THE JOURNAL OF BIOLOGICAL CHEMISTRY c ) 1991 by The American Society for Biochemistry and Molecular Biology, Inc,

Vol. 266. No. 9, Issue of March 25, pp. 5424-5429, 1991 Printed in U. S. A.

Catalytic Site of F1-ATPase of Escherichia coli LYS-155 AND LYS-201 OF THE /3 SUBUNIT ARE LOCATED NEAR THE 7-PHOSPHATE GROUP OF ATP IN THE PRESENCE OF M$+*

(Received for publication, August 20, 1990)

Kenji Ida, Takato Noumi, Masatomo Maeda, Toshio Fukui, and Masamitsu Futai From the Department of Organic Chemistry and Biochemistry, Institute of Scientific and Industrial Research, Osaka Uniuersity, Ibaraki, Osaka 567, Japan

The catalytic site of Escherichia coli F1 was probed using a reactive ATP analogue, adenosine triphospho- pyridoxal (AP3-PL). For complete loss of enzyme activ- ity, about 1 mol of AP3-PL bound to 1 mol of F1 was estimated to be required in the presence or absence of Mg2+. About 70% of the label was bound to the a subunit and the rest to the j3 subunit in the absence of Mg2+, and the aLys-201 and j3Lys-155 residues, respectively, were the major target residues (Tagaya, M., Noumi, T., Nakano, K., Futai, M., and Fukui, T. (1988) FEBS Lett. 233, 347-351). Addition of Mg2+ decreased the AP,-PL concentration required for half-maximal in- hibition, and predominant labeling of the j3 subunit (BLys-155 and j3Lys-201) with the reagent. ATP and ADP were protective ligands in the presence and ab- sence of Mg2+.

The a subunit mutation (aLys-201 + Gln or aLys- 201 deletion) were active in oxidative phosphoryla- tion. However, purified mutant Fls showed impaired low multi-site activity, although their uni-site cata- lyses were essentially normal. Thus aLys-201 is not a catalytic residue, but may be important for catalytic cooperativity. Mutant Fls were inhibited less by AP3- PL in the absence of Mg”, and consistent with this, modifications of their a subunits by AP3-PL were re- duced. AP3-PL was more inhibitory to the mutant en- zymes in the presence of Mg2’, and bound to the j3Lys- 155 and j3Lys-201 residues of mutant F1 (aLys-201 + Gln). These results strongly suggest that aLys-201, PLys-155, and j3Lys-201 are located close together near the y-phosphate group of ATP bound to the cata- lytic site, and that the two j3 residues and the y-phos- phate group become closer to each other in the presence of Mg”.

The H+-ATPase (FoFI) of Escherichia coli catalyzes the synthesis of ATP utilizing an electrochemical gradient of protons (for a review, see Refs. 1-4). The F1 portion, or F1- ATPase, consists of five nonidentical subunits, a, P, y, 6, and t, and has ATPase activity. The catalytic site is believed to be on the P subunit or at the interface between the a and /3 subunits.

Binding of about 1 mol of a reactive ATP analogue, aden-

* This research was supported in part by grants from the Ministry of Education, Science, and Culture of Japan. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

osine triphospho-pyridoxal (AP3-PL),’ to 1 mol of F1 caused nearly complete inactivation of the enzyme: about 70% of the AP3-PL bound to the a subunit and the rest to the P subunit in the absence of M$+ (5). The aLys-PO1 and PLys-155 residues of the a and P subunits, respectively, were the targets of this modification reagent (6). The PLys-155 residue is located in the glycine-rich sequence (&-Gly-Ala-Gly-Val- Gly-Lys-Thr, residues 149-156 of the E. coli P subunit) com- monly found in a group of nucleotide binding proteins (con- served residues are underlined), and mutations within the sequence cause drastic change in the kinetics of the enzyme (1, 7). The aLys-201 residue of the isolated a subunit was also labeled with adenosine diphospho-pyridoxal (8). These findings suggested that both residues are located at or near the catalytic site.

For further understanding of the roles of these residues in the catalytic process, we analyzed the modification of F1 with AP3-PL in the presence of M$+, since this cation is required for catalysis and our previous work (5, 6) was carried out in its absence. Furthermore, we constructed a recombinant plas- mid carrying the entire unc operon and introduced mutations (replacing the aLys-PO1 residue) into the uncA cistron. These protein chemical and molecular biological approaches dem- onstrated that Mg2f increased the accessibilities of PLys-155 and PLys-201 to AP3-PL, and that mutation of aLys-201 resulted in decreased modification of the a subunit and cata- lytic cooperativity of the enzyme. AP3-PL was more inhibitory in the presence of M$+, and less inhibitory to mutant en- zymes. On the basis of these results, we discuss the positional relationship of amino acid residues, including those in the glycine-rich sequence, and the y-phosphate group of ATP in the catalytic site of Fl-ATPase.

EXPERIMENTAL PROCEDURES

Bacterial Strains and Growth Conditions-Strains AN718 (unc- A401) (10) and DK8 (Aunc B-C, i1u::TnlO) (11) were as described previously. A rich medium (L-broth) supplemented with thymine, and a minimal medium supplemented with thymine, thiamine, and a carbon source (0.2% glucose, 0.4% succinate, or 0.5% glycerol) were used for genetic analyses (12) and the preparation of membranes and Fls (13). Ampicillin (50 pg/ml) was included in L-broth agar plates when necessary.

