Structure-Function Relationships of the C-Terminal End of the Saccharomyces cerevisiae ADP/ATP Carrier Isoform 2*

The adenine nucleotide carrier (Ancp) catalyzes the transport of ADP and ATP across the mitochondrial inner membrane, thus playing an essential role in the cellular energy metabolism. Two regions of Anc2p from Saccharomyces cerevisiae are specifically photolabeled using a photoactivable ADP derivative; they are the central matrix loop, m2, and the C-terminal end. To get more insights into the structure-function relationships of the C-terminal region during nucleotide transport, we have developed two independent approaches. In the first we have deleted the last eight amino acids of Anc2p (Anc2pΔCter) and demonstrated that the C-terminal end of Anc2p plays an essential role in yeast growth on a non-fermentable carbon source. This resulted from impaired nucleotide binding properties of the Anc2pΔCter variant in line with conversion of ADP binding sites from high to low affinity. In the second we probed the ligand-induced conformational changes of Anc2p C-terminal end (i) by assessing its accessibility to anti-C-terminal antibodies and (ii) by measuring intrinsic fluorescence changes of an Anc2p mutant containing only one tryptophan residue located at its C-terminal end (Anc2p3Y-u). We show that the C-terminal region is no further accessible to antibodies when Anc2p binds non-transportable analogues of ADP. Besides, Trp-316 fluorescence is highly increased upon ligand binding, suggesting large conformational changes. Taken together, our results highlight the involvement of the Anc2p C-terminal region in nucleotide recognition, binding, and transport.

The adenine nucleotide carrier (Ancp) is a nuclear-encoded protein located in the inner mitochondrial membrane that catalyzes the transmembrane exchange of ADP and ATP between the cytosolic and the matrix compartments (1). ADP/ATP transport is achieved by interconversion of Ancp between two conformational states that are fixed by the specific transport inhibitors carboxyatractyloside (CATR) 4 and bongkrekic acid (BA). These conformations, referred to as CATR and BA conformations, respectively, are characterized by different reactivities of well defined regions of the protein to chemical, enzymatic, and immunochemical reagents (1). Transition between the two conformations is very slow in the absence of transportable nucleotides but is markedly accelerated in the presence of ATP and ADP. Substrate-and inhibitor-dependent conformational changes have been observed not only with the membrane-bound ADP/ATP carrier but also with the detergent-isolated carrier (1). These results suggest that the nucleotide-induced conformational changes of the isolated Ancp are similar to those occurring in the membrane-bound carrier during ADP/ATP transport (1).
The beef Anc1p (BAnc1p) crystallized as the CATR-carrier complex is the only mitochondrial carrier for which high resolution structural data have been obtained (2). The 2.2-Å resolution model of the CATR conformation described six tilted transmembrane ␣-helices that form a wide cavity opening toward the intermembrane space. The CATR molecule is located in the cavity and strongly interacts with the peptidic chain of the carrier. At the bottom of the cavity a cationic cluster of arginines and lysines residues would contribute to the nucleotide translocation mechanism (2). Photolabeling of the ADP/ATP carrier substrate binding sites has been carried out with azido derivatives of ADP/ATP, and similar results were reported using different photoprobes and mitochondria from different sources (3,4). Two distinct peptidic regions of BAnc1p were specifically labeled by 2-azido-[␣-32 P]ADP (3). The first corresponded to a central domain (peptide Phe-153-Met-200) comprising residues from matrix loop m2 to helix H4. The second was located in the C-terminal region of the carrier (peptide Tyr-250 -Met-281) including residues from matrix loop m3 to helix H6. Two shortest distinct segments of the yeast mitochondrial ADP/ATP carrier were also labeled by 2-azido-3Ј-O-naphthoyl-[␤-32 P]ADP and identified (Fig. 1A), with peptide Ser-182-Arg-190 and peptide Ile-310 -Lys-317 belonging to the central part of the ADP/ATP carrier (matrix loop m2) and its C-terminal end, respectively (4). Plausible models in which the two distinct regions would be involved in organization of the nucleotide binding site(s) were proposed assuming the carrier is a dimer (3,4). However, it was recently shown that the yeast Ancp might be organized as a monomer in mitochondrial membranes and in detergent solution (5)(6)(7), suggesting that the two photolabeled regions would belong to the same monomer. Nevertheless, this hypothesis cannot accommodate the previous findings that one molecule of CATR or BA binds with high affinity to a carrier dimer, consistent with a multimeric organization of the carrier (8,9).
