Cloning of the yeastATP3gene coding for the γ-subunit of F1and characterization ofatp3mutants.

The binding of a new calcium sensitizer, levosimendan, to human cardiac troponin C (cTnC) is described. Fluorescence studies done on dansylated recombinant human cTnC and a site-directed mutant showed that levosimendan modulated the calcium-induced conformational change in cTnC, and revealed the role of Asp-88 in the binding of the drug to the NH2-terminal domain of cTnC. Furthermore, NMR studies performed on the NH2-terminal fragment of cTnC showed a spatial proximity between levosimendan and Met81, Met85, and Phe77 in the drug-protein complex. These data were used to build an optimized model of the drug-protein complex, in which levosimendan binds cTnC at the hydrophobic pocket of the NH2-terminal domain. The role of the binding of levosimendan to cTnC in the pharmacological action of this drug in vivo is discussed. (Less)


assembly/stability of F,.
Mitochondrial and bacterial proton-translocating ATPases are composed of two functionally and physically distinguishable parts (1,2). The F,-ATPase is a water soluble, heteroligomeric enzyme made up of five different subunits. F,-ATPase is attached to a hydrophobic group of proteins located in the inner membrane and jointly referred to as Fa. The F, component, depending on the source, is composed of 3-11 subunits (3, 4). Functionally, Fa is responsible for promoting energy-dependent proton translocation across the inner membrane. In Saccharomyces cereuisiae, the subunits of F, and all but three subunits of F, a r e encoded by nuclear genes (4). All the structural genes of yeast F, except for the y-subunit have been cloned, and the phenotypes elicited by null mutations in this set of ATPase constituents have been described (5)(6)(7).
To complete the characterization of the F, genes in yeast, we have screened pet' mutants for defects in mitochondrial ATPase. Complementation group G115 of the mutant collection * This research was supported by National Institutes of Health Grant HL22174. 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.
The nucleotide sequence(s1 reported in this paper has been submitted to the GenBankTM/EMBL Data Bank with accession numberrs) U09305.
The abbreviation used is: pet, mutants of S. cerevisiae defective in respiratory functions as a result of mutations in nuclear DNA.
consists of two independent mutants, both of which have abnormally low ATPase activity. This phenotype has previously been noted in strains that are unable to express functional F,-ATPase, because of either mutations in the structural subunits of the enzyme (5)(6)(7) or in ancillary factors necessary for assembly of the subunits into the functional oligomer (8). In this communication we show that the ATPase defect of G115 mutants is caused by mutations in the ATP3 gene for y-subunit of F,. The y-subunit mutants have been used to clone and characterize the ATP3 gene and to assess the role of this component in the assembly and catalytic activity of F,.
Cloning of the ATP3 Gene-The ATPase-deficient mutant C287LU1 (ol,leu2-3,112,ura3-l,atp3-1) was transformed by the method of Schiestl and Gietz (12) with a yeast genomic library consisting of a partial Sau3AI fragment of nuclear DNA ligated to the BamHI site of YEp24 (13). This library was kindly provided by Dr. Marian Carlson, Department of Human Genetics, Columbia University. Transformation of approximately 10, cells with 50 pg of plasmid DNA yielded 2 x lo4 uracilindependent clones of which four were also respiration competent based on their ability to grow on glycerol as a carbon source. Segregation tests indicated that the restoration of respiratory competence was plasmiddependent in only one of the four respiration-competent transformants. This clone, designated C287/LUl/Tl, was used to isolate plasmid pG115/T1 with the ATP3 gene.
Purification and Biochemical Analysis of y-Subunit-Purified y-subunit was obtained from the strain D273-10B/A/H/U (ol,met,ura3,his3). F, subunits were purified from a chloroform extract of mitochondria as described by Arselin et a2. (14). The dried y-subunit (1.5 nmol) was solubilized in 0.05% SDS, 1% NH,CO,, pH 8, by sonication for 5 s and incubation first at 37 "C for 5 h and then at 65 "C for 15 min. Alternatively, the protein was solubilized in 8 M urea, 0.4 M NH,CO,, pH 8. The soluble protein was carboxamidomethylated according to Stone et al. (15) and cleaved with L-1-tosylomido-2-phenylethyl chloromethyl ketone-treated trypsin (Sigma) for 18 h at 37 "C at a substrate/enzyme ratio of 20. The tryptic peptides were loaded onto a 5-pm Vydak C,, column and separated as described (15). Automated sequence analysis was performed with an Applied Biosystems liquid-phase sequenator.
