ATPase Activity of a Highly Stable a3b3g Subcomplex of Thermophilic F1 Can Be Regulated by the Introduced Regulatory Region of g Subunit of Chloroplast F1

A mutant F1-ATPase α3β3γ subcomplex from the thermophilicBacillus PS3 was constructed, in which 111 amino acid residues (Val92 to Phe202) from the central region of the γ subunit were replaced by the 148 amino acid residues of the homologous region from spinach chloroplast F1-ATPase γ subunit, including the regulatory stretch, and were designated as α3β3γ(TCT)(Thermophilic-Chloroplast-Thermophilic). By the insertion of this regulatory region into the γ subunit of thermophilic F1, we could confer the thiol modulation property to the thermophilic α3β3γ subcomplex. The overexpressed α3β3γ(TCT) was easily purified in large scale, and the ATP hydrolyzing activity of the obtained complex was shown to increase up to 3-fold upon treatment with chloroplast thioredoxin-f and dithiothreitol. No loss of thermostability compared with the wild type subcomplex was found, and activation by dithiothreitol was functional at temperatures up to 80 °C. α3β3γ(TCT) was inhibited by the ε subunit from chloroplast F1-ATPase but not by the one from the thermophilic F1-ATPase, indicating that the introduced amino acid residues from chloroplast F1-γ subunit are important for functional interaction with the ε subunit.

A mutant F 1 -ATPase ␣ 3 ␤ 3 ␥ subcomplex from the thermophilic Bacillus PS3 was constructed, in which 111 amino acid residues (Val 92 to Phe 202 ) from the central region of the ␥ subunit were replaced by the 148 amino acid residues of the homologous region from spinach chloroplast F 1 -ATPase ␥ subunit, including the regulatory stretch, and were designated as ␣ 3 ␤ 3 ␥ (TCT) (Thermophilic-Chloroplast-Thermophilic). By the insertion of this regulatory region into the ␥ subunit of thermophilic F 1 , we could confer the thiol modulation property to the thermophilic ␣ 3 ␤ 3 ␥ subcomplex. The overexpressed ␣ 3 ␤ 3 ␥ (TCT) was easily purified in large scale, and the ATP hydrolyzing activity of the obtained complex was shown to increase up to 3-fold upon treatment with chloroplast thioredoxin-f and dithiothreitol. No loss of thermostability compared with the wild type subcomplex was found, and activation by dithiothreitol was functional at temperatures up to 80°C. ␣ 3 ␤ 3 ␥ (TCT) was inhibited by the ⑀ subunit from chloroplast F 1 -ATPase but not by the one from the thermophilic F 1 -ATPase, indicating that the introduced amino acid residues from chloroplast F 1 -␥ subunit are important for functional interaction with the ⑀ subunit.
F 0 F 1 -ATP synthases ubiquitously occur in eucaryotic and procaryotic cells and synthesize ATP from ADP and inorganic phosphate at the expense of a trans-membrane electrochemical proton gradient (1)(2)(3)(4). The bacterial enzyme consists of a hydrophilic F 1 part with a subunit composition of ␣ 3 ␤ 3 ␥␦⑀ (5) and a hydrophobic F 0 part with a subunit composition of a 1 b 2 c 10 -12 (6). The subcomplex ␣ 3 ␤ 3 ␥ 1 is the minimum complex that is capable of hydrolyzing ATP (7). A high resolution x-ray structure of a major part of the bovine heart mitochondrial F 1 revealed an alternating hexagonal arrangement of the three ␣ and the three ␤ subunits with two ␣-helices of the ␥ subunit forming a coiled-coil in the central cavity (8). The crystal structure of an ␣ 3 ␤ 3 complex from the thermophilic Bacillus PS3 is completely symmetric (9), but the incorporation of the ␥ subunit into this complex induces a functional asymmetry among the three catalytic sites (10) residing on the ␤ subunits at the interface to the ␣ subunits. Rotation of the ␥ subunit relative to ␣ 3 ␤ 3 had been suggested from kinetic analyses (11), biochemical experiments (12), and biophysical measurements (13). Finally, by a single molecular observation technique, rotation of the ␥ subunit during the ATP hydrolysis reaction in the ␣ 3 ␤ 3 hexamer (14 -17) was shown.