The experiments described in this paper were carried out with the Saccharomyces cerevisiae Anc2p isoform of the ADP/ATP carrier. We have investigated the functional role of Anc2p C-terminal end using two different approaches. The first one was based on expression and biochemical characterization of a mutant of Anc2p of which we deleted the last eight amino acids (Anc2p⌬Cter) (Fig. 1A). In the second approach we probed the ligand-induced conformational changes of the C-terminal end of Anc2p by ELISA and measurements of intrinsic fluorescence changes of an Anc2p variant containing only one tryptophan residue at its C-terminal end (Fig. 1A). Taken together, our results highlight for the first time the involvement of the Anc2p C-terminal region in nucleotide recognition, binding, and transport. Our results are discussed in terms of participation of the C-terminal region in the nucleotide transport process and in the putative oligomerization state of the carrier.
Construction of the Mutated anc2 Genes-Oligonucleotidedirected mutagenesis of the ANC2 gene was performed using the standard QuikChange site-directed mutagenesis kit from Stratagene to delete the 3Ј end of the gene or to introduce a Trp residue at position 316. The pMD101 plasmid (TRP1/CEN6/ ARSH4)-containing ANC2 gene, coding for Anc2p, and its 3Јand 5Ј-flanking regions (between PstI and SalI restriction sites) was derived from pRS314 and was previously described (17). Deletion of the last eight codons (from position ϩ931 to ϩ954, position ϩ1 corresponding to A from the ATG codon) corresponding to the C-terminal end of Anc2p was obtained by PCR using the pMD101 plasmid as a template and the primers 5Ј-CAATGTACGACCAACTGCAAATGTAAGTCTAATC-TGGCTTGATTC-3Ј and 5Ј-GAATCAAGCCAGATTAGAC-TTACATTTGCAGTTGGTCGTACATTG-3Ј. The corresponding plasmid obtained was called pMD-ANC2⌬Cter. Two mutations at position ϩ950 and ϩ951 (Phe codon replaced with Trp codon) were introduced by PCR using the recombinant plasmid pMDANC2--⌬W (derived from pMD101) coding for a Trp-less variant of Anc2p (17) as a template and the primers 5Ј-GATCTTGTTTGGTAAGAAGTGGAAATAAGACT-AATCTGGCT-3Ј and 5Ј-AGCCAGATTAGTCTTATTTCC-ACTTCTTACCAAACAAGATC-3Ј. The resulting plasmid was named pMDANC2-3Y-W316. ANC2 open reading frame (ORF), mutated or not, were controlled by DNA sequencing (Genome Express) of plasmids pMD101 (wild type) pMD-ANC2-⌬Cter, pMDANC2-⌬W, and pMDANC2-3Y-W316 (mutated ORF), which were then used to transform the JL1⌬2⌬3u Ϫ strain according to the LiCl procedure (18). Transformants were selected, and yeast strains were named K317 (wild-type ANC2), M309, K317-⌬W, and W316, respectively.
Isolation of Mitochondria-Yeast cells grown on YPL were harvested in the late log phase (A 600 nm near 5). Mitochondria were prepared as previously described (19).