Construction of Hybrid Plasmids Carrying Wild-type and Mutant unc Genes-The NruI-NaeI fragment (1777 bp) from pMCR533 (14) was inserted into the EcoRV site of pBR322 to obtain plasmidpBWAl with the wild-type uncA gene. Recombinant plasmids carrying the uncA gene with mutations were constructed as follows: the EcoRI- Sal1 fragment (1414 bp) carrying most of the uncA (a) gene was prepared from pBWAl and ligated into the corresponding multi-

The abbreviations used are: AP3-PL, adenosine triphospho-pyri- doxal; HPLC, high-performance liquid chromatography; MOPS, 3- (N-morpho1ino)propanesulfonic acid; bp, base pair(s).

5424

Catalytic Site of Fl-ATPase 5425

cloning site of pUC119, and mutations were introduced by the method of Taylor et al. (15). The oligonucleotides used for introducing mu- tations (aLys-201 + Gln and aLys-201 deletion) were synthesized using a DNA Synthesizer model 381A (Applied Biosystems). An 430 bp XhoI-SnaBI fragment carrying either of the above mutations was substituted for the wild-type fragment of pBWA1. Mutations were confirmed by DNA sequencing (16). The resultingplasmids (pBMA1, trLys-201 + Gln and pBMA2, aLys-201 deletion) were introduced into a chromosomal uncA mutant AN718 (aSer-373 + Phe) to determine whether the mutant a subunit constructed could comple- ment the mutation.

A plasmid carrying the entire unc operon (pBWU1) was con- structed as follows: the HindIII-Hind111 fragment (4438 bp) carrying the upstream half of the unc operon (uncB(a), E(c) , F(b) , H(G),A(a), and 5'-half of uncG(y) genes) was prepared from pMCR533 and ligated into the Hind111 site of pTN1661(17) carrying the downstream half of the unc operon (3"half of uncG(r), uncD(0) and uncC(r) genes). The 1203 bp XhoI-Nsp7524V fragment of pBWUl was re- placed by the corresponding fragment from pBMAl (nLys-201 + Gln) or pBMA2 (aLys-201 deletion).

Preparation of F , and Its Modification with APR-PL-The hybrid plasmid pBWUl (wild), pBMUAl (aLys-201 + Gln), or pBMUA2 (nLys-201 deletion) was introduced into strain DK8, and the wild- type and mutant Fls released from the membranes were purified by a published method (13). The mutant F,s were homogeneous and had similar submit ratios to those of the wild-type enzyme (a&&) as shown by polyacrylamide gel electrophoresis. Uni- and multi-site ATPase activities were assayed as described previously (12, 13). Protein was measured by the methods of Lowry et al. (18) and Bradford (19) (Bio-Rad protein assay reagent).

For modification with APn-PL, purified F, was passed through a centrifuge column (Sephadex G-50, 0.5 X 5.5 cm) equilibrated with 50 mM MOPS-NaOH (pH 8.6) in the presence or absence of 2 mM MgCI,. The Fls contained five and four bound nucleotides in the presence and absence of Mg", respectively, as determined by a published procedure (20). The F1 (1 p ~ ) was incubated at 25 "C for 30 min with various concentrations of AP3-PL or ['HIAPI-PL in 50 pl of 50 mM MOPS-NaOH (pH 8.0) with or without 2 mM MgC12. Modification was terminated by adding 2 mM NaBH, and the ATPase activity of samples was assayed.

Modified F,s were subjected to 10% polyacrylamide gel electropho- resis in the presence of sodium dodecyl sulfate (21). The gel was dried and autoradiographed. The radioactive bands corresponding to the a and fi subunits were cut out and their radioactivities were measured in an Aloka LSC-700 liquid scintillation counter (5). The recoveries of the radiolabeled a and 0 subunits from polyacrylamide gel were about 80%.

Isolation of Peptides with Bound AP3-PL-Wild-type or mutant F1 was precipitated by ammonium sulfate and passed through a centri- fuge column as described above. The F1 (1 p ~ , 1.5 mg/4.0 ml) was modified with APo-PL at a concentration adjusted to cause 50% inhibition. After NaBH, treatment, the F, solution was concentrated with Centricon-10 (Amicon) and lyophilized. Modified F1 was dis- solved in 0.3 ml of 8 M urea, incubated for 2 h a t 37 "C, and diluted to 1.2 ml. This solution of denatured F,, adjusted to contain 2 M urea and 0.1 M NH,HCOs (pH 8.0), was incubated with trypsin (1 pg of trypsin/50 pg of F,) for 16 h a t 37 "C, and then for a further 4 h after a second addition of the same amount of trypsin. It was then lyoph- ilized and the residue was dissolved in 400 pl of 0.1% trifluoroacetic acid. Peptides modified with AP3-PL were purified by HPLC at room temperature using a Shimadzu Gradient-LC6A System (Kyoto, Ja- pan) and C4 reversed phase column (Vydac). The following buffers were used solution A, 0.1% trifluoroacetic acid in water; solution B, 0.1% trifluoroacetic acid and 80% acetonitrile in water; solution C, 10 mM acetic acid adjusted to pH 6.0 with triethylamine; solution D, 10 mM acetic acid/triethylamine (pH 6.0), and 75% acetonitrile. Amino acid sequencing was performed in an Applied Biosystems 477A gas-phase sequencer linked with an Applied Biosystems 120A phenylthiohydantoin analyzer.

