The F subunit of Thermus thermophilus V1-ATPase promotes ATPase activity but is not necessary for rotation.

V(1)-ATPase from the thermophilic bacterium Thermus thermophilus is a molecular rotary motor with a subunit composition of A(3)B(3)DF, and its central rotor is composed of the D and F subunits. To determine the role of the F subunit, we generated an A(3)B(3)D subcomplex and compared it with A(3)B(3)DF. The ATP hydrolyzing activity of A(3)B(3)D (V(max) = 20 s(-1)) was lower than that of A(3)B(3)DF (V(max) = 31 s(-1)) and was more susceptible to MgADP inhibition during ATP hydrolysis. A(3)B(3)D was able to bind the F subunit to form A(3)B(3)DF. The C-terminally truncated F((Delta85-106)) subunit was also bound to A(3)B(3)D, but the F((Delta69-106)) subunit was not, indicating the importance of residues 69-84 of the F subunit for association with A(3)B(3)D. The ATPase activity of A(3)B(3)DF((Delta85-106)) (V(max) = 24 s(-1)) was intermediate between that of A(3)B(3)D and A(3)B(3)DF. A single molecule experiment showed the rotation of the D subunit in A(3)B(3)D, implying that the F subunit is a dispensable component for rotation itself. Thus, the F subunit binds peripherally to the D subunit, but promotes V(1)-ATPase catalysis.


V 1 -ATPase from the thermophilic bacterium Thermus thermophilus is a molecular rotary motor with a subunit composition of A 3 B 3 DF, and its central rotor is composed of the D and F subunits. To determine the role of the F subunit, we generated an A 3 B 3 D subcomplex and compared it with A 3 B 3 DF. The ATP hydrolyzing activity of A 3 B 3 D (V max ‫؍‬ 20 s ؊1 ) was lower than that of A 3 B 3 DF (V max ‫؍‬ 31 s ؊1 ) and was more susceptible to MgADP inhibition during ATP hydrolysis. A 3 B 3 D was able to bind the F subunit to form A 3 B 3 DF. The C-terminally truncated F (⌬85-106) subunit was also bound to A 3 B 3 D, but the F (⌬69 -106) subunit was not, indicating the importance of residues 69 -84 of the F subunit for association with A 3 B 3 D. The ATPase activity of A 3 B 3 DF (⌬85-106) (V max ‫؍‬ 24 s ؊1 ) was intermediate between that of A 3 B 3 D and A 3 B 3 DF. A single molecule experiment showed the rotation of the D subunit in A 3 B 3 D, implying that the F subunit is a dispensable component for rotation itself.
Thus, the F subunit binds peripherally to the D subunit, but promotes V 1 -ATPase catalysis.
The pH within many intracellular compartments, such as the Golgi apparatus, endosomes, and lysosomes, is regulated by a family of H ϩ -pumping ATPases known as the vacuolar H ϩ -ATPases (V-ATPases) (1,2). V-ATPases exist in the membranes of these organelles of all eukaryotic cells and in the plasma membranes of some specific eukaryotic cells and are involved in a variety of physiological processes. V-ATPases are multisubunit enzymes arranged as a peripheral V 1 complex that is responsible for MgATP hydrolysis and that is attached to a membrane-embedded V 0 complex containing a proton pore. Based on the functional and structural similarity between V-ATPases and F-ATPases, it was assumed that V-ATPases use a rotary mechanism similar to that used by F-ATPases. Recent studies have shown that this is indeed the case (3,4).
A family of V-ATPases also exists in the plasma membranes of some bacteria (5-7). One example is V-ATPase from the thermophilic bacterium Thermus thermophilus (5, 8 -10). The T. thermophilus V-ATPase is capable of both ATP-driven proton translocation and proton-driven ATP synthesis in vitro and functions as an ATP synthase in vivo (4,11). The subunit structure of this V-ATPase is simpler than that of its eukaryotic counterpart. It is composed of nine subunits, A, B, D, F, C, E, G, I, and L (10), with several lines of evidence indicating that the D, F, C, and L subunits form a central rotor and that the I, E, and G subunits constitute a stator apparatus with an A 3 B 3 hexamer (3,4,12,13).
