Identification of the mitochondrial GTP/GDP transporter in Saccharomyces cerevisiae.

The genome of Saccharomyces cerevisiae contains 35 members of a family of transport proteins that, with a single exception, are found in the inner membranes of mitochondria. The transport functions of the 16 biochemically identified mitochondrial carriers are concerned with shuttling substrates, biosynthetic intermediates, and cofactors across the inner membrane. Here the identification and functional characterization of the mitochondrial GTP/GDP carrier (Ggc1p) is described. The ggc1 gene was overexpressed in bacteria. The purified protein was reconstituted into liposomes, and its transport properties and kinetic parameters were characterized. It transported GTP and GDP and, to a lesser extent, the corresponding deoxynucleotides and the structurally related ITP and IDP by a counter-exchange mechanism. Transport was saturable with an apparent K(m) of 1 microm for GTP and 5 microm for GDP. It was strongly inhibited by pyridoxal 5'-phosphate, bathophenanthroline, tannic acid, and bromcresol purple but little affected by the inhibitors of the ADP/ATP carrier carboxyatractyloside and bongkrekate. Furthermore, in contrast to the ADP/ATP carrier, the Ggc1p-mediated GTP/GDP heteroexchange is H(+)-compensated and thus electroneutral. Cells lacking the ggc1 gene had reduced levels of GTP and increased levels of GDP in their mitochondria. Furthermore, the knock-out of ggc1 results in lack of growth on nonfermentable carbon sources and complete loss of mitochondrial DNA. The physiological role of Ggc1p in S. cerevisiae is probably to transport GTP into mitochondria, where it is required for important processes such as nucleic acid and protein synthesis, in exchange for intramitochondrially generated GDP.

In the mitochondrial matrix, GTP is required as an energy source for protein synthesis; as a substrate for the synthesis of tRNA, mRNA, rRNA, and RNA primers; and as a phosphate group donor for the activity of GTP-AMP phosphotransferase (1) and G proteins (2,3). In several organisms, GTP is synthesized in the mitochondria by succinyl-CoA ligase, which catalyzes the conversion of succinyl-CoA to succinate with the generation of GTP, and by nucleoside diphosphate kinase, which catalyzes the transfer of the ␥ phosphate from ATP to a nucleoside diphosphate to yield nucleotide triphosphates. In Saccharomyces cerevisiae, however, succinyl-CoA ligase produces ATP instead of GTP (4), and the mitochondrial nucleoside diphosphate kinase is localized in the intermembrane space and absent in the matrix (5). These observations imply that in S. cerevisiae GTP has to be imported into the mitochondria probably via a carrier system embedded in the inner mitochondrial membrane.
Despite the importance of GTP in mitochondrial metabolism, the transport of guanine nucleotides has not been characterized in yeast mitochondria, nor has any mitochondrial protein responsible for this transport been identified. There are only two indirect observations that suggest that GTP is transported across the inner mitochondrial membrane of S. cerevisiae. First, mitochondrial protein synthesis is stimulated by the addition of external GTP (6), and second, on incubating yeast mitochondria with labeled GTP, the amount of radioactivity associated with mitochondria is time-dependent, inhibited by GDP and insensitive to carboxyatractyloside (7).
The inner membranes of mitochondria contain a family of proteins that transport various substrates and products into and out of the matrix (for a review see Ref. 8). These proteins are characterized by three tandem sequence repeats, each being approximately 100 amino acids in length and folded into two transmembrane ␣-helices joined by an extensive hydrophilic loop. The nuclear genome of S. cerevisiae encodes 35 members of this family. The functions of many members are unknown because the substrates transported have not yet been discovered. One of these, Ggc1p, i.e. the GTP/GDP carrier (encoded by YDL198c and previously known as Yhm1p) has been shown to be localized to mitochondria and to be a multicopy suppressor (by an unknown mechanism) of the ability of the abf2 null mutant to grow at 37°C on glycerol (9).