Binding Assays-Specific binding of [ 14 C]ADP or [ 3 H]ATR to the ADP/ATP carrier in mitochondria was assessed by the CATR chase procedures (20). Incubation was 30 min for [ 3 H]ATR to 2 h for [ 14 C]ADP at 4°C. The K d values and the total number of ligand-binding sites were calculated from Scatchard plot of the data. When Scatchard plot were curvilinear, they were arbitrarily fitted using Kaleidagraph 4.0 according to a model with two classes of an equal number of interacting binding sites (21) with high affinity (K d1 ) and low affinity constants (K d2 ). The number of bound ligands (N B ) is related to the total number of sites (S), N F (the number of free ligands), K d1 , and K d2 by the Adair equation, ADP/ATP Transport Assay in Isolated Mitochondria-ADP/ ATP transport was measured by means of a luminescence assay as described in Passarella et al. (22). Freshly prepared mitochondria were incubated at 25°C in 10 mM Tris-HCl, pH 7.4, 10 mM KH 2 PO 4 , 0.6 M mannitol, 0.1 mM EGTA, 2 mM MgCl 2 , 10 M Ap5A (an inhibitor of adenylate kinase), and 1 mM ␣-ketoglutarate (as respiratory substrate) in the presence of 0.1% (w/v) luciferin and 0.1% (w/v) luciferase (19).
Isolation of the ADP/ATP Carriers-The ADP/ATP carrier protein from yeast mitochondria was isolated by chromatography on hydroxylapatite (Bio-Rad) following the method described in Brandolin et al. (23). The final carrier protein concentration in the purified fraction ranged between 0.03 and 0.05 mg/ml.
Fluorescence Assays-Fluorescence assays were carried out with a high sensitivity spectrofluorometer (Biologic, Grenoble, France). Samples were introduced in a 1 ϫ 1-cm fluorescence quartz cuvette inserted in a temperature-controlled cell holder with continuous stirring. Routinely, the carrier preparation in 0.5 ml was diluted with 1.5 ml of 136 mM glycerol. Reagents were injected with Hamilton automatic syringes in small volumes (2-10 l). Fluorescence of the tryptophan residues was excited at 300 nm (1-nm band-pass) with a 150-watt xenon lamp. The emitted light was measured at right angles through a 0-54 Corning filter, the band-pass being centered at 297 nm. The temperature was set at 10°C.

SDS-PAGE and Western
Blotting-The protein concentrations were determined using a BCA protein assay kit (Sigma) and bovine serum albumin as a standard. For SDS-PAGE, samples were prepared as described in (19). Antibodies were used at the following dilutions: anti-C-terminal peptide of Anc2p, 1/3000; anti-SDS-treated Anc2p, 1/5000; anti-SDS-treated VDAC, 1/2000. Immunodetection was performed using horseradish peroxidase-coupled protein A and the ECL-enhanced chemiluminescence system (Amersham Biosciences).
Immunoreactivity Assays-The ability of anti-C-terminal antiserum to react with the membrane-bound carrier incubated with or without ligands was tested by ELISA, as previously described (24). Freeze-thawed mitochondria were suspended in 10 mM MOPS-NaOH, pH 6.5, containing 0.12 M KCl, 1 mM EDTA, 1 mM phenylmethylsulfonyl fluoride, 1 mM diisopropyl fluorophosphate, and Complete mini-EDTA-free antiproteases (according to supplier instructions). Because of the lower affinity of the carrier for nucleotides as compared with that for inhibitors, N-ADP and ADP were present at 50 or 100 M final concentration, respectively, during all steps of the assay.

Deletion of the Last Eight Amino Acids of Anc2p Severely Impairs Growth of Yeast Cells on Non-fermentable Carbon
Source-Photolabeling of Anc2p in mitochondria with 2-azido-3Ј-O-naphthoyl-[␤-32 P]ADP resulted in the identification of two specifically photolabeled sites located in the central matrix loop m2 and in the C-terminal end of the carrier, respectively (4). To assess the functional role of the C-terminal region, it was deleted from Anc2p by a genetic approach. The resulting carrier protein (ANC2⌬Cter) is 309-amino acids long and lacks its last eight amino acids at the C-terminal end. It has a predicted molecular weight of 33.3 kDa and is referred to as Anc2p⌬Cter in the following. The construct was cloned into the yeast pRS314 centromeric vector under the control of the native promoter and terminator of ANC2 (17). The resulting plasmid, pMD-ANC2-⌬Cter, and the control plasmid, pMD101 (17), were transformed in JL1⌬2⌬3u Ϫ strain in which the three ANC genes were inactivated (25) to give the resulting strains K317 and M309, respectively.
