The GxxxG motif of the transmembrane domain of subunit e is involved in the dimerization/oligomerization of the yeast ATP synthase complex in the mitochondrial membrane Genevie ` ve Arselin, Marie-France Giraud, Alain Dautant, Jacques Vaillier, Daniel Bre ` thes, Be ´ ne ´ dicte Coulary-Salin, Jacques Schaeffer and Jean Velours Institut de Biochimie et Ge ´ ne ´ tique Cellulaires du CNRS, Universite ´ Victor Segalen, Bordeaux, France A conserved putative dimerization GxxxG motif located in the unique membrane-spanning segment of the ATP syn- thase subunit e was altered in yeast both by insertion of an alanine residue and by replacement of glycine by leucine residues. These alterations led to the loss of subunit g and the loss of dimeric and oligomeric forms of the yeast ATP syn- thase. Furthermore, as in null mutants devoid of either subunit e or subunit g, mitochondria displayed anomalous morphologies with onion-like structures. By taking advant- age of the presence of the endogenous cysteine 28 residue in the wild-type subunit e, disulfide bond formation between subunits e in intact mitochondria was found to increase the stability of an oligomeric structure of the ATP synthase in digitonin extracts. The data show the involvement of the dimerization motif of subunit e in the formation of supra- molecular structures of mitochondrial ATP synthases and are in favour of the existence in the inner mitochondrial membrane of associations of ATP synthases whose masses are higher than those of ATP synthase dimers. Keywords: ATP synthase; oligomerization; subunit e; GxxxG motif; yeast. The F 0 F 1 -ATP synthase is a molecular rotary motor that is responsible for the aerobic synthesis of ATP. It exhibits a headpiece (catalytic sector), a basepiece (membrane sector) and two connecting stalks. The sector F 1 containing the headpiece is a water-soluble unit retaining the ability to hydrolyse ATP when in a soluble form. F 0 is embedded in the membrane and is mainly composed of hydrophobic subunits forming a specific proton conducting pathway. When the F 1 and F 0 sectors are coupled, the enzyme functions as a reversible H + -transporting ATPase or ATP synthase [1–4]. The two connecting stalks are constituted of components from F 1 and F 0 . The central stalk is a part of the rotor of the enzyme. The second stalk, which is part of the stator, connects F 1 and hydrophobic membranous components of the enzyme. High resolution X-ray crystal- lographic data have led to solving the structure of the F 1 [5–8] from different sources. Stock et al. [9] reported the 3.9 A ˚ resolution X-ray diffraction structure of the Saccharomyces cerevisiae F 1 associated with the c 10 -ring oligomer. In Escherichia coli,F 0 is composed of subunits a, b and c only. The mitochondrial F 0 of mammals is composed of 10 different subunits [10]. The same 10 components have been identifiedintheS. cerevisiae enzyme [11–13]. It has been shown that the yeast ATP synthase exists as dimeric and oligomeric forms in Triton X-100 and digitonin extracts and that the subunits of F 0 , e, g and 4(b) are essential for such a process [12,14,15]. In addition, the existence of the dimeric form in the inner mitochondrial membrane has been recently demonstrated [16]. Under its mature form, the yeast subunit e is composed of 95 amino-acid residues and displays a mass of 10 744 Da. It is an integral membrane protein anchored to the inner mitochondrial membrane by its unique membrane-spanning segment at its N-terminus, and which adopts, such as the mammalian ATP synthase subunit e, an N in –C out topology [11,17]. A stoichiometry of 2 mol of subunit e per mol of rat liver ATP synthase has been estimated [18]. In mammals, expression of the gene encoding subunit e is regulated in tissues and cells in response to physiological stimuli, thus suggesting that subunit e plays a regulatory role in the ATP synthase [19–21]. In yeast, subunit e is involved in the dimerization/oligomerization of ATP synthases, prob- ably in association with subunit g [12]. Surprisingly, mutant mitochondria devoid of either subunits e or g were found to have numerous digitations and onion-like struc- tures, thus suggesting a link between dimerization/oligo- merization of the ATP synthase and cristae morphology [14,15]. The purpose of the present work was to provide information on the involvement of subunit e in the dimerization/oligomerization of yeast ATP synthases in the inner mitochondrial membrane. Mutations were intro- duced into a putative membranous dimerization motif Correspondence to J. Velours, Institut de Biochimie et Ge ´ ne ´ tique Cellulaires du CNRS, UMR 5095, Universite ´ Victor Segalen, Bordeaux 2, 1, rue Camille Saint Sae ¨ ns, 33077 Bordeaux cedex, France. Fax: + 33 5 56999051, Tel.: + 33 5 56999048, E-mail: jean.velours@ibgc.u-bordeaux2.fr Abbreviations: BN/PAGE, blue native polyacrylamide slab gel electrophoresis; F 0 and F 1 , integral membrane and peripheral portions of ATP synthase; NEM, N-ethylmaleimide. (Received 3 February 2003, revised 3 March 2003, accepted 5 March 2003) Eur. J. Biochem. 