Chemicals-APR-PL and ['HIAPB-PL (2.4 X lo5 cpm/nmol) were synthesized by published methods (22,23). Trypsin (L-l-tosylamido- 2-phenylethyl chloromethyl ketone-treated) was obtained from Sigma. Restriction enzymes, T4 DNA ligase, and the large fragment of E. coli DNA polymerase I were purchased from Takara Shuzo Co., Kyoto, Japan and Nippon Gene, Toyama, Japan. [a-"PIdCTP (400 Ci/mmol) was obtained from Amersham Corp. All other chemicals used were of the highest grade available commercially.

RESULTS

Effect of Mf on the Modification of F, with AP3-PL- aLys-201 and PLys-155 of F1 were the sites of modification by APa-PL in the absence of Mg?+ (5). In the presence of MgZ+ (the divalent cation required for the enzyme reaction), enzyme inactivation by AP3-PL was much more pronounced (Fig. 1, A and B ) : half-maximal inhibition by APa-PL occurred a t 2.5 pM in the presence of Mp"', in contrast to 10 pM in its absence. Autoradiograms showed that the 0 subunit was pre- dominantly labeled with ['HIAPs-PL in the presence of M e and that the ratio of radioactivities associated with the a and P subunits changed from 4 1 without Mg?+ (Fig. 2 A ) to 1:3 with Mg2' (Fig. 2 B ) . 'This alteration in the pattern of prefer-

+ : 100 L

R

AP3-PL (PM)

FIG. 1. Loss of multi-site (steady state) ATPase activity of F1 modified with APs-PL in the presence and absence of M e . Wild-type or mutant F, (1 p ~ ) was incubated at 25 "C with various concentrations of APn-PL in 50 pl of 50 mM MOPS-NaOH (pH 8.0) with ( B ) or without (A) 2 mM MgC12. After 30 min, 2 mM NaBH, was added and the multi-site ATPase activities of samples of the mixtures were assayed. Results are expressed as percentages of the control without APo-PL 0, wild-type; ., aLys-201+ Gln; A, aLys- 201 deletion.

Wild

Lys-201- Gln

A ~ y s - 2 0 1

- 0 - %

- a - %

- a - %

FIG. 2. Binding of APa-PL to a and /3 subunits in the pres- ence and absence of Mg'. Wild-type or mutant F, (aLys-201 + Gln or aLys-201 deletion) (1 p ~ ) was incubated a t 25 "C with 5 p~ (with 2 mM MgC12, B ) or 10 pM (without M F , A ) ['HIAPn-PL in 50 pl of 50 mM MOPS-NaOH (pH 8.0). 1 mM of ATP was also added as indicated. After a 30-min incubation, 2 mM NaBH, was added and an aliquot (20 pl) was subjected to polyacrylamide gel electrophoresis in the presence of 0.1% sodium dodecyl sulfate. The gel was stained with Coomassie Brilliant Blue and autoradiographed. From top to bottom, wild type, aLys-201 + Gln and aLys-201 deletion. APnPL bound to only the a and B subunits. Essentially the same results were obtained with 2.5 p~ AP,-PL. The slightly greater incorporation of AP3-PL into the a subunit in the absence of M e than in the previous study (5) may be due to the lower concentration of F, used in the present study.

5426 Catalytic Site of F1-ATPase

entia1 modification with M$+ suggests that the residue(s) in the p subunit became more accessible to AP3-PL in the presence of Mg2+.

The modifications of the a and (3 subunits were both re- duced in the presence of ATP (Fig. 2) or ADP (data not shown) (both with and without Mg"). A plot of the residual ATPase activity against the total radioactivity of [3H]AP3- P L bound to the a and p subunits suggested that about 1 mol of AP3-PL binding resulted in complete inactivation of the enzyme both in the presence and absence of Mg2+ (Fig. 3A). Thus the modification of either the a or p subunit was responsible for the inactivation.

Predominant Binding of AP3-PL to PLys-155 (in the Gly- cine-rich Sequence) and /3Lys-201-The finding that modifi- cation of the p subunit was increased in the presence of M$+ prompted us to determine the residue modified. F1 modified in the presence of Mg2+ was digested with trypsin and the resulting peptides were separated by HPLC (Fig. 4, A-1): about 90% of the total fluorescent AP3-PL bound peptides injected was eluted from the column. The highest peak eluted at 80 min ( a r r o w h a d ) corresponded to 70% of the total fluorescent peptides (Fig. 4, A-2). The sample modified in the absence of Mg2+ gave the same peak, but of much reduced height (Fig. 4, A-4). The peak could not be detected, after modification of F1 in the presence of ATP (Fig. 4, A-3 and A- 5). Thus the fluorescent peak(s) contained a peptide(s) spe- cifically modified by AP3-PL in the presence of M e .

