Nuclear genes coding the yeast mitochondrial adenosine triphosphatase complex. Primary sequence analysis of ATP2 encoding the F1-ATPase beta-subunit precursor

The yeast nuclear gene ATP2 encodes a F1-ATPase beta-subunit protein of 509 amino acids with a predicted mass of 54,575 daltons. In contrast to the ATPase beta-subunit proteins determined previously from Escherichia coli and various plant sources, the yeast mitochondrial precursor peptide contains a unique cysteine residue within its immediate amino terminus. Expression of an in-frame deletion in ATP2 between residues 28 and 34 to eliminate this single cysteine residue located near the processing site of the matrix protease does not prevent the in vivo delivery of the subunit to mitochondria or its assembly into a functional ATPase complex. Thus, the import F1 beta-subunit into mitochondria does not require a covalent modification of the type utilized for the secretion of the major lipoprotein from E. coli. In addition, analysis of the level of the major F1-ATPase subunits in mitochondria prepared from an atp2- disruption mutant demonstrates that the in vivo import of these catalytic subunits is not dependent on each other. These data and additional studies, therefore, suggest that the determinants for mitochondrial delivery reside within the amino terminus of the individual precursors.     

The rate of import and assembly of F1-ATPase in Saccharomyces cerevisiae

Subunit specific antiserum can be employed to study the course of ATPase assembly in mitochondria isolated from bakers' yeast. Comparing rates of subunit import with rates of enzyme assembly indicated that no substantial pool of unassembled subunits exists for the three largest ATPase peptides (alpha, beta, and gamma). Blocking import of specific ATPase subunits, however, did reveal a possible accumulation of unassembled alpha and gamma subunits in isolated mitochondria. The kinetic experiments also revealed a lag in the import of beta subunit relative to the uptake of alpha and gamma precursors. Experiments conducted in yeast cells confirmed that beta subunit is assembled soon after it is imported, but did not indicate a delay in import relative to the other subunits of F1.     

Three subunit proteins of membrane enzymes in mitochondria of Neurospora crassa contain a pantothenate derivative

Three proteins of the inner mitochondrial membrane of Neurospora crassa were found to be covalently modified with a derivative of pantothenic acid. One of these proteins is a subunit of cytochrome c oxidase and two are subunits of the ATPase-ATP synthase. Cells of a pantothenate auxotroph of N. crassa were labeled with [14C]pantothenic acid, and mitochondrial proteins containing radiolabeled pantothenate were detected by electrophoresis of detergent-solubilized mitochondria. Mitochondria from cells that were colabeled with [14C]pantothenate and [3H]leucine were reacted with specific antisera against the cytochrome c oxidase and F1-ATPase enzyme complexes. Electrophoresis of the labeled subunits of these isolated complexes showed that the [14C]pantothenate-associated peptides corresponded to [3H]leucine-labeled subunit 6 of cytochrome c oxidase and two [3H]leucine-labeled subunits (tentatively identified as subunits 8 and 11) of the ATPase-ATP synthase. Pantothenate modification of these enzyme subunits, which are synthesized on extramitochondrial ribosomes, may contribute to their transport and assembly into mitochondria, or it may participate in the catalytic activity of the assembled enzymes.     

Precursor proteins are transported into mitochondria in the absence of proteolytic cleavage of the additional sequences

Many nuclear-coded mitochondrial proteins are synthesized as larger precursor polypeptides that are proteolytically processed during import into the mitochondrion. This processing appears to be catalyzed by a soluble, metal-dependent protease localized in the mitochondrial matrix. In this report we employ an in vitro system to investigate the role of processing in protein import. Intact Neurospora crassa mitochondria were incubated with radiolabeled precursors in the presence of the chelator o-phenanthroline. Under these conditions, the processing of the precursors of the beta-subunit of F1-ATPase (F1 beta) and subunit 9 of the F0F1-ATPase was strongly inhibited. Protease-mapping studies indicated that import of the precursor proteins into the mitochondria continued in the absence of processing. Upon readdition of divalent metal to the treated mitochondria, the imported precursors were quantitatively converted to their mature forms. This processing of imported precursors occurred in the absence of a mitochondrial membrane potential and was extremely rapid even at 0 degrees C. This suggests that all or part of the polypeptide chain of the imported precursors had been translocated into the matrix location of the processing enzyme. Localization experiments suggested that the precursor to F1 beta is peripherally associated with the mitochondrial membrane while the precursor to subunit 9 appeared to be tightly bound to the membrane. We conclude that proteolytic processing is not necessary for the translocation of precursor proteins across mitochondrial membranes, but rather occurs subsequent to this event. On the basis of these and other results, a hypothetical pathway for the import of F1 beta and subunit 9 is proposed.     

