ATP synthase. Conditions under which all catalytic sites of the F1 moiety are kinetically equivalent in hydrolyzing ATP.

Conditions have been reported under which the F1 moiety of bovine heart ATP synthase catalyzes the hydrolysis of ATP by an apparently cooperative mechanism in which the slow rate of hydrolysis at a single catalytic site (unisite catalysis) is enhanced more than 10(6)-fold when ATP is added in excess to occupy one or both of the other two catalytic sites (multisite catalysis) (Cross, R. L., Grubmeyer, C., and Penefsky, H. S. (1982) J. Biol. Chem. 257, 12101-12105). In the novel studies reported here, and in contrast to the earlier report, we have (a) monitored the kinetics of ATP hydrolysis of F1 by using nucleotide-depleted preparations and a highly sensitive chemiluminescent assay; (b) followed the reaction immediately upon addition of F1 to ATP, rather than after prior incubation with ATP; and (c) used a reaction medium with Pi as the only buffer. The following observations were noted. First, regardless of the source of enzyme, bovine or rat, and catalytic conditions (unisite or multisite), the rates of hydrolysis depend on ATP concentration to the first power. Second, the first order rate constant for ATP hydrolysis remains relatively constant under both unisite and multisite conditions declining only slightly at high ATP concentration. Third, the initial rates of ATP hydrolysis exhibit Michaelis-Menten kinetic behavior with a single Vmax exceeding 100 micromol of ATP hydrolyzed per min/mg of F1 (turnover number = 635 s-1) and a single Km for ATP of about 57 microM. Finally, the reaction is inhibited markedly by low concentrations of ADP. It is concluded that, under the conditions described here, all catalytic sites that participate in the hydrolysis of ATP within the F1 moiety of mitochondrial ATP synthase function in a kinetically equivalent manner.

Conditions have been reported under which the F 1 moiety of bovine heart ATP synthase catalyzes the hydrolysis of ATP by an apparently cooperative mechanism in which the slow rate of hydrolysis at a single catalytic site (unisite catalysis) is enhanced more than 10 6 -fold when ATP is added in excess to occupy one or both of the other two catalytic sites (multisite catalysis) (Cross, R. L., Grubmeyer, C., and Penefsky, H. S. (1982) J. Biol. Chem. 257, 12101-12105). In the novel studies reported here, and in contrast to the earlier report, we have (a) monitored the kinetics of ATP hydrolysis of F 1 by using nucleotide-depleted preparations and a highly sensitive chemiluminescent assay; (b) followed the reaction immediately upon addition of F 1 to ATP, rather than after prior incubation with ATP; and (c) used a reaction medium with P i as the only buffer. The following observations were noted. First, regardless of the source of enzyme, bovine or rat, and catalytic conditions (unisite or multisite), the rates of hydrolysis depend on ATP concentration to the first power. Second, the first order rate constant for ATP hydrolysis remains relatively constant under both unisite and multisite conditions declining only slightly at high ATP concentration. Third, the initial rates of ATP hydrolysis exhibit Michaelis-Menten kinetic behavior with a single V max exceeding 100 mol of ATP hydrolyzed per min/mg of F 1 (turnover number ‫؍‬ 635 s ؊1 ) and a single K m for ATP of about 57 M. Finally, the reaction is inhibited markedly by low concentrations of ADP. It is concluded that, under the conditions described here, all catalytic sites that participate in the hydrolysis of ATP within the F 1 moiety of mitochondrial ATP synthase function in a kinetically equivalent manner.
ATP synthase (F 0 F 1 -ATPase), the enzyme that synthesizes and hydrolyzes the ␥-phosphate bond of ATP, is crucial for the life of aerobic organisms. The enzyme resides in the inner mitochondrial membrane of animals, plants, yeast, and Neurospora; in the cytoplasmic membrane of bacteria and in the thylakoid membrane of chloroplasts in plants (see Refs. 1-6 for reviews). In accordance with the chemiosmotic hypothesis (7) the electrical energy of respiration is first conserved as a protonmotive force (⌬ H ϩ), which via the F 0 moiety (8) of the synthase delivers the accumulated energy to the ␤-subunits of the F 1 moiety (␣ 3 ␤ 3 ␥␦⑀), presumably through conformational changes in the stalk. There is a wealth of experimental evidence supporting the view that the step with the largest demand for energy is the one involved in the release of ATP from the catalytic sites of the enzyme. According to basic postulates of the "binding change" mechanism (9 -11), there is an equivalent participation of the three ␤-subunits in the synthesis of ATP as they proceed through a cycle of "open," "loose," and "tight" states. Thus, at any one time, all three catalytic sites are in different conformations, but all pass sequentially through the same conformations (9 -11). The alternating participation of catalytic sites was shown, for the first time by Adolfsen and Moundrianakis (12) in hydrolytic reactions catalyzed by bacterial F 1 -ATPase.