270, 1875–1884 (2003) Ó FEBS 2003 doi:10.1046/j.1432-1033.2003.03557.x GxxxG of subunit e. We demonstrate that such a motif is involved in both the edification of supramolecular ATP synthase species and in correct mitochondrial morphology. In addition, cross-linking experiments involving the endo- genous Cys28 residue of subunit e provided data in favour of the oligomerization of the yeast ATP synthase in the mitochondrial membrane. Experimental procedures Materials Digitonin was from Sigma. Oligonucleotides were pur- chased from MWG-BIOTECH. All other reagents were of reagent grade quality. Yeast strains and nucleic acid techniques The Saccharomyces cerevisiae strain D273–10B/A/H/U (MATa, met6, ura3, his3) [22] was the wild-type strain. The null mutant DTIM11 was constructed by a PCR-based mutagenesis and the kan r gene was removed [23]. The mutations 19A, G15L, G19L and C28S were introduced by a PCR mutagenesis procedure [24] into a plasmid pRS313 bearing an insert encoding the wild TIM11 gene and the kan r gene. The strains containing modified versions of subunit e were obtained by integration of the mutated versions of TIM11 gene at the chromosomic locus in the deleted-disrupted yeast strain and were selected for their resistance to Geneticin. The yeast mutants with a point mutation were named as (name of the subunit)(wild-type residue)(residue number)(mutant residue). The strain con- taining the subunit e(His) 6 was constructed according to the following strategy. Two partially complementary oligo- nucleotides 5¢-CGCGGAATTCTTAGTGATGGTGATG GTGATGTGTTGAAGCTTCCTTCAGGG-3¢ and 5¢-CAT CACCATCACCATCACTAAGAATTCCGCGATAGAA GCTTCAACATAAATAGGATACTA-3¢ were used to introduce the (His) 6 sequence into the C-terminus of subunit e by the PCR mutagenesis procedure. Biochemical procedures Cells were grown aerobically at 28 °C in a complete liquid medium containing 2% lactate as carbon source [25] and harvested in logarithmic growth phase. The rho – cell production in cultures was measured on glycerol plates supplemented with 0.1% glucose. Mitochondria were prepared from protoplasts as previously described [26] Protein amounts were determined according to Lowry et al. [27] in the presence of 5% SDS using bovine serum albumin as standard. Oxygen consumption rates were measured with NADH as substrate [28]. Phosphorylation rate was mea- sured in the respiratory buffer supplemented with 1 m M ADP by ATP formation measured by a bioluminescence technique [29]. ATP synthesis rate and oxygen consumption rate were measured at the same time in the oxygraph chamber. All reactions were stopped by adding an aliquot of the medium to perchloric acid. The ATP/O ratio stoichio- metries were determined from the yield of ATP synthesis rate vs. state 3 respiratory rate [30]. The ATPase activity was measured at pH 8.4 [31]. Cross-linking experiments Mitochondria isolated from wild-type and mutant cells were washed by centrifugation in 0.6 M mannitol, 50 m M Hepes pH 7.4 containing 0.25 m M of phenylmethylsulfonyl fluo- ride. The pellet was suspended at a protein concentration of 5mgÆmL )1 in 0.1 M mannitol, 50 m M Hepes pH 7.4 containing either 5 m M of EDTA and 5 m M of N-ethyl- maleimide (NEM) for the control experiment or 2 m M CuCl 2 for the cross-linking experiment. Incubation was perfomed at 4 °C for 30 min. The reaction was stopped upon addition of 5 m M EDTA and 5 m M of NEM. Mitochondrial membranes were then dissociated in the presence of 20 m M of NEM for SDS/gel electrophoresis and Western blot analysis. For BN/PAGE analyses of cross- linked products, mitochondrial membranes were centri- fuged at 10 000 g for 10 min at 4 °C after incubation with either 5 m M of NEM or 2 m M of CuCl 2 , Then, the digitonin solution containing 5 m M NEM was added to the mito- chondrial pellet to extract the ATP synthases. Electrophoretic and Western blot analyses SDS-gel electrophoresis was performed as described in [32,] Western blot analyses were described previously [33]. Nitrocellulose membranes (Membrane Protean BA83, 0.2 lm from Schleicher & Schuell) were used. Polyclonal antibodies against subunits e and g were raised against amino-acid residues 69–82 and 31–45, respectively. Anti- bodies against subunits e, g and i were used with dilutions of 1 : 10 000. Membranes were incubated with peroxidase- labeled antibodies and visualized with the ECL reagent of Amersham Pharmacia Biotech. Molecular mass markers (Benchmark Prestained Protein Ladder) were from Invitro- gen. BN/PAGE experiments were performed as described previously [34,35]. Mitochondria (1 mg of protein) were incubated for 30 min at 4 °C with 0.1 mL of digitonin solution with the indicated digitonin/protein ratio. The extracts were centrifuged at 4 °C for 15 min at 40 000 g and aliquots (40 lL) were loaded on the top of a 3–13% polyacrylamide slab gel. After electrophoresis the gel was incubated in a solution of 5 m M ATP, 5 m M MgCl 2 , 0.05% lead acetate, 50 m M glycine/NaOH pH 8.4 to reveal the ATPase activity [36,37]. Ultrastructural studies Freezing and freeze-substitution of yeast cell pellets were performed as previously described [14]. Results Presence of a dimerization motif in the membrane-spanning segment of subunit e The supernumerary subunit e is a component of the mitochondrial ATP synthase which is involved in the dimerization of ATP synthase [11,12]. Subunit e has been identified in many organisms. The multiple alignment of subunits e of different sources shows that five amino-acid residues are fully conserved (Fig. 1). These are Arg8, Ser10, Leu12, Gly15 and Gly19. The five conserved amino-acid 1876 G. Arselin et al. (Eur. J. Biochem. 270) Ó FEBS 2003 residues lay in a domain predicted to be a membrane- spanning segment beginning with Asn5 and ending at Leu26. Subunit e has an N in –C out orientation [11,17] which exposes the main part of the subunit to the intermembrane space, with the unique cysteine residue (Cys28) at the frontier between the membrane and the intermembrane space. The predicted membrane-spanning segment of sub- unit e displays the dimerization motif GxxxG of glyco- phorin A [38,39] at position Gly-Leu-Phe-Phe-Gly(15–19). The highly conserved Gly15 and Gly19 suggested their involvement in a transmembrane helix–helix interaction. To address whether the GxxxG motif and the two glycine residues are critical in a dimerization process which could be the basis of the dimerization/oligomerization of mito- chondrial ATP synthases, three mutants were constructed. An alanine residue was inserted after Phe18 in order to disrupt