Stabilization of rat liver mitochondrial F1-adenosine triphosphatase during chloroform-induced solubilization.

1. Isolation of ATPase from rat liver submitochondrial particles by chloroform treatment requires the presence of ATP or ADP during enzyme solubilization. In the absence of adenine nucleotides the enzyme activity is very low although all protein components of F1-ATPase are released. The low concentrations of ATP or ADP required (5 microM) indicate that the high affinity nucleotide-binding sites are involved in enzyme stabilization. Other nucleotides tested (ITP, GTP, UTP, CTP) were found to be less effective. 2. Polyacrylamide gel electrophoresis and immunodiffusion in agar plates revealed that in the absence of adenine nucleotides a fraction of F1-ATPase released by chloroform treatment is split into fragments. The part of the dissociated enzyme molecule has a molecular weight identical with that of a beta-subunit of F1-ATPase. 3. Dissociation of the F1-ATPase molecule could also be prevented by aurovertin. 4. Crude F1-ATPase solubilized by chloroform treatment can be further purified by Sepharose 6B gel filtration. Specific ATPase activity of the purified enzyme was 90 mumol Pi/min per mg protein and the enzyme was composed of five protein subunits (alpha, beta, gamma, delta, epsilon) with molecular weights 58 000, 55 000, 28 000, 13 000 and 8000, respectively. 5. Chloroform-released F1-ATPase from rat liver mitochondria displayed immunochemical cross-reactivity with that isolated from beef heart mitochondria.     

Purification and characterization of the F1 ATPase from Bacillus subtilis and its uncoupler-resistant mutant derivatives.

The F1 ATPase of Bacillus subtilis BD99 was extracted from everted membrane vesicles by low-ionic-strength treatment and purified by DEAE-cellulose chromatography, hydrophobic interaction chromatography, and anion-exchange high-performance liquid chromatography. The subunit structure of the enzyme was examined by sodium dodecyl sulfate-polyacrylamide gel electrophoresis in the absence and presence of urea. In the absence of urea, the alpha and beta subunits comigrated and the ATPase was resolved into four bands. The mobility of the beta subunit, identified by immunoblotting with anti-beta from Escherichia coli F1, was altered dramatically by the presence of urea, causing it to migrate more slowly than the alpha subunit. The catalytic activity of the ATPase was strongly metal dependent; in the absence of effectors, the Ca2+-ATPase activity was 15- to 20-fold higher than the Mg2+ -ATPase activity. On the other hand, sulfite anion, methanol, and optimally, octylglucoside stimulated the Mg2+ -ATPase activity up to twice the level of Ca2+ -ATPase activity (specific activity, about 80 mumol of Pi per min per mg of protein). The F1 ATPase was also isolated from mutants of B. subtilis that had been isolated and characterized in this laboratory by their ability to grow in the presence of protonophores. The specific activities of the ATPase preparations from the mutant and the wild type were very similar for both Mg2+- and Ca2+ -dependent activities. Kinetic parameters (Vmax and Km for Mg-ATP) for octylglucoside-stimulated Mg2+ -ATPase activity were similar in both preparations. Structural analysis by polyacrylamide gel electrophoresis and isoelectric focusing indicated that the five F1 subunits from ATPase preparations from the mutant and wild-type strains had identical apparent molecular weights and that no charge differences were detectable in the alpha and beta subunits in the two preparations. Thus, the increased ATPase activity that had been observed in the uncoupler-resistant mutants is probably not due to a mutation in the F1 moiety of the ATPase complex.     

Membrane ATPase of Vibrio alginolyticus. Ion transport activity and homology with F0F1-ATPase from E. coli]

F0F1-ATPase has been isolated from the marine alkali-resistant bacterium Vibrio alginolyticus. The enzyme subunits cross-reacted with antibodies against subunits alpha, beta, gamma, epsilon, and b of E. coli ATPase. The purified ATPase was reconstituted into liposomes effecting an ATP-dependent uptake of H+. Proton transport was inhibited by the ATPase blockers DCCD, triphenyltin, and venturicidin. Na+ ions had no effect on ATP-dependent proton transport. No ATP-dependent transport of Na+ was detected in proteoliposomes.     