The V 1 complex from T. thermophilus is ATPase-active and is thus known as V 1 -ATPase. The V 1 -ATPase is composed of four subunits with a presumed subunit composition of A 3 B 3 DF. Unlike the F 1 -ATPases, the structure and enzymatic properties of V 1 -ATPase are not well characterized. The central rotor in bacterial F 1 -ATPase is composed of two subunits, ␥ and ⑀ (14). The high resolution structure of F 1 -ATPase shows a hexameric arrangement of alternating ␣ and ␤ subunits that surround a highly ␣-helical ␥ subunit, forming the minimum rotary unit, ␣ 3 ␤ 3 ␥ (15). The F 1 -ATPase ⑀ subunit functions as an endogenous regulator of ATPase activity (16 -19). The central rotor of V 1 -ATPase is also composed of two different subunits, D and F (3). Electron microscopic and cross-linking experiments suggest that the D subunit is located in the central cavity of the A 3 B 3 hexamer and thus is a functional homolog of the ␥ subunit of F 1 -ATPase (10,20). Site-directed and random mutagenesis studies of the yeast D subunit also suggest an analogy between the V 1 -ATPase D subunit and the F 1 -ATPase ␥ subunit (21). In contrast, the function of the F subunit is, to date, still largely unknown. In yeast, the F subunit (Vma7p) is essential for the assembly of V 0 and V 1 to form the functional V 0 V 1 holoenzyme (22,23). Yeast strains lacking this subunit are unable to form the complete V 0 V 1 complex despite a normal amount of V 1 in the cytosol (22). The F subunit from T. thermophilus is a small hydrophilic protein with a molecular mass of 12 kDa and has 21% identity to the F subunit of yeast V-ATPase. However, there is no subunit in F-ATPase that has sequence similarity to the F subunit.
To elucidate the function of the F subunit, two different V 1 complexes (A 3 B 3 DF and A 3 B 3 D) were generated and characterized. The results reveal a novel role for the F subunit in V 1 -ATPase catalysis.

EXPERIMENTAL PROCEDURES
Chemicals-Pyrobest TM DNA polymerase and a DNA ligation kit were purchased from Takara (Kyoto, Japan). ATP and ADP were from Sigma. Restriction endonucleases, pyruvate kinase, lactate dehydrogenase, and NADH were obtained from Roche Applied Science. Oligonucleotides were from Hokkaido System Science (Sapporo, Japan). All other reagents were purchased from Wako Pure Chemical Industries Inc. (Osaka, Japan) unless otherwise noted.
Plasmids-DNA fragments containing genes coding for the F, A, B, and D subunits were amplified from T. thermophilus chromosomal DNA by PCR using oligonucleotide primers 5Ј-CCAGGTGAATTCGAG-GAAGAGGTGGTATGGTGCCCGTGAG-3Ј and 5Ј-AGGAAGAGGCG-GTGTACAAGC-3Ј. After digestion with EcoRI and HindIII, the amplified fragment was ligated between the EcoRI and HindIII sites of pUC118 to obtain pG1-FABD. A His 8 tag was introduced at the N terminus of the A subunit by PCR-based mutagenesis to obtain pG3-FABD. Plasmid pG3-FABD was used for expression of wild-type V 1 -ATPase. Plasmid pG1-DBA, which was used for expression of the A 3 B 3 D subcomplex, was constructed as follows. The gene for the D * 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. subunit was amplified from T. thermophilus chromosomal DNA by PCR using oligonucleotide primers 5Ј-GGAATTCAAGGAGGATTTAGATGA-GCCAGGTGAGCCCCAC-3Ј and 5Ј-CTCTAGACGCGTTAAAGGCCCG-CCCCGATCTCCACC-3Ј. The gene for the B subunit was amplified using oligonucleotide primers 5Ј-AGGCCACGCGTAAGGAGGTTATA-GAGATGGACCTTCTGAAGAAG-3Ј and 5Ј-GCTCTAGACTTAGTCCA-GGGCCTGGGGCGC-3Ј. The gene for the A subunit was amplified using oligonucleotide primers 5Ј-GCTCTAGAAGGAGGATATACATAT-GCACCACCACCACCACCACCACCACATGATCCAAGGGGTGATCC-AG-3Ј and 5Ј-CCTAGGATCGTCGACTTTAGGCCAGGGCCTTGAAGG-3Ј. The amplified D, B, and A fragment genes were digested with EcoRI and MluI, MluI and XbaI, and XbaI and SalI, respectively, followed by ligation between the EcoRI and SalI sites of pUC18 to obtain pG1-DBA, which contains genes for the D, B, and A subunits (in this order) under the control of the lac promoter. Plasmid pET-F, which was used for expression of the wild-type F subunit, was constructed as follows. The gene for the F subunit was amplified by PCR using oligonucleotide primers 5Ј-AGGAGGAGGTGGTCATATGGTGCCCGTGAG-3Ј and 5Ј-ATCATCGTCCCGGATCCATTCTACAGC-3Ј. The amplified DNA was digested with NdeI and BamHI, followed by ligation between the NdeI and BamHI sites of pET-17b (Novagen). The N terminus of the F subunit expressed by pET-F and pG3-FABD was modified by the addition of three amino acid residues (Met-Val-Pro) for efficient expression.