Here we report the identification and functional characterization of Ggc1p. This protein has been overexpressed in Escherichia coli, reconstituted into phospholipid vesicles, and identified from its transport properties as a carrier for GTP and GDP. Ggc1p operates in yeast mitochondria with transport properties similar to those observed with the recombinant protein. In addition, ggc1⌬ cells exhibit lower levels of GTP and increased levels of GDP in their mitochondria, are unable to grow on nonfermentable substrates, and have lost their mtDNA. The physiological role of Ggc1p in S. cerevisiae is probably to catalyze the exchange between external GTP and internal GDP to satisfy the need for GTP in the mitochondrial matrix, where this compound cannot be synthesized. This report presents the first information on the molecular properties of the mitochondrial GTP/GDP carrier and a definitive identification of its gene in S. cerevisiae. * This work was supported by grants from Ministero dell'Università e della Ricerca Scientifica e Tecnologica, University's Local Funds, the Italian National Research Council, National Research Council-Ministero dell'Università e della Ricerca Scientifica e Tecnologica project "Functional Genomics," the Centro di Eccellenza Geni in campo Biosanitario e Agroalimentare, and by the European Social Fund. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Bacterial Expression and Purification of Ggc1p-The coding sequence of ggc1 (open reading frame YDL198c) was amplified from S. cerevisiae genomic DNA by PCR. The oligonucleotide primers were synthesized corresponding to the extremities of the coding sequence, with additional BamHI and HindIII sites. The product was cloned into the pMW7 expression vector, and the construct was transformed into E. coli DH5␣ cells. Transformants were selected on 2ϫ TY plates containing ampicillin (100 g/ml) and screened by direct colony PCR and restriction digestion of plasmids. The overproduction of Ggc1p as inclusion bodies in the cytosol of E. coli was accomplished as described before (12), except that the host cells were E. coli C0214(DE3) (13,14). Control cultures with the empty vector were processed in parallel. Inclusion bodies were isolated, and Ggc1p was purified by centrifugation and washing steps as described previously (13,15). The proteins were separated by SDS-PAGE in 17.5% gels and either stained with Coomassie Blue dye or transferred to nitrocellulose membranes for immunodetection with a rabbit antiserum raised against bacterially expressed Ggc1p. The N terminus was sequenced, and the yield of purified Ggc1p was estimated by laser densitometry of stained samples (11).
Reconstitution into Liposomes and Transport Assays-The recombinant protein in sarkosyl was reconstituted into liposomes in the presence of substrates, as described before (16). External substrate was removed from proteoliposomes on Sephadex G-75 columns, pre-equilibrated with 50 mM NaCl and 10 mM PIPES-NaOH 1 at pH 7.0 (buffer A) or 1 mM PIPES-NaOH at pH 7.0 in the experiments reported in Table  II. Transport at 25°C was started by adding [8-3 H]GTP (Amersham Biosciences), [8,5Ј-3 H]GDP, or [␣-33 P]dGTP (PerkinElmer Life Sciences) to proteoliposomes and terminated by the addition of 15 mM bathophenanthroline and 30 mM pyridoxal 5Ј-phosphate (the "inhibitor stop" method (16)). In controls, the inhibitors were added at the beginning together with the radioactive substrate. All of the transport measurements were carried out in the presence of 10 mM PIPES at pH 7.0 in the internal and external compartments, except in the experiments reported in Table II, where 1 mM PIPES at pH 7.0 was used. The external substrate was removed, and the radioactivity in the liposomes was measured (16). The experimental values were corrected by subtracting control values. The initial transport rate was calculated from the radioactivity taken up by proteoliposomes after 20 s (in the initial linear range of substrate uptake). For efflux measurements, proteoliposomes containing 1 mM substrate were labeled with carrier free [8-3 H]GTP by carrier-mediated exchange equilibration (16). After 40 min, the external radioactivity was removed by passing the proteoliposomes through Sephadex G-75. Efflux was started by adding unlabeled external substrate or buffer A alone and terminated by adding the inhibitors indicated above.
Other Methods-GTP and GDP were determined in mitochondrial extracts by enzymatic assays (17). K ϩ diffusion potentials were generated by adding valinomycin (1.5 g/mg phospholipid) to proteoliposomes in the presence of KCl gradients. For the formation of an artificial ⌬pH (acidic outside), nigericin (50 ng/mg phospholipid) was added to proteoliposomes in the presence of an inwardly directed potassium gradient. The membrane potential of isolated mitochondria was assessed by recording the fluorescence changes of the voltage-sensitive dye 3,3Ј-dipropylthiadicarbocyanine iodide DiSC (3, 5) (Molecular Probes) as previously described (18). For DNA detection, the BY4741 and the isogenic ggc1⌬ strains were fixed with formaldehyde following growth on galactose to an A 600 of 2.0. Then the DNA was stained by incubation with 1 g/ml DAPI at 4°C overnight. DAPI fluorescence was detected using an inverted Zeiss Axiovert 200 epifluorescence microscope equipped with a CoolSNAP HQ CCD camera (Roper Scientific, Trenton, NJ) and the Metamorph software (Universal Imaging Corporation, Downington, PA).