The molecular structure of the F 0 F 1 -ATP synthase of chloroplasts may be basically the same as those of other F 0 F 1 , but this enzyme has a unique regulation system. The activity of chloroplast F 0 F 1 and F 1 (CF 1 ) is strongly regulated by the reduction and the oxidation of a disulfide bridge in the ␥ subunit formed between the two cysteines Cys 199 and Cys 205 (spinach chloroplast) (18); this regulation system is called thiol modulation (19). Reduction of the disulfide bond elicits the latent ATP hydrolyzing activity of the isolated CF 1 . In vitro reduction can be achieved by dithiothreitol (DTT) or other dithiols, but the natural reductant is reduced thioredoxin-f (Trx-f) (20 -22). Introduction of nine amino acids comprising the regulatory sequence into the cyanobacterial ␥ subunit induced thiol modulation in Synechocystis (23,24). Conversely, replacement of the two cysteines by serines in ␥ subunit of CF 0 CF 1 from Chlamydomonas reinhardtii (25,26) resulted in a non-modulated enzyme. However, due to the lack of suitable overexpression systems for the subunits of CF 1 and because of its insufficient ability for the reconstitution of the complete enzyme complex, little information is available about the details of this regulation mechanism. Recently we have succeeded in the reconstitution of a chimeric complex from recombinant ␣ and ␤ subunits of F 1 of the thermophilic Bacillus PS3 (TF 1 ) and the recombinant ␥ subunit from spinach CF 1 (27). The resulting chimeric ␣ 3 ␤ 3 ␥ complex, which had substantial ATPase activity, was clearly regulated by the disulfide/dithiol state of the two regulatory cysteine residues. We could demonstrate the importance of the region around the disulfide bridge of ␥ subunit for the regulatory interaction with ⑀ subunit which is known to inhibit activity (28). However, stability of this chimeric complex was still lower than that of wild type thermophilic ␣ 3 ␤ 3 ␥, and large scale preparation was difficult.
In order to understand the molecular mechanism of thiol regulation, we engineered a chimeric ␣ 3 ␤ 3 ␥ subcomplex of TF 1 in which the central half of the ␥ subunit was replaced by the equivalent part of the ␥ subunit from CF 1 including the regulatory cysteines (Fig. 1). The remaining parts derived from the TF 1 -␥ subunit mainly consists of the N-terminal and C-terminal ␣-helices forming the coiled-coil and a small additional ␣-helix. The chimeric ␣ 3 ␤ 3 ␥ subcomplex was successfully overexpressed in Escherichia coli and purified as an active ATPase. Then we investigated the enzymatic property of this complex under reduced and oxidized conditions.

MATERIALS AND METHODS
Chemicals-Restriction enzymes were purchased from Toyobo Co., Tokyo, Japan. DTT and MOPS were from Nacalai Tesque, Kyoto, Japan. TNP-ATP was synthesized and purified by the method as described (29,30). All other chemicals were of the highest grade commercially available.
Bacterial Strains-Plasmid amplification was carried out with E. coli strain JM 109, and single-stranded uracil-containing DNA was obtained from E. coli strain CJ 236, and for protein overexpression E. coli strain JM 103 ⌬uncB-D was used.
