Characterization and Localization of Mitochondrial Uncoupler Binding Sites with an Uncoupler Capable of Photoaffinity Labeling

Abstract 2-Azido-4-nitrophenol (NPA) is a potent, water-soluble uncoupler of oxidative phosphorylation. It is photochemically active and capable of covalent labeling of the mitochondrial components presumably concerned with the uncoupling process. Acrylamide gel electrophoresis of NPA-labeled mitochondria digested with sodium dodecyl sulfate and mercaptoethanol showed that 40% of NPA radioactivity was associated with protein bands in the molecular weight region of 20,000 to 30,000. Equilibrium binding studies with NPA, under nonphotolytic conditions, have shown that bovine heart mitochondria contain 0.56 ± 0.13 nmole of a uniformly reacting uncoupler binding site per mg of protein. The binding of NPA to these sites is competitively inhibited by other uncouplers, but not by antimycin A, rutamycin, valinomycin, or arsenate. The uncouplers tested against NPA were sodium azide, 2,4-dinitrophenol, pentachlorophenol, carbonylcyanide m-chlorophenylhydrazone, and 5-chloro-3-t-butyl-2'-chloro-4'-nitrosalicylanilide. The uncoupler binding sites are located in the inner membrane, and do not appear to involve F1 (ATPase). They are unaffected with respect to NPA binding by changes in the respiratory state or the energization state of mitochondria. Data regarding dissociation constants and the uncoupler potencies of NPA, 2,4-dinitrophenol, and azide have suggested that the uncoupler binding sites of mitochondria are functionally involved in the act of uncoupling.

drial components involved in this process. In addition to pursuing the other two approaches, we have embarked recently on the study of energy uncoupling in mitochondria with the novel technique of photoaffinity labeling (6). For this purpose, a new, photochemically active uncoupler, 2-azido-4-nitrophenol, was synthesized in radioactive form. When irradiated with visible or ultraviolet light, aromatic azides lose a molecule of nitrogen and form singlet nitrenes (7) which are highly reactive and capable of insertion into a neighboring molecule (Equation 1) (Reference 6, see also Reference 8).

R-N, hv R-a R'--H R-NH-E (1)
NPAi is a structural analogue of the classical uncoupler, 2,4dinitrophenol. It is also 2 to 3 times more potent than DNP. This fortunate circumstance, together with the fact that NPA is water soIuble, has brought together in this compound properties that have allowed two important, types of study to be performed regarding the uncoupler binding sites of mitochondria. These studies are (a) photoaffinity labeling of the mitochondrial uncoupler binding sites, and (b) equilibrium binding studies with NPA under nonphotolytic conditions. In addition to being crucial for photoaffinity labeling experiments, equilibrium binding studies have shown that mitochondria contain a finite number of uncoupler binding sites, which could be saturated with NPA and their concentration in mitochondria and submitochondrial particles estimated. These binding sites interact with NPA uniformly, thus indicating the same, or the same type of, uncoupler binding site for the three energy coupling sites of mitochondria. Competitive binding studies have shown that the NPA binding site interacts with other uncouplers, such as DNP, pentachlorophenol, carbonylcyanide m-chlorophenylhydrazone, 5-chloro-3-l-butyl-2'-chloro-4'-nitrosalilide, and aside. Furthermore, the same site was shown to interact covalently with NPA after photoactivation. METHODS

AND MATERIALS
Heavy bovine heart mitochondria and various types of submitochondrial particles were prepared according to published procedures (9-l 1). Protein was estimated by the biuret method 1 The abbreviations used are: NPA, 2-azido4-nitrophenol; DNP,2, PCP, pentachlorophenol; Cl-CCP, car- (12). Tritiated NPA was synthesized as described elsewhere.2 In a typical equilibrium binding experiment, 25 to 50 mg of mitochondrial protein suspended in 10 ml of 0.25 M sucrose, containing 10 mM Tris-HCl, pH 7.8, 1 rnM ATP, 1 mM MgC12, and 1 mM succinate, were incubated with NPA at varying concentrations for 5 min at 3" f 0.6". It has been ascertained that the duration of incubation was more than adequate for complete equilibration under all conditions reported in this study. The mixture was centrifuged for 10 min at 17,000 rpm in a number 40 rotor of a Spinco model L ultracentrifuge. The supernatants were saved for NPA determination, and the pellets were digested with NaOH, H202, and Soluene 100, and their radioactivity was counted in a Beckman LS-250 scintillation counter.