Although the equivalent, alternating participation of catalytic sites within ATP synthases does not require positive cooperativity among them, many investigators believe that such cooperativity does in fact occur. The genesis of this long held view derives primarily from studies (13,14) in which the F 1 moiety of the bovine heart enzyme was compared under unisite and multisite catalytic conditions, i.e. conditions in which the ATP/F 1 ratio was adjusted so that either one or all sites were operating. The resultant multisite/unisite rate enhancement ratio of 10 6 was interpreted as reflecting strong positive cooperativity among catalytic sites (13). The bovine heart F 1 preparation used in these experiments contained bound nucleotide (2.8 -3.5 mol/mol of F 1 ) and was prior incubated with ATP (13).
In studies reported here, and for the first time, the kinetics of ATP hydrolysis catalyzed by nucleotide-depleted F 1 have been determined by measuring the entire time course of individual reactions from the moment ATP (in the range from 1 nM to ϳ 20 M) enters in contact with the catalytic sites until it is totally hydrolyzed. Thus, catalytic sites are open when the experiment is commenced, and data are collected throughout the entire range from unisite through multisite reaction conditions. This study demonstrates that regardless of the ATP/F 1 molar ratio and enzyme source, bovine or rat, that all catalytic sites participating in ATP hydrolysis within F 1 can function in a kinetically equivalent manner.

EXPERIMENTAL PROCEDURES
Sources of Enzymes, Chemicals, and Materials-Bovine heart F 1 was obtained from Drs. William Allison and J. M. Jault (University of California, San Diego). The enzyme was prepared by a modification (15) of the procedure of Knowles and Penefsky (16) and depleted of nucleotides as described by Garrett and Penefsky (17). It was stored prior to use in the refrigerator at 4°C in 100 mM Tris-Cl, pH 8.0, 4 mM EDTA, and 50% glycerol (v/v). Rat liver F 1 was purified by a modification (18) of the procedure of Catterall and Pedersen (19). The purified enzyme, in 250 mM KP i and 5.0 mM EDTA, was divided into 100-l aliquots and lyophilized to dryness and stored at Ϫ20°C until use. Immediately before use the enzyme was redissolved in 100 l of water and precipi-tated twice with ammonium sulfate. The bovine heart and rat liver F 1 preparations contained, respectively, Ͻ0.04 (11 determinations) and 0.9 (7 determinations) tightly bound nucleotide/mol enzyme when determined after denaturation by the highly sensitive chemiluminescent assay described below. As it is known that such preparations can bind a total of 5-6 mol of nucleotide/mol of F 1 (20,21), these preparations are appropriately defined as "nucleotide-depleted." In fact, the bovine heart F 1 preparation employed here is essentially nucleotide free.
ATP was a product of Sigma and the 1243-200 ATP monitoring reagent, a mixture of luciferin and luciferase was a product of BioOrbit. The luminometer used in these studies to monitor ATP disappearance was a product of LKB (Wallac model 125), and the attached recorder was a product of Soltec (model 330). [␥-32 P]ATP was purchased from DuPont NEN and its radioactivity assessed in a Beckman LS600IC liquid scintillation counter using Budget solve complete counting mixture (Research Products International Corp.). Activated charcoal (number C4386), prewashed with HCl, was a product of Sigma, and the filtration device (Centricon 100, molecular weight cutoff ϭ 100,000) used to separate F 1 and charcoal-bound [␥-32 P]ATP from 32 P i was obtained from Amicon.
