Cross-linking of Two β Subunits in the Closed Conformation in F1-ATPase*

In the crystal structure of mitochondrial F1-ATPase, two β subunits with a bound Mg-nucleotide are in “closed” conformations, whereas the third β subunit without bound nucleotide is in an “open” conformation. In this “CCO” (β-closed β-closed β-open) conformational state, Ile-390s of the two closed β subunits, even though they are separated by an intervening α subunit, have a direct contact. We replaced the equivalent Ile of the α3β3γ subcomplex of thermophilic F1-ATPase with Cys and observed the formation of the β-β cross-link through a disulfide bond. The analysis of conditions required for the cross-link formation indicates that: (i) F1-ATPase takes the CCO conformation when two catalytic sites are filled with Mg-nucleotide, (ii) intermediate(s) with the CCO conformation are generated during catalytic cycle, (iii) the Mg-ADP inhibited form is in the CCO conformation, and (iv) F1-ATPase dwells in conformational state(s) other than CCO when only one (or none) of catalytic sites is filled by Mg-nucleotide or when catalytic sites are filled by Mg2+-free nucleotide. The α3β3γ subcomplex containing the β-β cross-link retained the activity of uni-site catalysis but lost that of multiple catalytic turnover, suggesting that open-closed transition of β subunits is required for the rotation of γ subunit but not for hydrolysis of a single ATP.

F 1 , together with the membrane-embedded F 0 part, constitutes ATP synthase which couples a transmembrane proton flow to ATP synthesis/hydrolysis. F 1 is easily and reversibly separated from F 0 part as a water soluble ATPase and is called F 1 -ATPase. The F 1 -ATPase is composed of five subunits, ␣, ␤, ␥, ␦, and ⑀ subunits, in a molar ratio of 3:3:1:1:1 (1)(2)(3). In the crystal structure of bovine mitochondrial F 1 -ATPase (MF 1 ) 1 (4), three ␣ and three ␤ subunits are arranged alternately like the segments of an orange around the central coiled-coil structure of the ␥ subunit. F 1 -ATPase contains six nucleotide binding sites; three of them are catalytic sites and mainly located on each ␤ subunit, whereas the other three do not bear a catalytic role and are mainly located on each ␣ subunit (4). The ␣ 3 ␤ 3 ␥ subcomplex of F 1 -ATPase has been recognized as a minimum ATPase-active complex with stability and characteristics similar to native F 1 -ATPase (5)(6)(7). F 1 -ATPase shows a complicated kinetic behavior in ATP hydrolysis. Three catalytic sites in F 1 -ATPase are not independent but are related to one another in a cooperative manner. Binding of ATP to the catalytic sites is governed by negative cooperativity, and inversely, hydrolysis of bound ATP at each catalytic site is governed by positive cooperativity. When substoichiometric amounts of ATP relative to the enzyme are added, ATP is hydrolyzed very slowly (uni-site catalysis) (8 -10). This reaction is accelerated by chase-addition of an excess amount of ATP (chase-promotion) (8 -10). As a whole, ATP hydrolysis reaction by F 1 -ATPase usually exhibits negative cooperativity as a function of ATP concentration (11)(12)(13)(14)(15)(16). These and other features had been unified into the binding change mechanism by Boyer (3). According to the mechanism, three catalytic sites interchange their roles alternately and sequentially during catalytic turnover accompanying the rotation of ␥ subunit in the center of the enzyme. Consistent with this rotary mechanism, the ␣ 3 ␤ 3 ␥ subcomplex of thermophilic Bacillus PS3 F 1 -ATPase (TF 1 ) containing two intact and one incompetent catalytic sites lost the ability to mediate catalytic turnover while it showed uni-site ATP hydrolysis and chase-promotion (17). The rotation of the ␥ subunit was supported by various methods (18 -20) and was directly proved by single molecule observation of the ␣ 3 ␤ 3 ␥ subcomplex of TF 1 (21).