Photochemical experiments, designed to induce covalent binding of NPA to mitochondria, were conducted in the same buffer system as above. About 60 mg heavy bovine heart mitochondria in 10 ml of buffer containing NPA (11 p~ in Fig. 6; 4.4 pM in Fig. 7) were placed in a loo-ml round bottom flask. The flask was immersed in a 5% CuSOl solution as light filter, and revolved with a rotation evaporator at about 250 rpm. A stream of argon was continuously blown into the flask at a rate of about 500 ml per min. The temperature of the bath was held at 4" using an immersion cooler. The light source (650 W, DVY 3400" K tungsten halogen projector lamp) was placed within about 8 inches of the middle of the flask, resulting in a calculated flux density of approximately 1 watt per cm2. During irradiation (1 to 3 min) the temperature rose to 6-7" and was allowed to fall to 4" before the next batch of mitochondria was treated. The irradiated mitochondria were freed from noncovalently bound NPA and other additions by washing three times with the buffer, once more with the buffer containing 10 mg per ml of bovine serum albumin, and once again with the buffer. In the absence of irradiation, it has been shown that this procedure removes the added NPA completely. Measurements of radioactivity corresponding to covalently bound NPA were conducted as described above. Binding data were analyzed by a least squares procedure for three variables (Ub, Ub/U,, and Vi) according to  ub/uf + 'Vl'u, where U is uncoupler concentration; KD, dissociation constant; and m, slope of the low affinity binding curve. Subscripts b, j, and max denote bound, free, and maximal, respectively. This equation is an expanded form of the Scatchard equation (13) in which the slope, m, is not equal to 0.
Soluene 100 was purchased from Packard Instrument Co., Protosol from New England Nuclear, pentachlorophenol from J. T. Baker, Cl-CCP from Calbiochem, and bovine serum albumin (four times recrystallized) from Nutritional Biochemical Corporation. Rutamycin, antimycin A, and S-13 were gifts, respectively, from Eli Lilly and Co., Kanegafuchi Chemical Industrial Co. of Japan, and Monsanto Commercial Products Co.

Characteristics of .%-A&o-Q-nitrophenol
Binding to Mitochon-d&-The binding of NPA to mitochondria is reversible and occurs in two distinct phases (Fig. 1A). At low uncoupler concentrations, there is a relatively large uptake of NPA. This high affinity phase is followed by a weaker binding, which is a linear function of NPA concentration. Whereas the high affinity phase was saturated at low levels of NPA (see below), the low affinity phase showed no such tendency even up to 600 pM NPA (Fig. 2). Thus, it appeared to have the characteristics of either very weak, but extensive, binding ( > 10 nmoles per mg of protein), or a partition equilibrium between the medium and the mitochondrial phases (the results shown in Figs. 1 and 2 might be a combination of both).a By subtracting this low affinity binding from the total binding of NPA to mitochondria, a hyperbolic saturation curve is obtained (Fig. lB), which is characteristic of the specific binding of NPA to mitochondria.
Maximal specific binding (i.e. the concentration of NPA binding sites), as determined from double reciprocal plots of a number of experimental curves similar to B of Fig. 1, is 0.56 f 0.13 nmole of NPA per mg of mitochondrial protein at 3" and pH 7.0. This value is of the same order of magnitude as the concentration of Fi molecules, estimated from electron microscopic (14) and aurovertin binding data (15), or the concentration of electron carriers in the vicinity of the three coupling sites (e.g., FMN + cl + as) (16,17). A Hill plot of the specific binding data is shown in Fig. 3. The slope of the straight line is 1.03 in a set of data extending from 157n to 89 '% saturation of the binding sites. This indicates that essentially all the specific binding sites have nearly equal aflinity for NPA, and bind NPA without appreciable cooperativity (18). The dissociation constant (KD) is 6 & 3 PM at 3', and is pH independent between pH 7 and pH 9.4. The lack of an appreciable pH effect on KD together with our thermodynamic studies (AH = -8 & 1 Cal, A unitary entropy (AS,) = 2 f 6 e.u.) suggest that NPA binds as the anion, and that its binding neither generates increased hydrophobic interactions nor involves major conformational changes of the membrane. For best preservation of mitochondria, the binding studies described above were conducted in a medium which contained both succinate and ATP. However, as shown in Fig  binding.