Expression and Purification of V 1 -ATPase and the A 3 B 3 D Subcomplex-Escherichia coli BL21-Codon Plus-RP cells (Stratagene) harboring pG3-FABD or pG1-DBA were cultured for 20 h at 37°C in Super broth (32 g/liter Tryptone, 20 g/liter yeast extract, and 5 g/liter sodium chloride) containing 100 g/ml ampicillin, 20 g/ml chloramphenicol, and 2 mM isopropyl-␤-D-thiogalactopyranoside. The expressed cells were suspended in 0.1 M sodium phosphate (pH 8.0), 20 mM imidazole HCl, and 0.3 M NaCl and disrupted by sonication, followed by heat treatment at 65°C for 30 min. After removal of denatured E. coli proteins by centrifugation at 19,000 ϫ g for 60 min, the solution was applied to a nickel-nitrilotriacetic acid (Ni 2ϩ -NTA) 1 Superflow column (QIAGEN Inc.), which was then washed thoroughly and eluted with 0.1 M sodium phosphate (pH 8.0), 0.2 M imidazole HCl, and 0.3 M NaCl. The buffer was exchanged with 20 mM Tris-HCl (pH 8.0) containing 1 mM EDTA by ultrafiltration (Amicon Ultra, Millipore Corp.), and the solution was applied to a UNO Q column (Bio-Rad). The protein was eluted with a linear gradient of NaCl (0 -400 mM), and the fractions containing the enzyme were collected. Unfortunately, purified A 3 B 3 DF and A 3 B 3 D had only low activity, as most of the protein was in the MgADPinhibited form (11). Tightly bound nucleotides were removed as follows. The enzyme solution was dialyzed against 100 mM sodium phosphate (pH 8.0) and 5 mM EDTA and heated at 65°C for 10 min, followed by cooling on ice for 30 min; the process was repeated five times. The enzyme solution was then applied to an NAP-10 column (Amersham Biosciences) equilibrated with 100 mM sodium phosphate (pH 8.0) and 5 mM EDTA. After repeating this application procedure four times, the enzyme was dialyzed against 50 mM Tris-HCl (pH 8.0), concentrated with an Amicon Ultra filter unit, and then stored at 4°C until used.
Expression and Purification of the F Subunit-E. coli BL21-Codon Plus (DE3)-RP cells (Stratagene) harboring pET-F, pET-F (⌬85-106) , or pET-F (⌬69 -106) were grown overnight in 2ϫ YT medium (16g/liter Tryptone, 10 g/liter yeast extract, 5 g/liter NaCl) containing 100 g/ml ampicillin and 20 g/ml chloramphenicol at 37°C. Ten ml of overnight culture was then transferred to 1 liter of 2ϫ YT medium containing 100 g/ml ampicillin and 20 g/ml chloramphenicol, and the cells were cultured at 37°C for 4 h. The T7 promoter was subsequently induced by the addition of 1 mM isopropyl-␤-D-thiogalactopyranoside. The cells were cultured for an additional 6 h at 37°C, collected by centrifugation, and suspended in 50 mM Tris-HCl (pH 8.0) and 1 mM EDTA. The suspension was sonicated and then incubated at 65°C for 30 min, and denatured E. coli proteins were removed by centrifugation at 19,000 ϫ g for 60 min. The resultant supernatant was mixed with 50 mM Tris-HCl (pH 8.0) containing 1 mM EDTA and 0.6 M ammonium sulfate and then applied to a butyl-Toyopearl column (Tosoh Corp.) equilibrated with 50 mM Tris-HCl (pH 8.0) containing 0.3 M ammonium sulfate. The protein was eluted with a linear gradient of ammonium sulfate (300 to 0 mM). The fractions containing the F subunit were collected, concentrated with an Ultrafree filter unit (Millipore Corp.), and then applied to a Superdex HR-200 column (Amersham Biosciences) equilibrated with 20 mM Tris-HCl (pH 8.0) and 150 mM NaCl. The fractions contain-ing the F subunit were concentrated with an Ultrafree filter unit and then stored at 4°C until used.