Plasmid Construction-A DNA fragment coding for ␣, ␥ (S106C), and ␤ (carrying an N-terminal tag of 10 histidine residues) subunits of TF 1 on the expression plasmid, pkkHisCys5 (14), was cloned into the multicloning site of the M13mp18 vector, and single-stranded uracil-containing DNA was obtained by the method of Ref. 31. The location for two silent mutations to construct NspV restriction sites was determined with the program MOTOJIMAN provided by Fumihiro Motojima, Tokyo Institute of Technology, and introduced by the method as described (31) using the primers 5Ј-CACGAGGCGCAGCACGTTCGAATTG-TACGCGCCAGC and 5Ј-TTCTTCTTGCGACGGTTCGAATTCGTA-CACCGTGCG. Then the DNA fragment coding the deduced ␥ subunit was cut out with the restriction enzymes NheI and BglII and cloned into the equivalent position between the genes for the ␣ and ␤ subunits on the pkkHisCys5 vector.
A fragment of spinach chloroplast DNA coding for amino acid residues Leu 96 -Phe 243 of the ␥ subunit of CF 1 was amplified by the polymerase chain reaction method using the phosphorylated primers 5Ј-CGG-GATCCCGTTAGAAGTTGCTGAAGAAGGCTGAGTCTAGG and 5Ј-CGGGATCCCGTTCGAATTCCAGAATTGGGGAAAATGCTGG with the vector containing the gene for the ␥ subunit of spinach CF 1 (27) as a template. The obtained DNA fragment was inserted by blunt-end ligation into pBluescript vector, which was previously digested with SmaI. Then the vector was digested with NspV to obtain a cohesive-end DNA fragment. Finally the above described pkkHisCys5 vector with two introduced NspV sites was digested with NspV and ligated with the cohesive-end DNA fragment to obtain the expression plasmid for ␣ 3 ␤ 3 ␥ (TCT) (pkkTCT vector). The correct orientation of the DNA fragment inserted into pkkTCT was confirmed by digestion with NheI and AflII.
Overexpression and Protein Purification of ␣ 3 ␤ 3 ␥ (TCT) and ␣ 3 ␤ 3 ␥ (␤-His,␥S106C)-pkkTCT was transformed into E. coli JM 103 strain, cultivated on LB medium agar plates containing 100 g/ml ampicillin, and then inoculated into the liquid culture. A 2-liter liquid culture in the same medium was harvested after 9 h cultivation. Cells were collected by centrifugation and disrupted by sonication. The superna-tant was directly applied to a nickel-NTA (nitrilotriacetic acid) superflow column (Qiagen, Hilden, Germany) equilibrated with 50 mM MOPS-KOH, pH 7.0, 100 mM KCl, 5 mM MgCl 2 , and 10 mM imidazole. The column was washed with the same buffer containing 25 mM imidazole and then the enzyme was eluted with the same buffer containing 250 mM imidazole.
The recombinant ␣ 3 ␤ 3 ␥ subcomplex of TF 1 containing a histidine tag on the N terminus of the ␤ subunit and the substitution of Ser 106 of the ␥ subunit to Cys (␣ 3 ␤ 3 ␥ (␤-His, ␥S106C)) was expressed by using the pkkHisCys5 vector and purified by the method described (11).
Preparation of the ⑀ Subunits-⑀ subunit of TF 1 was overexpressed in E. coli and isolated as described (32). The recombinant ⑀ subunit of CF 1 was expressed as an inclusion body and purified (27). Prior to use, the folded ⑀ subunit of CF 1 was prepared as follows: the inclusion body was solved with 8 M urea, 50 mM Tris-Cl, pH 8.0, 1 mM EDTA, and 0.5 mM DTT and dialyzed against 50 mM MOPS-KOH, pH 7.0, 100 mM KCl, and 5 mM MgCl 2 for 6 h.
Preparation of the Recombinant Trx-f-Spinach chloroplast Trx-f was overexpressed in E. coli and isolated as described (21). The concentration of the purified Trx-f was determined by the measurement of the absorbance at 278 nm using the published molar absorption coefficient value of 16,830 M Ϫ1 ⅐cm Ϫ1 (33).