Therefore, it may be concluded that NPA binding by mitochondria is an equilibrium process and not due to active accumulation driven by suceinate oxidation or ATP hydrolysis. Equilibrium binding studies such as shown above are not possible with water-insoluble uncouplers (e.g. Cl-CCP, S-13), because they are preferentially accumulated in mitochondria. Such studies are also very difficult with DNP, which is water-soluble and structurally very similar to NPA.
The reason appears to be related to the fact that, as an uncoupler, DNP is 2 to 3 times less potent than NPA, and its dissociation constant is larger (see below).
Thus, when a binding curve such as A of Fig. 1 was plotted for DNP, the slope of the initial, high affinity phase was not sufficiently different from the slope of the low-affinity phase to permit accurate separation and analysis of the two phases. In A of Fig. 1, the initial slope of the total NPA binding curve is nearly 5 times the slope at higher NPA concentrations.
A similar experiment with radioactive DNP showed a slope difference of less than a-fold.
Specificity of Mitochondrial Uncoupler Binding Sites-The specificity of the NPA binding site for uncouplers is illustrated in Tables I and II Table I lists the effects of several uncouplers and inhibitors on the over-all binding of NPA to mitochondria.
It is seen that uncouplers such as DNP, PCP, Cl-CCP, and aside inhibit the binding of NPA, whereas rutamycin and antimycin A, which are specific inhibitors of phosphorylation and electron transport respectively, have no effect on NPA binding.
Table I also shows that the ionophore valinomycin, which is another modifier of membrane function, did not alter the extent of NPA binding to mitochondria.
The finding that arsenate does not inhibit NPA binding (Table I) is particularly interesting in view of the fact that arsenate uncouples oxidative phosphorylation, but apparently between the sites of oligomycin inhibition and ATP synthesis (19)(20)(21)(22) rather than between the oligomycin inhibition site and the respiratory chain. Therefore, the ineffectiveness of arsenate suggested that inhibition of NPA binding by other uncouplers might involve site specificity.
An example of such competitive inhibition is shown in Fig. 5. In these experiments, the uncoupler, S-13, which has been considered to act stoichiometritally (23)(24)(25), was deliberately used at a concentration (0.4 nmole per mg of protein) which is substoichiometric for total uncoupling.
In addition, the considerable competitive inhibition of NPA binding by such small amounts of S-13 suggests that the mitochondrial site concerned with NPA binding is also functionally involved in the act of uncoupling.
This point is further supported by the data of Table II, which allow a comparison of NPA with DNP in terms of their Ko, @I/Z (concentration of uncoupler necessary for 50% uncoupling), and &, i values. The latter is the DNP dissociation constant determined from the competitive inhibition of NPA binding by DNP. Thus, it is seen that Kn and +1/z values are essentially the same for eech uncoupler, and that both KD and @l/2 for DNP are larger than those for NPA by a factor of 2 to 3. In addition, the similarity of the KD and Kc, i values for DNP further indicate that this compound competes with NPA for the same binding site, and excludes the possibility of an allosteric effect of DNP on the affinity of mitochondria for NPA. Location of Uncoupler Binding Sites-The NPA binding site is present in mitochondria, and in phosphorylating submitochondrial particles. It is also present to the same extent in submitochondrial particles, which are deficient in Fi (4), incapable of phosphorylation or ATP utilization, but still competent. in the presence of oligomycin, for transhydrogenation driven by energy derived from substrate oxidation (II). These results indicate that the uncoupler binding site is located in the inner membrane but apparently not in Fi. This is indeed to be expected, because uncoupling is generally believed to occur between the electron transport system and the site of oligomycin inhibition (4).
The NPA binding site appears not to involve the bulk of the mitochondrial lipids, as prior removal of more than 80% of mitochondrial lipids with aqueous acetone (26) did not change the extent of NPA binding. Preparations of the electron transfer complexes (27) did not bind NPA. However, these negative results might be misleading because preparations of the complexes contain deoxycholate, which (a) uncouples oxidative phosphorylation, and (b) might inhibit NPA binding as a result of its unspecific interaction with the mitochondrial hydrophobic regions.
Covalent Labeling of Uncoupler Binding Sites with .%A.zido+ nitrophenol-Three important questions with regard to covalent labeling of the mitochondrial uncoupler binding sites with NPA are as follows. (a) Does NPA bind to mitochondria covalently when irradiated? (b) Is the labeling of mitochondrial components specific or diffuse? (c) Does covalent labeling occur at the uncoupler binding sites?