Existence of stoichiometric amounts of an intrinsic ATPase inhibitor and two stabilizing factors with mitochondrial ATP synthase in yeast

Two proteinaceous factors, 15K and 9K proteins, which acted together to stabilize the inactivated yeast F1F0-ATPase-inhibitor complex [Hashimoto, T., et al. (1984) J. Biochem. 95, 131-136] were hardly distinguishable from the sigma and epsilon subunits, respectively, of yeast F1-ATPase by sodium dodecyl sulfate (SDS) polyacrylamide gel electrophoresis. However, they were clearly distinguishable from these subunits by analyses of the sequences at their amino terminals and by immunoblotting combined with SDS polyacrylamide gel electrophoresis. The two stabilizing factors and an ATPase inhibitor existed in mitochondria in equimolar ratios to F1-ATPase. These three protein factors were not present in purified F1-ATPase or in F1F0-ATPase preparations, but remained in the mitochondrial membranes after extraction of F1F0-ATPase with Triton X-100. These observations strongly suggest that the two stabilizing factors and the ATPase inhibitor form a regulatory substructure of mitochondrial ATP synthase, in addition to the F1 and F0 subunits.     

Identification of the subunits of F1F0-ATPase from bovine heart mitochondria.

An oligomycin-sensitive F1F0-ATPase isolated from bovine heart mitochondria has been reconstituted into phospholipid vesicles and pumps protons. this preparation of F1F0-ATPase contains 14 different polypeptides that are resolved by polyacrylamide gel electrophoresis under denaturing conditions, and so it is more complex than bacterial and chloroplast enzymes, which have eight or nine different subunits. The 14 bovine subunits have been characterized by protein sequence analysis. They have been fractionated on polyacrylamide gels and transferred to poly(vinylidene difluoride) membranes, and N-terminal sequences have been determined in nine of them. By comparison with known sequences, eight of these have been identified as subunits beta, gamma, delta, and epsilon, which together with the alpha subunit form the F1 domain, as the b and c (or DCCD-reactive) subunits, both components of the membrane sector of the enzyme, and as the oligomycin sensitivity conferral protein (OSCP) and factor 6 (F6), both of which are required for attachment of F1 to the membrane sector. The sequence of the ninth, named subunit e, has been determined and is not related to any reported protein sequence. The N-terminal sequence of a tenth subunit, the membrane component A6L, could be determined after a mild acid treatment to remove an alpha-N-formyl group. Similar experiments with another membrane component, the a or ATPase-6 subunit, caused the protein to degrade, but the protein has been isolated from the enzyme complex and its position on gels has been unambiguously assigned. No N-terminal sequence could be derived from three other proteins. The largest of these is the alpha subunit, which previously has been shown to have pyrrolidonecarboxylic acid at the N terminus of the majority of its chains. The other two have been isolated from the enzyme complex; one of them is the membrane-associated protein, subunit d, which has an alpha-N-acetyl group, and the second, surprisingly, is the ATPase inhibitor protein. When it is isolated directly from mitochondrial membranes, the inhibitor protein has a frayed N terminus, with chains starting at residues 1, 2, and 3, but when it is isolated from the purified enzyme complex, its chains are not frayed and the N terminus is modified. Previously, the sequences at the N terminals of the alpha, beta, and delta subunits isolated from F1-ATPase had been shown to be frayed also, but in the F1F0 complex they each have unique N-terminal sequences.(ABSTRACT TRUNCATED AT 400 WORDS)     

Binding properties of an intrinsic ATPase inhibitor and occurrence in yeast mitochondria of a protein factor which stabilizes and facilitates the binding of the inhibitor to F1F0-ATPase