ATP Hydrolysis Activity-ATP hydrolysis activity was measured in the presence of an ATP regenerating system (34) with a spectrophotometer model U-3100 (Hitachi, Tokyo, Japan) equipped with a stirrer. The reaction mixture containing 50 mM MOPS-KOH, 100 mM KCl, 5 mM MgCl 2 , 2 mM phosphoenolpyruvate, 2 mM ATP, 50 g/ml pyruvate kinase, 50 g/ml lactate dehydrogenase, and 0.2 mM NADH was used. After a base line was monitored for 1-2 min, the reaction was initiated by the addition of 5-10 g of ␣ 3 ␤ 3 ␥ (TCT) , and the activity was measured by monitoring the decrease of NADH absorption at 340 nm at 25°C.
In the case where the ATP regenerating system was not applicable, the activity was measured by quantifying the amount of produced phosphate (35). 80 l of 50 mM MOPS-KOH, pH 7.0, 100 mM KCl, 5 mM MgCl 2 , and 2 mM ATP were incubated in a water bath for 2 min, and then the reaction was started by the addition of 20 l (2 g) of ␣ 3 ␤ 3 ␥ (TCT) . After 2 min, the reaction was quenched by the addition of 100 l of 2.4% (w/v) trichloroacetic acid. The reaction was carried out at 25°C unless stated otherwise.
Inhibition by the ⑀ Subunit-␣ 3 ␤ 3 ␥ (TCT) or ␣ 3 ␤ 3 ␥ (␤-His, ␥S106C) were mixed with the ⑀ subunit of CF 1 or TF 1 and incubated for 1 h at room temperature. The mixture was then directly applied to the ATP hydrolysis assay using an ATP regenerating system as described above.
ATP Hydrolysis Under Uni-site Conditions-␣ 3 ␤ 3 ␥ (TCT) was oxidized with 50 M CuCl 2 or reduced with 15 mM DTT for 1 h at 30°C. DTT in the sample was removed by an HPLC gel filtration column (TSK 3000SW XL , Tosoh Co., Tokyo, Japan) equilibrated with 50 mM MOPS-KOH, pH 7.0, 100 mM KCl, and 5 mM MgCl 2 because DTT directly affected the following quantitative measurement of nucleotides. This step also reduced the amount of intrinsically bound nucleotides from 0.8 mol ADP/mol ␣ 3 ␤ 3 ␥ (TCT) to Ͻ0.1 mol ADP/mol ␣ 3 ␤ 3 ␥ (TCT) . Each of the complexes was diluted to a protein concentration of 700 nM. A 50-l aliquot was preincubated in a 25°C water bath and then mixed with the same volume of 150 nM TNP-ATP. At the indicated period the reaction was terminated by addition of 5 l of 24% (w/v) trichloroacetic acid. The amounts of TNP-ATP and TNP-ADP were determined by reversedphase HPLC (35). In case of chase acceleration experiments, 10 l of 100 M ATP was added instead of trichloroacetic acid and incubated for further 10 s. Then the reaction was stopped by the addition of trichloroacetic acid as stated above.
Gel Electrophoresis-Polyacrylamide gel electrophoresis in the presence of SDS was performed according to Laemmli (36). Protein bands were visualized by the staining with Coomassie Brilliant Blue R-250.

Preparation and Purification of
coli JM 103 and purified by Ni-NTA affinity chromatography. The purity of the protein and the occurrence of the individual subunits in the complex were confirmed by SDS-gel electrophoresis (Fig. 2). The yield of ␣ 3 ␤ 3 ␥ (TCT) complex was about 80 mg of protein from a 2-liter E. coli culture, comparable to the yield of wild type ␣ 3 ␤ 3 ␥ (7) or ␣ 3 ␤ 3 ␥ (␤-His, ␥S106C) (14).