The peak material was further purified using a second solvent system to obtain a major fluorescent peak (Fig. 5A, lower trace) contained two peptides (Table I). These could not be separated by rechromatography at a slower flow rate and with a more gradual gradient. Peptide I corresponded to the sequence between Val-145 and Arg-164 of the p subunit, and Peptide I1 to that between Thr-183 and Arg-216 of the same subunit. The residues between Gly-207 and Arg-216 of Peptide I1 could not be detected. No phenylthiohydantoin derivative of p-Lys-155 or PLys-201 was detectable at the corresponding cycles, suggesting that these residues were modified with AP3-PL. The amounts of both peptides were essentially similar, and they were not cross-linked by disulfide bonds, because the @ subunit has only 1 cysteine residue at position 137. F1 modified in the absence of M$+ gave a fluorescent peak with an identical retention time to that of the above peptides (Fig. 5B), but this peak contained only Peptide I (Table I). These results suggested that @Lys-201 was accessible to the formyl group of AP3-PL only in the presence of M$+, whereas PLys-155 was accessible in both

Bound AP3-PL (mol /mol F1)

FIG. 3. Estimation of the amount of AP8-PL binding re- quired for complete inactivation of the enzyme. Wild-type and mutant (aLys-201 + Gln and aLys-201 deletion) Fls were modified with [3H]AP:r-PL (2.5, 5, 10, and 50 PM) as described in the legend to Fig. 2, and multi-site ATPase activities were measured. The radioac- tivities associated with the a and 0 subunits after gel electrophoresis were determined, and the total amounts of AP,-PL bound to F, (moles/mol) are plotted against the percentage residual ATPase activity. A , wild type; B , aLys-201 + Gln; C, OtLys-201 deletion.

I . . . . I . . . . , I 0

U 50 100 80

T i m e ( rn in )

FIG. 4. Purification of peptides modified with AP3-PL by reversed phase HPLC. Wild-type F, (1.5 mg) modified with AP3- PL in the presence of 2 mM MgC1, was concentrated with Centricon- 10 (Amicon) and trypsinized. A portion of the digest (equivalent to 200 pg of Fl) was subjected to HPLC on a C4 reversed phase column equilibrated with solution A. The column was washed with solution A for 8 min and then peptides were eluted with a linear gradient of 0-50% (v/v) solution B in solution A in a period of 100 min a t a flow rate of 1 ml/min. The absorbance a t 214 nm ( A - 1 ) and fluorescence (arbitrary units) at 395 nm (excitation wave length = 335 nm) (A-2 ) were monitored. The elution profile of the peptide from wild-type F1 modified in the presence of 1 mM ATP was essentially the same as A - l (not shown) and only peaks monitored by fluorescence were shown in A-3. A-4 and A-5 show the fluorescent peaks from a digest of wild-type F1 modified without MgZ+ in the absence and presence of ATP, respectively. The fluorescent peak at about 4 min was due to free adenosine triphospho-pyridoxine derived from AP,-PL. Mu- tant Fls ( B , OtLys-201 +. Gln; C, aLys-201 deletion) gave similar fluorescent peaks to A - 2 and A-4 irrespective of the presence (B-I and C-1) or absence ( B - 3 and C-3) of M e . Only the recording the region of 80 min is shown for the mutants. Fluorescent materials could not be detected in a methanol wash of the column after gradient elution. The fluorescence patterns of the peptide from F, incubated with AP3-PL and ATP are also shown: E-2 and (2-2, plus M e ; B-4 and C-4, minus M e .

the presence and absence of the cation. Effect of aLys-201 Mutations on Modification of Fl with

AP3-PL-Next we examined the inhibitory effect of AP3-PL on F1 with a mutant a subunit in which the aLys-201 residue (the main AP3-PL binding site in the absence of M e ) (6) was substituted (Fig. 6). Consistent with the lack of aLys- 201, treatment of mutant F1 with AP3-PL (in the absence of Mg2+) caused only weak inhibition compared with that of the wild-type F1: half-maximal inhibitions of aLys-201 "-* Gln substitution and aLys-201 deletion Fls were observed at 40 and 50 I . ~ M AP3-PL, respectively, (Fig. lA), and the incorpo- rations of [3H]AP3-PL into the mutant a subunits were much less than that into the wild-type F, (Fig. 2 A ) . The ratios of the radioactivities incorporated into the a and /3 subunits were constant (1:2) in the presence of 2.5-10 jtM APs-PL. These results were consistent with the previous finding and aLys-201 of the wild-type F1 was the major target of AP3-PL in the absence of M$+.