Assays-The protein concentrations of the A 3 B 3 DF and A 3 B 3 D subcomplexes were determined from UV absorbance calibrated by quantitative amino acid analysis; 1 M gave 0.23 A units at 280 nm. The protein concentration of the F subunit was determined using the BCA protein assay kit (Pierce) with bovine serum albumin as a standard. ATPase activity was measured at 25°C in the presence of an ATPregenerating system as described (11). Briefly, the reaction was initiated by the addition of enzyme solution (which was kept on ice for at least 30 min) to 2 ml of assay mixture consisting of 50 mM Tris-HCl (pH 8.0), 100 mM KCl, 2 mM MgCl 2 , 2 mM phosphoenolpyruvate, 0.2 mg/ml NADH, 0.1 mg/ml pyruvate kinase, 0.1 mg/ml lactate dehydrogenase, and a range of MgATP concentrations. The rate of ATP hydrolysis was monitored continuously as the rate of oxidation of NADH, which was determined by the absorbance decrease at 340 nm. The initial ATP hydrolysis rates for both A 3 B 3 DF and A 3 B 3 D did not fit a simple Michaelis-Menten model, but did fit a typical substrate inhibition model (Scheme 1). Kinetic parameters and error estimates were determined from least-squares fits to the following equation: is the enzyme concentration in the assay mixture. The amount of nucleotide bound to V 1 -ATPase and the A 3 B 3 D subcomplex was determined as described previously (11). Role of the F Subunit in V 1 -ATPase and 2 mM phosphoenolpyruvate). Rotation of the beads was observed at 25°C with an Olympus IX70 bright-field microscope (magnification ϫ1000). Images of A 3 B 3 DF and A 3 B 3 D were recorded with a CCD camera at frame rates of 125 and 30 s Ϫ1 , respectively. Analysis of rotation was performed as described previously (24).

The Kinetics of the A 3 B 3 D Subcomplex Differ from Those of A 3 B 3 DF-Wild-type A 3 B 3 DF (V 1 -ATPase) and A 3 B 3 D from T.
thermophilus were expressed in E. coli and purified to homogeneity (Fig. 1). After purification, the wild-type A 3 B 3 DF and A 3 B 3 D subcomplexes exhibited almost no ATPase activity due to tightly bound MgADP, which inactivates V 1 (11). To reactivate the enzymes, the bound nucleotide was removed by phosphate/EDTA treatment as described under "Experimental Procedures." One mol of the final A 3 B 3 DF and A 3 B 3 D subcomplexes used in this work contained 0.33 and 0.30 mol of ADP, respectively. Fig. 2 shows the time course of MgATP hydrolysis catalyzed by A 3 B 3 DF and A 3 B 3 D with the ATPregenerating system. After the addition of A 3 B 3 DF or A 3 B 3 D to the assay mixture, an apparent short initial lag was observed. As observed previously (11), the initial lag phase was rapidly followed by a second phase of a rapid rate of hydrolysis, which decelerated slowly. Although the overall profiles of ATP hydrolysis of A 3 B 3 DF and A 3 B 3 D were similar, the ATPase activity of A 3 B 3 D was apparently lower than that of A 3 B 3 DF (Fig. 2a). To compare the steady-state kinetic parameters of A 3 B 3 DF and A 3 B 3 D, the rate of rapid ATP hydrolysis after the initial lag phase was measured at various concentrations of MgATP.
[S] Ϫ v plots for both A 3 B 3 DF and A 3 B 3 D exhibited typical kinetics for substrate inhibition rather than simple Michaelis-Menten kinetics (Fig. 2b). The V max , K m , and K i values for A 3 B 3 DF were calculated to be 31 s Ϫ1 , 0.26 mM, and 24 mM, respectively ( Table  I). The V max and K m values were consistent with previously reported values (34 s Ϫ1 and 0.24 mM, respectively) (11). In contrast, the V max of A 3 B 3 D was significantly lower than that of A 3 B 3 DF (Table I). In addition, the K m and K i values were increased in A 3 B 3 D (Table I). These results suggest that the F subunit promotes V 1 activity by increasing catalytic turnover and affinity for MgATP.