Preparation of Yeast Mitochondria and Submitochondrial Particles-Yeast was grown in YPGal or YPD at 30 "C to early stationary phase. Mitochondria were prepared, according to Faye et al. (16) except that zymolyase 20,000 (ICN Biomedicals) instead of Glusulase was used to digest cell walls. Submitochondrial particles were prepared by sonic irradiation of a suspension of mitochondria (4 mg/ml) for 10 s using a Branson model 185 sonifier equipped with a microprobe. After centrifugation at 40,000 rpm for 20 min in a 50Ti Beckman rotor, the membranes were resuspended in 20 m M Tris-HC1, pH 7.5, 0.1 m M EDTA.
Preparation of Antibodies Directed against ATPase Subunit 4 and Subunit d-A part of the ATP4 and ATP7 reading frames, coding for subunits 4 (17) and d (18) ofthe yeast ATPase, respectively, were ligated in frame to the amino terminus of the Escherichia coli DpE gene in the expression vectors PATH 2 and PATH 11 (19). Bacterial transformants were used to express the fusion proteins as described previously (19). The partially purified proteins were used as antigens to raise antibodies in rabbits.
Cloning of the atp3 Gene from C287/LUl-Chromosomal DNA purified from C287LUl was used as a template for polymerase chain reaction amplification of the mutant atp3 gene. One of the two synthetic primers had the sequence matching the sense strand from nucleotides -45 to -15 (see Fig. 2) except for two base changes that were introduced to create an EcoRI site. The second primer was complementary to the sense strand from nucleotides +932 to +955 of the sequence except for two base changes forming a BgZII site. The polymerase chain reaction product obtained from the synthesis was digested with a combination of EcoRI and BglII and was ligated to yEp352 (20) linearized with EcoRI and BamHI.
Miscellaneous Procedures-Standard techniques were used for restriction endonuclease analysis of DNA, purification and ligation of DNA fragments, and transformation and recovery of plasmid DNA from E. coli (21). DNA was sequenced by the method of Maxam and Gilbert (22). For Western analysis of a-and p-subunits, mitochondrial proteins were separated on 10% polyacrylamide gels run in the electrophoretic system of Laemmli (23) except that the separation buffer was adjusted to pH 8 and the composition of the running buffer was 0.05 M Tris, 0.38 M glycine, and 0.1% SDS. Following electrophoretic transfer of the proteins to nitrocellulose, the blot was reacted with antibodies as described by Schmidt et al. (24). Mitochondrial proteins separated on 15% polyacrylamide gels containing 15% glycerol were also detected by staining with silver nitrate.
ATPase activity was measured by the colorimetric determination of inorganic phosphate released from ATP (25). Protein concentration was estimated by the method of Lowry et al. (26). Proteolipid components of the ATPase were extracted from mitochondria as previously described (27).

RESULTS
Cloning and Disruption of ATP3-The mutants reported in this study were selected from a pet mutant collection screened for alterations in oligomycin-sensitive ATPase activity. Strains assigned to complementation group G115, henceforth referred to as atp3 mutants, were found to have a low ATPase activity (Table I) indicative of a lesion in F,. Since the atp3 mutants were complemented by testers bearing mutations in the genes for the a-, p-, and +subunits, the most likely explanation for the ATPase-deficient phenotype was a mutation in either the structural gene for the y-or h u b u n i t of F, or an ancillary protein needed for proper assembly of the F, oligomer.
To identify the primary biochemical lesion responsible for the low mitochondrial ATPase activity, the gene affected by the mutation was cloned by transformation of the atp3 mutant C287/LU1 with a genomic library. The transformation yielded a single clone, C287/LUl/Tl, whose restoration of respiratory competence and mitochondrial ATPase was found to segregate during vegetative growth and, therefore, to be dependent on the presence of a plasmid. This plasmid, designated pG115PT1, was further characterized by subcloning various regions of the nuclear DNA insert. The results of the complementation tests carried out with the new constructs indicated that the smallest plasmid (pG115/ST5) capable of restoring wild type growth to the mutant contained a 2.4-kilobase region of the original insert located between the unique XbaI and Hind111 sites (Fig. 1). The sequence of the pG115BT5 insert disclosed the presence of a 933-nucleotide-long reading frame initiated by a methionine codon 373 nucleotides downstream of the XbaI site (Fig.  2). The deduced primary structure of the protein encoded by this gene was homologous to the y-subunits of bacterial, plant, and mammalian F,-ATPases (Fig. 3). Two regions of the protein defined by the amino-terminal 27 residues (30% identical residues) of the mature protein and the carboxyl-terminal 50 residues (30%) showed the highest extent of sequence similarity.