The ADP/ATP transport activity of Anc2p⌬Cter and Anc2p expressed from the two strains was first assessed by their ability to rescue growth of JL-1⌬2⌬3u Ϫ on a non-fermentable carbon source (YPL) (Fig. 2). All of the strains developed similarly when plated onto a solid synthetic glucose medium lacking Trp ( Fig.  2A). However, the M309 strain grew much more slowly on solid medium than K317 with lactate as the sole carbon source ( Fig.  2A). This behavior was confirmed by examining cell growth at 28°C in liquid YPL medium (Fig. 2B). Indeed, M309, exhibited a doubling time of 16 h with A 600 nm ϭ 6 for the saturation phase, whereas that of K317 was 2.5 h with A 600 nm ϭ 11 for the saturation phase (Table 1). We, thus, concluded that the poor growth of M309 on the non-fermentable carbon source reflects a severe impairment of the carrier functioning.
Deletion of Anc2p C-terminal End Does Not Modify the Carrier Content in Yeast Mitochondria-The ADP/ATP carrier contents of K317 and M309 mitochondrial lysates were evaluated by Western blot analyses after SDS-PAGE to check out if impairment of M309 growth could be related to a lower Anc2p mitochondria content. VDAC was used as a control for the amount of mitochondrial proteins loaded on the gel. Both strains exhibited similar Anc2p content (Fig. 3A, lanes 1 and 2). As one could expect, Anc2p⌬Cter was not immunodetected by the anti-C-terminal peptide antiserum (Fig. 3A, lane 2). A more accurate determination of the carrier content was performed by binding experiments of [ 3 H]ATR, a specific inhibitor of Ancp. As shown in Fig. 3B, binding of [ 3 H]ATR increased as a function of added ligand and reached a plateau, with no significant difference between Anc2p and Anc2p⌬Cter. K d values and maximum number of binding sites were calculated from the Scatchard plots (not shown) and were similar for both the variant and the wild-type carriers (Table 1). Similar results were obtained from binding experiments of [ 3 H]BA (data not shown). Therefore, the C-terminal end is not involved in the inhibitor binding sites. Consequently, the poor M309 growth on YPL may be explained by a lower ADP/ATP exchange activity of Anc2p⌬Cter.
Anc2p-⌬Cter Mitochondria Are Deficient in ADP/ATP Transport Activity-Measurements of the ADP/ATP transport of Anc2p and Anc2p⌬Cter were carried out on mitochondria isolated from the corresponding strains by the luciferase/luciferin system (cf."Experimental Procedures"). As shown in Table 1, Anc2p and Anc2p⌬Cter exhibit very similar apparent V max and k cat values. In contrast, the apparent K m of external free ADP for Anc2p⌬Cter was ϳ10 times higher than for Anc2p (Table 1). This result suggests that the C-terminal end of Anc2p could be involved in ADP binding step as already hypothesized (4).
C-terminal End Deletion Dramatically Modifies the Affinity of Anc2p for ADP-Specific binding of [ 14 C]ADP to Anc2p and Anc2p⌬Cter in mitochondria was assessed by a differential method based on the use of CATR (20). Scatchard representation of data exhibits a curvilinear plot for Anc2p, indicating the presence of high and low affinity binding sites (Fig. 4A). The total number of binding sites for ADP (S ADP ) determined by extrapolation of the Scatchard plot was about 600 pmol/mg of protein, a value similar from the number of ATR binding sites (500 pmol/mg of protein, Table 1). The curvilinear Scatchard plot was fitted with a two interactive binding site model according to the Adair equation (cf."Experimental Procedures"), leading for Anc2p to the ADP binding parameters summarized in Table 1: S ADP ϭ 631 pmol/mg of protein, K d1 ϭ 0.6 M, and K d2 ϭ 32 M. Interestingly, a linear Scatchard plot was obtained for Anc2p⌬Cter, leading to S ADP ϭ 539 pmol/mg of protein and a different apparent K d value, K d3 ϭ 21 M (Table 1). As in the case of Anc2p, the S ADP value was in the same range as the ATR maximum number of binding sites. This indicates that the C-terminal end of the carrier does not constitute a whole binding site but participates in it. Unlike inhibitors and ADP, nontransportable nucleotide analogues modify C-terminal end reactivity toward anti-C-terminal antibodies.