The A 3 B 3 D Subcomplex Is More Susceptible to Inhibition at Higher MgATP Concentrations-As described previously (11), the ATPase activity of A 3 B 3 DF decelerated during ATP hydrolysis due to MgADP inhibition (Fig. 2a). A similar inactivation has also been observed for CaATP hydrolysis of yeast V 1 (25). It has been suggested that the active form of F 1 -ATPase is converted to the MgADP-inhibited form (which is completely inactive) during MgATP hydrolysis. At the same time, the inhibited form is also converted to the active form (26). We assumed that this kind of interconversion also occurred in V 1 -ATPase during MgATP hydrolysis. Based on this scheme, the rate constants for A 3 B 3 DF deceleration at 0.05, 0.2, and 2 mM MgATP were estimated to be 0.0009, 0.0020, and 0.0024 s Ϫ1 , respectively ( Table I). The estimated rate constant for deceleration of A 3 B 3 D ATPase activity at 2 mM MgATP was significantly higher than that for A 3 B 3 DF (Table I). In contrast, the rate constants at 0.05 and 0.2 mM MgATP were very similar to those for A 3 B 3 DF (Table I).

The A 3 B 3 D Subcomplex Binds to the Isolated F Subunit-
A 3 B 3 D was mixed with a 10-fold molar excess of the purified F subunit, followed by gel filtration chromatography to remove the free F subunit (Fig. 3a). As shown in Fig. 3 (b and c, lanes  5), the F subunit eluted together with the A 3 B 3 D subcomplex from the gel filtration column, clearly demonstrating that A 3 B 3 D was able to bind to the isolated F subunit. To assay which regions of the F subunit are important for binding to A 3 B 3 D, we constructed six expression plasmids with deletion mutants of the F subunit (F (⌬1-8) , F (⌬1-13) , F (⌬1-24) , F (⌬1-44) , F (⌬85-106) , and F (⌬69 -106) ). Among them, only F (⌬85-106) and F (⌬69 -106) were expressed in E. coli. MALDI-TOF mass spectrometry showed that the F (⌬85-106) and F (⌬69 -106) mutants as well as the intact F subunit were expressed as designed (data not shown). However, four mutants with deletions of N-terminal residues (F (⌬1-8) , F (⌬1-13) , F (⌬1- 24) , and F (⌬1-44) ) were not expressed in E. coli (data not shown). This suggests that the N-terminal region of the F subunit is important for folding and/or expression. Interactions between the A 3 B 3 D subcomplex and F (⌬85-106) and F (⌬69 -106) were also assayed by reconstitution. As shown in Fig. 3 (b and c, lanes 6 and 7), F (⌬85-106) bound to A 3 B 3 D, whereas F (⌬69 -106) did not. These results indicate that the C-terminal 38 residues of the F subunit are involved in binding to A 3 B 3 D, with residues 69 -84 being essential.
Steady-state Kinetics and MgADP Inhibition of Reconstituted V 1 -ATPases-The ATP hydrolyzing activity of the A 3 B 3 D subcomplex was measured in the presence of a 20-fold molar excess of the isolated F subunit or F (⌬85-106) . Fig. 4a shows the time course of the ATPase assay for A 3 B 3 D in the presence of the intact F subunit or F (⌬85-106) at 2 mM MgATP. MgATP hydrolysis by the A 3 B 3 D subcomplex was significantly enhanced by the addition of either the F subunit or F (⌬85-106) . The isolated F subunit did not show ATPase activity (data not shown). The addition of the F subunit to A 3 B 3 D resulted in a significant increase in V max (30 s Ϫ1 ) and a decrease in K m (0.24 mM) and K i (29 mM) ( Fig. 4b and Table I). These values are comparable with those for A 3 B 3 DF, suggesting that the observed differences between the A 3 B 3 DF and A 3 B 3 D subcomplexes are due principally to the F subunit. However, the inhibition rate for A 3 B 3 D bound to the F subunit at 2 mM MgATP was slightly higher than that for A 3 B 3 DF (Table I). The reason for this higher inhibition rate is not clear at present. Binding of F (⌬85-106) to A 3 B 3 D also increased V max (24 s Ϫ1 ) and decreased K m (0.41 mM), but the effect was less than that seen for the intact F subunit (Fig. 4b and Table I). These results suggest that the C-terminal 22 residues of the F subunit may play a role in catalytic events.