These two domains are also the most highly conserved among all the other known y-subunits.
The identity of the reading frame as the ATP3 gene for the y-subunit was confirmed by the partial sequence of peptides obtained after tryptic digestion of the mature protein isolated from yeast F,. The sequence of the amino-terminal 28 residues of the y-subunit matched the sequence starting from Ala34 of the product encoded by ATP3. The sequences of four other internal tryptic peptides were also consistent with the sequence inferred from the gene except for residue 207, which is specified by a TCT codon for serine in the DNA sequence but was confirmed to be an isoleucine by protein sequence analysis. This discrepancy may stem from a polymorphism between the strains of S. cereuisiae used as sources of the genomic DNA and of the purified y-subunit. The absence of the first 33 amino acids in the mature y-subunit is in agreement with earlier evidence indicating the presence of a long amino-terminal mitochondrial targeting sequence in the primary translation product (31,32).
Two lines of evidence indicated that the phenotype of C287 and derivative mutants is due to a mutation in the y-subunit. The cloned ATP3 gene was used to institute a deletion in the chromosomal copy of the gene. The region between the two BclI sites of the coding sequence in pG115/ST5 was replaced by a 1.7-kilobase BarnHI fragment containing the yeast HIS3 gene. The deleted allele (atp3SZS3 was isolated as linear 4-kilobase XbaI-Hind111 fragment and used to transform the wild type strain W303-1B by the one-step gene replacement procedure (33). The respiration-deficient and histidine-independent clones issued from the transformation were further screened by genetic crosses to known testers. The respiratory deficiency of one such clone, W303AATP3, was complemented by a p" strain but not by the atp3 mutant B287. Southern analysis of genomic DNA from W303AATP3 confirmed the presence of the atp3:BZS3 allele. The lack of complementation by W303AATP3, therefore, suggests that B287 has a mutation in ATP3. This was confirmed by allelism tests. C287LU1 was transformed with an integrative plasmid containing URA3 and ATP3. This plasmid was linearized at the unique BstXI site in ATP3 to direct integration at the chromosomal atpd locus. A respiration-competent and Ura+ transformant was crossed to W303-1A and W303AATP3. Diploid cells formed in each cross were sporulated for tetrad dissections. In the cross to the wild type parental strain W303-1A the meiotic progeny of 10 complete tetrads dissected showed 2:2 segregation of the uracil auxotrophy and were 100% respiratory competent indicating that ATP3 had integrated at the locus of the mutation in C2871 LU1. The meiotic progeny issued from the cross to W303AATP3 also showed 2:2 segregation of the uracil auxotrophy. In this case, however, the Ura' spores were respiratory competent, and the Ura-spores were respiratory deficient. The cosegregation of the Ura' and respiration-competent phenotypes in this cross further confirms the genetic linkage of the mutant allele in C287LU1 to the atp3:flIS3 allele in W303AATP3.
To characterize the mutation in the atp3 gene of C287LU1, the coding sequence was amplified in two independent polymerase chain reactions. The two products were separately cloned in the shuttle vector YEp352 (20). In both cases the sequence of the insert showed the same single base change (C + T) at nucleotide 818 of the gene leading t o a substitution ~f A l a '~~ by a valine. The identity of this amino acid change as the mutation in C287 was also confirmed by the inability of the mutant gene to complement W303AATP3 either on a multicopy plasmid or in single copy after integration into chromosomal DNA.