There are evidences that the accessibilities of different regions of Ancp depend on its conformational state (19, 24, 26 -32). However, so far no data are available about the Anc2p C-terminal end. Therefore, in a first step we assessed Anc2p C-terminal immunoreactivity in the presence of various ligands by ELISA. The antibodies were directed against a peptide corresponding to the last 14 Anc2p amino acids and were used with freeze-thawed mitochondria from strain K317 (Fig. 1). As a control, we examined the specificity of the antibodies toward mitochondria from strain M309. The results presented in Fig.  5A show that the anti-C-terminal antibodies react specifically with Anc2p in a protein concentration-dependent manner without any significant difference between mitochondria incu- were spotted onto YNB Glc W Ϫ (fermentable carbon source) or YPL (non-fermentable) solid media. Plates were incubated under aerobiosis at 28°C for 3 days (YNB Glc W Ϫ and YPL media) or 6 days (only YPL medium). The JL1⌬2⌬3u Ϫ strain transformed by pR314 vector, carrying only the promoter and terminator sequences of the ANC2 gene, was used as a control (pRS314⌬ANC2). B, growth curves of strains K317 and M309 in liquid YPL-rich medium monitored by absorbance at 600 nm (A 600 nm ) for 45 and 150 h, respectively. The cultures were inoculated with cells previously grown on YNB Glc W Ϫ medium to obtain an initial A 600 nm ϭ 0.1. Values are the means of at least three independent experiments. bated with or without inhibitors. Thus, accessibility of Anc2p C-terminal end to the antibodies does not depend on the CATR-or BA-induced conformations.
Similar experiments were performed in the presence of either the natural substrate ADP or the nontransportable N-ADP (12). Both ligands were used at saturating concentrations. ADP did not modify C-terminal end reactivity (Fig. 5B), whereas an inhibition of ϳ60 -70% was observed in the presence of N-ADP in a CATR-sensitive manner (Fig. 5B). We have checked that this effect was not merely induced by the naphthoyl part of the nucleotide analogue. Indeed, no inhibition of the C-terminal reactivity toward antibodies was observed with 3Ј-Onaphthoyl-AMP tested at a concentration twice higher than N-ADP (data not shown). Moreover, similar inhibitions were obtained with other non-transportable nucleotide analogues such as 2Ј(3Ј)-O-(2,3,6-trinitrophenyl)-ADP or 8-Br-ADP (data not shown). Thus, the formation of a stable N-ADP-carrier complex would restrict access of the C-terminal epitope to the antibodies. This hypothesis is in line with the specific photolabeling of Anc2p C-terminal end by an N-ADP analogue, the 2-azido-3Ј-O-naphthoyl-[␤-32 P]ADP (4). We conclude that the N-ADP/carrier and the ADP-carrier complexes display distinct topographies at the C-terminal end.
Trp-316 Can Probe the Ligand-induced Conformational Changes of the C-terminal Region of Anc2p-The Anc2p isoform contains three Trp residues located at position 87, 126, and 235 in the polypeptide chain. In a previous work we replaced all the Trp residues with Tyr residues by site-directed mutagenesis (17). The Anc2p⌬W mutant was able to bind inhibitors and to catalyze the ADP/ATP exchange with properties similar to those of Anc2p (17). To probe the ligand-induced conformational changes of Anc2p C-terminal end by measuring intrinsic fluorescence changes, we substituted Phe-316 by a Trp residue in Anc2p⌬W (Fig. 1A). This was achieved by site-directed mutagenesis using the pMDANC2-3Y plasmid. The strain transformed with the resulting plasmid was named W316 (cf."Experimental Procedures"). The K317-⌬W and W316 strains grew on YPL, indicating that the mutated genes encoded active ADP/ATP carriers (not shown). Both strains displayed similar growth characteristics with a doubling time of ϳ3 h and A 600 ϭ 11 at the growth plateau. Moreover, the amounts of Anc2p⌬W and Anc2p3Y-W316 in the corresponding mitochondria were similar, as evaluated by Western blot analyses of mitochondrial lysates and ATR binding experiments (not shown). Thus, replacement of Phe-316 with a tryptophan residue did not impair the biochemical properties of Anc2p⌬W.