Another complication of F 1 -ATPase kinetics is caused by the "Mg-ADP inhibited form" (22)(23)(24)(25). In general, F 1 -ATPases from mitochondria, chloroplasts, and bacteria are prone to develop turnover-dependent inactivation; Mg-ADP trapped transiently in a catalytic site causes the slow transition from an active form to an inhibited form called as Mg-ADP inhibited form (26 -28). The slow transition during catalysis is accelerated by the simultaneous occupation of two catalytic sites by Mg-nucleotides (25,29). Mg-ATP bound to the noncatalytic nucleotide binding site promotes dissociation of inhibitory Mg-ADP from the affected catalytic site (30). Further, when the enzyme is preincubated with stoichiometric Mg-ADP, the Mg-ADP inhibited form is generated. An inhibitor of F 1 -ATPase, azide stabilizes the Mg-ADP inhibited form (23)(24)(25), and an activator of F 1 -ATPase, N,N-dimethyldodecylamine-N-oxide (LDAO) destabilizes it (31).
The crystal structure of MF 1 (4) revealed that three ␤ subunits in the MF 1 molecule are in different states; one ␤ (␤ TP ) has an ATP analog, Mg-AMP-PNP, at its catalytic site; another ␤ (␤ DP ) has Mg-ADP; the third ␤ (␤ E ) has none. The structures of ␤ TP and ␤ DP are very similar to each other, and they are in the "closed" conformation, in which the carboxyl-terminal domain is lifted close to the nucleotide binding domain. In contrast, ␤ E adopts the "open" conformation, in which the crevice * 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.
‡ To whom correspondence should be addressed: Research Laboratory of Resources Utilization, R-1, Tokyo Institute of Technology, 4259 Nagatsuta, Yokohama 226, Japan. Tel.: ϩ81 45 924 5233; Fax: ϩ81 45 924 5277; E-mail: myoshida@res.titech.ac.jp. 1 The abbreviations used are: MF 1 , F 1 -ATPase from mitochondria; TF 1 , F 1 -ATPase from thermophilic Bacillus strain PS3; PAGE, polyacrylamide gel electrophoresis; LDAO, N,N-dimethyldodecylamine-Noxide; DTT, dithiothreitol; TNP-AT(D)P, 2Ј,3Ј-O-(2,4,6-trinitrophenyl) derivatives of AT(D)P; HPLC, high performance liquid chromatography; NEM, N-ethylmaleimide; AMP-PNP, adenosine 5Ј-(␤,␥-imino)triphosphate; TNP-ATP, 2Ј(3Ј)-O-(2,4,6-trinitrophenyl)adenosine 5Ј-triphosphate. for substrate binding is open. This crystal structure, characterized by two closed and one open ␤ subunits ("CCO" conformational state) has been assumed to be a "snapshot" of MF 1 cycling the catalytic turnover along the pathway predicted by the binding change mechanism. Whether this assumption is really the case has not been established by experiments. To address this question, we need a specific probe to detect the CCO conformational state of F 1 -ATPase in solution. Looking at the structure of MF 1 carefully, we noticed that the two closed ␤ subunits, even though separated by an intervening ␣ subunit, have a direct contact at the position of Ile-390 of each ␤ subunit (see Fig. 1, A and B). If this Ile is replaced with Cys, two closed ␤ subunits in F 1 -ATPase in the CCO conformation would be cross-linked by a disulfide bond, and it would fix two ␤ subunits in the closed conformation. Indeed, when we examined this experiment using ␣ 3 ␤ 3 ␥ subcomplex of TF 1 , the cross-link was formed in an Mg-nucleotide-dependent, azide-facilitated manner with concomitant loss of the activity of catalytic turnover.