RESULTS
Bacterial Expression of Ggc1p-Ggc1p was expressed at high levels in E. coli C0214(DE3) (Fig. 1, lane 4). It accumulated as inclusion bodies and was purified by centrifugation and washing (Fig. 1, lane 5). The apparent molecular mass of the recombinant protein was 33.5 kDa (the calculated value with initiator methionine was 33,215 Da). The identity of the purified protein was confirmed by N-terminal sequencing. Approximately 80 -90 mg of purified protein were obtained per liter of culture. The protein was not detected in bacteria harvested immediately before induction of expression (Fig. 1, lane 2) nor in cells harvested after induction but lacking the coding sequence in the expression vector ( Fig. 1, lane 3).
Functional Characterization of Recombinant Ggc1p-Ggc1p was reconstituted into liposomes, and its transport activities for a variety of potential substrates were tested in homoexchange experiments (i.e. with the same substrate inside and outside). Using external and internal substrate concentrations of 1 and 10 mM, respectively, the reconstituted protein catalyzed an active [8-3 H]GTP/GTP exchange, inhibitable by a mixture of bathophenanthroline and pyridoxal 5Ј-phosphate. It did not catalyze homoexchanges for phosphate, ATP, ADP, AMP, pyruvate, malate, oxoglutarate, citrate, glutamate, aspartate, proline, histidine, lysine, arginine, serine, threonine, tryptophan, glutathione, carnitine, and choline. No [8-3 H]GTP/GTP exchange activity was observed with Ggc1p that had been boiled before incorporation into liposomes nor by reconstitution of sarkosyl-solubilized material from bacterial cells either lacking the expression vector for Ggc1p or harvested immediately before the induction of expression.
The substrate specificity of Ggc1p was examined in greater detail by measuring the uptake of [␣-33 P]dGTP into proteoliposomes that had been preloaded with various potential substrates ( Fig. 2A). High rates of [␣-33 P]dGTP uptake into proteoliposomes were observed with internal GDP, GTP, dGDP, dGTP, IDP, and ITP. Much smaller activities were found with internal guanosine 5Ј-tetraphosphate and (deoxy)nucleoside diand triphosphates of U and T. No activity was detected with internal (d)NDP and (d)NTP of A and C, with GMP and all the other (deoxy)nucleoside monophosphates tested, with NaCl and (not shown) with adenosine, adenine, NMN, FMN, thiamine pyrophosphate, NADH, NADPH, FAD, phosphate, pyrophosphate, and malate.
The [8-3 H]GTP/GTP exchange reaction catalyzed by reconstituted Ggc1p was inhibited strongly by pyridoxal 5Ј-phosphate and bathophenanthroline (inhibitors of many mitochondrial carriers), tannic acid, and bromcresol purple (inhibitors of the mitochondrial glutamate carrier) and only partially by the sulfydryl reagents mercuric chloride and mersalyl and by 1,2,3-benzenetricarboxylate (inhibitor of the mitochondrial citrate carrier) (Fig.  3). Carboxyatractyloside and bongkrekate (powerful inhibitors of the ADP/ATP carrier) had little effect at much higher concentrations than those that completely inhibit the ADP/ATP carrier. Very little inhibition was observed with p-hydroxymercuribenzoate, p-hydroxymercuribenzene sulfonate, N-ethylmaleimide, butylmalonate, phenylsuccinate, and ␣-cyano-4-hydroxycinnamate (inhibitors of other mitochondrial carriers).