a helix–helix packing interface involving the amino- acid residues on both sides of the insertion [40]. In the other two mutants, the small amino-acid residues Gly15 and Gly19 were replaced by leucine residues. Phenotypic analyses of mutant strains and oxidative phosphorylation properties of isolated mitochondria are reported in Table 1. The three mutants displayed a slight increase in the doubling time with lactate as carbon source. An interesting point was the low amount of rho – cells in cultures in comparison with the null mutant DTIM11, which is devoid of subunit e. From these data, there appears to be a correlation between the increase in spontaneous rho – cell conversion (rho – cells are unable to grow with lactate as carbon source) and the increase in the generation time of yeast strains. The ATPase activity of mutant mitochondria displayed a low sensitivity toward the F 0 inhibitor oligo- mycin, thus showing a decreased stability of F 0 under the experimental conditions of ATPase activity measurements (pH 8.4 and Triton X-100). In contrast, under oxidative phosphorylation conditions, the ATP/O ratio value of e19A mitochondria indicated that the efficiency of the oxidative phosphorylation machinery was not altered, as in mutants devoid of either subunits e or g [14]. Loss of dimerization/oligomerization of e19A, eG15L and eG19L ATP synthases We next sought whether the mutations in the dimerization motif of subunit e affected the dimerization/oligomerization of the ATP synthases. Therefore, the presence of Table 1. Phenotypic analysis of yeast strains used. Yeast cells were grown at 28 °C on complete medium containing lactate as carbon source. rho– production was measured on glycerol plates supplemented with 0.1% glucose. Mitochondria were prepared from protoplasts. ATPase activities and the sensitivity to oligomycin (6 lgÆmL )1 ) were measured at pH 8.4 in the presence of Triton X-100 to remove the F 1 inhibitor. ATP/O ratios were determined with NADH as substrate. ND, not determined. Strains Doubling time (min) Rho – cells in cultures (%) ATPase activity ATP/O lmol PiÆmin )1 Æmg protein )1 Inhibition– Oligomycin + Oligomycin Wild-type 161 0.9 5.26 ± 0.31 0.31 ± 0.05 94 1.09 ± 0.13 DTIM11 229 41.0 5.79 ± 0.29 2.99 ± 0.03 48 0.89 ± 0.03 e19A 190 1.6 4.05 ± 0.16 1.01 ± 0.6 75 1.05 ± 0.07 eG15L 210 11.2 4.68 ± 0.22 2.90 ± 0.23 38 ND eG19L 192 7.5 4.66 ± 0.02 2.83 ± 0.24 39 ND Fig. 1. Multiple alignment of subunits e from different sources. The conserved amino-acid residues between subunits e from human (H. s.) (P56385, Swiss-Prot), pig (S. s.) (Q06185, Swiss-Prot), bovine (B. t.) (Q00361, Swiss-Prot), hamster (C. l.) (P12633, Swiss- Prot), mouse (M. m.) (Q06185, Swiss-Prot), rat (R. n.) (P29419, Swiss-Prot), Drosophila (D. m.) (AY060656, GenBank), Neurospora crassa (N. c., AW710731, GenBank), Botryo- tinia fuckelians (B. f., AL111090, EMBL) and Saccharomyces cerevisiae (P81449, Swiss-Prot) are in bold. The numbering of the yeast sub- unit e begins at the initiating methionine. The star indicates the position of the unique cys- teine residue of the yeast subunit e (position 28). The putative transmembrane segment of subunit e (TM) is boxed. Ó FEBS 2003 Supramolecular species of ATP synthase (Eur. J. Biochem. 270) 1877 supramolecular species of the ATP synthase in the mito- chondrial digitonin extracts of mutant strains was examined by BN/PAGE. The digitonin extracts were loaded on a 3– 13% acrylamide slab gel and the mitochondrial complexes were separated under native conditions. The gel was incubated with ATP-Mg 2+ and Pb 2+ to reveal the ATPase activity (Fig. 2). The wild-type digitonin extracts contained the dimeric and oligomeric forms of the enzyme that were destabilized upon increasing the digitonin-to-protein ratio, as shown previously [14]. The mitochondrial digitonin extracts of mutant strains did not display any oligomeric forms of the ATP synthase. Whatever the digitonin-to- protein ratio used, the monomeric form of the enzyme was predominant, although a small amount of dimeric form was found, as already observed for null mutants devoid of either subunit e or subunit g [14,15]. It was previously shown that in the absence of subunit e, subunit g is not present in mitochondrial membranes [12]. As a consequence, the presence of both subunits in strains mutated in subunit e was checked by Western blot analyses of SDS-solubilized mitochondrial membranes. Despite the presence of altered subunits e, the amount of subunit g was highly decreased in e19A mitochondrial membranes and the subunit was not detectable in eG15L and eG19L mito- chondrial membranes (Fig. 3). The e19A, eG15L and eG19L mutants are defective in the mitochondrial morphology It has been previously reported that the null mutants in either TIM11 or ATP20 genes have anomalous mitochon- drial morphologies [14,15]. Thus, transmission electron microscopy of yeast cell sections was performed to examine the effect of mutations in the dimerization motif of subunit e on the ultrastructure of mitochondria. Figure 4 shows that cells of e19A, eG15L and eG19L strains had abnormal mitochondria such as onion-like structures similar to those observed in mutant cells devoid of either subunits e or g. The subunit e of the e19A mutant dimerizes spontaneously via Cys28 in a form which is loosely or not associated to the yeast ATP synthase To gain more insight into the behaviour of mutant subunits e, Western blot analyses of SDS-dissociated wild-type and mutant mitochondria were performed. Figure 3 shows that polyclonal antibodies against subunit e revealed the pre- sence of subunit e and a 21.4-kDa band in e19A, eG15L and eG19L mutant mitochondria. The 21.4-kDa band was observed upon oxidation of wild-type mitochondria with CuCl 2 (Fig. 5A) but it was absent in a mutant devoid of Cys28 (not shown), thus indicating the involvement of Cys28 in the formation of the adduct. The 21.4-kDa band corresponded to a homodimer of subunit e resulting from the formation of a disulfide bond between two subunits e. This result was obtained upon incubation with CuCl 2 of wild-type mitochondria complemented with a pRS313 shuttle vector bearing a gene encoding a subunit e having a(His) 6 sequence at its C-terminus (wild-type + eHis 6 ). In this case, Western blot analysis of CuCl 2 -treated mitochon- dria displayed three bands in the 21.4-kDa region which could be attributed to e + e, e + eHis 6 and eHis 6 +eHis 6 dimers because of their respective apparent molecular masses (Fig. 5A). This result is in full agreement with that of Brunner et al.