Functional transitions of F 0 F 1 -ATPase mediated by the inhibitory peptide IF1 in yeast coupled submitochondrial particles Mikhail Galkin 1, *, Rene ´ e Venard 1 , Jacques Vaillier 2 , Jean Velours 2 and Francis Haraux 1 1 Service de Bioe ´ nerge ´ tique & CNRS-URA 2096, Gif-sur-Yvette, France; 2 Institut de Biochimie et Ge ´ ne ´ tique Cellulaires du CNRS, Bordeaux, France The mechanism of inhibition of yeast F 0 F 1 -ATPase by its naturally occurring protein inhibitor (IF1) was investi- gated in submitochondrial particles by studying the IF1- mediated ATPase inhibition in the presence and absence of a protonmotive force. In the presence of protonmotive force, IF1 added during net NTP hydrolysis almost completely inhibited NTPase activity. At moderate IF1 concentration, subsequent uncoupler addition unexpect- edly caused a burst of NTP hydrolysis. We propose that the protonmotive force induces the conversion of IF1- inhibited F 0 F 1 -ATPase into a new form having a lower affinity for IF1. This form remains inactive for ATP hydrolysis after IF1 release. Uncoupling simultaneously releases ATP hydrolysis and converts the latent form of IF1-free F 0 F 1 -ATPase back to the active form. The rela- tionship between the different steps of the catalytic cycle, the mechanism of inhibition by IF1 and the interconver- sion process is discussed. Keywords: ATP synthase; catalytic state; inhibitory peptide; latent ATPase; protonmotive force; yeast. Energy-driven ATP synthesis in energy-transducing mem- branes is carried out by membrane-bound F 0 F 1 -ATPase complex or ATP synthase [1]. The extrinsic F 1 subcomplex is composed of five types of subunits in the stoichiometry a 3 b 3 cde. Three fast-exchangeable nucleotide-binding sites, presumably catalytic, and three slow-exchangeable nucleo- tide-binding sites, certainly noncatalytic, reside in a/b interfaces [2,3] (an alternative point of view can be found in [4]). The membranous F 0 subcomplex promotes proton translocation, energetically coupled to ATP synthesis/ hydrolysis via a stalk connecting F 0 and F 1 .F 1 can be biochemically separated from F 0 and remains competent for uncoupled ATP hydrolysis. Numerous data, including direct observations of the rotation of the central axis of F 0 F 1 relative to a 3 b 3 crown during ATP hydrolysis [5,6], indicate that F 0 F 1 is a rotary motor, with an asymmetrical rotor composed, in mitochondria, of c, d and e subunits, anchored to a membranous decameric ring of c subunits thought to compose a proton-driven turbine [7]. Mitochondrial F 0 F 1 is regulated by a naturally occurring inhibitor protein called IF1 [8]. IF1 is a small acid- and heat-stable protein, which stoichiometrically binds to the F 1 sector of ATP synthase and inhibits ATP hydrolysis [8–19]. D ~ l H þ is believed to favour the release of IF1 from ATP synthase [10–17] or, alternatively, to shift this regulatory peptide from an inhibitory to a silent position on the enzyme [10,18,19]. The IF1 binding site was proposed to be located close to the C-terminus proximal DELSEED loop of the b-subunit [20], close to the catalytic site [21], at the a/b interface [22], or at the a/c interface [23]. A recent tridimensional model of the bovine MF 1 –IF1 complex drawn from radiocristallographic data showed IF1 bound at the a/b interface, and interacting weakly with c [24]. The F 0 F 1 –IF1 interaction is modulated by a number of factors such as pH, ionic strength, D ~ l H þ and nucleotides [8–16,25–31]. Interaction of F 0 F 1 –ATPase with adenine nucleotides is very complex in itself. In addition to simple competitive inhibition of ATP hydrolysis, ADP causes hysteretic inhibition of ATPase activity in contrast to nonadenylic nucleotides IDP and GDP [32–37]. Likewise, MgADP-inhibited enzyme can undergo an energy-depend- ent transformation changing its functional properties [32–34]. For all these reasons, it is very difficult to discriminate the roles of D ~ l H þ , nucleotide occupancy and enzyme turnover in the IF1-related regulatory processes. However, it is well established that ATPase turnover is necessary for IF1 binding [12], and a complex relationship was found between ATP concentration and rate of IF1 binding to