Mg2+ reduced the AP3-PL concentrations required for half- maximal inhibition of the mutant enzymes (5 and 3 j t ~ AP3- PL for aLys-201 +- Gln substitution and aLys-201 deletion, respectively) (Fig. 1B). Furthermore, AP3-PL binding to the p subunits became predominant in the presence of M2+: 75 and 86% of the radioactivities were bound to the /3 subunits of the aLys-201 +- Gln substitution and aLys-201 deletion mutants, respectively (Fig. 2B). A plot of the residual ATPase activity against the total radioactive AP3-PL bound to the a and p subunits suggested that 1 mol of AP3-PL binding was demonstrated to be sufficient for complete inactivation of the mutant Fls in the presence or absence of M$+ (Fig. 3, B and

Catalytic Site of F1-ATPase 5427

30 40 50 60 T i m e (rnin)

FIG. 5. Rechromatography of fluorescent peptide(s). The fluorescent peak fractions indicated by arrowheads in Fig. 4 ( A - 2 and A - 4 ) were concentrated and dissolved in 10 mM CH3COOH-triethyl- amine (pH 6.0). Samples (corresponding to 4 mg of F,) were subjected to HPLC using the same column as for Fig. 4. Peptides were eluted at a flow rate of 1 ml/min with 20% solution D in solution C for 20 min, a gradient of 20-40% solution D in solution C in 40 min, and a gradient of 40-80% solution D in solution C in 5 min. Fluorescent peak was eluted during washing with the second solvent system. Absorbance at 214 nm (upper) and fluorescence (arbitrary units) a t 395 nm (lower) were monitored. A , peptides from F, modified in the presence of Mg’+; E , peptides from F, modified in the absence of Mgy+.

C) and ATP (Fig. 2) and ADP (data not shown) protected the mutant F,s from modification.

Binding of AP,-PL to PLys-155 and PLys-201 of Mutant F1-Mutant Fls (aLys-201 + Gln and aLys-201 deletion) modified with AP,-PL were digested with trypsin and sub- jected to HPLC as described for the wild-type: essentially the same fluorescent peak was obtained in the presence or absence of Mg2+ (Fig. 4, B-1, B-3, C-I, and C-3), and this peak was not obtained from F1 that had been incubated with ATP and AP,-PL (Fig. 4, B-2, B-4, C-2, and C - 4 ) . On further purifica- tion of the fluorescent peak from mutant F1 (aLys-201 + Gln), a single major peak with the same retention time as that of the wild-type was obtained (not shown). Sequence analysis demonstrated that the peak from F1 modified with Mg2+ contained Peptide I and Peptide 11, and that PLys-201 and PLys-151 were modified, while the peak from F, modified without Mg’+ contained only Peptide I in which PLys-155 was modified (Table I). These findings indicated that the PLys- 155 and PLys-201 residues (identical residues to those of wild- type F,) in mutant F1 became more accessible to the formyl group of APa-PL in the presence of Mg2+.

Properties of Fls with aLys-201 Mutations-When strain DK8 lacking the unc operon was transformed with recombi- nant plasmids carrying the entire unc operon with a mutant or wild-type uncA gene, the transformants synthesized FoF, with a mutant or wild-type a subunit. The cell with altered uncA gene (aLys-201 + Gln or aLys-201 deletion) was able

to grow by oxidative phosphorylation, indicating that both mutant a subunits were functionally active. The growth yields of the transformants with succinate and glucose were essen- tially the same as those of the wild type (data not shown). Consistent with these observations, recombinant plasmids carrying only the mutant uncA gene (pBMA1 and pBMA 2) also complemented strain AN718 of a defective a subunit (aSer-373 + Phe).

Membranes of the cells with mutant FoF1 showed significant ATPase activities (Table 11). Translocation of H+ coupled with ATP hydrolysis was observed with mutant membrane vesicles by monitoring quenching of dye fluorescence, and F1- depleted vesicles showed passive H+ translocation through Fo (25) (not shown). Specific activities (multi-site or steady state ATP hydrolyses) of the purified enzymes with mutant a subunits were lower than that of the wild-type F,, consistent with their membrane bound ATPase activities (Table 11). However, uni-site (single site) ATP hydrolysis by F1 with the alys-201 + Gln substitution or aLys-201 deletion was not significantly different from that of the wild-type F,. These results suggest that Fls with mutant a subunits were active and that aLys-201 itself is not a catalytic residue, although replacement of this residue or structural alteration in its vicinity lowered the catalytic cooperativity.

DISCUSSION

Incorporation of about 1 mol of AP,-PL (either at the aLys- 201 or PLys-155 residue) per mol F, in the absence of Mg2+ (5) was sufficient for complete enzyme inactivation, suggest- ing that both residues are located at or near the catalytic ATP-binding site and possibly in the vicinity of the y-phos- phate group of ATP. Substrate analogues, such as 3’-0-(4- benzoy1)benzoyl ATP (26), 4-azido-2-nitrophenyl pyrophos- phate (27), and 3’-arylazido-P-alanyl-2-azido ATP (28), are known to modify both the a and /3 subunits of mitochondrial and thermophilic bacterial F,s.

However, PLys-155 and PLys-201 were more reactive with AP3-PL in the presence of Mg’+. Judging from the yields of peptides with modified PLys-155 and PLys-201 residues, AP:,- PL reacted mainly with these residues in the presence of Mg2+. It is noteworthy that PLys-155 is in the Gly-rich se- quence (Gly-X-X-X-X-Gly-Lys-Thr/Ser) commonly found in a group of nucleotide binding proteins (7) including F, (Gly- Gly-Ala-Gly-Val-Gly-Lys-Thr, E. coli residues 149-156), sug- gesting that the region is important for catalysis. Mutational studies also supported this notion (7). Furthermore, a syn- thetic peptide corresponding to the sequence between residue 134 and 183 (E. coli numbering system) of the mitochondrial P subunit carrying the Gly-rich sequence could bind ATP (29). These results strongly suggest that the Gly-rich sequence of the P subunit is part of the catalytic site.