Kinetic Characteristics of Recombinant Ggc1p-In Fig. 4, the kinetics are compared for the uptake of 0.5 mM [8-3 H]GTP into proteoliposomes either in the presence or in the absence of internal 5 mM GTP. The uptake of GTP by exchange followed a first order kinetics (rate constant, 22.6 min Ϫ1 ; initial rate, 1.9 mmol/min/g protein), isotopic equilibrium being approached exponentially (Fig. 4A). In contrast, no [8-3 H]GTP uptake was observed without internal substrate, suggesting that Ggc1p does not catalyze a unidirectional transport (uniport) of GTP. The uniport mode of transport was further investigated by measuring the efflux of [8-3 H]GTP from prelabeled active proteoliposomes because it provides a more convenient assay for unidirectional transport (16). In the absence of external substrate no efflux was observed even after incubation for 20 min (Fig. 4B). However, upon the addition of external GTP or GDP, an extensive efflux of radioactivity occurred, and this efflux  (Table I) because they increased the apparent K m without changing the V max (not shown). These results confirm that GTP is the highest affinity external substrate (K i , 0.9 M). The K i values of all of the NTPs are lower than those of their corresponding NDPs. Furthermore, the affinity of Ggc1p for GTP and GDP is approximately 1 order of magnitude higher than for dGTP and dGDP and approximately 2 orders of magnitude higher than for ITP and IDP.
Influence of Membrane Potential and pH Gradient on the Ggc1p-mediated GTP/GDP Exchange-In view of the different charges carried by GTP and GDP at physiological pH levels, we investigated the influence of the membrane potential on the [8,5Ј-3 H]GDP/GTP or [8-3 H]GTP/GDP heteroexchanges catalyzed by the recombinant Ggc1p. A K ϩ diffusion potential was generated across the proteoliposomal membranes with valinomycin/KCl (calculated value was approximately 100 mV, positive inside) ( Table II). The rates of the GDP out /GTP in and GTP out /GDP in heteroexchanges as well as of the GDP/GDP and GTP/GTP homoexchanges were unaffected by valinomycin in the presence or absence of the K ϩ gradient. In contrast, the aspartate out /glutamate in exchange, mediated by the recombinant and reconstituted aspartate/glutamate carrier, which is known to catalyze an electrophoretic exchange between aspartate Ϫ and glutamate Ϫ ϩ H ϩ (13,19), was stimulated by valinomycin in the presence of a K ϩ gradient of 1/50 (mM/mM, in/out) (not shown). These results indicate that the GTP/GDP heteroexchange catalyzed by Ggc1p is not electrophoretic. We therefore became interested in the question as to whether the charge imbalance of the GTP/GDP heteroexchange is compensated by proton movement in the same direction as GTP. A pH difference across the liposomal membranes (basic inside the vesicles) was created by the addition of the K ϩ /H ϩ exchanger nigericin to proteoliposomes in the presence of a potassium gradient of 1/50 (mM/mM, in/out) (Table II). Under these conditions the uptake of [8,5Ј-3 H]GDP in exchange for internal GTP decreased, and the uptake of [8-3 H]GTP in exchange for internal GDP increased, whereas no effect on the GDP/GDP and GTP/GTP homoexchanges was observed. Therefore, the GTP/ GDP heteroexchange in either direction is driven by the ⌬pH, indicating that the charge imbalance of the exchanged substrates is compensated by the movement of protons. Furthermore, no uptake of [8,5Ј-3 H]GDP or [8-3 H]GTP by unloaded liposomes was observed even in the presence of an energy input (either membrane potential or pH gradient). In other experiments it was found that the rate of 1 M  Mitochondria Lacking ggc1 Are Impaired in GTP Uptake and Contain Reduced Levels of GTP-Having established the transport function of Ggc1p by in vitro assays, the effect of deleting its gene on yeast cells was investigated. We first measured the contents of GTP and GDP in the mitochondria of wild-type and mutant cells. The amount of GTP was approximately 7-fold lower in ggc1⌬ mitochondria than in the organelles from wildtype cells (Fig. 5A). Vice versa the amount of GDP was approximately 5.5-fold higher in ggc1⌬ than in wild-type mitochondria (Fig. 5A). These results are consistent with Ggc1p controlling the entry of GTP and the exit of GDP.