Chemiluminescent Method for Monitoring ATP Hydrolysis-The standard reaction medium, in 1.0 ml final volume at 24°C, consisted of 200 mM sucrose, 50 mM KCl, 10 mM NaP i , pH 7.05, 2 mM MgS0 4 , and 50 l of a 5-ml solution in distilled water of the ATP monitoring reagent. ATP in amounts indicated in the legends to the tables and figures was then added to the stirred reaction medium to elicit the chemiluminescent response, followed by the addition of F 1 to initiate the ATPase reaction. The reverse (adding ATP to suspensions of F 1 ) was also possible without affecting the steady-state kinetics of the reaction. Changes in ATP concentration were monitored with an LKB Wallac model 125 luminometer. The electrical signal elicited by light emission was suitably amplified up to 10,000 times by changes in the current (from 10 Ϫ6 to 10 Ϫ8 A) and the voltage (from 10 V to 1 mV), and recorded using a Soltec model 330 multichannel recorder, usually run at a chart speed of 120 cm/min. The contents of the cell were stirred with a magnetic bar rotating at speeds of near 1000 rpm by means of an electrical device placed on the side of the reaction chamber. The concentration of ATP, in different standard solutions, was spectrophotometrically determined from the absorbance at 259 nm using a millimolar extinction coefficient of 15.4.
As the observed initial rates of decrease of the chemiluminescent signal were always first order with respect to the concentration of ATP (Ϫd[ATP]/dt ϭ k[ATP]), the rate constant "k," was calculated from the integrated form of the equation (ln[ATP] t ϭ ln[ATP] 0 Ϫ kt) by plotting 1n[ATP] t versus t at 400-ms intervals. Only steady-state segments (Ͼ70% of the reaction) were considered in these calculations (Fig. 2).
Radioactive Method for Monitoring ATP Hydrolysis-The entire reaction was carried out in a Centricon-100 filtration device. The standard reaction medium, again in 1-ml final volume at 24°C, was identical to that described above except that the ATP monitoring reagent was omitted and [␥-32 P]ATP (2.1 ϫ 10 6 cpm) rather than "cold" ATP was included in the assay. The reaction was initiated with F 1 , quenched at 2 s by addition of 1 ml of 100 mg/ml activated charcoal, and immediately centrifuged at 3000 rpm (1086 ϫ g) in a SS-34 rotor in a Sorvall RC-2B centrifuge at 4°C. The filtrate containing 32 P i released in the ATPase reaction was then assessed for radioactivity in 10 ml of Budget Solve as indicated above.
Determination of Protein-Protein was determined by the method of Lowry et al. (22) after first precipitating with 5% trichloroacetic acid.

RESULTS AND DISCUSSION
Response Time and Accuracy of the Luciferin-Luciferase Chemiluminescent Assay in Detecting ATP-In kinetic studies described below ATP hydrolysis was monitored by following the disappearance of the chemiluminescent signal induced by adding ATP to an assay medium containing luciferin and luciferase. Prior to commencing these experiments, it was important to define the response time of the system for detecting ATP and to establish whether concentrations of ATP could be accurately detected at ratios of ATP/F 1 in the assay ranging from less than 1 to much greater than 1 (i.e. from unisite to multisite conditions). Confirming previous studies of DeLuca et al. (23,24), the results depicted in Fig. 1A (traces a to c) show that, regardless of the amount of ATP (1-20,000 pmol), the response time of the luciferin-luciferase system is close to 300 ms, much faster than the time course of the ATPase reaction at concentrations of ATP and F 1 used in this study. The accuracy of the technique depicted in Fig. 1B, the data of which were derived from over 70 different experiments, shows that the correlation coefficient between light emission and ATP concentration is better than 0.999 at ATP concentrations of Յ15-17 M ATP. As very low F 1 concentrations are used in the studies described below, the technique allows for accurate ATP measurements to be made with assays in which the ATP/F 1 ratio ranges from ϳ0.1 to ϳ1900. Above 15-17 M ATP, the extent of the chemiluminescent signal rapidly decreases as the concentration of ATP is increased (Fig. 1), due most likely to the accumulation of dehydroluciferin (23). For this reason, and for the reason indicated below, the hydrolytic reaction was initiated by injecting F 1 into the reaction cell already containing ATP.