[41]. Despite the presence of NEM during solubilization of mutant mitochondria by SDS, the amount of e + e dimer was large in mutant mitochondria (Fig. 3), whereas wild- type mitochondria did not display such a dimer, thus showing that pre-existing e + e dimers were present in mutant mitochondrial membranes. It was possible to increase considerably the e + e dimer formation by oxidation with CuCl 2 . As shown in Fig. 5B, incubation of intact mutant mitochondria with CuCl 2 ledtonearlyfull Fig. 2. Lack of dimerization/oligomerization of the yeast ATP synthase upon alteration of the GxxxG dimerization motif of subunit e. Mito- chondria were isolated from wild-type, e19A, eG15L and eG19L strains. Digitonin extracts were obtained with the indicated digitonin/ protein ratios and analysed by BN/PAGE. The gels were incubated with ATP-Mg 2+ and Pb 2+ to reveal ATPase activity. Fig. 3. Mitochondria isolated from e19A, eG15L and eG19L strains are deficient in subunit g. Mitochondria isolated from wild-type (lane 1), e19A (line 2), eG15L (lane 3) and eG19L (lane 4) were treated with NEM as described in the Material and methods section to prevent disulfide bond formation during the dissociation with SDS. Aliquots (30 lg of protein) were analysed by Western blot. The blots were incubated either with antibodies raised against subunits e and i or with antibodies raised against subunits g and i. 1878 G. Arselin et al. (Eur. J. Biochem. 270) Ó FEBS 2003 conversion of mutant subunits e into the dimeric form. This was not the case with wild-type mitochondria, as densito- metric analyses revealed that 45 ± 8% (mean of five experiments) of wild-type subunit e led to the dimeric form in the presence of CuCl 2 . This result reflected a different behaviour of mutant and wild-type subunits e, thus suggesting different relationships between subunit e and the other F 0 components of wild-type and mutant ATP synthases. To verify this point, BN/PAGE analyses were performed with mitochondrial digitonin extracts obtained with a digitonin-to-protein ratio of 0.75 gÆg )1 (Fig. 6). Under these conditions, the wild-type ATP synthase displayed only dimeric and oligomeric forms in BN/PAGE analysis [14]. After migration, the slices of gel were cut and submitted to an SDS/gel electrophoresis in the second dimension. The proteins were transferred onto a nitro- cellulose membrane which was probed with polyclonal antibodies directed against subunit e. Polyclonal antibodies directed against subunit i were also used as control to detect the position of the different forms of ATP synthases, as subunit i is strongly associated to the yeast enzyme at a digitonin-to-protein ratio of 0.75 gÆg )1 .WhenNEM-treated or copper-treated wild-type mitochondria were solubilized with digitonin, subunits e and i comigrated with the dimeric and oligomeric forms of the ATP synthase during native electrophoresis (Figs 6A,B). However, the e + e dimer, which resulted from the incubation of wild-type mitochon- dria with CuCl 2 , was found in the oligomeric form of the ATP synthase but not in the dimeric form. The subunit e and the dimer of subunit e of NEM- and copper-treated e19A mitochondria were observed mainly at a position corresponding to the front of the native gel (right side of the SDS-gel electrophoresis) (Fig. 6D). However, upon oxida- tion of e19A mitochondria, a faint amount of e + e dimer was also observed at molecular masses higher than that of the remaining dimeric form of the ATP synthase (Fig. 6D). Owing to the destabilization of supramolecular ATP synthase forms of e19A mitochondria, a continuous band of subunit i was observed stretching from the top of the native gel to the position of the monomeric form of the enzyme. From these data, it was concluded that the mutation e19A altered the relationship between subunit e and the other F 0 components of the ATP synthase and that under the experimental conditions used, subunit e was highly dissociated from the mutant enzyme. Subunit e is involved in the oligomerization of the yeast ATP synthase An interesting result shown in Fig. 6 is the presence of e + e dimers in the oligomeric forms of the wild-type ATP synthase upon incubation of wild-type mitochondria with CuCl 2 , thus suggesting that subunits e participate in an interface allowing ATP synthase oligomers to exist. To test this hypothesis, wild-type and eC28S mitochondria were incubated either in the presence or absence of CuCl 2 .BN/ PAGE analyses of digitonin extracts (Fig. 7A) revealed the presence of the oligomeric form of the CuCl 2 -treated wild- type ATP synthase migrating at an acrylamide concentra- tion of 4.8%, despite a digitonin-to-protein ratio of 2 gÆg )1 , i.e. conditions which highly destabilize the oligomeric forms of wild-type mitochondria. With the same digitonin- to-protein ratio of 2 gÆg )1 , the eC28S extract did not display this oligomeric form. This result indicates an increased stabilization of the wild-type oligomeric form by the disulfide bond formation between two subunits e. The monomeric, dimeric and oligomeric forms of ATP synthase Fig. 4. Mitochondria isolated from e19A, eG15L and eG19L strains are defective in mitochondrial morphology. Transmission electron micro- scopy of yeast cell sections of e19A (A), eG15L (B) and eG19L (C) strains. m, mitochondria. The bar indicates 0.5 lm. Ó FEBS 2003 Supramolecular species of ATP synthase (Eur. J. Biochem. 