547-655227. Tokuyasu, K. T. (1980) Histochem. J. 12, 381-40328. Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989) Molecular Cloning: A Laboratory Manual, 3nd

Ed., Cold Spring Harbor Laboratory Press, Cold Spring Harbor29. Lagerstedt, J. O., Voss, J. C., Wieslander, A., and Persson, B. L. (2004) FEBS Lett. 578, 262-26830. Tjaden, J., Schwoppe, C., Mohlmann, T., Quick, P. W., and Neuhaus, H. E. (1998) J. Biol. Chem.

273, 9630-963631. Weber, A. P., Schwacke, R., and Flugge, U. I. (2005) Annu. Rev. Plant Biol. 56, 133-16432. Sun, Q., Emanuelsson, O., and van Wijk, K. J. (2004) Plant Physiol. 135, 723-73433. Kleffmann, T., Russenberger, D., von Zychlinski, A., Christopher, W., Sjolander, K., Gruissem, W.,

and Baginsky, S. (2004) Curr. Biol. 14, 354-36234. Nury, H., Dahout-Gonzalez, C., Trezeguet, V., Lauquin, G. J., Brandolin, G., and Pebay-Peyroula, E.

(2006) Annu. Rev. Biochem. 75, 713-74135. Anderson, J. M., and Andersson, B. (1982) Trends Biochem. Sci. 7, 288-29236. Laloi, M. (1999) Cell. Mol. Life. Sci. 56, 918-94437. Magnusson, R. P., and McCarty, R. E. (1976) J. Biol. Chem. 251, 7417-742238. McCarty, R. E. (2005) J. Bioenerg. Biomembr. 37, 289-29739. Smith, D. J., and Boyer, P. D. (1976) Proc. Natl. Acad. Sci. U. S. A. 73, 4314-431840. Abbott, M. S., Czarnecki, J. J., and Selman, B. R. (1984) J. Biol. Chem. 259, 12271-1227841. Winkler, H. H., Bygrave, F. L., and Lehninger, A. L. (1968) J. Biol. Chem. 243, 20-2842. Winkler, H. H., and Neuhaus, H. E. (1999) Trends Biochem. Sci. 24, 64-6843. Kunji, E. R. (2004) FEBS Lett. 564, 239-24444. Klingenberg, M. (2006) Biochim. Biophys. Acta 1757, 1229-123645. Zimmermann, P., Hirsch-Hoffmann, M., Hennig, L., and Gruissem, W. (2004) Plant Physiol. 136,

2621-263246. Bertrand, M., Bereza, B., and Dujardin, E. (1988) Z. Naturforsch. C43, 443-44847. Meurer, J., Plucken, H., Kowallik, K. V., and Westhoff, P. (1998) EMBO J. 17, 5286-529748. Borisjuk, L., Rolletschek, H., Radchuk, R., Weschke, W., Wobus, U., and Weber, H. (2004) Plant

Biol. (Stuttg) 6, 375-38649. Palmieri, L., Arrigoni, R., Blanco, E., Carrari, F., Zanor, M. I., Studart-Guimareas, C., Fernie, A. R.,

and Palmieri, F. (2006) Plant Physiol. 142, 855-86550. Che, F. S., Watanabe, N., Iwano, M., Inokuchi, H., Takayama, S., Yoshida, S., and Isogai, A. (2000)

Plant Physiol. 124, 59-7051. Inoue, K., Baldwin, A. J., Shipman, R. L., Matsui, K., Theg, S. M., and Ohme-Takagi, M. (2005) J.