isolated bovine MF 1 subcomplex [31]. In this work, we have investigated the influence of D ~ l H þ on IF1-mediated inhibition, using yeast IF1 and coupled submitochondrial particles (SMP) with ATP and GTP as substrates. Unexpectedly, in the presence of D ~ l H þ ,IF1was found to promote the conversion of active F 0 F 1 –ATPase to a latent form which remains inactive in ATP hydrolysis even after IF1 release. Finally, we propose a scheme where Correspondence to F. Haraux, Service de Bioe ´ nerge ´ tique & CNRS- URA 2096, DBJC, CEA Saclay, F91191 Gif-sur-Yvette, France. Fax: + 33 1 69 08 87 17, Tel.: + 33 1 69 08 98 91, E-mail: francis.haraux@cea.fr Abbreviations: IF1, inhibitor peptide of mitochondrial ATPase; SMP, submitochondrial particles; FCCP, carbonyl-cyanide p-(trifluoromethoxy)phenylhydrazone. Enzymes: mitochondrial ATP synthase complex (EC 3.6.3.14). *Present address: Southern Methodist University, Department of Biological Sciences, PO Box 750376, Dallas TX 75275-0376, USA. (Received 3 February 2003, revised 18 March 2004, accepted 25 March 2004) Eur. J. Biochem. 271, 1963–1970 (2004) Ó FEBS 2004 doi:10.1111/j.1432-1033.2004.04108.x specific microscopic states and catalytic steps play an explicit role in IF1 locking and in the interconversion between active and latent ATPase. Materials and methods Preparation of phosphorylating SMPs Frozen mitochondria (10–20 mg proteins) prepared as described by Gue ´ rin et al. [38] were thawed [39] and diluted to 5–8 mgÆmL )1 in 10 m M Tris/HCl, 100 l M EDTA pH 8.1. The suspension was saturated with argon for 2 min, then sonicated on ice using a W-225R tip sonicator (Ultrasonics Inc.). Two 20 s cycles (output control 2; 60% duty cycle) separated by 30 s interval were applied. Then the suspension was centrifuged at 23 000 g for 20 min at 4 °C. The supernatant containing SMP was centrifuged at 100 000 g for 40 min at 4 °C. The pellet of SMP was suspended in 0.65 M mannitol. ATPase and ATP synthase activities by these SMP were stable for at least 1–2 days when stored at room temperature. SMP could also be rapidly frozen in liquid nitrogen for future use. Protein content was deter- mined according to Lowry et al. [40] in the presence of 5% (w/v) SDS using BSA as a standard. The average yield of SMP was 10–20% of the total mitochondrial protein. Conditioning of phosphorylating SMP Before an experiment a suspension of SMP was thawed rapidly (at about 85 °C), then diluted (to 1 mgÆmL )1 )ina medium containing (final concentrations) 0.65 M mannitol, 10 m M Hepes, 5 m M potassium phosphate pH 8.0, 100 l M EDTA, 2 m M malonate, and a substoichiometric amount of oligomycin (0.1 lgÆmg )1 protein). The suspension was incubated for 1 h at room temperature before the assays for complete activation of succinate dehydrogenase [41] and ATP synthesis [32]. Standard SMP preparations coupled by substoichiometric oligomycin had the following activities (lmolÆmin )1 Æmg protein )1 ): coupled ATP hydrolysis, 1–1.5; uncoupled ATP hydrolysis, 2–3; succinate-mediated ATP synthesis, 0.3–0.5; coupled succinate oxidation, 0.3; coupled NADH oxidation, 0.37; uncoupled NADH oxidation, 1.5 (respiratory control with NADH, % 4). ATP hydrolysis measurements The reaction was monitored continuously at 25 °Cina stirred cuvette as H + release [42], with phenol red as pH indicator (A 557 ) in 1 or 2 mL of the standard mixture containing 0.65 M mannitol, 10 m M phosphate pH 8.0, 10 m M succinate, 100 l M EDTA (potassium salts), 17 m M KCl, 5 m M MgCl 2 ,60 l M Phenol red, and 1 mgÆmL )1 BSA. (Mg)ATP and (Mg)GTP were added as indicated in the figures. NTP concentrations > 0.8 m M were practically saturating. Carbonyl-cyanide p-(trifluoromethoxy)phenyl- hydrazone (FCCP; 5 l M )or2.5lgÆmL )1 gramicidin D was used as an uncoupler. The same results were obtained with either uncoupler. The different additions (nucleotides, IF1, uncouplers) had negligible effects on the pH of the medium containing no SMP or SMP treated with excess of oligomycin. First-order kinetics of inhibition by IF 1 [39] was fitted to the equation: yðtÞ¼V 1 tþ½ðV 0 ÀV 1 Þ=k app ð1Àe Àkt app Þþy 0 Eqn ð1Þ where y 0 and y(t)areA 557 at zero time and t time after IF1 addition, respectively, V 0 is