The content of an intrinsic ATPase inhibitor in mitochondria was determined by a radioimmunoassay procedure which showed the molar ratio of the inhibitor to ATPase to be 1:1. The ratio in submitochondrial particles, where half of the enzyme was activated, was the same as that of mitochondria, indicating that the inhibitor protein has affinity for the mitochondrial membrane as well as for F1-ATPase. The inhibitor protein could be removed from the mitochondrial membrane by incubation with 0.5 M Na2SO4 and concomitantly the enzyme was fully activated. The enzyme fully activated by the salt treatment was inactivated again by the externally added ATPase inhibitor in the presence of ATP and Mg2+. The enzyme-inhibitor complex (inactive) on the mitochondrial membrane was more stable than the solubilized enzyme-inhibitor complex but gradually dissociated in the absence of ATP and Mg2+. However, in mitochondria, the enzyme activity was inhibited even in the absence of the cofactors. A protein factor stabilizing the enzyme-inhibitor complex on the mitochondrial membrane was isolated from yeast mitochondria. This factor stabilized the inhibitor complex of membrane-bound ATPase while having no effect on that of purified F1-ATPase. It also efficiently facilitated the binding of the inhibitor to membrane-bound ATPase to form the complex, which reversibly dissociated at slightly alkaline pH.     

Biogenesis of mitochondria: Defective yeast H+-ATPase assembled in the absence of mitochondrial protein synthesis is membrane associated

We have investigated the extent to which the assembly of the cytoplasmically synthesized subunits of the H⁺-ATPase can proceed in a mtDNA-less (rho°) strain of yeast, which is not capable of mitochondrial protein synthesis. Three of the membrane sector proteins of the yeast H⁺-ATPase are synthesized in the mitochondria, and it is important to determine whether the presence of these subunits is essential for the assembly of the imported subunits to the inner mitochondrial membrane. A monoclonal antibody against the cytoplasmically synthesized -subunit of the H⁺-ATPase was used to immunoprecipitate the assembled subunits of the enzyme complex. Our results indicate that the imported subunits of the H⁺-ATPase can be assembled in this mutant, into a defective complex which could be shown to be associated with the mitochondrial membrane by the analysis of the Arrhenius kinetics of the mutant mitochondrial ATPase activity.This paper is No. 61 in the seriesBiogenesis of Mitochondria. For paper No. 60, see Novitskiet al. (1984).     

The biogenesis of rat liver mitochondrial ATPase. Subunit composition of the normal ATPase complex and of the deficient complex formed when mitochondrial protein synthesis is blocked.

1. An ATPase complex containing 12 subunits was isoalted from rat liver mitochondria. 2. In vivo inhibition of mitochondrial protein synthesis by the chloramphenicol analogue thiamphenicol leads to the formation of an oligomycin-insensitive membrane-bound ATPase complex in mitochondria of regenerating rat liver. 3. This oligomycin-insensitive, membrane-bound ATPase was isolated by the same procedure as the ATPase complex from regenerating livers of untreated animals. 4. SDS-polyacrylamide gel electrophoresis of in vivo labelled ATPase complexes from control and from thiamphenicol-treated rats reveals that three subunits out of the 12 are not synthesized or assembled when the mitochondrial translation activity is blocked. 5. From the subunits synthesized and assembled when mitochondrial pror (Fo) of the ATPase complex (subunit 5). 6. The oligomycin sensitivity-conferring protein seems absent in the ATPase complex formed in the presence of thiamphenicol.     

Aurovertin binds to the beta subunit of yeast mitochondrial ATPase.

Isolated beta subunit of ATPase (F1) from yeast mitochondria does not catalyze an ATPase reaction but still binds the specific F1 inhibitor aurovertin. Binding was measured by enhancement of aurovertin fluorescence; it was as tight as that to F1-ATPase. No binding was observed with F1 or with isolated beta subunit from a single-gene nuclear yeast mutant whose F1-ATPase was resistant to aurovertin.     

Structure of the yeast mitochondrial adenosine triphosphatase. Results of trypsin degradation

A comparison was made of the subunit sensitivities of the F1-ATPase and the Triton-solubilized ATPase complex to trypsin degradation. The dissociation of the F1-ATPase from ATPase complex increased the trypsin sensitivity of subunit 3 by a factor of 2 and increased the sensitivity of a particular trypsin site (or group of sites) on subunit 1 by 7-fold. The overall degradation of subunits 1 and 2 appears to be the same in solubilized ATPase complex and the F1-ATPase. Implications of these findings for structural models of the ATPase complex are discussed.     