Cell. Biol. 171, 425-43052. Gao, H., Sage, T. L., and Osteryoung, K. W. (2006) Proc. Natl. Acad. Sci. U.S.A. 103, 6759-676453. Lundin, B., Thuswaldner, S., Shutova, T., Eshaghi, S., Samuelsson, G., Barber, J., Andersson, B., and

by guest on July 12, 2018http://w

ww

.jbc.org/D

ownloaded from

A plant thylakoid ATP/ADP carrier

13

Spetea, C. (2006) Biochim. Biophys. A c t a Epub ahead of print 7 Nov 2006,doi:2010.1016/j.bbabio.2006.2010.2009

FOOTNOTES

* This work was supported by the Swedish Research Council (to C.S. and B.L.P.), the Graduate ResearchSchool in Genomics and Bioinformatics and Hagberg Foundation (to C.S.), the Helge Ax:son JohnsonFoundation (to C.S., J.O.L. and S.T.), the Human Frontier Science Organization (to B.L.P.), DeutscheForschungsgemeinschaft grant (AD92/7-1; AD92/7-2) and the Konstanz University grant (to I.A.), theFrench Ministère de l’Education Nationale de l’Enseignement Supérieur et de la Recherche, the InstitutNational de la Recherche Agronomique, the Centre National de la Recherche Scientifique (Dijon group).§Present address: Departments of Biochemistry and Molecular Medicine, and Internal Medicine,University of California, School of Medicine, Davis, 95616 California, USA.|| Both authors equally contributed to this work.1Abbreviations: AAC, ADP/ATP carrier; CCCP, m-chlorocarbonyl-cyanide phenylhydrazone; Chl,chlorophyll; GFP, green fluorescent protein; IPTG, isopropyl 1-thio-b,D-galactopyranoside; MCF,mitochondrial carrier family; NDPK, nucleoside diphosphate kinase; PS, photosystem; TAAC, thylakoidATP/ADP carrier; TLC, thin layer chromatography; TMD, transmembrane domain; WT, wild type.

FIGURE LEGENDS

Fig. 1. In silico analyses of the predicted TAAC protein from Arabidopsis. A, Sequence alignment of theTAAC precursor protein (TrEMBL #Q9M024) with the bovine AAC (bAAC, Swiss-Prot P02722), asperformed by ClustalW. The transit peptide cleavage site of TAAC (as predicted by TargetP) is markedby an arrow. Identical and conserved residues are indicated in yellow and pink, respectively. A glycinerepeat is highlighted (*). Four residues composing the selectivity filter are marked by borderlines. B,Alignment of three repeated homologous regions in the TAAC amino acid sequence. Identical residuesbetween the repeats are highlighted in yellow. The putative TMDs (H1 to H6) and the short amphipatichelices (h12, h34, h56) are underlined. The consensus MCF motif and the AAC signature are indicatedbelow the alignment. C , Predicted folding model of the TAAC protein, as based on the membranetopology of bAAC. Boxes H1 to H6 denote the putative TMDs. N and C denote the protein termini, bothlocated on the lumenal side of the thylakoid membrane. D, Structural homology model of TAAC(residues 110 to 384), as based on the X-ray structure of bAAC (PDB # 1okc A). The three repeatsunderlined in B, are here shown in blue (residues 110-207), yellow (residues 208-300), and pink (residues301-384), respectively. The side-chains of the selectivity filter residues marked with borderlines in A, arehere highlighted in red.Fig. 2. Targeting of the TAAC-GFP fusion protein to chloroplasts in tobacco leaf protoplast, as revealedby confocal microscope imaging. These are representative data of protoplasts that transiently expressedthe fusion protein 24 h after electroporation. The bars indicated in the figures represent 5 mm. A,Differential interference contrast picture. B, GFP green fluorescence. C , Chl red autofluorescence. D,Overlay of images A, B, and C.Fig. 3. Intra-chloroplast location of the TAAC protein in Arabidopsis. A, Western blot analysis usingpreimmune and anti-TAAC antibodies was performed for protein extracts (30 mg/lane) from isolatedchloroplasts (C), thylakoids (T), and envelope (E). The positions of the markers are shown on the left.The position of the TAAC protein (36.5 kDa) is indicated. As references, the distribution of Lhcb2(thylakoid marker) and TIC110 (envelope marker) are shown. B, Thylakoids were non-treated (Ctrl) ortreated with 1 M NaCl, 0.1 or 1% (w/v) Triton X-100, and the distribution of the TAAC protein wasanalyzed by Western blotting in the corresponding soluble (S) and membrane (M) fractions (50 mg