the initial rate of A 557 change (at zero time), V 1 is the final rate of A 557 change (at infinite time), and k app is the apparent deactivation rate constant. k app did not depend on NTP concentrations from 0.5 m M to 2m M Nonlinear least-square minimization was carried out using the solver of Microsoft EXCEL software. Reagents and peptides ADP, ATP and GDP were from Roche-Boehringer Mannheim. ATP contained 0.65% (w/w) ADP, and GTP contained 2.4% (w/w) GDP and no detectable adenine nucleotides (by HPLC analysis). Oligomycin, gramicidin and FCCP were from Sigma. They were prepared as stock solutions in methanol. All other chemicals were of analytical grade from Sigma or Merck. Synthetic IF1 from Neosystem (Strasbourg, France) as well as IF1 purified from yeast were used, and gave identical results [39]. Results Effect of protonmotive force on interaction of IF 1 with ATPase The effect of D ~ l H þ on the IF 1 –enzyme interaction was studied using SMPs coupled with substoichiometric amount of oligomycin. Fig. 1 shows hydrolysis of ATP in the presence of D ~ l H þ generated by succinate oxidation (a respiratory pathway which does not generate or consume scalar protons). As expected, FCCP addition stimulates Fig. 1. Time-course of ATP hydrolysis in coupled SMPs. The reaction was monitored by scalar H + release as described in Materials and methods. The reaction mixture contained 0.9 m M ATP (curves a–c), or 0.9 m M ATP and 2 m M malonate (curve d). Arrows indicate different additions: SMP 9 lg (a–c) or 20 lg (d), FCCP 5 l M (a–d) and IF 1 0.85 l M (b–d). Labels near the curves give rates of hydrolysis in lmolÆmin )1 Æmg protein )1 . Ordinate scales, indicated by vertical arrows and given in l M H + , are different for a–b, c and d. 1964 M. Galkin et al. (Eur. J. Biochem. 271) Ó FEBS 2004 ATP hydrolysis (curve a), and IF1 addition inhibits FCCP- uncoupled ATP hydrolysis (curve b). Curve c shows the time-course of ATP hydrolysis in the presence of succinate- dependent D ~ l H þ , before and after addition of IF1 at the same concentration as in curve b. Upon IF1 addition, fast ATPase inhibition occurred. The most remarkable obser- vation was made after addition of FCCP to these IF1- inhibited SMP (third part of curve c). Quite unexpectedly, FCCP addition resulted in a transient recovery of ATPase activity, reaching 70–90% of the control activity, which is given by the activity in the presence of FCCP in curves a–b. This recovered ATPase activity decayed afterwards, at approximately the same rate as after addition of IF1 to SMP previously uncoupled with FCCP (compare terminal parts of curves b and c). In the experiment shown in Fig. 1 curve d, inhibition by IF1 was studied using SMP in the presence of malonate, which fully inhibited the respiratory chain (control not shown). Addition of IF1 to SMP during ATP hydrolysis resulted in fast inhibition of ATPase activity, as previously. However, in contrast with respiring SMP, addition of FCCP to IF1-inhibited material resulted in only poor transient ATPase reactivation. This shows that the extent of recovery of ATPase activity following uncoupler addition strongly depends on the magnitude of the D ~ l H þ ,whichwas generated here only by ATP hydrolysis and which presum- ably became dramatically low after IF1 addition. The fact that an almost complete inhibition of ATPase by IF 1 in the presence of D ~ l H þ is followed by an almost full recovery of ATPase activity just after D ~ l H þ collapse, cannot be readily interpreted. It suggests that a special enzyme state, formed in the presence of D ~ l H þ from the IF1-inhibited state and remaining ATPase-inactive as long as D ~ l H þ is maintained, immediately becomes active upon uncoupling. After uncoupling, the reactivated form of the enzyme deactivates again. The rate of deactivation is apparently the same as that observed after adding IF1 to SMP previously treated with FCCP (compare curves b and c, Fig. 1), suggesting that it is also controlled by IF1 binding. To check this hypothesis, we studied IF1-inhibition of coupled ATPase and recovery of its activity using different IF1 concentrations. Relationship between IF1 concentration and uncoupler-induced ATPase reactivation