EXPERIMENTAL PROCEDURES
Strains, Plasmids, and Preparation of Subcomplexes-Escherichia coli strains used were JM109 (32) for preparation of plasmids, CJ236 (33) for generating uracil-containing single-stranded plasmid for sitedirected mutagenesis, and JM103D (uncB-uncD) (34) for expression of the wild-type and mutant ␣ 3 ␤ 3 ␥ subcomplexes of TF 1 . Plasmid pTABG3 (6), which carried genes for the ␣, ␤, and ␥ subunit of TF 1 was used for mutagenesis and expression. A helper phage M13KO7 was obtained from Amersham Pharmacia Biotech, Tokyo. The expression plasmid for the mutant subcomplex was constructed as follows. The mutation (Ile 3 Cys at ␤-386) was introduced into pTABG3 (6) by using a synthetic oligonucleotide: 5Ј-TTGTCTTCATCCGACAGTTCATCCATCCCCAAG-CATGCGATGATGTCTTGCAATTCTTTATAACGTTCGA-3Ј (changed bases are underlined) (33). The EcoRI-BglII fragment from the resultant plasmid was ligated into the EcoRI-BglII site of pKABG3 to produce the expression plasmid, pKAG3-␤-Ile386Cys. Recombinant DNA procedures were performed as described in the manual (35). The wild-type and the mutant ␣ 3 ␤ 3 ␥ subcomplexes were purified and stored as described previously except that 2 mM dithiothreitol (DTT) was added in all buffers (6,17). Just before use, they were subjected to gel-filtration HPLC with a TSK-G3000SWXL column (Tosoh, Japan) equilibrated with 50 mM Tris-HCl (pH 7.0), 200 mM NaCl (Tris-NaCl buffer). Gelfiltrated preparation of the mutant subcomplex did not contain a detectable amount of endogenously bound adenine nucleotide (Ͻ0.1 mol/ mol of subcomplex). In the presence of LDAO, the mutant subcomplex showed the steady-state ATPase activity of 15.2 mol of ATP hydrolyzed/min/mg at 25°C, which is ϳ70% of that of the wild-type subcomplex (21.3 mol of ATP hydrolyzed/min/mg). In the absence of LDAO, the activity was 4.6 (mutant subcomplex) and 7.7 (wild-type subcomplex) mol of ATP hydrolyzed/min/mg. Disulfide Cross-link Formation of the Mutant Subcomplex-The subcomplexes were incubated at 25°C in Tris-NaCl buffer containing indicated components. At the indicated time, an aliquot was taken out and formation of the cross-link was analyzed with 10.5% polyacrylamide gel electrophoresis in the presence of sodium dodecyl sulfate without exposure to a reducing reagent (non-reducing SDS-PAGE) (36). To prevent further oxidation during electrophoresis, all sample solutions were treated with 10 mM N-ethylmaleimide (NEM) for 5 min prior to electrophoresis. The gels were stained with Coomassie Brilliant Blue R-250. The yields of the cross-linked product were determined from the staining intensity of bands with a photodensitometer (model SM3, Howtec Inc.) using the ␣ subunit band as an internal standard.
Analysis of Bound Adenine Nucleotide-The wild-type and mutant subcomplexes were incubated in 20 M ATP and 2 mM MgCl 2 for 2.5 h at 25°C. The yield of cross-linked ␤ dimer of the mutant subcomplex estimated from non-reducing SDS-PAGE was nearly 100% after this incubation. As a control, the samples treated with the same procedures, except that 10 mM DTT was included in the solution, were also analyzed. To remove free adenine nucleotide, the samples were passed through an HPLC column TSK-G3000SWXL. The protein fractions were recovered, and perchloric acid was added to precipitate proteins. The amounts of adenine nucleotides contained in the supernatant were measured with reversed-phase HPLC on a Cosmosil 5C18-AR-II (Nacarai Tesque, Japan), equilibrated, and eluted with 0.1 M sodium phosphate buffer (pH 6.9).
Other Assays-ATPase activity was measured at 25°C in the pres-ence of an ATP-regenerating system (37). Unless stated, a neutral detergent, LDAO, was included in the assay mixture to eliminate the effect of the initial lag phase caused by the Mg-ADP inhibition (31). The assay mixture contained 50 mM Tris-HCl (pH 8.0), 100 mM KCl, 2.5 mM phosphoenolpyruvate, 2.5 mM ATP, 2 mM MgCl 2 , 0.2 mM NADH, 50 g/ml pyruvate kinase, 50 g/ml lactate dehydrogenase, and 0.1% LDAO. Uni-site ATPase activity and chase promotion were measured at 26 and 8°C using TNP-ATP as a substrate as described previously (38). A reaction mixture (50 l) containing 50 mM Tris-HCl (pH 8.0), 4 mM MgCl 2 , 200 mM KCl, and 0.3 M TNP-ATP was incubated at 26°C. The reaction was initiated by the addition of an equal volume of 1 M subcomplexes in 50 mM Tris-HCl (pH 8.0). The reaction was quenched by the addition of 7.5 l of ice-cold 24% perchloric acid. In chasepromotion experiments, 10 l of 30 mM ATP-Mg was added instead of perchloric acid. After 5 s, the reaction was quenched by the addition of 7.5 l of ice-cold 24% perchloric acid. The amounts of TNP-ATP and TNP-ADP were measured by HPLC (38). For uni-site and chase-promotion experiments, the mutant subcomplex with the cross-link lacking bound nucleotide was prepared. The mutant subcomplex was treated with 100 M CuCl 2 for 24 h at 25°C, and the subcomplex was separated from CuCl 2 with gel filtration (Sephadex G-25 column, 2 cm ϫ 5 cm). Formation of the cross-link at 100% yield was confirmed with SDS-PAGE before use. Protein concentration of ␣ 3 ␤ 3 ␥ subcomplex was determined by UV absorbance using the factor of 0.45 at 280 nm as 1.0 mg/ml.