Time Course of ATP Hydrolysis Catalyzed by either Nucleotide-depleted Bovine Heart or Rat Liver F 1 -Three conditions were adhered to in the performance of all experiments. First, nucleotide-depleted preparations of F 1 (bovine heart or rat liver) were used to avoid any possible allosteric effects resulting from nucleotides bound to noncatalytic sites, i.e. sites located predominantly on ␣-subunits (25). Second, the reaction was initiated by adding F 1 to the reaction mixture already containing ATP to allow the nucleotide free catalytic sites to bind and hydrolyze ATP immediately upon contacting ATP. Third, and in contrast to the earlier studies (13,14), prior incubation of F 1 with ATP ("aging") was avoided both to allow for detection of the actual initial rates of ATP hydrolysis and to avoid product (ADP) inhibition that might give rise to "apparent" cooperative kinetics upon addition of excess ATP. Fig. 2 shows that under the above conditions the initial rates of ATP hydrolysis depend on ATP concentration to the first power as precisely defined by the first order rate equation . This is true regardless of the source of enzyme (bovine heart or rat liver) or whether the reaction takes place under "unisite" (ATP/F 1 ϭ 0.02, Fig. 2A) or "multisite" (ATP/F 1 ϭ 48.5, Fig. 2B) catalytic conditions. Plots (insets in Fig. 2) of the integrated form of the first order rate equation rate constant, k, which in turn was used to calculate the turnover number. For example, for the hydrolysis of 1 pmol of ATP by bovine heart F 1 (45.9 pmol) the first order rate constant, k, obtained from the plot ln[ATP] versus t ( Fig. 2A, inset) was 0.475 s Ϫ1 , and the calculated turnover number (1 pmol of ATP ϫ 0.475 s Ϫ1 /45.9 pmol of F 1 ) was 0.010 s Ϫ1 .
Initial Rates of ATP Hydrolysis by F 1 in the Substrate Concentration Range Spanning 1 nM to 17.1 M ATP-It has been reported that the F 1 moiety of bovine heart F 1 catalyzes the hydrolysis of ATP by a mechanism in which the slow rate of hydrolysis at a single catalytic site is enhanced by ϳ10 6 -fold when the ATP concentration is increased by more than 3 ϫ 10 3 -fold, a process interpreted to result in strong cooperative interactions between catalytic sites (13). The important data leading to this conclusion is actually derived from two separate experiments. The first was conducted under unisite conditions (see Fig. 5 in Ref. 14) in which a turnover number of 3.6 ϫ 10 Ϫ4 s Ϫ1 was determined by monitoring the dissociation of ADP remaining bound to F 1 after prior incubation (aging) of ATP and excess F 1 (ATP/F 1 ϭ 0.2) The second experiment was conducted under multisite conditions (Fig. 1 in Ref. 13) where a turnover number of 300 s Ϫ1 was determined by monitoring release of 32 P i from [␥-32 P]ATP after adding ATP (2.5 ml, 10 mM) in excess to a unisite mixture of F 1 (3 M in 2.5 ml) and [␥-32 P]ATP (0.3 M in 2.5 ml) that had been prior incubated for 2 s. (The multisite turnover number is somewhat higher (600 s Ϫ1 ) when measured by the authors under steady-state conditions in the presence of an ATP regenerating system.) The ϳ10 6 -fold enhancement factor was calculated from the simple multisite/unisite ratio of 300 (or 600 s Ϫ1 )/3.6 ϫ 10 Ϫ4 s Ϫ1 .
In experiments reported here (Tables I and II, Fig. 3), and in contrast to those described above, nucleotide-depleted F 1 preparations (bovine and rat) were employed; a single method (chemiluminescent assay) was used to measure both unisite and multisite turnover rates; the entire time course of the reaction was monitored at each ATP concentration as indicated in Fig. 2; and ATP concentration was varied over a wide range. Table I and Table II summarize kinetic data obtained for typical experiments with bovine and rat liver F 1 , respectively. The first order rate constant for ATP hydrolysis remains relatively constant under both unisite and multisite conditions declining only slightly at high concentrations of ATP. Turnover numbers under unisite conditions range from 0.01 to 0.06 s Ϫ1 at the lowest ATP/F 1 assay ratios, i.e. from 0.11 to 0.55. Fig. 3, A and B, summarize V versus ATP and 1/V versus 1/ATP (Lineweaver-Burk) plots of the data presented in Table I for bovine heart F 1 . Significantly in neither plot, which spans the range from unisite condition, with an ATP/F 1 assay ratio as small as 0.11, to multisite conditions, with an ATP/F 1 ratio as high as 1900, is there any indication of sigmoidicity characteristic of cooperative behavior. Rather, it is clear that bovine heart F 1 assayed under these conditions exhibits strictly Michaelis-Menten kinetic behavior consistent with the view that all participating catalytic sites are kinetically equivalent. The extrapolated single K m is 57 M, and the extrapolated V max is 103 mol of ATP hydrolyzed per min/mg of F 1 . The turnover number of 635 s Ϫ1 is very close to the value of 600 s Ϫ1 obtained under multisite conditions in the earlier study (Ref. 13, see discussion above). Similar results were obtained with rat liver F 1 (Table II, Fig.  3B), which exhibits a V max nearly identical to bovine heart F 1 and a slightly higher K m of 79 M.