Catalytic Properties of Fl in atpd Mutants-The effects of mutations in the y-subunit on the structure and activity of F,-ATPase were studied in both W303AATP3, a mutant construct with a deletion spanning half of the coding sequence of ATP3, and in the point mutant C287iLUl. The mitochondrial contents of a-and P-subunits of F, in the mutants and in the wild type strain were estimated immunochemically with subunit-specific antibodies. The results of the Western blot analyses indicated these subunits to be present in both mutants at approximately the same concentrations as in wild type (Fig. 4). Neither the point mutation in the y-subunit nor the absence of the protein altogether appears to compromise the synthesis, import, or stability of the a-and @subunits that bear the catalytic sites for ATPase hydrolysislsynthesis (1, 2, 4). Similar results were obtained when the immunochemical analyses were done on fractions enriched in F, (data not shown). Notwithstanding the presence of nearly wild type concentrations of these F, subunits, the mitochondrial ATPase activity in both mutants was significantly reduced compared with the wild type ( Table 11). The total ATPase activity in W303AATP3 is approximately three times lower than in the parental wild type strain. The fraction of ATPase activity inhibited by oligomycin, which reflects more accurately the F,-dependent hydrolysis, is only 10% of the wild type. The Ala273 + Val substitution also causes a greater than 2-fold decrease in total ATPase activity, and in this mutant, oligomycin-sensitive activity is five times lower than in the wild type.
The effect of the mutations on F,-catalyzed ATP hydrolysis is also revealed by the activities measured in mitochondrial fractions enriched for F,. Approximately equivalent amounts of aand P-subunits were extracted from wild type and mutant mitochondria by extensive sonic irradiation as judged by Western blot analysis. Measurements of ATPase indicated absence of activity in the extract of W303AATP3 and only 10% of wild type activity in the extract of C287LU1 (Table 111).
Membrane Association of Fl Subunits in y-Subunit Mutants-Several approaches were used to study the role of the y-subunit in assembly of oligomeric F, and in the interaction of F, with the hydrophobic F, membrane unit of the larger ATPase complex. The F, subunits in the active oligomer are present in a a3P3y& stoichiometry (34). Native F, is tightly bound to the F, subunits, its dissociation from the membrane requiring prolonged sonic irradiation (25) or other physical shearing procedures (35).
Brief sonic irradiation of mitochondria, conditions commonly used in the preparation of submitochondrial vesicles, caused only a small percentage of the a-and P-subunits to be solubilized from either wild type or C287LU1 (Fig. 41, suggesting the point mutation in the y-subunit does not prevent the interaction of these F, subunits with the membrane components. Strikingly different results were obtained with W303AATP3, the mutant with a deletion in ATP3. In this case, sonic irradiation resulted in release of a significant part of F, subunits in a soluble form, similar to that observed with mitochondria of a po strain. The latter type of mutant is capable of assembling F, subunits into the catalytically active oligomer, but the enzyme fails to bind to the membrane because of the absence of the mitochondrially encoded subunits 6, 8, and 9 of the F, component (36). It is not clear at present whether the a-and P-subunits in W303AATP3 are loosely bound to the membrane or exist as dissociated proteins in the matrix. While atp3 mutations increase the frequency of spontaneous deletions in mitochondrial DNA, growth of W303AATP3 on rich glucose medium (YPD) tends to minimize such events, and in most experiments the percentage of po and p-mutants is of the order of 20% of the total population in a stationary phase culture. The nearly quantitative recovery of F, subunits as soluble proteins following sonic disruption of W303AATP3 mitochondria cannot be solely the consequence of the accumulation of po mutations by this strain, since in the experiment of Fig. 4 only 20% of the culture used for the preparation of mitochondria was converted to po" mutants.  The sedimentation properties of the a-and P-subunits in suextracted from wild type and from C287LU1 mitochondria have crose gradients differ depending on whether they exist as monomeric or oligomeric proteins (37). Their behavior in such gradients can, therefore, be used to gauge the state of assembly of F,. To assess the physical state of F, in the y-subunit mutants, mitochondria of C287LU1 and W303AATP3 were sonically irradiated, and the soluble proteins released by this treatment were separated from small membrane vesicles by differential centrifugation. The sonic extracts were applied to sucrose gradients, and the sedimentation of the F, subunits was determined by Western blot analysis of the gradient fractions (Fig. 5). The results of this experiment indicate that the a-and p-subunits sedimentation properties expected of the F, oligomer. Even though the Ala273 + Val mutation abolishes catalytic function as evidenced by the low ATPase activity in the mutant extract, the mutation does not affect assembly or stability of F,. Again this was not true of W303AATP3, a strain in which synthesis of the y-subunit is prevented by the long deletion in ATP3. The F, subunits extracted by sonic treatment of mitochondria from this mutant were detected in regions of the gradient corresponding to monomers or partial oligomers. The absence of y-subunit in this strain, therefore, appears either to prevent formation of the F, oligomer or causes the enzyme to have an unstable quater-

s -L S P K G E I C D I N G K C V D A A M D E L F R L T T K E G K L T V E
- nary structure that is lost during the extraction or subsequent sucrose gradient centrifugation. Effect of the Ala273 4 Val Mutation on Interaction of Fl with F,--The concentration of F, components in atp3 mutants was examined by electrophoretic separation and staining of the proteolipid components (subunits 6 and 9) extracted from mitochondria with chloroform-methanol (27) and by immunochemical detection of subunits 4 and d among total mitochondrial proteins. These assays indicated that the mitochondrial concentrations of the F, subunits were reduced by more than 70% in W303AATP3 but not in C287LU1 in which the proteins were present in amounts comparable with those of wild type (data not shown).