With regard to a and b, our preliminary results have shown that under appropriate conditions, mitochondria could be labeled with radioactive NPA to the extent of 0.5 nmole per mg of protein. Furthermore, acrylamide gel electrophoresis of NPAlabeled mitochondria in the presence of sodium dodecyl sulfate and mercaptoethanol showed that more than 40% of radioactivity was associated with a protein band in the region of molecular weight 20,000 to 30,000. These results are shown in Fig.  6 (see the region of slices 10 to 15). Further work on the isolation and characterization of the NPA-labeled material is currently in progress.
With regard to c, covalent binding experiments were carried out in the presence of several concentrations of DNP, which was shown to be a competitive inhibitor of NPA binding under nonphotolytic conditions. Thus, NPA, mitochondria, and increasing concentrations of DNP were irradiated for 3 min and the particles were analyzed for covalently bound uncoupler. Duplicate experiments were carried out in the dark, and the extent of equilibrium binding of NPA was determined. The results are presented in Fig. 7 as a normalized Dixon-type plot. It is seen that DNP inhibits competitively both the equilibrium and the covalent binding of NPA, thus indicating that the same un-at 5 10 15 20 25   coupler binding site as characterized above has been labeled under photolytic conditions. In Fig. 7, the slope of each line is a measure of the apparent affinity of DNP for the binding site, in competition with NPA. An analysis of the data showed that DNP is 50 to 60% less effective as a competitor for NPA under photolabeling conditions than under equilibrium binding conditions. This is to be expected, since DNP is not able to compete with covalently bound NPA. We have also tested the capacity of mitochondria covalently labeled with cold NPA to bind tritiated NPA under equilibrium binding conditions. Mitochondria were first irradiated in the presence of 20 PM cold NPA for 4 min. They were then washed free of unreacted NPA and photodissociated products, and subjected to equilibrium binding of radioactive NPA as described above. A similar control experiment was carried out with a portion of the same preparation of mitochondria involving all the treatments and washing steps, except that the cold NPA was added after irradiation of mitochondria. The control mitochondria, which did not involve covalently bound, cold NPA, showed a specific uptake of radioactive NPA of 0.34 nmole per mg of protein, which is in agreement with previous results (29) obtained at the pH 7.8 used for this experiment.
By comparison, the mitochondria, which were prelabeled with cold NPA, showed only 50% as much capacity for the specific binding of tritiated NPA.
These results further indicate that the same finite number of binding sites appear to be involved in equilibrium and covalent binding of NPA to mitochondria.

DISCUSSION
Stoichiomety of Uncoupler Binding-The chemical hypothesis of oxidative phosphorylation assumes the presence of a high energy intermediate with which uncouplers interact. Therefore, the stoichiometry of uncoupling under various conditions has been a subject of interest in several laboratories (23-25, 31, 32). Potent, water insoluble uncouplers, such as carbonylcyanide phenylhydrazones and S-13, have been used for this purpose, and functional parameters, such as loss of phosphorylation activity, loss of respiratory control, stimulation of ATPase activity, and release of the inhibition of respiration by azide, have been studied as a function of uncoupler concentration. Margolis et al. (31) showed that the 4112 of carbonylcyanide m-chlorophenylhydrazone for uncoupling oxidative phosphorylation was about 5 x lo-l1 moles per mg of protein of bovine heart mitochondria,4 and calculated that for 100% uncoupling it was sufficient to add 1 uncoupler molecule per 9 respiratory chains, or 27 coupling sites. These authors also showed that the effective concentration of uncouplers was directly related to the rate of respiration and the number of coupling sites involved. Similar relationships were found by Tsou and van Dam (34) for DNP, dicoumarol, and tetrachlorotrifluoromethylbenzimidazole, and by Kaplay et al. (25) for S-13. The latter group calculated that in rat liver mitochondria, at comparable oxidation rates, the number of moles of S-13 per mole of cytochrome a needed for complete uncoupling were 0.39, 0.81, and 1.1 when the substrates were, respectively, ascorbate plus tetramethyl-p-phenylenediamine, succinate, and glutamate-malate.
Using functional parameters other than P : 0 values, Wilson (23) showed that in rat liver mjtochondria the amounts of S-13 needed per respiratory chain for complete effect were 0.6, 1.0, and 1.35 molecules when the assays were, respectively, the loss of respiratory control, the stimulation of ATPase, and the release of azide-inhibited respiration. He has concluded, in contrast to the above authors, that the S-13 titer is independent of the rates of electron transfer and generation of high energy intermediates.