270) 1879 of CuCl 2 -treated wild-type mitochondria extracted with a digitonin-to-protein ratio of 2 gÆg )1 were cut from the BN/ PAGE slab and the proteins they contained were separated by SDS-gel electrophoresis. The gel was transferred to a nitrocellulose sheet which was probed with polyclonal antibodies raised against subunits i and e (Fig. 7B). An intense band corresponding to the e + e dimer was found only in the oligomeric form of the ATP synthase as in Fig. 6B, whereas the monomeric subunit e was present only in the dimeric forms. As described in [12], subunit e was absent from the monomeric form of the yeast ATP synthase. In control experiments, incubation of wild-type digitonin extracts (digitonin-to-protein ratio of 2 gÆg )1 )with CuCl 2 did not promote oligomer formation and in addition, the CuCl 2 -treated eC28S mitochondria extracted with either a digitonin-to-protein ratio of 0.75 or 2 gÆg )1 displayed only the monomer of subunit e in the oligomeric and dimeric forms of the yeast ATP synthase (not shown). Discussion Yeast mutants altered in the dimerization motif of subunit e are devoid of subunit g and ATP synthases neither dimerize nor oligomerize The subunits e, g and 4 are three components of the yeast ATP synthase F 0 which are involved in the dimerization/ oligomerization of ATP synthases. The purpose of the present paper was to provide information on the involve- ment of a putative dimerization motif located in the membranous domain of subunit e in the dimerization/ oligomerization of yeast ATP synthases. From the analysis of yeast mutants altered in this motif, we show that this conserved dimerization motif of subunit e has an essential role in the cohesion of an interface between ATP synthases, as shown by BN/PAGE analysis. However, subunit e was still present in mutant mitochondria, as shown by Western blot analysis of whole mitochondrial membranes, but subunit e was loosely or not bound to the ATP synthase, as observed by SDS/gel electrophoresis followed by Western blot analysis of the mitochondrial digitonin extracts separ- ated by electrophoresis under native conditions. This result indicates an alteration of the relationships between subunit e and other F 0 components. For instance, an interesting point was the absence of subunit g, a small hydrophobic protein of F 0 , which has been identified as a near neighbour of subunit e in bovine submitochondrial particles [17]. It has also been shown that in the absence of subunit e, subunit g is not present in mitochondria whereas the absence of subunit g in the null mutant DATP20 does not preclude the presence of subunit e. The lack of subunit g has also been described in a mutant devoid of the first membrane-spanning segment of subunit 4, whereas subunit e was still present. On the basis of cross-linking data, the absence of subunit g in the latter mutant was attributed to the loss of interaction between subunits 4 and g at their membranous levels [15]. Taken together, these observations indicate (a) a close relationship of subunits e, g and 4 in the membranous F 0 domain and (b) that subunit g is a very unstable protein which disappears from the ATP synthase upon alterations of either subunits e Fig. 5. Oxidation of cysteine 28 promoted the dimerization of subunit e of wild type and mutant mitochondria. (A) Mitochondria iso- lated from wild-type, eHis 6 cells and wild- type cells complemented with the plasmid pRS313 encoding eHis 6 (wild-type + eHis 6 ) were incubated in the presence or absence of CuCl 2 as described in the experimental pro- cedure. The control experiment (in the absence of CuCl 2 ) was performed in the presence of NEM instead of CuCl 2 .(B)Mito- chondria isolated from wild-type, e19A, eG15L and eG19L strains were incubated in the presence or absence of CuCl 2 . The control experiment (in the absence of CuCl 2 )was performed in the presence of NEM instead of CuCl 2 . Cross-linking conditions are described in the experimental procedure. After dissoci- ation of samples with SDS in the presence of 20 m M of NEM, aliquots (30 lgofprotein) were analysed by Western blot. The blots were incubated with polyclonal antibodies raised against subunit e. 1880 G. Arselin et al. (Eur. J. Biochem. 270) Ó FEBS 2003 or 4. This has two consequences; the loss of dimerization/ oligomerization of ATP synthases as revealed by BN/ PAGE analysis and the presence of anomalous mitochon- dria with onion-like structures. This underlines the relation- ship between the dimerization/oligomerization of the mitochondrial ATP synthase and the yeast mitochondrial morphology, a point which has been reported in previous papers [14,15]. As the above phenotypes have a common feature, i.e. the absence of subunit g, we propose that by its presence, subunit g exerts a central role in the interfaces allowing the dimerization/oligomerization of the yeast ATP synthases. This role and these interfaces will be more precisely studied by site-directed mutagenesis of the ATP20 gene encoding subunit g to identify the interaction domains between subunit g and subunits e and 4. The oligomeric forms of the yeast ATP synthase in the inner mitochondrial membrane The unique cysteine residue of the wild-type subunit e, which is located in the intermembrane space, is an accessible target to chemical reagents that allows the environmental study of this subunit in the inner mitochondrial membrane. Western blot analysis of proteins separated by SDS-gel electrophoresis originating from native