by guest on July 12, 2018http://w

ww

.jbc.org/D

ownloaded from

A plant thylakoid ATP/ADP carrier

14

protein/lane).Fig. 4. Cryosections prepared from Arabidopsis leaf tissue were incubated with the anti-TAAC antibodyfollowed by immunogold labeled secondary antibody and electron microscope imaging. The barsindicated in the figures represent 0.1 mm. A, Gold particles (arrows) specifically labeled the chloroplastthylakoid membrane (T), and were absent in the soluble stroma (St), cytosol (Cy), mitochondria (M) andvacuole (V). Asterisk - plastoglobule. B, Labeling of the thylakoid membrane system. C, Labeling of thestroma-exposed regions is highlighted. D, Few particles were observed within the grana stacks. E, Controlsection (no anti-TAAC antibody) showed no labeling.Fig. 5. Transport kinetics of the recombinant TAAC protein in E. coli. A, TAAC-expressing (pos.) or non-expressing (neg.) E. coli cells were IPTG-induced for protein synthesis. Western blot analysis withXpress antibodies was carried out for protein extracts obtained during subfractionation of E. coli cells (45mg): cytosol (Cyt), and membrane (M), as well as during the two-step purification process (2 mg): eluatefrom the Ni-column (M1) and eluate from the FLAG-column (M2a). Western blot with the anti-TAACantibody of the second eluate (M2b) confirmed the identity of the recombinant protein. B, IPTG-inducedTAAC-expressing cells were incubated with 50 mM [a-32P]ATP (n) or [a-32P]ADP (l) for the indicatedperiods of time. The uptake was terminated by rapid filtration. C, Uptake of 50 mM [a-32P]ATP wascarried out in the presence of 2.5 mM nonlabeled nucleotides for 1 min. Values represent the percentageof remaining uptake compared to the value obtained in the absence of nonlabeled nucleotides. D, TAAC-expressing E. coli cells were preloaded with 50 mM [a-32P]ATP followed by disruption or back exchangewith buffer (-), 200 mM ATP or ADP, in the absence or presence of 100 mM CCCP. After sedimentationand stop of enzymatic activities, an aliquot of the supernatant was analyzed by TLC andphosphorimaging. Upper panel: plot of the amounts of exported nucleotides as a percentage of totalradioactivity detected in the disrupted cells. Lower panel: representative phosphorimaged TLC plate usedfor quantification. The data presented in panels B-D are the mean of at least three independentexperiments. SE<10% of the mean value.Fig. 6. Functional characterization of a TAAC Arabidopsis knockout mutant. A, Thylakoids were isolatedfrom the taac mutant and wild type (WT) plants and analyzed by Western blotting with anti-TAAC (30mg protein/lane) and anti-Lhcb2 (1 mg protein/lane) antibodies. B, Thylakoids were incubated in darknesswith 10 mM [a-32P]ATP for 1 or 5 min, washed and the bound radioactivity counted by scintillation. C,Thylakoids were incubated with 1 mM [g-32P]ATP and 1 mM GDP for the indicated periods of time, andthe nucleotides were analyzed by TLC followed and phosphorimaging.Fig. 7. TAAC expression pattern in Arabidopsis. The total RNA (5-10 mg) was isolated, and expression ofthe TAAC transcript was investigated by Northern blotting. As a reference, the 23S rRNA (ribosomalRNA) pattern in the gel was visualized by staining with ethidium bromide. The corresponding totalprotein extracts (15 mg) were analyzed by Western blotting with the anti-TAAC antibody. A, Arabidopsisseedlings were grown on liquid MS media for 10 days in darkness, and then illuminated with 50 µmolphotons m-2 s-1 for the indicated periods of time. B, Leaves at various developmental stages: (D)Developing; (Y) Young; (M) Mature fully developed; (OG) Old still green, and (OY) Senescent yellow.C , Detached leaves were exposed for 0, 1 or 2 h to various stress conditions: (W, wounding stress) cutinto 1 cm2 segments; (HL, high light stress) floated on water and illuminated with 1,000 µmol photons m-