Then we examined the intrinsic fluorescence properties of the Anc2p3Y-W316 carrier purified as previously described    Fig. 6 shows the time-course of the ATP-induced fluorescence change of Anc2p3Y-W316, measured at 10°C. The addition of ATP at a subsaturating concentration (5 M) resulted in a very rapid increase of fluorescence (⌬F/F ϭ 1.9%). Upon the addition of CATR, the fluorescence decreased to a level lower than the initial level of fluorescence of the unliganded carrier (⌬F/F ϭ Ϫ0.9%). As illustrated in Fig. 6, the fluorescence signal increased with the concentration of added ATP to reach a plateau at about 8 -10 M. ADP induced a fluorescence increase similar to that obtained with ATP (not shown). When added before ATP or ADP, CATR used at saturating concentrations induced approximately a 1% decrease of the fluorescence of the carrier (⌬F/F ϭ Ϫ0.8%), and it totally prevented the ATP-induced fluorescence increase (Fig. 6). The effect of BA, which modified intrinsic fluorescence only in the presence of ATP (or ADP), led to a large increase in fluorescence (⌬F/F ϭ 6.8%) (Fig.  6). Taken together, these results show that Trp-316 probes conformational changes of the carrier induced by ligand-binding (ADP, ATP, CATR, and BA). It remains to determine whether the fluorescence variations result from either a direct and local modification of the molecular environment of the peptide chain at the C-terminal end or a long distance reflection of global conformational changes undergone by the protein upon ligand binding and indirectly affecting this part of Anc2p.

DISCUSSION
Anc2p C-terminal region was identified as a putative or participating to a nucleotide binding site by previous photoaffinity labeling experiments (4). Unfortunately no structural data are available so far about this region since the last five amino acids of the BAnc1p are not resolved in the three-dimension structure (2) (Fig. 1B). To ascertain the involvement of Anc2p C-terminal end in the carrier functioning, possibly in the early nucleotide binding steps, we first considered deleting this part of the carrier by a genetic approach.
As a first result, yeast growth was severely impaired on nonfermentable carbon source, indicating a impaired Anc2p function. It was determined that Anc2p⌬Cter was properly imported into mitochondria and that the ATR binding properties were not modified as compared with that of the wild-type Anc2p. This suggests that although shortened, the Anc2p⌬Cter variant is properly folded in the mitochondrial inner membrane. We then carried out ADP/ATP transport assays with isolated mitochondria. Our results show that Anc2p⌬Cter exhibits V max and k cat values very similar to that of Anc2p, but the apparent K m for external free ADP is about 10 times higher for Anc2p⌬Cter than for Anc2p. This is the first time a role in the nucleotide transport function can be clearly assigned to the C-terminal region of Anc2p. However, considering an unchanged V max and a higher K m ADP , we can hypothesize that the Ile-310 -Lys-317 peptide would not participate in the translocation process per se but, rather, in the productive nucleotide binding step. This is consistent with the photolabeling experiment results reported by Dianoux et al. (4). Two regions were photolabeled, indicating either participation of two different regions of the carrier in a nucleotide binding site or the existence of two different binding sites.