complexes separated by BN/PAGE allowed us to examine the behaviour of subunit e in the different forms of the yeast ATP synthase extracted by digitonin (at a digitonin-to-protein ratio of 0.75 gÆg )1 ). The dimeric and oligomeric forms of ATP synthase of NEM-treated wild-type mitochondria contained only monomeric subunit e. In contrast, CuCl 2 incubation of wild-type mitochondria led to the formation of the e + e dimer that was found only in the oligomeric form of the yeast ATP synthase migrating at an acrylamide concentra- tion of 4.8%, i.e. at an apparent molecular mass which corresponds to at least a tetrameric form of the enzyme [14]. In addition, while a digitonin-to-protein ratio of 2 gÆg )1 destabilized the oligomeric forms of the enzyme, the disulfide bond formation between two subunits e via the Cys28 residues increased the stability of the oligomeric form migrating at an acrylamide concentration of 4.8%. The existence of supramolecular structures of the yeast ATP synthase in Triton X-100 and digitonin extracts and in the inner mitochondrial membrane has been well documented. Biochemical evidence has shown that the dimeric forms of the yeast ATP synthase are not due to the aggregation of the monomeric form of the enzyme as (a) dimerization and oligomerization of the ATP synthase are dependent on the Fig. 6. Upon oxidation, the subunit e dimer of e19A mitochondria was loosely associated with the ATP synthase, whereas the subunit e dimer of wild- type mitochondria was associated only with the oligomeric forms of the wild-type enzyme. Wild-type and e19A mitochondria were incubated either with NEM or with CuCl 2 as described above. Mitochondrial digitonin extracts were obtained with a digitonin-to-protein ratio of 0.75 gÆg )1 and analysed by BN/PAGE. A part of the gel was revealed by the ATPase activity and slices of the BN/PAGE are shown on the top of each figure (first dimension). Corresponding slices were cut, incubated with 1% SDS and submitted to SDS/gel electrophoresis (second dimension). The proteins of the gel were transferred onto a nitrocellulose membrane, which was probed with polyclonal antibodies raised against subunits e and i. NEM-treated wild-type (A) and NEM-treated e19A (C) mitochondria (control experiments). CuCl 2 -treated wild-type (B) and CuCl 2 -treated e19A (D) mito- chondria. Ó FEBS 2003 Supramolecular species of ATP synthase (Eur. J. Biochem. 270) 1881 presence of subunits e, g and the first membrane-spanning segment of the b-subunit (subunit 4) and (b) inter-ATP synthase cross-linking with a bis-maleimide reagent and involving a cysteine residue introduced into the C-terminal part of subunit i located in the intermembrane space has been reported both in detergent extracts which preserve the dimeric form of the ATP synthase and in intact mitochon- dria, thus showing the existence of such dimers in the inner mitochondrial membrane [16]. Whether oligomeric forms of the ATP synthase exist in the inner mitochondrial mem- brane is still a matter of discussion. Using freeze-fracturing, deep-etching and replicates, Allen et al.[42]showedthe presence of double rows of ATP synthases on cristae of Paramecium multimicronucleatum mitochondria. Subse- quently, Allen proposed a model that described the association of ATP synthase dimers as generating the tubular cristae [43]. In yeast, oligomeric forms of ATP synthase have been found in mitochondrial digitonin extracts obtained with digitonin-to-protein ratios of 0.75– 1gÆg )1 , but they were absent at higher ratios. However, by the formation of disulfide bonds between subunits e in intact wild type mitochondria, we have now found that it is possible to increase slightly the stability of an oligomeric form of the ATP synthase in digitonin extracts. The association of ATP synthases in supramolecular structures higher than dimeric forms by oxidation could result from the Brownian lateral diffusion of proteins in the inner mitochondrial membrane. However (a) the experiments were performed at 4 °C to decrease the diffusion, (b) no other cross-links between the ATP synthase and other mitochondrial complexes have been identified and (c) these oligomeric structures exist without cross-linking in mito- chondrial digitonin extracts. Therefore, we consider that these data are in favour of the existence in the inner mitochondrial membrane of oligomeric forms such as those observed by Allen et al. [42]. Such oligomeric forms of ATP synthase imply the existence of two different interfaces between ATP synthase monomers. On the basis of cross- linking data on mitochondrial membranes [44], it was proposed that two subunits 4 belonging to two neighbour- ing ATP synthases participate at one interface [14]. The second interface involves subunits e and g [12] and it appears from the above data that the dimerization motif in subunit e is essential for the stability of this interface. Subunits e have different environments While oxidation of mutant mitochondria led to nearly full conversion of altered subunit e under a dimeric form, only 50% of wild type subunit e was converted under its dimeric form in the presence of CuCl 2 . On the other hand, whatever the digitonin-to-protein ratio used, the e + e dimer was only found in wild type ATP synthase oligomer upon oxidation, whereas the monomeric form of subunit e was associated with both oligomeric and dimeric forms of the enzyme at a digitonin-to-protein ratio of 0.75 gÆg )1 .Witha digitonin-to-protein ratio of 2 gÆg )1 , which destabilizes the ATP synthase oligomers, the monomeric form of subunit e was removed, while an e + e dimer was still present in the CuCl 2 -induced ATP synthase oligomer. As a stoichiometry of two subunits e has been established in rat ATP synthase, it appears that the two subunits e of each wild type enzyme react differently and