In the next step, the [8-3 H]GTP/GTP exchange was measured in proteoliposomes that were reconstituted with Triton X-100 extracts of wild-type and ggc1⌬ mitochondria. No GTP/GTP exchange activity was detected upon reconstitution of the mitochondrial extracts from the knock-out strain (Fig. 5B). In contrast, an active GTP/GTP exchange was observed using parental mitochondrial extracts (Fig. 5B). The reconstituted GTP/GTP exchange was inhibited markedly by 2 mM pyridoxal 5Ј-phosphate, 2 mM bathophenanthroline, or 0.05% tannic acid and by the Ggc1p substrates GDP and dGTP (but not by ADP, ATP, CDP, and CTP) added at a concentration of 20 M together with 1 M [8-3 H]GTP (data not shown). Similar results were obtained by measuring the [8,5Ј-3 H]GDP/GDP and the [␣-33 P]dGTP/dGTP exchanges in proteoliposomes reconstituted with mitochondrial extracts from wild-type and ggc1⌬ cells. Therefore, the Ggc1p present in the mitochondria exhibits the same specificity and inhibitor sensitivity than the reconstituted protein. As a control, Fig. 5B shows that, compared with the proteoliposomes reconstituted with wild-type extracts, in those reconstituted with ggc1⌬ extracts the phosphate/phosphate exchange and (not shown), the ADP/ADP exchange fell by approximately 60% but was not abolished. When analyzing the levels of several mitochondrial proteins, we found that Ggc1p was completely absent in the mutant mitochondria (Fig. 5C). However, the amounts of the phosphate and the ADP/ATP carriers and (not shown) the oxaloacetatesulfate and the thiamine pyrophosphate carriers and of cytochrome c 1 and subunit 9 of complex III but not of porin were 50 -60% lower in ggc1⌬ mitochondria than in the organelles from wild-type cells (Fig. 5C). Therefore, in the mutant mitochondria, Ggc1p and GTP transport are completely absent, whereas the activities of other transporters are diminished because the transporters are present in smaller amounts in ggc1⌬ than in wildtype mitochondria.
Because the presence of a membrane potential (⌬) across the inner membrane is a prerequisite for any protein transport into or across this membrane ((import) (21), we assessed the membrane potential of ggc1⌬ mitochondria by using the fluorescent dye DiSC (3,5,22). The difference between the fluorescence after the addition of mitochondria and substrates and that after the subsequent addition of the potassium ionophore valinomycin (in the presence of external K ϩ , leading to a complete dissipation of ⌬) is taken as an assessment of the mitochondrial membrane potential (18). The decrease in valinomycin-sensitive fluorescence observed with ggc1⌬ mitochondria was only approximately 4% of that observed with wild-type mitochondria (Fig. 5D), demonstrating that in vitro the membrane potential of ggc1⌬ mitochondria was very low.

Ggc1⌬ Yeast Cells Are Not Able to Grow on Nonfermentable Carbon Sources and Have Lost Their DNA-
The ggc1⌬ mutant was also tested for its ability to utilize different carbon sources. Yeast cells lacking ggc1 showed substantial growth on YP medium containing fermentable carbon sources (glucose or galactose), similarly to the wild-type strain. However, they did not grow on the same medium containing nonfermentable substrates (glycerol, lactate, ethanol, acetate, pyruvate, or oxaloacetate) (data not shown). It should be noted that, in wild-type yeast cells, we found by quantitative immunoblotting that Ggc1p was expressed at similar levels on fermentable (glucose and galoctose) and nonfermentable (lactate, glycerol, ethanol, and acetate) carbon sources (data not shown). The abundance of Ggc1p in mitochondria from yeast cells fed on galactose was 210 Ϯ 47 pmol/mg of protein, in four determinations. The lack of growth on lactate was observed previously upon YDL198c deletion in the YPH499 strain (23), whereas no substantial defect on glycerol was found upon disruption of YDL198c in the W303 strain (9).