2 s-1; (O, oxidative stress) incubated in 2% H2O2 solution; (S, salt stress) incubated in 0.4 M NaClsolution; (D, desiccation stress) kept on Whatman paper at room temperature; (C, cold stress) Detachedleaves were floated on water at 4°C; and (HS, heat shock) floated on water at 42°C. D, Various plantorgans: (R) roots; (S) stems; (RL) rosette leaves, (CL) cauline leaves; (FB) flower buds; (F) flowers, and(GS) green siliques.Fig. 8. Schematic model of adenine nucleotide exchange by AAC and TAAC. The native andrecombinant AAC proteins export endogenously synthesized ATP and import external ADP intomitochondria and E. coli cells, respectively, as indicated by previous uptake and back exchange

by guest on July 12, 2018http://w

ww

.jbc.org/D

ownloaded from

A plant thylakoid ATP/ADP carrier

15

experiments (17,41). The native TAAC transports stromal ATP into the lumenal space (ref. 10 and Fig.S3, this work). The recombinant TAAC protein transports external ATP in exchange for cytosolic ADPacross E. coli membrane (Fig. 5, this work). Thus, the direction of adenine nucleotide exchange by therecombinant AAC and TAAC across E. coli membrane is similar to the direction of transport by thenative proteins across the mitochondrial and thylakoid membrane, respectively.

by guest on July 12, 2018http://w

ww

.jbc.org/D

ownloaded from

TAAC MGEEKSLLQFRSFPSLKTSDFALTEEPSWRLENNVSSNRRRGNKRSGGVFTNFASLSVAI 60bAAC ------------------------------------------------------------

TAAC RRDRRESTFNGRNGGGGGAFASVSVVIPKEEDEFAPTSAQLLKNPIALLSIVPKDAALFF 120bAAC ----------------------------------SDQALSFLKD--------------FL 12 TAAC AGAFAGAAAKSVTAPLDRIKLLMQTHGVRAGQQSAKKAIGFIEAITLIGKEEGIKGYWKG 180bAAC AGGVAAAISKTAVAPIERVKLLLQVQHASKQISAEKQYKGIIDCVVRIPKEQGFLSFWRG 72 TAAC NLPQVIRIVPYSAVQLFAYETYKKLF-----RGKDGQLSVLGRLGAGACAGMTSTLITYP 235bAAC NLANVIRYFPTQALNFAFKDKYKQIFLGGVDRHKQFWRYFAGNLASGGAAGATSLCFVYP 132

TAAC LDVLRLRLAVEPG-------YRTMSQVALNMLREEGVASFYNGLGPSLLSIAPYIAINFC 288bAAC LDFARTRLAADVGKGAAQREFTGLGNCITKIFKSDGLRGLYQGFNVSVQGIIIYRAAYFG 192 TAAC VFDLVKKSLPEKYQQKTQSSLLTAVVAAAIATGTCYPLDTIRRQMQLKGT------PYKS 342bAAC VYDTAKGMLPDPKNVHIIVSWMIAQTVTAVAGLVSYPFDTVRRRMMMQSGRKGADIMYTG 252 TAAC VLDAFSGIIAREGVVGLYRGFVPNALKSMPNSSIKLTTFDIVKKLIAASEKEIQRIADDN 402bAAC TVDCWRKIAKDEGPKAFFKGAWSNVLRGMG-GAFVLVLYDEIKKFV-------------- 297 TAAC RKKASPNTIDEQT 415bAAC -------------