The F Subunit Is Not Essential for Rotary Movements of V 1 -ATPase-The central rotor of V 1 -ATPase from T. thermophilus is composed of two subunits, D and F, which rotate relative to the A 3 B 3 hexamer (3). However, to date, it has been unclear if the F subunit is an indispensable component for rotation of V 1 -ATPase. To answer this question, we attempted to observe rotation of the D subunit in A 3 B 3 D using a single molecule technique. All intrinsic Cys residues were replaced with Ser except for Cys 255 of the A subunit, as it was found that the A(C255S) mutation led to a decrease in ATPase activity (Table  II). Other substitutions did not significantly affect ATPase activity (Table II). The D subunit was specifically labeled with biotin (data not shown), and A 3 B 3 D or A 3 B 3 DF was immobilized on an Ni 2ϩ -NTA-coated glass surface via a His 8 tag. Streptavidin-coated beads (0.56 m) attached to the D subunit were observed under a bright-field microscope after infusion of buffer containing 4 mM MgATP into the flow cell. Beads attached to A 3 B 3 DF were observed to rotate at Ϸ6.1 rotation/s (Fig. 5), which is much higher than the previously reported value of Ϸ2.6 rotation/s (3). When A 3 B 3 D was used, bead rotation was also observed. Like A 3 B 3 DF, the direction of the rotation was always counterclockwise when viewed from the membrane side (V 0 side). Bead rotation was also observed for A 3 B 3 D, although at a lower rate of Ϸ1.8 rotation/s (Fig. 5).
These results indicate that the F subunit is not an essential subunit for rotary catalysis of V 1 -ATPase and that A 3 B 3 D is the minimum ATP-driven rotary unit of V-ATPase rather than A 3 B 3 DF. These results also suggest that the F subunit binds peripherally to the D subunit rather than being located within the central cavity of the A 3 B 3 hexamer. (⌬69 -106) . a, shown is the elution profile of A 3 B 3 D mixed with the intact F subunit on a Superdex HR-200 column. b, the first peak from the Superdex HR-200 column was analyzed by SDS-PAGE. c, the F subunit was detected by Western blot analysis using anti-F subunit antiserum (13). Lanes 1, A 3 B 3 D subcomplex (20 pmol); lanes 2, F subunit (90 pmol); lanes 3, F (⌬85-106) (90 pmol); lanes 4, F (⌬69 -106) (90 pmol); lanes 5, first peak from A 3 B 3 D and the F subunit (20 pmol); lanes 6, first peak from A 3 B 3 D and F (⌬85-106) (20 pmol); lanes 7, first peak from A 3 B 3 D and F (⌬69 -106) (20 pmol). mAU, absorbance milliunits.

DISCUSSION
The F subunit is a component of the V 1 complex of V 0 V 1 -ATPase, forming the central rotor of V 1 -ATPase together with the D subunit (3). However, little is known about its function. To investigate the function of the F subunit, the A 3 B 3 DF (V 1 -ATPase) and A 3 B 3 D subcomplexes of T. thermophilus V-ATPase were prepared and characterized. The ATPase activity of A 3 B 3 D was significantly lower than that of A 3 B 3 DF. Both the V max and affinity for MgATP in A 3 B 3 D were lower compared with those in A 3 B 3 DF. In addition, the MgADP inhibition rate for A 3 B 3 D at the highest MgATP concentration was ϳ3 times greater than that for A 3 B 3 DF. The reconstituted V 1 -ATPase from A 3 B 3 D and the isolated F subunit exhibited the same ATPase activity kinetic parameters as V 1 -ATPase. These results indicate that the F subunit functions as an intrinsic activator of the ATP hydrolysis reaction of V 1 -ATPase. It is not clear at present, however, if the F subunit affects V 1 -ATPase catalysis by directly interacting with the catalytic A subunit or by changing the conformation of the D subunit. The ATP hydrolysis reaction catalyzed by V 1 -ATPase is thought to proceed via the following steps: 1) ATP binding to V 1 -ATPase, 2) hydrolysis of ATP to ADP and P i , and 3) release of ADP and P i . The F subunit should affect Steps 2 and/or 3 because the rate of MgATP hydrolysis by A 3 B 3 DF is ϳ1.5-fold higher than that of A 3 B 3 D under V max conditions in which Step 1 is no longer rate-limiting. Moreover, it seems that the F subunit also participates in Step 1 by changing the affinity of the A subunit for ATP because the K m for A 3 B 3 D is ϳ2-fold higher than that for A 3 B 3 DF. The result of the single molecule assay that the rotation speed of A 3 B 3 D was lower than that of A 3 B 3 DF indicates that binding of the F subunit to A 3 B 3 D also affects rotary catalysis in the A 3 B 3 D rotary unit. However, the rotation speed in the single molecule experiment was significantly lower than that estimated from the ATP hydrolysis rate in the bulk phase assay (assuming that V 1 hydrolyzes 3 molecules of ATP/rotation) for both complexes. This indicates that the rotation speed of the bead-attached enzyme depends largely on both the torque generated by ATP hydrolysis and the frictional load due to attached beads in this experiment. Thus, the lower rotation rate for A 3 B 3 D compared with A 3 B 3 DF cannot be solely accounted for by the lower ATP hydrolysis rate for A 3 B 3 D in the bulk phase assay. One possible reason for the lower rotation speed of A 3 B 3 D is that the torque is decreased in the absence of the F subunit.