Two important points arise from the above stoichiometries and the effects of electron flux and number of coupling sites involved. (a) There was agreement in findings when the assay used for uncoupling was the same. This is the case with regard to the results of Margolis et al. (31) and Kaplay et al. (25), even though the former used bovine heart mitochondria and the uncouplers Cl-CCP and carbonylcyanide p-trifluoromethoxyphenylhydrazone, and the latter used rat liver mitochondria and S-13. (b) There was disagreement in findings when the assays were not the * Muraoka and Terada (33) have shown with rat liver mitochondria that the uncoupler 3,5-di-tert-butyl-4-hydroxybenzylidene malononitrile also shows maximal effects at 2 to 4 X IO-" moles per mg of protein.
same. This involves the results of Wilson (23) and Kaplay et al. (25), even though the source of mitochondria and the nature of the uncoupler were the same. In a, the parameter studied as a function of uncoupler concentration was P:O ratio. In this assay, the effectiveness of a given amount of uncoupler might be expected to depend on the steady state concentration of the high energy intermediate, which in turn would depend on both the rate of electron flux and the number of coupling sites involved. In b, the assays used by Wilson (23) were not a direct measure of the loss of P:O ratio, and a correlation such as above was not observed. This is probably related to the kinetics of the different reactions measured by Wilson. The above data clearly show that each particular assay condition yields a different uncoupling titer. This is to be expected, because in general kinetic studies yield a functional titer, not a binding site titer.
Consider, for example, the enzyme-catalyzed sequence A --t B --) C in which the rate of C production is being measured, but the first reaction is rate limiting and can be stimulated by uncouplers.
In such a system, study of the kinetics of C production as a function of increasing uncoupler concentration does not necessarily yield information regarding the titer of the uncoupler with respect to the first reaction. This is because long before saturation of the first reaction with respect to uncoupler is achieved, the second reaction might become rate limiting.
Consequently, the result will not be a thermodynamic saturation titer, but a kinetic effectiveness titer, which could vary with the enzymatic reactions under consideration.
Because of these complications, it is important to distinguish clearly between the stoichiometry of uncoupling as studied by kinetic methods and the stoichiometry of uncoupler molecules bound per respiratory chain or coupling site.
In our studies, the concentration of uncoupler binding sites was determined from equilibrium binding experiments. Thus the results showed a constant saturation level and an invariable KD regardless of whether the mitochondria were in the oxidized or reduced state, energized by addition of ATP (Fig. 4), deenergized by addition of arsenate (Table I), inhibited at the level of the respiratory chain with antimycin A, or inhibited at the level of high energy transfer reactions with rutamycin. The latter two conclusions also arise from the data of Table I in  which antimycin A and rutamycin neither decreased nor increased the extent of NPA binding.
The concentration of uncoupler binding sites found in bovine heart mitochondria by the equilibrium binding method is 0.56 f 0.13 nmole per mg of protein.
In these mitochondria, the concentration of electron carriers ranges from 0.15 to 0.7 nmole per mg of protein (except for coenzyme Q, which may be a mobile carrier), and the concentration of F1 molecules is approximately 0.2 nmole per mg of protein (calculated from the data of Reference 14). These comparable stoichiometries between electron carriers, phosphorylation enzymes, and the components at or near the coupling sites suggest that energy conservation, energy coupling and dissipation by uncouplers, and high energy transfer are all related molecular interactions rather than electrochemical or conformational effects transferred from one bulk phase to another.
For the latter hypotheses, a stoichiometric relationship of the component of these three systems would not be particularly necessary.
E$ect of Uncoupler Binding on Membrane Structure-It has been shown that uncouplers change the gross structure of mitochondrial cristae in the energized state (35-37), and reverse the ATP-induced fluorescence enhancement of bound 8-anilino-lnaphthalenesulfonic acid (38). These results have suggested that uncoupling of energized mitochondria is associated with the conformation changes of cristae detectable both by electron microscopy and by a probe for membrane hydrophobic-hydrophilic interphases.
Our data (small AS,; invariable K, in the absence and presence of succinate plus ATP, antimycin A, rutamycin) indicate, however, that uncoupler binding (as distinct from uncoupling) causes neither a significant net conformation change (i.e. in terms of ordering and disordering) of the membranes, nor a specific conformation change of the uncoupler binding sites. These conclusions are in agreement with the results of Zimmer et al. (39) obtained with a spin label.