y -Y I S I L Y N R T R Q A V I T N E L V D I I T G A I S S m G
The marginal effect of the Ala273 mutation on the abundance of F, subunits combined with the observed tight association of an oligomer-sized F, t o the membrane implied the existence of an inactive but physically intact F,-F, complex in the mutant. This was tested by analyzing the sedimentation properties of the ATPase extracted from mutant mitochondria with Triton X-100. Addition to mitochondria of low concentrations of this  W303MTP3 ( M T P 3 ) , and C287LU1 and suspended at a protein concentration of 4 mg/ml in 10 mM "is-HCI, pH 7.5, 1 mM EDTA, and 2 mM ATP (TEA). The suspensions were sonically irradiated for 10 s and centrifuged a t 40,000 rpm for 20 min (23 "C) in a Beckman 50Ti rotor. The supernatants were removed, and the pellets consisting of submitochondrial particles were resuspended in the starting volume ofTEA. Equivalent volumes of mitochondria (MI, submitochondrial particles ( P ) , and the supernatant fractions ( S ) were separated on a 10% polyacrylamide gel and transferred to nitrocellulose. The Western blots were probed with a mixture of antibodies against a-and P-subunits. ATPase activity of mitochondrial extracts from wild type and atp3 mutants Mitochondria suspended in TEA at a protein concentration of 6 mdml were sonically irradiated for a total of 30 s and centrifuged at 40,000 rpm for 20 min. The supernatants (30 pl) were assayed for ATPase in the presence of 10 mM ATP a t 37 "C for 12 min. The activities reported are based on duplicate assays.

Strain ATPase
Exp. detergent solubilizes the oligomycin-sensitive F1-Fo complex, which sediments with an apparent size of 550 kDa in sucrose gradients (37). Analysis of such an extract from C287LU1 mitochondria confirmed that the F, subunits in this mutant cosedimented with Fo as a high molecular weight complex (Fig.  6). In this experiment proteins were visualized by silver staining following separation by electrophoresis. Fractions of the gradient expected to contain F1-Fo had a pattern of proteins characteristic of the wild type complex. The identity of some Fo subunits was confirmed by immunochemical probings with subunit-specific antibodies and of the F, subunits with an antibody against the holoenzyme (Fig. 6). These data provide additional evidence that the Ala273 + Val mutation in the y-subunit exerts its effect predominantly on the catalytic activity of F, rather than on its assembly or interaction with F , .

DISCUSSION
Previous biochemical screens of a collection of respirationdeficient pet strains of s. cerevisiae have helped to identify mutants in the a-and P-subunits of the F,-ATPase (5,6,38). In the present study we report another group of mutants whose low mitochondrial ATPase activity is shown to be caused by mutations in the y-subunit of F,. C287LU1, a representative of LU1. The ATPase complex was extracted from submitochondrial particles of the C287lLU1 with Triton X-100 as described previously (37). The Triton extract was layered on top of 4.5 ml of a linear 5-20% (wh) linear sucrose gradient containing 5 mM Tris acetate, pH 7.5, and 0.2% Triton X-100. The gradient was centrifuged in a Beckman SW 65 rotor a t 65,000 rpm for 3 h (23 "C), and 13 fractions were collected from the bottom of the tube. A sample of each fraction was separated on a 15% polyacrylamide gel, and the proteins were visualized either by staining with silver nitrate or with antibodies against different subunits of the F, and F, after transfer to nitrocellulose. The results are shown only for fractions 2-5, representing the lower third of the gradient. A , gradient fractions 2-5 were stained with silver nitrate. The identified components of the F,-F, are marked in the left-hand margin. Subunit 6 was also detected but because of its poor staining is not seen on the print. B, the proteins of the peak ATPase fractions 3 and 4 were transferred to nitrocellulose and probed with antibodies to a mixture of subunits d and 4 of the F,. C, the peak gradient fractions 3 and 4 were probed with antibodies to yeast F,. The migration of size standards is marked on the righthand side of eachpanel. The Western blot was overloaded in order to visualize the poorly antigenic 6 and +subunits. The proteins marked with an asterisk co-purified with the F, preparation (25) used to raise antibodies.