Thuswaldner et al.: Figure 1

A

B C

D

H1 H2 H3 H4 H5 H6

NC

stroma

lumen

lumenal view

lumen

N

C

stroma

*****

H1, H3, H5 1 SIVPKDAALFFAGAFAGAAAKSVTAPLDRIKLLMQTHGVRAGQQSAKKA 158 2 GQLSVLGRLG-AGACAGMTSTLITYPLDVLRLRLAVEPGYR-------- 250 3 YQQKTQSSLLTA-VVAAAIATGTCYPLDTIRRQMQLKGTPY-------- 340 PxDxxRRRMMM E K h12, h34, h56 H2, H4, H6 1 IGFIEAITLIGK-EEGIKGYWKGNLPQVIRIVPYSAVQLFAYETYKKLFRGKD 210 2 --TMSQVALNMLREEGVASFYNGLGPSLLSIAPYIAINFCVFDLVKKSLPEK- 300 3 KSVLDAFSGI-IAREGVVGLYRGFVPNALKSMPNSSIKLTTFDIVKKLIAASE 392 EGxxxaGKG D R

by guest on July 12, 2018http://w

ww

.jbc.org/D

ownloaded from

by guest on July 12, 2018http://w

ww

.jbc.org/D

ownloaded from

A

B

C T E 2 10 2 10 10 50 µg protein

Lhcb 2

TIC110

C T E 10 50 10 50 2 5 µg protein

Triton X-100 Ctrl 0.1 % 1 % NaCl S P S P S P S P

Thuswaldner et. al.: Figure 3

TAAC

70 55 40 35

25 15

10

C T E C T Epre-immune anti-TAAC

TAAC

by guest on July 12, 2018http://w

ww

.jbc.org/D

ownloaded from

by guest on July 12, 2018http://w

ww

.jbc.org/D

ownloaded from

Thuswaldner et al.: Figure 5

A

B

C

D

neg. pos. Cyt M M1 M2a M2b

TAAC

0

20

40

60

80

100

No add. ATP ADP ADP AMP GTP GDP -glucose

Upt

ake

of [32

P]-A

TP(%

of m

ax.)

ADP

ATP

ADPATP

0

20

40

60

Tran

spor

t of [

32P

] Nuc

leot

ide

(% o

f tot

al)

- ATP ADP ATP ADP- - - CCCP CCCP

Disruptedcells

Back exchange

Time (min)

Upt

ake

of [32

P]-N

ucle

otid

e(p

mol

mg-

1 pr

otei

n) ATP

ADP

neg. ATP

neg. ADP

by guest on July 12, 2018http://w

ww

.jbc.org/D

ownloaded from

0 15 30 45 15 30 45 min

0

30

60

90

120

150

1 5Time (min)

WTtaac

Thuswaldner et al.: Figure 6

A

B

C

Upt

ake

of [32

P]-A

TP

Lhcb2

TAAC

WT taac

ATP

GTP

36.7 71.5 100 29 45.8 77 %

WT taac

(pm

ol m

g-1 C

hl)

by guest on July 12, 2018http://w

ww

.jbc.org/D

ownloaded from

by guest on July 12, 2018http://w

ww

.jbc.org/D

ownloaded from

Thuswaldner et al.: Figure 8

by guest on July 12, 2018http://w

ww

.jbc.org/D

ownloaded from

Adamska, Bengt L. Persson and Cornelia SpeteaDer, Nathalie Leborgne-Castel, Arti Mishra, Francis Marty, Benoît Schoefs, Iwona

Sophie Thuswaldner, Jens O. Lagerstedt, Marc Rojas-Stütz, Karim Bouhidel, Christophefrom Arabidopsis

Identification, expression and functional analyses of a thylakoid ATP/ADP carrier

published online January 29, 2007J. Biol. Chem. 

  10.1074/jbc.M609130200Access the most updated version of this article at doi:

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

Supplemental material:

  http://www.jbc.org/content/suppl/2007/01/30/M609130200.DC1

by guest on July 12, 2018http://w

ww

.jbc.org/D

ownloaded from