Two ␤ Subunits in the Closed Form Have a Contact
Site-In the crystal structure of MF 1 , two ␤ subunits with a bound nucleotide, ␤ TP and ␤ DP , take the closed conformation and contact each other at the carboxyl-terminal region near the central axis (Fig. 1, A and B, shown by an arrow). The residues in the contact position are Ile-390s of each of ␤ subunits (Fig. 1,  A and B, shown by red). The distance between ␣ carbons of the two Ile-390s is 0.79 nm. The distance between the nearest carbon atoms (C ␤ of ␤ TP -Ile-390 and C ␥2 of ␤ DP -Ile-390) is 0.42 nm, and the two residues actually appear to have hydrophobic interaction. On the contrary, the distance between ␣ carbons of ␤ TP -Ile-390 and ␤ E -Ile-390 and that between ␤ DP -Ile-390 and ␤ E -Ile-390 are 2.40 nm and 2.58 nm, respectively. This residue is located in a region just preceding the "DELSEED sequence" (39) and is highly conserved in F 1 -ATPases from various sources.
The ␣ 3 ␤ 3 ␥ Subcomplex with Cys at the Contact Sites Formed ␤-␤ Cross-link in the Presence of Both Mg 2ϩ and AT(D)P-We generated a mutant ␣ 3 ␤ 3 ␥ subcomplex of TF 1 in which ␤-Ile-386, an equivalent residue to MF 1 ␤-Ile-390, was replaced with Cys. ATPase activity of the mutant subcomplex was about 70% of that of the wild-type subcomplex (see "Experimental Procedures"). The cross-linked ␤ dimer in the mutant subcomplex was detected by non-reducing SDS-PAGE after incubation at 25°C for 2 h in the Tris-NaCl buffer containing 0.25 M CuCl 2 and indicated compounds (Fig. 2). No new band other than three subunit bands, ␣, ␤, and ␥ was found in a reference sample, incubated alone or with DTT (Fig. 2, lanes 1 and 2). The Enzyme Containing the Cross-link Was Inactive in Catalytic Turnover-As mentioned above, cross-linking occurs in Mg-ATP where the enzyme is catalyzing hydrolysis of ATP. When we added 100 M CuCl 2 during continuous assay of ATPase, ATP hydrolysis stopped almost immediately (Ͻ15 s), but the full activity was recovered by addition of 150 mM DTT (Fig. 3A, trace a). It should be noted that the solutions contained LDAO so that generation of the Mg-ADP inhibited form was avoided. ATPase activity of the mutant subcomplex pretreated with NEM (trace b) and that of the wild-type subcomplex (trace c) were not affected either by CuCl 2 or DTT. This suggests that the cross-linking causes the loss of the activity to cycle the catalysis. Indeed, the time-course of the ATPase inactivation mirrored that of the yield of the cross-linked ␤-␤ dimer (Fig. 3B), and the analysis of their correlation showed that the subcomplex containing a cross-linked ␤ dimer completely lost the ability to mediate catalytic turnover of ATP hydrolysis (Fig. 3C). The inactivated subcomplex recovered a full activity by incubation with a reducing reagent, for example 100 mM DTT (data not shown).
Cross-linking Occurred Most Rapidly in Mg-ADP ϩ Azide-Rates of cross-linking were assessed by measuring the rates of ATPase inactivation in the presence of nucleotide and azide (Fig. 4). To ensure that the residual ATPase activities at each time point reflect correctly the fraction of the native subcomplexes without the cross-link, they were measured in the ATPase assay solutions containing LDAO in which subcomplexes showed uninhibited linear activities without an initial lag. Inactivation of ATPase activities proceeded with single exponential curves (solid lines) most rapidly in Mg-ADP ϩ azide, next in Mg-ADP, Mg-ATP ϩ azide, and most slowly in Mg-ATP. The half-decay times were 3.6, 5.8, 10.2, and 21.0 min under the described conditions. Because the Mg-ADP inhibited form is produced most efficiently when F 1 -ATPase is incubated with Mg-ADP and it is further facilitated by azide (23)(24)(25), the conformation of the Mg-ADP inhibited form appears to be favorable for the cross-linking.