this complementation group (G115), has enabled us to clone the ATP3 gene that restores respiratory function to this mutant. Several lines of evidence indicate that ATP3 codes for the y-subunit of the yeast mitochondrial ATPase. Deletions of half of the coding sequence from the chromosomal copy of the gene elicits a deficiency in the F,-ATPase activity. In addition to being homologous to the y-subunits of a large number of different bacterial, mitochondrial, and chloroplast F,-ATPases, the identity of the ATP3 product as the y-subunit has been confirmed by direct sequencing of several tryptic peptides of the mature yeast protein.
Information concerning the role of the y-subunit in the function and structure of F, has been derived almost exclusively from studies of ECF, from E. coli (39) and TF, of the thermophilic bacterium PS3 (40) largely because of the feasibility of reconstituting the oligomeric enzymes from the separated bacterial subunits. The E. coli ATPase also lends itself to genetic manipulation, thereby allowing biochemical investigations to be extended to mutants with either natural or designed modifications in the primary structure of the different subunits.
Reconstitution experiments with the ECF, subunits (39, 41) indicate that a3P3y is the minimal complex capable of catalyzing ATP hydrolysis with high specific activity. A complex with high multisite catalytic activity has been reconstituted from the a-, P-, and &subunits of the thermophilic TF, ATPase   (40, 42). The a3P36 complex, however, is not stable and shows altered nucleotide and divalent metal specificity (42). Both stability and native catalytic properties are restored in the TF, complex by the addition of y-subunit. These studies support the notion that the y-subunit not only contributes to maintaining the basic aaP3 core but also exerts more subtle effects on subunit interactions needed for catalytic activity. A dual structural and functional role of the y-subunit is also supported by the properties of the mutants reported here. Even though the mitochondrial concentrations of a-and P-subunits are almost normal in a mutant with a large deletion in ATP3, they are not part of an F, type of oligomer. Brief sonic treatment of mitochondria from the mutant leads to a quantitative release of the F, subunits into the soluble fraction. The extracted proteins, however, sediment either as a monomer (@subunit) or as a heterodisperse oligomer (a-subunit). The absence of a n F, complex in the y-less mutant suggests that the y-subunit is necessary either for assembly of the a-and P-subunit core structure or for preventing this complex from dissociating. The latter interpretation seems more likely in view of the in vitro reconstitution of a stoichiometric a3&6 complex from the isolated subunits of the thermophilic enzyme (40,421.
In contrast, the mutation in the highly conserved alanine residue near the carboxyl-terminal end of the protein has a much less deleterious effect on the structure of F,. This mutation does not preempt the association of the F, subunits into an oligomeric structure of a size comparable with F,. The mutant F, is tightly associated with the F, membrane unit as evidenced by the composition of the F,-F, complex extracted with detergent and purified by centrifugation on a sucrose sizing gradient. This purified complex contains all the subunits of F, including the mutant y-subunit. Dissociation of the mutant F, from the F, constituents by extensive sonic treatment of mitochondrial membranes does not appear to affect its stability since the a-and @-subunits in the extract are associated in a native size F, complex. The ATPase activity of the mutant F,, however, is only 10% of that measured in wild type. The alanine residue, therefore, is essential in establishing a catalytically competent conformation of the enzyme.
The carboxyl-terminal 50 residues constitute the most conserved region of the sequenced y-subunits. Earlier evidence from studies of E. coli mutants suggested that this region is important for both ATPase activity and assembly of F, (43).
Nested truncations of the region comprised by the carboxylterminal 26 residues of the E. coli protein indicate a progressive loss of catalytic activity followed by the failure of the subunits to form the oligomer. For example, deletion of the carboxyl-terminal 10 residues as well as various amino acid replacements in this region reduces the activity by more than 80% without affecting assembly (44). A longer deletion inclusive of the carboxyl-terminal 26 amino acids, however, causes loss not only of function but also of a stable oligomer (44). The mutant allele of the yeast enzyme reported in this study suggests that the region of the y-subunit critical for enzyme activity extends considerably further and may include the entire conserved carboxyl-terminal domain.