Comparison of the Chemiluminescent Assay for Monitoring ATP Hydrolysis with an Assay That Monitors Release of 32 P i from [␥-32 P]ATP-Although the turnover number of 635 s Ϫ1 for multisite catalytic conditions reported here for the bovine heart F 1 is nearly identical to that reported in the earlier study (13), the unisite turnover numbers at the lowest ATP/F 1 ratios are much higher, in the range of 0.01-0.06 s Ϫ1 (Tables I and II), rather than near 10 Ϫ4 s Ϫ1 (13, 14). Consequently, data obtained here result in multisite/unisite ratios near 10 4 rather than near ATP concentration Reactions were initiated by injecting 9.16 pmol of nucleotide-depleted bovine heart F 1 into a 1 ml chemiluminescent reaction system containing the indicated amount of ATP. Assays were carried out exactly as described under "Experimental Procedures." The first order rate constant (k) of the reaction was calculated from the slope of the line that results from plotting the integrated form of the first order rate equation: ϪdATP/dt ϭ k [ 2. Time course of the ATPase reaction catalyzed by F 1 under unisite and multisite catalytic conditions. Assay conditions are exactly as described under "Experimental Procedures." In A., the reaction was initiated by adding 45.9 pmol of bovine heart F 1 to a medium containing only 1 pmol of ATP. In B, the reaction was initiated by adding 20.6 pmol of rat liver F 1 to a medium containing 1000 pmol of ATP. The first order rate constant, k, was calculated from the slope of the line that results from plotting the integrated form of the first order rate equation at 0.4-s intervals (see "Experimental Procedures"). ATP concentration ϭ light intensity in arbitrary units.

6 (13) as reported earlier.
For this reason, it might be argued that the chemiluminescent assay, which monitors ATP disappearance, may not accurately report ATP hydrolytic rates at low ATP/F 1 assay ratios (unisite conditions). To address this question, we compared the chemiluminescent assay for monitoring ATP hydrolysis under unisite conditions with that of an assay that monitors 32 P i release from [␥-32 P]ATP. Two experimental conditions were chosen, one in which the ATP/F 1 assay ratio was only 0.11 (lowest data point in Table I), and one in which the ratio was 1.1. The assay medium was identical to that used for monitoring ATP disappearance except that luciferin and luciferase were omitted. Significantly, the specific activities of 2.9 Ϯ 0.21 and 14.5 Ϯ 2.9 nmol of P i released per min/mg of protein obtained using the 32 P i release assay (Table  III) compare favorably with those of 1.8 and 18 nmol of ATP disappeared per min/mg of protein obtained using the chemiluminescent assay (Table I), for ATP/F 1 ratios, respectively, of 0.11 and 1.1. Therefore, it seems clear that the chemiluminescent assay is a reliable indicator of the ATP hydrolytic rate under unisite as well as multisite conditions. Summary and Mechanistic Implications-In studies described here, the bovine heart F 1 preparation was essentially nucleotide-free, and the rat liver F 1 preparation contained less than 1 mol/mol of F 1 . Both enzymes exhibited strictly Michaelis-Menten kinetic behavior with maximal turnover numbers of 635 s Ϫ1 and multisite/unisite rate enhancement ratios near 10 4 , consistent with the view that F 1 -ATPases can catalyze ATP hydrolysis by a mechanism in which all participating catalytic sites are kinetically equivalent. In earlier studies (13,14), the bovine heart F 1 preparations used contained 2.8 -3.5 mol of bound nucleotide/mol of F 1 . These preparations exhibited maximal turnover numbers of 600 s Ϫ1 and multisite/unisite rate enhancement ratios of 10 6 , consistent with the view that F 1 -ATPases can catalyze ATP hydrolysis by a mechanism in which positive cooperativity occurs between catalytic sites. Perhaps the simplest interpretations of the two studies is that, depending on nucleotide content and its subunit distribution, F 1 ATPases can exist in different conformational states, one in which all participating catalytic sites are kinetically equivalent and one in which they are kinetically nonequivalent. In support of this view are two different x-ray structures of F 1 (25,26), one in which the catalytic ␤-subunits appear more structurally equivalent (25) than in the other (26).