thus likely have different environment in F 0 . It has not yet been established whether the dimeri- zation motif of subunit e mediates homodimer formation between subunits e of each enzyme or between subunits e of two interacting enzymes or heterodimer formation with another component of F 0 in the wild type ATP synthase. The combination of BN/PAGE, SDS-gel electrophoresis of isolated supramolecular complexes and cross-linking Fig. 7. The disulfide bond formation between two subunits e stabilizes an oligomeric form of the yeast ATP synthase. Mitochondria isolated from wild-type and eC28S cells were incubated in the absence or in the presence of CuCl 2 . (A) Digitonin extracts obtained with the indicated digitonin-to-protein ratios were submitted to BN/PAGE. The gel was incubated with ATP-Mg 2+ and Pb 2+ to reveal the ATPase activity. NEMwasaddedinthecontrolexperimentsinsteadofCuCl 2 .(B) Mitochondria isolated from wild-type cells were incubated with CuCl 2 and solubilized with a digitonin-to-protein ratio of 2 gÆg )1 . After BN/ PAGE analysis the bands were revealed by the ATPase activity. The oligomeric form (lane 1), the high band (lane 2), the low band (lane 3) of the dimeric forms and the monomeric form (lane 4) of the ATP synthase were cut and submitted to SDS-gel electrophoresis. The slab gel was transferred onto nitrocellulose which was probed with poly- clonal antibodies against subunits i and e. T, acrylamide concentration. 1882 G. Arselin et al. (Eur. J. Biochem. 270) Ó FEBS 2003 experiments with engineered targets in subunits e, g and 4 will allow the identification of contact areas of the different partners involved in the interfaces between ATP synthases. Acknowledgements We are grateful to Drs C. Napias and R. Cooke for their contribution to the editing of the manuscript. This research was supported by the Centre National de la Recherche Scientifique (Programme Dynamique et Re ´ activite ´ des Assemblages Biologiques), the Universite ´ Victor Segalen, Bordeaux 2 and the Etablissement Public Re ´ gional d’Aquitaine. References 1. Fillingame, R.H. (1999) Molecular rotary motors. Science 286, 1687–1688. 2. Pedersen, P.L., Ko, Y.H. & Hong, S. (2000) ATP synthases in the year 2000: evolving views about the structures of these remarkable enzyme complexes. J. Bioenerg. Biomembr. 32, 325–332. 3. Stock, D., Gibbons, C., Arechaga, I., Leslie, A.G.W. & Walker, J.E. (2000) The rotary mechanism of ATP synthase. Curr. Opin. Struct. Biol. 10, 672–679. 4. Senior, A.E., Nadanaciva, S. & Weber, J. (2002) The molecular mechanism of ATP synthesis by F(1)F(0)-ATP synthase. Biochim. Biophys. Acta 1553, 188–211. 5. Abrahams, J.P., Leslie, A.G.W., Lutter, R. & Walker, J.E. (1994) Structure at 2.8 angstrom resolution of F-1-ATPase from bovine heart mitochondria. Nature 370, 621–628. 6. Bianchet, M.A., Hullihen, J., Pedersen, P.L. & Amzel, L.M. (1998) The 2.8-angstrom structure of rat liver F-1-ATPase: configuration of a critical intermediate in ATP synthesis/hydrolysis. Proc. Natl Acad. Sci. USA 95, 11065–11070. 7. Hausrath, A.C., Gru ¨ ber, G., Matthews, B.W. & Capaldi, R.A. (1999) Structural features of the gamma subunit of the Escherichia coli F-1 ATPase revealed by a 4.4-angstrom resolution map obtained by x-ray crystallography. Proc. Natl Acad. Sci. USA 96, 13697–13702. 8. Gibbons, C., Montgomery, M.G., Leslie, A.G.W. & Walker, J.E. (2000) Thestructure of the central stalk in bovine F1-ATPase at 2.4 A ˚ resolution. Nat. Struct. Biol. 7, 1055–1061. 9. Stock, D., Leslie, A.G.W. & Walker, J.E. (1999) Molecular architecture of the rotary motor in ATP synthase. Science 286, 1700–1705. 10. Collinson, I.R., Runswick, M.J., Buchanan, S.K., Fearnley, I.M., Skehel, J.M., van Raaij, M.J., Griffiths, D.E. & Walker, J.E. (1994) Fo membrane domain of ATPsynthase from bovine heart mitochondria: purification, subunit composition, andreconstitu- tion with F1-ATPase. Biochemistry 33, 7971–7978. 11. Arnold, I., Bauer, M.F., Brunner, M., Neupert, W. & Stuart, R.A. (1997) Yeastmitochondrial F1F0-ATPase: the novel subunit e is identical to Tim11. FEBS Lett. 411, 195–200. 12. Arnold,I.,Pfieffer,K.,Neupert,W.,Stuart,R.A.&Scha ¨ gger, H. (1998) Yeastmitochondrial F1F0-ATP synthase exists as a dimer: identification of three dimer-specific subunits. EMBO J. 17, 7170– 7178. 13. Velours, J. & Arselin, G. (2000) The Saccharomyces cerevisiae ATP synthase. J. Bioenerg. Biomembr. 32, 383–390. 14. Paumard, P., Vaillier, J., Coulary, B., Schaeffer, J., Soubannier, V., Mueller, D.M., Bre ` thes, D., di Rago, J.P. & Velours, J. (2002) The ATP synthase is involved in generating mitochondrial cristae morphology. EMBO J. 21, 221–230. 15. Soubannier, V., Vaillier, J., Paumard, P., Coulary, B., Schaeffer, J. & Velours, J. (2002) In the absence of the first membrane-spanning segment of subunit 4(b), the yeast ATP synthase is functional but does not dimerize or oligomerize. J. Biol. Chem. 277, 10739–10745. 16. Paumard, P., Arselin, G., Vaillier, J., Chaignepain, S., Bathany, K., Schmitter, J.M., Bre ` thes, D. & Velours, J. (2002) Two ATP syn- thases can be linked through subunits i in the inner mitochondrial membrane of Saccharomyces cerevisiae. Biochemistry 41, 10390– 10396. 17. Belogrudov, G.I., Tomich, J.M. & Hatefi, Y. (1996) Membrane topography andnear-neighbor relationships of the mitochondrial ATP synthase subunits e, f, and g. J. Biol. Chem. 271, 20340– 20345. 18. Arakaki, N., Ueyama, Y., Hirose, M., Himeda, T., Shibata, H., Futaki, S., Kitagawa, K. & Higuti, T. (2001) Stoichiometry of subunit e in rat liver mitochondrial H(+)-ATP synthase and membrane topology of its putative Ca(2+)-dependent regulatory region. Biochim. Biophys. Acta 1504, 220–228. 19. Higuti, T., Kuroiwa, K., Kawamura, Y. & Yoshihara, Y. (1992) Complete amino acid sequence of subunit e of rat liver mito- chondrial H(+)-ATP synthase. Biochemistry 31, 12451–12454. 