The present study suggests that PLys-201 may also be present near the catalytic site. The vicinity of this residue is suggested to be important for catalysis: replacement of PMet- 209 by Ile resulted in defective catalysis (30) and binding of dicyclohexylcarbodiimide to PGlu-181 or PGlu-192 inhibited the enzyme activity (31, 32). The Mg2+-dependent AP,-PL modification of PLys-155 and PLys-201 observed in this study and the change of the divalent cation dependence of the catalysis due to a PSer-174 -+ Phe substitution (33) suggest that Mg”-binding affects the conformation of the region between PLys-155 and PLys-201. I t will be interesting to determine whether this region participates directly in Mg2+ binding.

Deletion or substitution of aLys-201 reduced the reactivity of the a subunit with AP3-PL, and the inhibition of the

Catalytic Site of Fl-ATPase TABLE I

Amino acid sequences of AP3-PL binding peptides AP,-PL binding peptides purified from wild-type and mutant (aLys-201-+ Gln) Fls (modified in the presence

and absence of M P ) were used for amino acid sequence analyses. The amino acid residues identified (phenylthio- hydantoin derivatives) and their recoveries are shown. Phenylthiohydantoin derivative of the residue not detectable is represented as ND. Peptide I corresponds to the sequence between Val-145 and Arg-264 of the f l subunit. The amino acid residues of Peptide I1 after cycle 24 or 21 could not be detected, but this peptide corresponds to the sequence between Thr-183 and Arg-216 judging from the amino acid sequence of the @ subunit (24). In cycle 12, only the phenylthiohydantoin derivative of the Thr residue could be detected. Thus the values shown in the table are sums of the residues derived from Peptides I and 11.

Wild Cycle No. +M$+

Peptide I -Mg?+ Peptide I

Peotide I1

aLys-201+ Gln

+ M e - M P Peptide I

Peptide I Peptide I1 pmol

1 Val 2 GlY 3 Leu 4 5

Phe GlY

6 7

GlY

8 Ala

9 GlY Val

10 Gly 11 LYS 12 Thr 13 Val 14 Asn 15 Met 16 Met 17 Glu 18 Leu 19 Ile 20 Arg 21 22 23 24

464 519 482 507 577 487 510 434 544 445 ND 262 297 225 283 335 158 211 156 41

Thr Ark? Glu G ~ Y Asn ASP Phe Tyr His Glu Met Thr ASP Ser Asn Val Ile ASP LYS Val Ser Leu Val Tvr

82 33

151 228 202 138 216 63 59

173 179 262 83 27

128 143 96 73

ND 99 19 42 55 10

pmol Val 132 GlY 221 Leu 89 Phe 89 Gly 162 GlY 183 Ala 160 GlY 93 Val 92 GlY 88 LYS ND Thr 10 Val 28 Asn 36 Met 24 Met 24 Glu 33 Leu 5 Ile 9

194 201 I1. TYP V.1 A l a I l e Glu G l n L y e Ala Ser Thr I l e Scr Asn

207

w i l d ATC TAT GTC GCT ATC CGC CAD AAA GCG TCC ACC ATT TCT AAC

LYS-201 - C h *TC TAT GTC GCT A$ GCC CAC f E CCG TCC ACC ATT TCT x"

A ~ r s - 2 0 1 ATC IAT CTC WT AT? mi CE . . . $ xcc ACc A I T TCT S I

FIG. 6. Construction of hybrid plasmids carrying an altered uncA gene. Recombinant plasmid pBWAl carrying the wild-type uncA gene was constructed as described under "Experimental Pro- cedures." The EcoRI-Sal1 fragment was transferred to pUC119 to introduce site-directed mutations. The oligonucleotides (39-mer) used are shown in the figure. Base substitutions (upper dots), and new restriction sites (Ban and StuI) (underlined) were introduced. XhoI- SnaBI fragments carrying the mutation (aLys-201-+ Gln and aLys- 201 deletion) were substituted for the wild-type fragment of pBWA1, giving the plasmids pBMAl and pBMA2, respectively. A plasmid

pMCR533 and pTN1661. The XhoI-Nsp7524V fragment of pBWUl carrying the entire unc operon (pBWU1) (43) was constructed from

was replaced by the corresponding fragments from pBMAl (cuLys- 201 + Gln) and pBMA2 (aLys-201 deletion). The resulting plasmids (pBMUA1 and pBMUA2) were introduced into DK8 for preparation of F,. Abbreviations of restriction sites: Ba, BalI; E, EcoRI; Na, NaeI; Nr, NruI; Ns, Nsp7524V; S, SalI; Sn, SnaBI; St, StuI; X, XhoI.

pmol Val GlY Leu Phe GlY GlY Ala GlY Val G ~ Y LY s Thr Val Asn Met Met Glu Leu Ile Arg

164 160 562 125 149 231 150 140 140 82

ND 67 52 51 30 41 26 40 25 48

Thr Arg Glu GlY Asn ASP Phe TYr His Glu Met Thr ASP Ser Asn Val Ile ASP LY s Val Ser