The molecular mass of the ␥ subunit in the ␣ 3 ␤ 3 ␥ (TCT) complex was about 35 kDa, clearly larger than the ␥ subunit of ␣ 3 ␤ 3 ␥ (␤-His, ␥S106C) but very similar to the ␥ subunit of CF 1 (Fig. 2), indicating the successful expression and the formation of the complex with the mutant ␥ subunit.
Modulation of Activity by Oxidation and Reduction-The activity of ATP hydrolysis catalyzed by the purified ␣ 3 ␤ 3 ␥ (TCT) was measured spectrophotometrically by the pyruvate kinase/ lactate dehydrogenase assay that provides continuous regeneration of ATP. Upon addition of oxidized and reduced ␣ 3 ␤ 3 ␥ (TCT) to the reaction mixture, ATP hydrolysis proceeded in a linear manner for at least 10 min without time-dependent inactivation (Fig. 3A), but the rates were different as follows: ␣ 3 ␤ 3 ␥ (TCT) was clearly activated by preincubation with DTT. When DTT alone was used for reduction, up to 2-fold activation was obtained (Fig. 3, A and B). Activation was saturated at 15 mM DTT (Fig. 3B). Extension of the preincubation time for more than 1 h did not significantly increase activation. Interestingly, the activation was higher if reduction was carried out with a combination of DTT and Trx-f (Fig. 3, A and B). When ATP hydrolysis reaction was initiated by addition of oxidized ␣ 3 ␤ 3 ␥ (TCT) and 15 mM DTT plus 500 nM Trx-f (final concentration) was added to the reaction mixture after 5 min, remarkable activation could be observed within several minutes (Fig.  3A, inset). The concentration of Trx-f for half-maximal activation was about 200 nM (Fig. 3B, inset), considerably lower than that for authentic CF 1 , which was more than 1 M (21). The specific activity of ␣ 3 ␤ 3 ␥ (TCT) in the oxidized state was 1.0 -1.2 units/mg, and upon reduction with 15 mM DTT it increased to 2.2-2.6 units/mg, and with 15 mM DTT plus 500 nM Trx-f 3.0 -3.3 units/mg were measured. Quantification of thiol groups indicated additional 1.6 -1.9 mol of SH/mol of ␣ 3 ␤ 3 ␥ (TCT) after reduction with 15 mM DTT and additional 1.7-2.0 mol of SH/ mol of ␣ 3 ␤ 3 ␥ (TCT) after reduction with 15 mM DTT plus 500 nM Trx-f.
Interaction with the ⑀ Subunit-The effect of the inhibitory ⑀ subunit of CF 1 and that of TF 1 on ␣ 3 ␤ 3 ␥ (TCT) was investigated (Fig. 5A). ATPase activity of ␣ 3 ␤ 3 ␥ (TCT) in its oxidized state was inhibited to 50 -60% by the ⑀ subunit of CF 1 . Interestingly, ␣ 3 ␤ 3 ␥ (TCT) in its reduced state was nearly insensitive to the CF 1 -⑀ subunit, resulting in about a 4-fold higher activity compared with ␣ 3 ␤ 3 ␥ (TCT) in the oxidized state in the presence of the CF 1 -⑀ subunit. As reported previously, neither the CF 1 -⑀ subunit nor the TF 1 -⑀ subunit had any significant effect on the activity of ␣ 3 ␤ 3 ␥ (␤-His, ␥S106C) when the ATP concentration was relatively high (32). However, a slight but significant increase of activity was observed upon incubation of oxidized ␣ 3 ␤ 3 ␥ (TCT) with the TF 1 -⑀ subunit (Fig. 5A). Again, this effect was less pronounced for ␣ 3 ␤ 3 ␥ (TCT) in the reduced form.