Assembly of imported subunit 8 into the ATP synthase complex of isolated yeast mitochondria

This study concerns the assembly into a multisubunit enzyme complex of a small hydrophobic protein imported into isolated mitochondria. Subunit 8 of yeast mitochondrial ATPase (normally a mitochondrial gene product) was expressed in vitro as a chimaeric precursor N9L/Y8-1, which includes an N-terminal-cleavable transit peptide to direct its import into mitochondria. Assembly into the enzyme complex of the imported subunit 8 was monitored by immunoadsorption using an immobilized anti-F1-beta monoclonal antibody. Preliminary experiments showed that N9L/Y8-1 imported into normal rho+ mitochondria, with its complement of fully assembled ATPase, did not lead to an appreciable assembly of the exogenous subunit 8. With the expectation that mitochondria previously depleted of subunit 8 could allow such assembly in vitro, target mitochondria were prepared from genetically modified yeast cells in which synthesis of subunit 8 was specifically blocked. Initially, mitochondria were prepared from strain M31, a mit- mutant completely incapable of intramitochondrial biosynthesis of subunit 8. These mit- mitochondria however were unsuitable for assembly studies because they could not import protein in vitro. A controlled depletion strategy was then evolved. An artificial nuclear gene encoding N9L/Y8-1 was brought under the control of a inducible promoter GAL1. This regulated gene construct, in a low copy number yeast expression vector, was introduced into strain M31 to generate strain YGL-1. Galactose control of the expression of N9L/Y8-1 was demonstrated by the ability of strain YGL-1 to grow vigorously on galactose as a carbon source, and by the inability to utilize ethanol alone for prolonged periods of growth. The measurement of bioenergetic parameters in mitochondria from YGL-1 cells experimentally depleted of subunit 8, by transferring growing cells from galactose to ethanol, was consistent with the presence in mitochondria of a mosaic of ATPase, namely fully assembled functional ATPase complexes and partially assembled complexes with defective F0 sectors. These mitochondria demonstrated very efficient import of N9L/Y8-1 and readily incorporated the imported processed subunit 8 protein into ATPase. Comparison of the kinetics of import and assembly of subunit 8 showed that assembly was noticeably delayed with respect to import. These findings open the way to a new systematic analysis of the assembly of imported proteins into multisubunit mitochondrial enzyme complexes.     

Influence of the alpha-, beta- and gamma-subunits of the energy-transducing adenosine triphosphates from Micrococcus lysodeikticus in the immunochemical properties of the protein and in their reconstitution studied by a radioimmunoassay method.

We developed a sensitive and specific radioimmunoassay of the energy-transducing adenosine triphosphatase (F1-ATPase, EC 3.6.1.3) of Micrococcus lysodeikticus and extended the assay to the alpha-, beta- and gamma-subunits of the enzyme. We isolated these subunits and studied cross-reactions. We found the immunochemical properties of alpha- and beta-subunits to differ, and gamma-subunits showed an intermediate behaviour between that of alpha- and beta-subunits. Our findings indicate that each subunit of M. lysodeikticus F1-ATPase has its own identity and that conformational antigenic determinants and/or co-operative antigenic sites-arise from subunit assembly. Equimolecular amounts of alpha- and beta-subunits (up to three copies of each) reconstituted partially the immunochemical properties of the ATPase molecule, and addition of 2 mol of gamma-subunit per mol of alpha 3 beta 3 complex improved reconstitution. Our findings describe the first reconstitution of biological activity of this ATPase by assembly of the isolated subunits, and provide support for earlier proposals on the stoicheiometry of the alpha 3 beta 3 gamma 2 type for M. lysodeikticus F1-ATPase. The radioimmunoassay method affords opportunities to study the homologies between different energy-transducing ATPases and their constituent polypeptides before the primary structure of these complex proteins has been determined.     

Monoclonal antibodies to Escherichia coli F1-ATPase. Correlation of binding site location with interspecies cross-reactivity and effects on enzyme activity