20. Swartz, D.A., Park, E.I., Visek, W.J. & Kaput, J. (1996) The e subunit gene of murine F1F0-ATP synthase – Genomic sequence, chromosomal mapping and diet regulation. J. Biol. Chem. 271, 20942–20948. 21. Levy, F.H. & Kelly, D.P. (1997) Regulation of ATP synthase subunit eexpression by hypoxia: cell differentiation stage-specific control. Am. J. Physiol. 272, C457–C465. 22. Paul, M.F., Velours, J., Arselin de Chateaubodeau, G., Aigle, M. &Gue ´ rin, B. (1989) The role of subunit, 4, a nuclear-encoded protein of the F0 sector of yeast mitochondrial, ATP synthase, in the assembly of the whole complex. Eur. J. Biochem. 185, 163–171. 23. Gu ¨ ldener, U., Heck, S., Fielder, T., Beinhauer, J. & Hegemann, J.H. (1996) Anew efficient gene disruption cassette for repeated use in budding yeast. Nucleic Acids Res. 24, 2519–2524. 24. Salleh, H.M., Patel, M.A. & Woodard, R.W. (1996) Essential cysteines in 3-deoxy-D-manno-octulosonic acid 8-phosphate syn- thase from Escherichia coli: analysis by chemical modification and site-directed mutagenesis. Biochemistry 35, 8942–8947. 25. Arselin de Chateaubodeau, G., Gue ´ rin, M. & Gue ´ rin, B. (1976) Perme ´ abilite ´ de la membrane interne des mitochondries de levure: e ´ tude des relations entre structure et activite ´ . Biochimie (Paris) 58, 601–610. 26. Gue ´ rin, B., Labbe, P. & Somlo, M. (1979) Preparation of yeast mitochondria (Saccharomyces cerevisiae)withgoodP/Oandres- piratory control ratios. Methods Enzymol. 55, 149–159. 27. Lowry, O.H., Rosebrough, N.J., Farr, A.L. & Randall, R.J. (1951) Protein measurement with the folin reagent. J. Biol. Chem. 193, 265–275. 28. Rigoulet, M. & Gue ´ rin, B. (1979) Phosphate transport and ATP synthesis in yeast mitochondria: effect of a new inhibitor: the tri- benzylphosphate. FEBS Lett. 102, 18–22. 29. Lundin, A., Richardsson, A. & Thore, A. (1976) Continuous monitoring of ATP-converting reactions by purified firefly luci- ferase. Anal. Biochem. 75, 611–620. 30. Devin, A., Gue ´ rin, B. & Rigoulet, M. (1997) Control of oxidative- phosphorylation in rat liver mitochondria: effect of ionic media. Biochim. Biophys. Acta 1319, 293–300. 31. Velours, J., Vaillier, J., Paumard, P., Soubannier, V., Lai-Zhang, J. & Mueller, D.M. (2001) Bovine coupling factor 6, with just 14.5% shared identity, replaces subunit h in the yeast ATP synthase. J. Biol. Chem. 276, 8602–8607. 32. Scha ¨ gger, H. & von Jagow, G. (1987) Tricine-sodium dodecyl sulfate-polyacrylamide gel electrophoresis for the separation of proteins in the range from 1 to100 kDa. Anal. Biochem. 166, 368– 379. 33. Arselin, G., Vaillier, J., Graves, P.V. & Velours, J. (1996) ATP synthase of yeast mitochondria. Isolation of the subunit h and disruption of the ATP14 gene. J. Biol. Chem. 271, 20284–20290. Ó FEBS 2003 Supramolecular species of ATP synthase (Eur. J. Biochem. 270) 1883 34. Scha ¨ gger, H. & von Jagow, G. (1991) Blue native electrophoresis for isolation of membrane protein complexes in enzymatically active form. Anal. Biochem. 199, 223–231. 35. Scha ¨ gger, H., Cramer, W.A. & von Jagow, G. (1994) Analysis of molecular masses and oligomeric states of protein complexes by blue native electrophoresis and isolation of membrane protein complexes by two-dimensional native electrophoresis. Anal. Bio- chem. 217, 220–230. 36. Yoshida, M., Sone, N., Hirata, H. & Kagawa, Y. (1975) A highly stable adenosine Triphosphatase from a thermophillie bacterium. Purification, properties, and reconstitution. J. Biol. Chem. 250, 7910–7916. 37. Grandier-Vazeille, X. & Gue ´ rin, M. (1996) Separation by blue native and colorless native polyacrylamide gel electrophoresis of the oxidative phosphorylation complexes of yeast mitochondria solubilized by different detergents: specific staining of the different complexes. Anal. Biochem. 242, 248–254. 38. Senes, A., Gerstein, M. & Engelman, D.M. (2000) Statistical analysis of amino acid patterns in transmembrane helices: the GxxxG motif occurs frequently and in association with b-bran- ched residues at neighboring positions. J. Mol. Biol. 296, 921–936. 39. Russ, W.P. & Engelman, D.M. (2000) The GxxxG motif: a fra- mework for transmembrane helix–helix association. J. Mol. Biol. 296, 911–919. 40. Mingarro, I., Whitley, P., Lemmon, M.A. & von Heijne, G. (1996) Ala-insertion scanning mutagenesis of the glycophorin A trans- membrane helix: a rapid way to map helix–helix interactions in integral membrane proteins. Protein Sci. 7, 1339–1341. 41. Brunner, S., Everard-Gigot, V. & Stuart, R.A. (2002) Subunit e of the yeast F1F0ATP synthase forms homodimers. J. Biol. Chem. 277, 48484–48489. 42. Allen, R.D., Schroeder, C.C. & Fok, A.K. (1989) An investigation of mitochondrial inner membranes by rapid-freeze deep-etch techniques. J. Cell. Biol. 108, 2233–2240. 43. Allen, R.D. (1995) Membrane tubulation and proton pumps. Protoplasma 189, 1–8. 44. Spannagel, C., Vaillier, J., Arselin, G., Graves, P.V., Grandier- Vazeille, X. & Velours, J. (1998) Evidence of a subunit 4 (subunit b) dimer in favor of the proximity of ATP synthase complexes in yeast inner mitochondrial membrane. Biochim. Biophys. Acta 1414, 260–264. 1884 G. Arselin et al. (Eur. J. Biochem. 270) Ó FEBS 2003 . The GxxxG motif of the transmembrane domain of subunit e is involved in the dimerization/oligomerization of the yeast ATP synthase complex in the mitochondrial membrane Genevie ` ve Arselin,. enzyme. Subunit e is involved in the oligomerization of the yeast ATP synthase An interesting result shown in Fig. 6 is the presence of e + e dimers in the oligomeric forms of the wild-type ATP synthase upon. and Saccharomyces cerevisiae (P81449, Swiss-Prot) are in bold. The numbering of the yeast sub- unit e begins at the initiating methionine. The star indicates the position of the unique cys- teine residue of