Effect of Mg-Nucleotide Concentrations on the Cross-linking-To estimate how many nucleotides were required for the formation of the cross-link at 0.25 M CuCl 2 , the mutant subcomplex was incubated with various concentrations of Mg-ADP and Mg-TNP-ADP, and the final yield of the cross-linked ␤ dimer was measured. At low concentrations of nucleotides, cross-linking proceeded slowly and it was safe to wait for 24 h to reach the final, maximum yield (Fig. 5). Formation of the cross-link was saturated when the concentrations of Mg-nucleotide, expressed as a molar ratio Mg-nucleotide:subcomplex, reached 3:1. It should be noted that at molar ratio 1:1, crosslinked ␤ dimer was formed in only a small fraction of the subcomplex (20%, Mg-ADP; 3%, Mg-TNP-ADP) but at molar ratio 2:1, 50% or more populations of the subcomplexes had a cross-linked ␤ dimer. Similar results were obtained when the subcomplex was incubated with Mg-ATP and Mg-TNP-ATP (not shown). It was demonstrated previously from difference absorption spectrum titration that, when added with 1:1 molar ratio at micromolar range, almost all Mg-TNP-ADP bound to a catalytic site of the subcomplex (7). Also, nearly all Mg-ADP added at 1:1 molar ratio was bound to the subcomplex (22,40). Therefore, for cross-linking to occur, occupation of a single catalytic site by Mg-nucleotide is not sufficient but at least two catalytic sites need to be filled by Mg-nucleotides.
Subcomplex Containing the Cross-link Had Two Bound Nucleotides-The amount of bound nucleotide in the mutant subcomplex was measured after the cross-link was formed in Mg-ATP (Table I). Unbound nucleotide was removed by gel filtration, and nucleotide bound to the subcomplex in a stable manner was analyzed. For comparison, two samples, the wildtype subcomplex incubated with Mg-ATP and the mutant subcomplex incubated with Mg-ATP ϩ DTT were also analyzed. The mutant subcomplex incubated with Mg-ATP ϩ DTT had only a trace amount of bound nucleotide. The mutant subcomplex after formation of the cross-link retained about 2 mol/mol bound ADP (Mg 2ϩ was not analyzed). This result indicates that when the carboxyl-terminal domains of the two ␤ subunits are fixed in the closed conformation by the cross-link, two Mg-ADPs remain trapped in their catalytic sites. The wild-type subcomplex had about 1 mol/mol bound ADP (Table I). It can bind as much as 3 mol/mol Mg-ADP when analyzed with equilibrium dialysis (5), but 2 mol of Mg-ADP probably dissociated upon separation from unbound Mg-ADP during gel filtration. Related to our observation, it was reported that when ␣ subunit and ␥ (or ⑀) subunit of E. coli F 1 -ATPase were cross-linked, entrapped nucleotide (ATP or ADP) cannot be released (41).
The Enzyme Containing the Cross-link Catalyzed Uni-site Catalysis and Chase-promotion-Hydrolysis of a substoichiometric amount of substrate (uni-site catalysis) and its acceleration by chase-added ATP (chase-promotion) of the subcomplex with the cross-link were measured. To avoid confusion arising from bound Mg-ADP, nucleotide-free subcomplex with the cross-link was prepared in 100 M CuCl 2 in the absence of nucleotide. The subcomplex thus prepared did not have the activity of steady-state ATP hydrolysis, but it retained the ability to catalyze uni-site hydrolysis of TNP-ATP (Fig. 6). The rates of uni-site catalysis of the subcomplexes with and without the cross-link were almost the same at 26°C (Fig. 6, A and B, closed circles), and chase-promotion was also observed for both (open circles). This result indicates that a single high affinity catalytic site and its communication with the second catalytic site of the subcomplex are not lost by the cross-linking. However, the efficiency of the communication was diminished because at low temperature, 8°C, chase-promotion observed for the subcomplex with the cross-link was very poor compared with that of the subcomplex without the cross-link (Fig. 6, A  and B, insets).