The inability of ggc1⌬ mutant cells to grow on nonfermentable media led us to check whether these cells had lost their mitochondrial genome. To address this problem, mutant cells were stained with the DNA-specific dye DAPI and examined by fluorescence microscopy. The major fluorescence source in the central region of the cells, corresponding to DAPI-stained nuclear DNA, was observed both in wild-type and in ggc1⌬ cells (Fig. 6). In contrast, the small and weak fluorescent spots in the cell periphery, corresponding to mtDNA, were observed only in wild-type cells, indicating that ggc1⌬ cells were devoid of mitochondrial DNA (Fig. 6). DISCUSSION In this work, overexpression in E. coli of a hitherto unidentified mitochondrial carrier, reconstitution of the recombinant protein in liposomes, and phenotype analysis of yeast knockout cells have been employed to investigate the function of the S. cerevisiae Ggc1p (encoded by YDL198c). The results ob- The exchanges were started by the addition of 5 M ͓8,5Ј-3 H͔GDP or 1 M ͓8-3 H͔GTP to proteoliposomes containing 2 mM GDP or GTP. K in ϩ was included as KCl in the reconstitution mixture, whereas K out ϩ was added as KCl together with the labeled substrate. The differences in osmolarity were compensated for by the addition of NaCl. 1 mM PIPES-NaOH at pH 7.0 was present in the internal and external compartments. Valinomycin or nigericin was added in 10 l ethanol/ml proteoliposomes. In the control samples the solvent alone was added. The exchange reactions were stopped after 20 s. Similar results were obtained in four independent experiments. tained here, together with the targeting of Ggc1p to mitochondria reported before (9), demonstrate that this protein is the mitochondrial transporter for GTP and GDP. This is the first time that a mitochondrial carrier for guanine nucleotides has been identified from any organism. Ggc1p does not show significant sequence homology with any other mitochondrial carrier functionally identified until now in yeast, mammals, and plants (Refs. 8, 20, and 24 -27 and references therein). In a phylogenetic tree of the S. cerevisiae members of the mitochondrial carrier family (28,29), Ggc1p clusters together with transporters for nucleotides or nucleotide analogs (the three isoforms of the ADP/ATP carrier (30 -32) and the carriers for coenzyme A (33) and for thiamine pyrophosphate (20) and with YDL119c (20% identity), which has not yet been identified). Some proteins encoded by the genomes of lower eukaryotes, such as AL031525 from Schizosaccharomyces pombe (67% of identical amino acids), q8wzw8 from N. crassa (70%), AN5132.1 from Aspergillus nidulans (63%), and AE017168 from Trypanosoma brucei (43%), are likely orthologs of Ggc1p. It is doubtful, however, that there is an orthologous carrier in higher eukaryotes as the closest sequences to Ggc1p in Caenorhabditis elegans (F55C1, 26% of identical amino acids), Drosophila melanogaster (AE003693, 25%), Arabidopsis thaliana (AT2G26360, 20%), and Homo sapiens (the uncoupling protein 2, UCP2, 20%) exhibit a low degree of similarity with Ggc1p as compared with the basic homology existing between the different members of the mitochondrial carrier family. Furthermore, in these organisms the presence of the GTP/GDP carrier is not strictly necessary, because with the exception of A. thaliana, they possess a mitochondrial GTP-producing succinyl-CoA ligase. Besides transporting GTP and GDP with high efficiency, reconstituted Ggc1p also transports the corresponding deoxynucleotides, the structurally related ITP and IDP, and, to a much lesser extent, the (deoxy)nucleoside di-and triphosphates of U and T, but none of the many other compounds tested. The substrate specificity of Ggc1p is distinct from that of any other previously characterized member of the mitochondrial carrier family. In particular, Ggc1p differs markedly from the well known ADP/ATP carrier (34,35), because both the yeast and the human ADP/ATP carrier isoforms transport-only (deoxy)adenine nucleotides are strongly inhibited by carboxyatractyloside and bongkrekic acid and share only 9 -14 and 16 -18% of identical amino acids, respectively, with the Ggc1p. Ggc1p is also quite different from the human deoxynucleotide carrier (36) and its most closely related protein in S. cerevisiae (the thiamine pyrophosphate carrier (Tpc1p (20)), because deoxynucleotide carrier transports all (deoxy)NDPs and less efficiently the corresponding (deoxy)NTPs, whereas Tpc1p transports all of the (deoxy)nucleotides with the following order of efficiency: NMPs Ͼ NDPs Ͼ NTPs. Further- more, at variance with Ggc1p, Tpc1p catalyzes both the uniport and the exchange modes of transport.