69 238 100 112 72

104 100 44 32 43 53 67 46 15 32 49 63 40

ND 21 8

P m l Val 240 GlY 172 Leu 32 Phe 106 Gly 243 G ~ Y 211 Ala 228 GlY 139 Val 221 GlY 97 LYS ND Thr 41 Val 85 Asn 54 Met 46 Met 54 Glu 18 Leu 9 Ile 28 Arg 3

-

TABLE I1 Effects of replacement or deletion of cuLys-201 on ATPase activities of

membrane bound and purified Fi DK8 cells were transformed by the plasmids (pBWU1, wild type;

pBMUA1, aLys-201 -+ Gln; pBMUA2, aLys-201 deletion) con- structed in this study as described. Membrane vesicles were prepared and multi-site ATPase activity was assayed as described previously (13). One unit of the enzyme was defined as the amount hydrolyzing 1 pmol of ATP/min. F,s were purified and multi-site (steady state) and uni-site (single site) ATP hydrolyses were assayed (12, 13). The kl is the rate constant of binding of ATP to Fl in mi-site hydrolysis and was calculated from the results of a cold-chase experiment.

Purified F,-ATPase Mutation Membrane

ATPase Multi-site Uni-site V.".. k,

units/mg unitslmg M" s-l Wild 6.4 80-90 9 X lo4 Lys-201+ Gln 2.1 20 7 x 104 Lys-201 deletion 1.7 8.5 5 X 104

mutant enzyme by AP3-PL in the absence of M P . As modi- fication of CYLYS-201 in the wild type caused loss of enzyme activity, and in the isolated CY subunit this residue was modi- fied by a similar nucleotide analogue AP2-PL (adenosine diphosphopyridoxal) (8), this residue may be located near the y-phosphate group of ATP. However, this residue could be substituted without impairing the oxidative phosphorylation (in vivo) significantly, suggesting that aLys-201 itself is not a catalytic residue, although the mutation lowered the multi- site activity requiring conformational interaction between the

h

Catalytic Site

FIG. 7. Model of the catalytic site of FI. APB-PL binding sites (aLys-201, PLys-155, and PLys-201) (black), binding sites of other chemical modification reagents (hatched) and loci of catalytic muta- tions (black) are shown.- a helix; e, /3 sheet. *, unique orientation of the putative a helix and /3 sheet in this model.

a and P subunits. In this regard Senior (34) suggested that the nucleotide binding site in the a subunit is a noncatalytic site and AP3-PL or AP2-PL disrupt subunit-subunit interac- tion.

The residues in the catalytic sites of Fls from various sources have been probed by chemical modification. The region PThr-287 + Tyr-297 (Thr-287, Ile-290, and Tyr-297, E. coli numbering system) was a target of photoaffinity label- ing reagent 8-azido-ATP (35). Moreover, residues in this region reacted with 7-chloro-4-nitrobenzofurazan (PTyr-297 (36)) and 4-azido-2-nitrophenyl phosphate (PIle-290, Gln-294 and Tyr-297 d37)). Since the nitrobenzofurazan moiety is transferred from PTyr-297 to BLys-155 (38), the regions in their vicinities may be close together. 2-Azido-ADP reacted with the region PLeu-328 + Pro-332 (Leu-328, Ile-330, Tyr- 331, and Pro-332) (39). myr-331 and @Asp-338 are the targets for 3’-0-(4-benzoyl)benzoyl ATP (40), and myr-331 was also modified with 5’-p-(fluorosulfonyl)benzoyl inosine (41). A model of catalytic site was drawn based on these previous results (Fig. 7). A similar model was presented previously (29, 30) based on the crystal structure of adenylate kinase (42), but the unique orientation of the putative a! helix and /3 sheet (marked with asterisks) in the present model is noteworthy. The @Lys-201 residue was located far from the y-phosphate group of ATP in the previous model (30) in contrast to its position in this model. In this model we also include the mutation sites that are known to cause catalytic defects (1). As these important residues are mainly localized in the region between Gly-142 and Arg-246, this region may be especially important for catalytic processes: not only substrate binding, but also M e binding, ATP hydrolysis, and catalytic coop- erativity.

Acknowledgment-We are grateful to K. Furukawa for carrying out amino acid sequence analyses.

REFERENCES 1. Futai, M., Noumi, T., and Maeda, M. (1989) Annu. Reo. Biochem.