␤-␤ Cross-linking as a Specific Probe to Detect the CCO Conformational
State of F 1 -ATPase-Intersubunit cross-linking has been proved to be a useful method in the study of F 1 -ATPase to know relative location and motion of subunits (19,41,42). In the crystal structure of the ␣ 3 ␤ 3 subcomplex of TF 1 , the whole structure is arranged in an exact three-fold  3. The cross-linking inactivates ATPase activity. A, traces of ATP hydrolysis catalyzed by the subcomplexes. Hydrolysis was monitored by the oxidation of NADH with an ATP-regenerating system as described under "Experimental Procedures." The reaction mixture contained 0.1% LDAO to avoid the influence of the Mg-ADP inhibited form. The reactions were initiated by addition of the subcomplex at the time indicated by arrowheads (a, 1.1 g of the mutant subcomplex; b, 1.1 g of mutant subcomplex treated with 1 mM NEM for 20 min at 25°C; c, 0.8 g of wild type subcomplex). Then 100 M CuCl 2 or 150 mM DTT were added at the time indicated by arrowheads, respectively. These high concentrations of CuCl 2 and DTT were used to ensure the rapid reactions. B, time courses of formation of the cross-linked ␤-␤ dimer and inactivation of ATPase activity. The mutant subcomplex was incubated at 25°C with 2 mM ATP and 2 mM MgCl 2 . At indicated times, aliquots were taken out, and ATPase activity (q) and the yield of the cross-linked ␤-␤ dimer (E) were analyzed. C, correlation between the cross-linked ␤-␤ dimer and residual ATPase activity. symmetry, and all three ␤ subunits are in the open conformation (43). The distance between Ile-386s of the two ␤ subunits in this subcomplex is 3.27 nm, and a disulfide bond would not be possible even if ␤-Ile-386s are replaced with Cys. By substituting the closed ␤ subunit in the crystal structure of MF 1 with the open ␤ E subunit with the aid of a computer, we generated the structure of the ␣ 3 ␤ 3 ␥ subcomplex with one closed and two open ␤ subunits (COO conformational state) (Fig. 1C) and that with three open ␤ subunits (OOO conformational state) (Fig.  1D). ␤-Ile-390s in the COO conformational state are distant from each other. The distances between Ile390 of the closed ␤ and that of each of the two open ␤s are 2.42 and 3.65 nm, too far for cross-linking. This is also the case for the OOO conforma-tional state; ␤-Ile390s are far away from each other, just similar to ␣ 3 ␤ 3 subcomplex of TF 1 (43). Only when two ␤ subunits take the closed conformation (CCO conformational state) do their lifted carboxyl-terminal domains bring ␤-Ile-390s into the contact position (Fig. 1, A and B). One might think of a CCC conformational state, but when we generated CCC conformational state based on the MF 1 structure, ␥ subunit could not be accommodated in the center of the molecule without steric collision. Therefore, it is unlikely that all three ␤ subunits take closed conformations. Thus, ␤-␤ cross-linking between the introduced Cys at the position ␤-386 of TF 1 is a specific means to identify the CCO conformational state of F 1 -ATPase in solution. This rationale should be valid for F 1 -ATPases from other sources.