In ggc1⌬ yeast cells the mitochondrial content of GTP is drastically decreased, indicating that Ggc1p catalyzes the uptake of GTP into the mitochondrial matrix. In the mitochondria, GTP is converted to GDP (in protein synthesis for the formation of the initiation complex and for the elongation of the polypeptide chain, by GTP-AMP phosphotransferase and by GTPases) or incorporated into the various types of RNA present in the mitochondria, including the RNA primers, which are required for the initiation of DNA replication and repair. Therefore, because in S. cerevisiae GTP is not synthesized in the mitochondrial matrix, Ggc1p appears to be essential for a number of major processes occurring in the mitochondria that depend on the availability of intramitochondrial GTP, such as the initiation of DNA replication and repair, protein synthesis, and recovery of AMP. Because Ggc1p functions by a strict exchange mechanism, the carrier-mediated uptake of GTP requires the efflux of a counter-substrate. On the basis of our transport measurements, GDP, that is produced from GTP intramitochondrially and is phosphorylated by nucleoside diphosphate kinase in the intermembrane space, may serve as the counter-substrate of Ggc1p for GTP. Therefore, the main physiological role of Ggc1p is most likely to catalyze the uptake of GTP into the mitochondrial matrix in exchange for internal GDP. The results obtained by imposing an external energy gradient on our simple liposomal system show that the charge imbalance of the GTP/GDP heteroexchange is compensated by H ϩ carried by Ggc1p in the same direction as GTP. The physiological consequence of this finding is that in mitochondria the GTP out /GDP in exchange catalyzed by Ggc1p is driven by the ⌬pH component of the proton motive force generated by electron transport and is unaffected by its electrical component. The H ϩ -compensated electroneutral type of mechanism of the newly identified GTP/GDP carrier is consistent with its main function (to import GTP in exchange for matrix GDP) and is different from that of the well studied ADP/ATP carrier, which is electrophoretic and ⌬pH-insensitive (37). Another function of Ggc1p may be to catalyze the uptake of dGTP into the mitochondria where it is required for DNA replication and repair. However, the affinity of dGTP for Ggc1p is more than 1 order of magnitude lower than that of GTP. Furthermore, deoxynucleotides can be imported by Tpc1p (see Ref. 20), which is the protein encoded by the yeast genome with the highest sequence similarity to the human deoxynucleotide carrier.
The importance of Ggc1p is highlighted by the observation that the ggc1 null mutant does not grow on any nonfermentable carbon source and has no mitochondrial DNA. The complete loss of mtDNA in ggc1⌬ cells shows that the ggc1 gene product is essential for the maintenance of mtDNA. The involvement of Ggc1p in mtDNA maintenance is in agreement with the previous observation that Ggc1p is a multicopy suppressor of the temperature-sensitive defect of the abf2 null mutant (9). Abf2p is a histone-like mitochondrial DNA-binding protein that is required for maintenance of the yeast mitochondrial genome at 37°C. Our results suggest that Abf2p plays an accessory role to Ggc1p in a GTP-dependent reaction involved in mtDNA maintenance, and hence its inactivation is rescued by the presence of Ggc1p at 25-28°C and by Ggc1p overexpression at 37°C (9). One possibility is that Ggc1p promotes the replication of the mtDNA by stimulating the synthesis of the RNA primers by the enzymes mtRNA polymerase and primase. Interestingly, the mammalian counterpart of Abf2p has been shown to be a limiting factor for mtDNA replication (38). The identification of the mitochondrial GTP/GDP carrier reported here provides a new tool for gaining further insight into the molecular mechanisms underlying the regulation of mtDNA maintenance and metabolism in yeast.
During the revision of this work an accumulation of iron in the mitochondria of yeast cells lacking the yhm1 gene, i.e. the ggc1 gene, was published (39). There are no data available on the role of intramitochondrial guanine nucleotides on iron metabolism in yeast and higher eukaryotes. However, it may be speculated that the Ggc1p-mediated import of GTP into the mitochondrial matrix is required for or regulates a reaction involved in entry/exit of iron into/from the mitochondria or in the synthesis of heme and of iron-sulfur clusters. It is interesting that the bacterial membrane protein FeoB, which is essential for Fe(II) uptake in bacteria, contains a G protein similar to small regulatory G proteins found in eukaryotes and that the function of the G protein is required for Fe(II) uptake through the FeoB-dependent system (40). It is also possible that iron accumulation in mitochondria of yhm1⌬ yeast cells is a secondary effect of the mitochondrial lesions mentioned above caused by the shortage of GTP in the mitochondria. Further studies are necessary to clarify how the Ggc1p-catalyzed transport of guanine nucleotides across the mitochondrial membrane influences iron metabolism.