The F subunit of V-ATPase resembles the ⑀ subunit of bacterial F-ATPase in the sense that both F and ⑀ subunits bind to a central shaft subunit that forms the central rotor of V 1 and F 1 , respectively. In addition, the F and ⑀ subunits have similar molecular masses (12-14 and 13-15 kDa, respectively). However, there are significant differences between the two subunits. Unlike the F subunit, the ⑀ subunit acts as an endogenous inhibitor of ATP hydrolysis by both F 1 -and F 0 F 1 -ATPases (16 -19). Recent studies indicate that the ⑀ subunit regulates ATP hydrolysis/synthesis of F-ATPase through a drastic conformational rearrangement of the C-terminal ␣-helical domain (27)(28)(29). Moreover, it has been predicted that the structure of FIG. 4. MgATP hydrolysis catalyzed by A 3 B 3 D bound to the F subunit and F (⌬85-106) . a, time course of ATP hydrolysis catalyzed by A 3 B 3 D and A 3 B 3 DF at 25°C and at 2 mM MgATP. The reaction was started by the addition of 20 l of enzyme solution containing 2 M A 3 B 3 D and 40 M F subunit or F (⌬85-106) (arrowhead) to 2 ml of assay mixture (see "Experimental Procedures" for details). b, [S] Ϫ v plot of MgATP hydrolysis. The rate of rapid MgATP hydrolysis after the initial lag phase (see "Results" for details) was measured at various MgATP concentrations. E, A 3 B 3 D ϩ F; q, A 3 B 3 D ϩ F (⌬85-106) . The solid lines show fitting functions with V max ϭ 29.7 s Ϫ1 , K m ϭ 0.24 mM, and K i ϭ 28.7 mM (A 3 B 3 D ϩ F) and with V max ϭ 23.7 s Ϫ1 , K m ϭ 0.41 mM, and K i ϭ 34.5 mM (A 3 B 3 D ϩ F (⌬85-106) ). The inset shows a magnification of the low [MgATP] range. mAU, absorbance milliunits.  the yeast F subunit has an ␣␤-fold and lacks a long C-terminal ␣-helix (30), although the ⑀ subunit consists of an N-terminal ␤-sandwich and a C-terminal ␣-helical domain (14,31). Taken together, these data strongly suggest that the F subunit of V-ATPase is not a functional homolog of the ⑀ subunit of bacterial F-ATPase.
We have recently demonstrated that the central stalk region of V-ATPase is quite different from that of F-ATPase (13). In F-ATPase, the central rotor of the F 1 complex directly interacts with the proteolipid ring (32). In contrast, it seems that the central rotor of V 1 interacts directly with the C subunit (d subunit in eukaryotic V-ATPase), not with the proteolipid ring (12,13). Cross-linking experiments based on the crystal structure of the C subunit from T. thermophilus have shown that the F subunit is in close proximity to the C subunit (13). Experiments on the yeast enzyme also suggest that the F subunit is located at the interface between V 0 and V 1 (22). The distinct central stalk architecture at the interface between V 0 and V 1 of V-ATPase could be relevant to the unique reversible dissociation/association of V 1 and V 0 (1), and the F subunit might be one of the candidates for a regulator of the dissociation/association of V 0 and V 1 . Although the F subunit is a very small subunit, it might be a multifunctional subunit involved in assembly, catalysis, and regulation.