F 1 -ATPase Takes the CCO Conformation in Catalytic Cycle and in the Mg-ADP Inhibited Form-The cross-link was formed in the solution containing Mg-ATP and ATP regeneration system where catalysis was going on (Fig. 3A). Therefore, at least one of the intermediates in the catalytic cycle of the subcomplex takes the CCO conformation. 2 Because azide stabilizes the Mg-ADP inhibited form (23)(24)(25), accelerated formation of cross-linking in Mg-ATP by azide (Fig.  4) indicates that the Mg-ADP inhibited form, which is probably derivatized from the catalytic intermediate(s) with a CCO conformation as mentioned above, also takes a CCO conformation. This contention is supported by the fact that cross-linking occurs even more efficiently in Mg-ADP and Mg-ADP ϩ azide, conditions where the Mg-ADP inhibited form is produced efficiently. Requirement for 2 mol of Mg-nucleotide per mol of the subcomplex for the cross-linking (Fig. 5) is consistent with the observation that the rate of development of azide inhibition during ATP hydrolysis was saturated at an ATP concentration of about 10 M, a concentration range of bi-site catalysis where the enzyme operates with two catalytic sites being occupied by substrates (22). 2 One can argue that the Mg-ADP inhibited form, but not an active intermediate, is the molecular species in which cross-link is formed under the conditions in Fig. 3A because the Mg-ADP inhibited form, which also takes the CCO conformation, is generated more or less in a dynamic equilibrium during catalysis. However, this argument may not be the case. The solution contained LDAO, an activator known to keep F 1 -ATPases from falling into the Mg-ADP inhibited form during catalysis and ATP regenerating system to convert the free ADP into ATP. Therefore, the fraction of the Mg-ADP inhibited form should be minimal in the solution. The inactivation of ATPase activity, which was caused by cross-linking, occurred immediately when 100 M CuCl 2 was added. If the cross-link had been formed only from the Mg-ADP inhibited form, the inactivation would have proceeded very slowly in the presence of LDAO.   Existence of COO (or OOO) Conformational State of F 1 -ATPase Is Suggested-It has been known that Mg-free AT(D)P can bind to catalytic sites of F 1 -ATPase but not to noncatalytic nucleotide binding site (44,45). The ␣ 3 ␤ 3 ␥ subcomplex can bind 3 mol of Mg-free ADP per mol (5). Nonetheless, the cross-linked ␤ dimer was not formed in Mg-free AT(D)P (Fig. 2). This indicates that subcomplex with bound Mg-free AT(D)P at catalytic site(s) dwells in a conformational state, probably either in a COO or OOO conformational state, in which two or three ␤ subunits are in open conformations. In this respect, the liganding of Mg 2ϩ to nucleotide at the catalytic site has a great effect on the conformation of the ␤ subunit; shifting the carboxylterminal domain from the open position to the closed one. Probably related to this, the profound effect of Mg 2ϩ on the binding affinity of AT(D)P to the catalytic sites of F 1 -ATPase was reported by Senior's group (44). Also, we previously observed that chemical labeling of ␤-Glu-190, a catalytic residue acting as a general base in ATP hydrolysis reaction (17), by dicyclohexylcarbodiimide was completely blocked when Mg 2ϩ was liganded in a catalytic site (46).
␥ Subunit Cannot Rotate without Open-Closed Transition of ␤ Subunits-The mutant subcomplex cannot catalyze multiple catalytic turnover when the cross-link was formed (Fig. 3). Because multiple catalytic turnover of F 1 -ATPase couples with rotation of ␥ subunit, indication of this result is that, when two ␤ subunits are fixed in the closed conformation, the rotation of the ␥ subunit is blocked. The open-closed conformational transition of each of the ␤ subunits and exchange of the CC (␤closed, ␤-closed) pair among three ␤ subunits appear to be necessary for the rotation of the ␥ subunit and mechanical prevention of this transition by the cross-linking results in blocking the rotation and, thereby, turnover of ATPase cycle. Probably by the same reason, E. coli F 1 -ATPase in which the residue ␤-381, an equivalent residue of TF 1 ␤-E391, is labeled by a bulky fluorescent dye is inactive in ATPase catalytic turnover (42). In a sense, the mechanism of inactivation caused by cross-linking may be similar to that of the Mg-ADP inhibition, blocking sequential transition of the intermediate states in catalysis of the enzyme. The difference is that the former blocks physically and the latter does kinetically. Once the cross-link is formed in the Mg-ADP inhibited form, the transition is now physically blocked.
Based on the experiments of the ␤-␥ cross-linked E. coli F 1 -ATPase, Capaldi's group has proposed that uni-site catalysis does not accompany the rotation (47). Our result, that the subcomplex with the ␤-␤ cross-link can catalyze uni-site TNP-ATP hydrolysis (Fig. 6), again reinforces this conclusion. Mere hydrolysis of a single ATP probably does not require the openclosed transition of the ␤ subunits. Furthermore, because the uni-site catalysis by the subcomplex with the ␤-␤ cross-link was promoted by chase-added ATP (Fig. 6), communication between two ␤ subunits in the enzyme is still possible without open-closed transition of the ␤ subunits. Probably the openclosed transition of the ␤ subunits is needed to drive the critical catalytic step, i.e. simultaneous exchange of the role of each catalytic site.