Alternative interpretations are possible and require further investigation. One possibility is that F 1 -ATPases normally function during ATP hydrolysis as simple Michaelis-Menten enzymes and that deviations from this behavior (i.e. sigmoid kinetic behavior), rather than reflecting positive catalytic co-  Table I for bovine heart F 1 . See legend to Table I for details of the assay. Note that the first 7 of the 14 data points are plotted on an expanded scale in the inset. Values for V max and K m were obtained from a Lineweaver-Burk plot shown in B. B, Lineweaver-Burk plots of the data tabulated in Table I for bovine heart F 1 (A) and in Table II for Rat Liver F 1 ( B). See legends to Tables I and II   Reaction conditions were identical to those used in the chemiluminescent assay (see "Experimental Procedures") except that the luciferin-luciferase reagent was omitted and [␥-32 P]ATP rather than "cold" ATP was present. Upon addition of 9.16 pmol of nucleotide-depleted bovine heart F 1 to the assay medium contained within a Centricon-100 filtration device, the reaction was allowed to proceed for 2 s followed by immediate quenching with activated charcoal. Separation of 32 P i from charcoal bound [ 32 P]ATP was performed by centrifugation (see "Experimental Procedures"). Specific activity values are reported as mean Ϯ the standard deviation.  operativity, simply reflect inhibitory ADP bound at a catalytic site that is displaced as ATP concentration is raised. As shown in Fig. 4, ADP is a potent inhibitor of bovine heart F 1 with over 90% inhibition being observed when the ADP concentration equals the ATP concentration at values Ն5 M. Significantly, F 1 preparations are frequently stored or isolated in ATP, the hydrolysis of which results in F 1 ⅐ADP complex formation. Specifically, as it applies to the earlier studies (13,14) leading to the view that bovine heart F 1 exhibits positive catalytic cooperativity, it is interesting to note that the unisite turnover number of 10 Ϫ4 s Ϫ1 was based neither on the release of P i nor ADP following ATP hydrolysis per se, but rather on the dissociation of ADP remaining bound to F 1 following hydrolysis (Fig.  5 in Ref. 14). Thus, the possibility exists that the unisite turnover number of 10 Ϫ4 s Ϫ1 obtained from this earlier analysis is not an accurate reflection of the actual value and corresponds to release of ADP from a noncatalytic site. It is important to note that the same investigators in a separate study ( Fig. 1 in Ref. 13) find that incubation of [␥-32 P]ATP (0.3 M in 2.5 ml) under unisite conditions with F 1 (3 M in 2.5 ml) for 2 s prior to adding excess ATP results in 20% hydrolysis of the total ATP present. Had these investigators calculated the unisite turnover number under these conditions, they would have obtained a value of 0.01 s Ϫ1 (i.e. 0.15 M ATP ϫ 0.20/1.5 M F 1 ϫ 2 s). This value is almost identical to the unisite turnover number of 0.011 s Ϫ1 reported here as the first entry in Table I and consistent with Michaelis-Menten rather than cooperative kinetic behavior.
Finally, consistent with the studies reported here, recent work on nucleotide-depleted Escherichia coli F 1 (27) show that in the ATP concentration range of 1 M to 1 mM a single K m value of 38 M is sufficient for an adequate description of the ATP hydrolytic behavior of the enzyme. However, these authors believe that all catalytic sites (presumably three) must be occupied to achieve significant rates of ATP hydrolysis. Clearly, this is not the case with the nucleotide-depleted F 1 preparations from the animal systems examined here, as significant rates of ATP hydrolysis are readily detected at ATP/F 1 assay ratios as low as 0.02 ( Fig. 2A). Moreover, at 17 M ATP, well below the K m of 57 M, the catalytic turnover number is already 149 s Ϫ1 (ϳ24 mol of ATP hydrolyzed per min/mg of F 1 ), as shown in Table I. Thus, nucleotide-depleted animal F 1 preparations, in accordance with simple Michaelis-Menten kinetic behavior, show significant rates of ATP hydrolysis at all concentrations of ATP tested. FIG. 4. Inhibition of ATP hydrolysis catalyzed by bovine heart F 1 . The chemiluminescent assay conditions are exactly as described under "Experimental Procedures" except that the reaction was initiated by adding ATP to a medium containing 9.16 pmol of F 1 and an amount of ADP equal to that of the added ATP.