It should be pointed out that, as in the case of the yeast Anc2p, two different regions of the Banc1p were specifically labeled with photoactivable nucleotide analogue (3). The first one corresponds roughly to the C-terminal half of the m2 matrix loop and comprises Lys-162 and -165 (Lys-179 and -182 for Anc2p). In contrast, the second one does not correspond to the protein C-terminal end but to the end of matrix loop m3 and the N-terminal region of helix H6, Val-254 -Met-281 (Phe-269 and Val-296 for Anc2p). This may result from the fact that the C-terminal region of BAnc1p is shorter than that of the yeast protein (Fig. 1B), enlightening some differences of the bovine and the yeast ADP/ATP carriers folding (3,4).
Involvement of the C-terminal region of Anc2p in ADP binding was confirmed from binding experiments of external ADP to isolated mitochondria. As observed previously for the mem- brane-bound BAnc1p (21), the data obtained with Anc2p are analyzed with a two-binding site model, with K d values similar to that measured for BAnc1p (21). Interestingly, the Anc2p⌬Cter mutant still binds ADP, with the same number of maximum binding sites as Anc2p, but analysis of binding data show a single class of binding sites with a K d value corresponding to the lowest ADP binding site affinity. This suggests that the C-terminal end of Anc2p would not constitute per se a nucleotide binding site but would contribute to set up the high affinity binding site for an efficient nucleotide translocation. Indeed, when the C-terminal region is deleted, the K m value is increased in such a way that nucleotide transport is dramatically decreased. Two cooperative binding site classes were formerly hypothesized for the BAnc1p (21). However, it is interesting to note that rectilinear Scatchard plots were obtained for the binding of non-transportable nucleotides, such as N-ADP and 8-bromo-ADP, to BAnc1p (21) or N-ADP to Anc2p (17), suggesting that the observed cooperativity is linked to the "transportable" property of the ligand.
Several regions of the ADP/ATP carrier undergo CATR-or BA-induced conformational changes as evidenced by ELISA (24), limited proteolysis (19), intrinsic fluorescence changes (23), and cysteine labeling (19, 26 -32). The BAnc1p N-terminal end is masked upon BA binding and exposed to anti-Nterminal antibodies upon CATR binding (24). Unfortunately, anti-C-terminal antibodies did not react with Banc1p carrier (24). In the case of the yeast Anc2p, the C-terminal end is freely accessible to antibodies whatever the inhibitor present as well as in the presence of ADP. Surprisingly, it is masked only in the presence of a non-transportable nucleotide analogue, N-ADP, and this effect is reversed in the presence of CATR. Therefore, we can postulate that binding of transportable nucleotides to Anc2p triggers a conformational change for the carrier to achieve a transport-competent conformation, in which the C-terminal end is freely accessible to the anti-C-terminal peptide antibodies. Also, intrinsic fluorescence measurements are a powerful tool to follow conformational changes and are more precise than ELISA, which is an all-or-nothing technique. We have mutagenized Phe at position 316 into Trp in an Anc2p variant that contained no tryptophan residues (17). The observed intrinsic fluorescence changes of Anc2p3Y-W316 provided evidence that Anc2p C-terminal end undergoes environmental modifications upon interactions with ADP or ATP (enhanced in the presence of BA) and CATR. The effect of CATR is surprising given the fact that the CATR binding site is located in the central cavity of BAnc1p (2) probably pretty far from the C-terminal end. It could also be the case for BA, but nothing is known about the structural determinants of its binding site. These effects may merely reflect long distance changes upon ligand binding. Thus, our data clearly establish that Anc2p C-terminal end is involved in nucleotide binding path but not in the translocation step.
As a result two different regions of the carrier participate in ADP/ATP binding. This could be related to the negative cooperativity between two classes of nucleotide binding sites hypothesized first for BAnc1p (21). Such a hypothesis raises the widely debated problem of the functional oligomeric unit of mitochondrial carriers.
In summary, we have demonstrated in this paper that the last eight amino acids of the Anc2p C-terminal region are not critical for the import of the carrier protein within the mitochondrial inner membrane but, in contrast, play a prominent functional role for the interactions between carrier and nucleotides. We suggest a role of Anc2p C-terminal end during the nucleotide binding step of the transport mechanism that would be the setting-up of a high affinity nucleotide binding site.