To clarify the cause of the weakly stimulating effect of the ⑀ subunit of TF 1 on ATPase activity of ␣ 3 ␤ 3 ␥ (TCT) , we examined its binding to the complex. After incubation with ⑀ subunit of TF 1 , the complex was separated from unbound ⑀ subunit by gel filtration HPLC. The ␣ 3 ␤ 3 ␥ (TCT) fraction was then applied to SDS-PAGE (Fig. 5B). ␣ 3 ␤ 3 ␥ (TCT) in the oxidized state as well as in the reduced state could bind the ⑀ subunit of TF 1 in an approximately stoichiometric ratio, comparable to ␣ 3 ␤ 3 ␥ (␤-His, ␥S106C) under identical conditions.

DISCUSSION
In this study, an expression plasmid for an ␣ 3 ␤ 3 ␥ subcomplex of TF 1 containing a TF 1 /CF 1 chimeric ␥ subunit (␣ 3 ␤ 3 ␥ (TCT) ) was constructed, and the desired protein complex was successfully overexpressed and purified with a yield comparable to those of wild type ␣ 3 ␤ 3 ␥ (7) or ␣ 3 ␤ 3 ␥ (␤-His, ␥S106C) (14). This proves that combining two halves of ␥ subunits from different organisms and the insertion of additional amino acid residues from the ␥ subunit of CF 1 into the central part of the ␥ subunit of TF 1 are no major obstacles for biosynthesis and in vivo stability of this protein complex in the employed expression system. Thus, although the homology between the transferred part of CF 1 -␥ subunit and the replaced part of the TF 1 -␥ sub- unit was only about 15% on the amino acid level, proper folding of the ␥ subunit and insertion into the ␣ 3 ␤ 3 hexagon appears to be unaffected. This is consistent with a recently published intermediate-resolution x-ray structure of E. coli F 1 , which indicated that four ␣-helices, thought to be comprised by the central amino acid residues of the ␥ subunit (roughly E. coli F 1 ␥100 -200, corresponding to residues 100 -235 of ␥ (TCT) ) are located at the tip of the enzyme and not in interaction with the ␣ 3 ␤ 3 hexagon (40).
The ATPase activity of the ␣ 3 ␤ 3 ␥ (TCT) complex was modulated by the reduction or oxidation of the two cysteines that were introduced by the insertion of the respective part of the CF 1 -␥ subunit. An up to 3-fold activation could be observed by the complete reduction of these cysteines (Fig. 3). As this activation ratio is in the same range as that reported for CF 1 (3-7-fold) (19 -22), most likely the mechanism of thiol activation is the same in the ␣ 3 ␤ 3 ␥ (TCT) complex. Sokolov et al. (41) recently proposed that the conformation of the ␥ subunit of CF 1 may be different from other F 1 -␥ subunits. Their conclusion is based on former results of energy transfer measurements be-tween ␥ subunit and the catalytic sites on ␤ subunits (42,43) and their investigation of the ATPase activity of a complex formed from the isolated CF 1 ␣ 3 ␤ 3 subcomplex and recombinant ␥ subunit of which the C terminus was partially deleted. In their model, the tip of the C terminus of the ␥ subunit does not act as a spindle for rotation. However, our results indicate that a bacterial ␥ subunit can be transformed into a functional chloroplast-like ␥ subunit. Thus the essential structural features and basic functions of both types of ␥ subunits should be the same. This is also consistent with previous reports that upon introduction of the chloroplast ␥-like regulatory stretch into the ␥ subunit of cyanobacterial F 1 conferred thiol modulation on this enzyme (23,24). In this mutant both ATP hydrolysis and ATP synthesis are controlled by thiol modulation. Furthermore, ATP hydrolysis activity of a reconstituted chimeric complex containing the ␣ and ␤ subunits of TF 1 , and the ␥ subunit of CF 1 (␣ T3 ␤ T3 ␥ C ), was regulated by thiol modulation (27,28) although the complex showed reduced stability.
Interestingly, a higher degree of activation of the complex ␣ 3 ␤ 3 ␥ (TCT) could be achieved with a mixture of DTT and Trx-f than with DTT alone (Fig. 3B), although we could not detect a significant difference in the amount of reduced thiol groups. This is consistent with our recently proposed model that a conformational change, which affects the enzyme activity, is induced by the interaction with Trx-f (21).