Specificity of Uncoupler Binding
Sites-The diversity of structure and size of the uncouplers used in this work (S-13, Cl-CCP, PCP, DNP, and azide) is difficult to reconcile on traditional "active site" grounds with the fact that all these uncouplers are competitors of NPA binding, and therefore appear to bind to the same site. Wilson et al. (24) have suggested a general acid-base catalysis for the mechanism of action of uncouplers, in which certain uncouplers are considered to be active in protonated (acidic) form and others in anionic (basic) form. These differences also appeared to have no effect on the recognition of uncouplers by the binding sites described above, since both the "acidic" uncoupler, Cl-CCP, and the "basic" uncoupler, S-13, were competitive inhibitors of NPA binding.
The one common feature in the above uncouplers is that in all of them the negative charge is delocalized in a 7r electron system. This common characteristic may have functional significance. but can hardly account for a conventional site specificity. Therefore, we feel that our competitive binding data suggest a unique structure for the uncoupler binding sites of mitochondria.
The isolation and characterization of the mitochondrial component covalently labeled with NPA should provide considerable insight into this puzzle.

Nature of Uncoupler Binding
Component-It has been shown that the mitochondrial uncoupler binding sites are located in the inner membrane and do not appear to involve F1 and the bulk of mitochondrial lipids.
The preliminary results of the covalent binding studies suggest protein involvement in the range of molecular weight 20,000 to 30,000. This provisional molecular weight range excludes a number of electron carriers (e.g. DPNH dehydrogenase, cytochrome cl, and cytochrome c) as possible candidates for uncoupler binding.
The ineffectiveness of rutamycin and antimycin A on NPA binding also excludes oligomytin sensitivity-conferring protein (40) and the antimycin-binding site of complex III.
Since S-13 releases the inhibition of respiration by aside, Wilson (23) has suggested that the site of action of S-13 is associated with cytochrome oxidase. The subunits of bovine heart cytochrome oxidase (41) do not fit well i&the molecular weight range of 20,000 to 30,000. However, the suggestion of Wilson is now amenable to direct test, since Ragan and Racker (42) have reconstituted site I oxidative phosphorylation from complex I and coupling factors, none of which contains cytochrome oxidase. This reconstituted system was shown to be sensitive to the uncoupler, dihexafluoroacetone-acetone.
An important question with regard to the uncoupler binding sites described above is whether these sites are also involved in the act of uncoupling.
We have no rigorous proof for this, but three pieces of information favor this possibility: (a) KD values for NPA, DNP (Table II), and azide (not, shown) are very close to their respective 41/Z values, (b) the specific binding data as analyzed by a Hill plot (Fig. 3) indicate only one type of binding (i.e. in terms of affinity) from 15y0 to 89% saturation, and (c) all the uncouplers tested to date were competitive inhibitors of NPA binding.
Mechanism of Uncoupling-As stated earlier, the chemical hypothesis of oxidative phosphorylation assumes the presence of a high energy intermediate (X*)5 with which uncoupler + interacts and dissipates its energy. Thus the interaction of the uncoupler with X* may be written as x*+o~x*-@-+x+~ (3) Our results on NPA binding and competitive inhibition of binding by other uncouplers indicate that uncouplers must also be capable of reversible interaction with the de-energized form of X as shown in Equation 4 x*+*2 x*-~+x--9~x+~ (4) This is because the concentration of X*, and therefore of X* -@', would be expected to vary considerably in the presence and absence of ATP plus succinate, antimycin A, oligomycin, or arsenate, whereas the concentration of X can be independent of these conditions.
In these considerations, we do not imply that @ interacts with the high energy intermediate site on X. Uncouplers might interact with an allosteric site on X or X*, but have a destabilizing effect, on X*. An important consideration with regard to the mechanism of uncoupling is the properties of mitochondria containing covalently bound NPA.
This problem is currently under investigation.
Our preliminary results seem to suggest that mitochondria partially labeled with NPA show essentially no change in P : 0 values and state 4 rates with succinate or 3-hydroxybutyrate as substrate. However, state 3 rate and oligomycin-sensitive ATPase activity were both decreased to approximately the same extent.
The decreased state 3 rate could not, be increased by the addition of uncouplers to the reaction mixture. These effects could be due either to damage or specific labeling of mitochondria.
The latter possibility is compatible with a mechanism in which a component required for oxidative phosphorylation is prevented from turning over by photoaffinity labeling.