Twenty-one hybridoma cell lines which secret antibodies to the subunits of the Escherichia coli F1-ATPase were produced. Included within the set are four antibodies which are specific for alpha, six for beta, three for gamma, four for delta and four for epsilon. The antibodies were divided into binding competition subgroups. Two such competition subgroups are represented for the alpha, beta, and epsilon subunits, one for delta and three for gamma. The ability to bind intact F1-ATPase was demonstrated for some of the antibodies to alpha and beta, and for all of those to delta, while the antibodies to gamma and epsilon gave unclear results. All of the antibodies to alpha and beta which bound ATPase were found to have effects on the ATPase activity of purified E. coli F1-ATPase. One of those to alpha inhibited activity by about 30%. Another anti-alpha was mildly stimulatory. The four antibodies to beta which bound ATPase inhibited activity by 90%. In contrast, membrane-bound ATPase was hardly affected by the antibodies to alpha, but was inhibited by 40-60% by the antibodies to beta. The other antibodies to alpha and beta bound only free subunits, or partially dissociated ATPase, suggesting that their epitopes are buried between subunits in ATPase. These antibodies had no effects on activity. The ability of the antibodies to recognize ATPase subunits present in crude extracts from mitochondria, chloroplasts, and a variety of bacteria was tested using nitrocellulose blots of sodium dodecyl sulfate-polyacrylamide gels. One anti-beta specifically recognized proteins in the range of 50,000-60,000 daltons in each of the extracts, although the reaction with mitochondrial beta was weak. Some of the other antibodies had limited cross-reaction, but most were specific for the E. coli protein. In some species, those proteins which were recognized by the anti-beta ran with a higher apparent molecular weight than proteins which were recognized by an anti-alpha. All antibodies which exhibited cross-reactivity were found to recognize sites which were not exposed in intact ATPase, implying that the surfaces which lie between subunits are most highly conserved.     

Atp10p assists assembly of Atp6p into the F0 unit of the yeast mitochondrial ATPase

The F(0)F(1)-ATPase complex of yeast mitochondria contains three mitochondrial and at least 17 nuclear gene products. The coordinate assembly of mitochondrial and cytosolic translation products relies on chaperones and specific factors that stabilize the pools of some unassembled subunits. Atp10p was identified as a mitochondrial inner membrane component necessary for the biogenesis of the hydrophobic F(0) sector of the ATPase. Here we show that, following its synthesis on mitochondrial ribosomes, subunit 6 of the ATPase (Atp6p) can be cross-linked to Atp10p. This interaction is required for the integration of Atp6p into a partially assembled subcomplex of the ATPase. Pulse labeling and chase of mitochondrial translation products in vivo indicate that Atp6p is less stable and more rapidly degraded in an atp10 null mutant than in wild type. Based on these observations, we propose Atp10p to be an Atp6p-specific chaperone that facilitates the incorporation of Atp6p into an intermediate subcomplex of ATPase subunits.     

Energy-linked reactions catalyzed by the purified ATPase complex (F0F1) from Rhodospirillum rubrum chromatophores

1. The isolation of F0F1-ATPase complex from Rhodospirillum rubrum chromatophores by the use of taurodeoxycholate is described. 2. The enzyme preparation contains about 12 polypeptides; five are subunits of the F1 moiety. 3. The ATPase activity of the purified enzyme is dependent on the addition of phospholipids. 4. Km-vales for Mg2+-ATP and Ca2+-ATP are similar to the values obtained for the membrane-bound enzyme. 5. The F0F1-ATPase complex is more than 70% inhibited by oligomycin and N,N'-dicyclohexylcarbodiimide. 6. The F0F1-ATPase complex was integrated into liposomes. The reconstituted proteoliposomes catalyzed energy transduction as shown by ATP-dependent quenching of acridine dye fluorescence and ATP-32Pi exchange.     

Guanidine-induced dissociation of mitochondrial F1-ATPase

The effect of guanidine hydrochloride on ATPase activity, gel filtration, turbidity, and the fluorescence emission intensity of mitochondrial F1-ATPase was examined. Purified F1 from bovine heart mitochondria was slowly inactivated at low denaturant concentration, and inactivation was associated with delta and epsilon subunit dissociation. delta and epsilon subunits were bound together to form a stable and soluble heterodimer. In parallel, appearance of turbidity was observed. This was caused by the formation of alpha3beta3gamma non-covalent aggregates, as analyzed by SDS-PAGE. Short periods of exposition of the F1 complex to high concentrations of guanidine hydrochloride (0.8-3 M) again induced deltaepsilon dissociation as a heterodimer and the formation of an inactive alpha3beta3gamma subcomplex. This eventually dissociated progressively into single subunits caused by partial unfolding, as evidenced through changes of the protein intrinsic fluorescence emission. Our results suggest that the delta and epsilon subunits are loosely bound to alpha3beta3gamma , and play an important role in determining structural stability to isolated mitochondrial F1-ATPase.