The measurement of the temperature optimum of ATPase activity shows that ␣ 3 ␤ 3 ␥ (TCT) , similar to ␣ 3 ␤ 3 ␥ (␤-His, ␥S106C), is more stable than a chimeric ␣ T3 ␤ T3 ␥ C , which displayed the highest activity at 55-60°C (Fig. 4, compare with Fig. 2 of Ref. 27). This result implies that the introduction of the 148 amino acid residues from a mesophilic enzyme does not affect thermostability of the complex significantly. The most probable explanation is that the interaction between the coiledcoil structure of ␥ subunit and the inner surface of the cavity of ␣ 3 ␤ 3 hexagon is especially important for the stability of the complex. Although the homology of this coiled-coil region of the ␥ subunit is relatively high (27), some varying amino acid residues in the coiled-coil region might be very important for the stability of the complex. Activation by DTT was not affected by the temperature, so although thiol modulation in vivo usually occurs at temperatures Ͻ30°C in plants, it is also functional at temperatures as high as 80°C.
The ⑀ subunit is known to be an inhibitory subunit of F 1 - ATPases. It has been proposed that the C-terminal helical domain of the ⑀ subunit is mainly responsible for the inhibitory function (44 -48). As the ⑀ subunit could be cross-linked to the ␤ subunit (49,50), the inhibition may be achieved by binding to the DELSEED region of the ␤ subunit. We found that the ⑀ subunit of CF 1 can efficiently inhibit the ATPase activity of ␣ 3 ␤ 3 ␥ (TCT) but not that of ␣ 3 ␤ 3 ␥ (␤-His, ␥S106C), although the ␣ and ␤ subunits are identical (Fig. 5). This induces the conclusion that an interaction of the ⑀ subunit with the ␥ subunit may be more crucial for inhibition, possibly by providing proper binding and orientation of the ⑀ subunit. As the sequence of the amino acid residues 92-239 of the ␥ subunit of ␣ 3 ␤ 3 ␥ (TCT) is different from ␣ 3 ␤ 3 ␥ (␤-His, ␥S106C), it can be concluded that this region confers the sensitivity to inhibition by the ⑀ subunit. Consistent with this finding, E. coli F 1 residue ␥106 and the region ␥202-212 have previously been found to be involved in binding of the ⑀ subunit by cross-linking and chemical labeling experiments (51). E. coli F 1 -␥106 corresponds to ␥ (TCT) 101, near the N-terminal part of the insertion from CF 1 , ␥202-212 corresponds to ␥ (TCT) 234 -244, located at the C terminus of the insertion (Fig. 1). Taken together, the part of the ␥ subunit that is of chloroplast origin in ␥ (TCT) appears to be crucial for binding of and inhibition by the ⑀ subunit.
As supposed from the suggested mechanism for the uni-site catalysis (29,30), no significant differences could be detected in the rate of uni-site hydrolysis by oxidized or by reduced ␣ 3 ␤ 3 ␥ (TCT) . However, we found a difference in chase acceleration of hydrolysis of TNP-ATP by the addition of excess amounts of ATP (Fig. 6). The range of the chase acceleration for the reduced ␣ 3 ␤ 3 ␥ (TCT) was comparable to that published for TF 1 (35), wild type ␣ 3 ␤ 3 ␥ (32), and ␣ 3 ␤ 3 ␥ (␤I389C) (39). However, significantly lower acceleration was observed for the oxidized ␣ 3 ␤ 3 ␥ (TCT) , indicating an impaired interplay between the catalytic sites in the suppressed activity state caused by the formation of the disulfide bridge between Cys 199 and Cys 205 of the ␥ subunit. The regulation of the enzyme activity might be attributed to this difference observed at the chase acceleration level.