2. Futai, M., and Kanazawa, H. (1983) Microbiol. Reu. 47, 285-312 3. Senior, A. E. (1988) Physiol. Reu. 6 8 , 177-231

58, 111-136

3f Fl-ATPuse 5429

4. Walker, J. E., Saraste, M., and Gay, N. J. (1984) Biochim.

5. Noumi, T., Tagaya, M., Miki-Takeda, K., Maeda, M., Fukui, T.,

6. Tagaya, M., Noumi T., Nakano, K., Futai, M., and Fukui, T.

7. Hsu, S.-H., Noumi, T., Takeyama, M., Maeda, M., Ishibashi, S.,

8. Rao, R., Cunningham, D., Cross, R. L., and Senior, A. E. (1988)

9. Kanazawa, H., Horiuchi, Y., Takagi, M., Ishino, Y., and Futai,

10. Noumi, T., Futai, M., and Kanazawa, H. (1984) J. Biol. Chem.

11. Klionsky, D. J., Brusilow, W. S. A,, and Simoni, R. D. (1984) J.

12. Noumi, T., Taniai, M., Kanazawa, H., and Futai, M. (1986) J.

13. Futai, M., Sternweis, P. C., and Heppel, L. A. (1974) Proc. Natl.

14. Kanazawa, H., Tamura, F., Mabuchi, K., Miki, T., and Futai, M.

15. Taylor, J. W., Ott, J., and Eckstein, F. (1985) Nuckic Acids Res.

16. Sanger, F., Coulson, A. R., Barell, B. G., Smith. A. J. H., and

17. Kuki, M., Noumi, T., Maeda, M., Amemura, A., and Futai, M.

18. Lowry, 0. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J.

19. Bradford, M. M. (1970) Anal. Biochem. 7 2 , 248-254 20. Hanada, H., Noumi, T., Maeda, M., and Futai, M. (1989) FEBS

21. Laemmli, U. K. (1970) Nature 227,680-685 22. Tagaya, M., and Fukui, T. (1986) Biochemistry 25,2958-2964 23. Raibaud, O., and Goldberg, M. E. (1974) FEBS Lett. 40,41-44 24. Kanazawa, H., and Futai, M. (1982) Ann. N. Y. Acad. Sci. 402 ,

45-64 25. Eya, S., Noumi, T., Maeda, M., and Futai, M. (1988) J. Biol.

Chem. 263,10056-10062 26. Williams, N., and Coleman, P. S. (1982) J. Bwl. Chem. 257 ,

2834-2841 27. Michel, L., Garin, J., Issartel, J.-P., and Vignais, P. V. (1989)

Biochemistry 28 , 10022-10028 28. Schafer, H.-J., Rathgeber, G., Dose, K., and Kagawa, Y. (1989)

FEBS Lett. 253,264-268 29. Garboczi, D. N., Shenbagamurthi, P., Kirk, W., Hullihen, J., and

Pedersen, P. L. (1988) J. Biol. Chem. 263,812-816 30. Parsonage, D., Duncan, T. M., Wilke-Mounts, S., Kironde, F. A.

S., Hatch, L., and Senior, A. E. (1987) J. Biol. Chem. 262 ,

31. Yoshida, M., Poser, J. W., Allison, W. S., and Esch, F. S. (1981) J. Biol. Chem. 256 , 148-153

32. Yoshida, M., Allison, W. S., Esch, F. S., and Futai, M. (1982) J. Biol. Chem. 257,10033-10037

33. Noumi, T., Mosher, M. E., Natori, S., Futai, M., and Kanazawa, H. (1984) J. Biol. Chem. 259,10071-10075

34. Senior, A. E. (1990) Annu. Reu. Biophys. Biophys. Chem. 19 , 7- 41

35. Hollemans, M., Runswick, M. J., Fearnley, I. M. and Walker, J. E. (1983) J. Biol. Chem. 258,9307-9313

36. Andrews, W. W., Hill, F. C., and Allison, W. S. (1984) J. Biol. Chem. 259,8219-8225

37. Garin, J., Michel, L., Dupuis, A., Issartel, J.-P., Lunardi, J., Hoppe, J., and Vignais, P. (1989) Biochemistry 28,1442-1448

38. Andrews, W. W., Hill, F. C . , and Allison, W. S. (1984) J. Biol.

39. Garin, J., Boulay, F., Issartel, J. P., Lunardi, J., and Vignais, P.

40. Admon, A., and Hammes, G. G. (1987) Biochemistry 2 6 , 3193-

41. Bullough, D. A., and Allison, W. S. (1986) J. Biol. Chem. 2 6 1 ,

42. Sachsenheimer, W., and Schulz, G. E. (1977) J. Mol. Biol. 114 ,

43. Takeyama, M., Ihara, K., Moriyama, Y., Noumi, T., Ida, K., Tomioka, N., Itai, A., Maeda, M., and Futai, M. (1990) J. BioL Chem. 265,21279-21284

Biophys. Acta. 7 8 6 , 164-200

and Futai, M. (1987) J. Biol. Chem. 262,7686-7692

(1988) FEBS Lett. 233,347-351

and Futai, M. (1987) FEBS Lett. 218 , 222-226

J. Biol. Chem. 263,5640-5645

M. (1980) J. Biochem. (Tokyo) 88,695-703

259,10076-10079

Bacteriol. 160 , 1055-1060

Biol. Chem. 261,9196-9201

Acad. Sci. U. S. A. 71,2725-2729

(1980) Proc. Natl. Acad. Sci. U. S. A. 7 7 , 7005-7009

13,8765-8785

Roe, B. A. (1980) J. Mol. Biol. 143,161-178

(1988) J. Biol. Chem. 2 6 3 , 17437-17442

(1951) J. Biol. Chem. 193,265-275

Lett. 257,465-467

6301-6307

Chem. 259,14378-14382

V. (1986) Biochemistry 26,4431-4437

3197

14171-14177

23-36