Fig. 2A shows the time course of ATP hydrolysis in coupled SMP, with successive additions of IF1 (0.5 l M )andFCCP separated by 2 min. Fig. 2B shows the time course of ATP hydrolysis by coupled SMP to which IF1 was previously added, as above, with focusing on the last step of reactivation by FCCP. Curves 1 and 2 show FCCP-induced reactivation of ATPase previously inhibited by IF1 2 l M and 0.5 l M , respectively. The higher the IF1 concentration, the lower FCCP-induced ATPase activity. However, it is difficult to discriminate between the effects of IF1 concentration on the initial ATPase activity and on the subsequent rate of decay. To make the picture clearer, we have made new experiments using GTP instead of ATP for two reasons: (a) in contrast to ADP, GDP does not produce hysteretic inhibition [33]; (b) more generally, GDP can accumulate to significant amounts without slowing down the rate of hydrolysis. Accordingly, coupled GTPase activity was per- fectly stable for at least 5 min, unlike ATPase activity, and subsequent addition of uncoupler caused twofold stimula- tion of the activity which remained constant for at least 5 min; furthermore, the rate of IF1-dependent inhibition of uncoupled GTP hydrolysis was not sensitive to GDP accumulation (data not shown). Figure 3A shows typical kinetics of GTP hydrolysis by coupled SMP in the presence of succinate. IF1 at various concentrations was added during GTP hydrolysis (curves 1– 2) and FCCP was added 2.5 min later (control experiments revealed that the rate of coupled ATP hydrolysis no longer varied between 2.5 and 4 min after IF1 addition). Curve 3 shows what happened when IF1 (at the same concentration as in curve 2) was added after FCCP, and comparison of curves 2 and 3 shows that at a low IF1 concentration, the extent of GTPase recovery after FCCP addition is almost 100%. As in Fig. 2, the recovery decreases with IF1 concentration. This is more visible on Fig. 3B which focuses on a short time range before and after FCCP addition and shows how rates of GTP hydrolysis are computed. Rates just before and after FCCP additions are plotted vs. IF1 concentration in Fig. 4A. These two rates obviously follow different patterns: half-inhibition of the coupled activity is reached at about 0.2 l M IF1 (curve 1), and half-inhibition of the uncoupler-recovered activity at about 1 l M IF1 (curve 2). Curve 3 (dashed) shows what GTP hydrolysis rate after FCCP addition should be if it obeyed the same pattern as coupled GTP hydrolysis. Curve 4 (dashed) shows what Fig. 2. IF1-dependent inhibition and FCCP-induced recovery of ATPase activity in SMPs. The reaction was monitored as in Fig. 1. The reac- tion mixture contained 1.4 m M ATP. Additions (SMP 60 lg, IF1, FCCP 5 l M ) are indicated by arrows. (A) ATPase inhibition by IF1 0.5 l M in the presence of D ~ l H þ and after FCCP-induced recovery. (B) ATPase inhibition after FCCP addition to SMP previously inhibited with IF1 2 l M (curve 1) or 0.5 l M (curve 2), as in (A). Only the final stage is shown. Ó FEBS 2004 IF1-mediated transitions of F1-ATPase in yeast SMP (Eur. J. Biochem. 271) 1965 GTP hydrolysis after FCCP addition should be if it was only due to a subpopulation of tightly coupled, IF1-insensitive SMP (see Discussion). Kinetics of GTP hydrolysis after FCCP addition were fitted to a mono- exponential decay [39], and the resulting rate constants of deactivation k app were plotted in Fig. 4B. k app is propor- tional to IF1 concentration [39], which confirms that the final decay of GTPase activity is due to IF1 rebinding. Kinetics of ATPase reactivation in IF1-pretreated SMPs To study ATPase reactivation further we used SMP preincubated with MgATP in succinate-free medium, with or without highly concentrated IF1 (50 l M ). SMP were then diluted 100-fold in the reaction medium containing succi- nate and checked for ATP hydrolysis under different conditions. This allowed the experiment to be started with ATPase fully inhibited by IF1 in a reactivation medium containing a limited concentration of IF1 (0.5 l M ). Fig. 5A indeed shows that in FCCP-containing medium, IF1- pretreated SMP were initially fully inactive (trace 2), compared to SMP preincubated without IF1 (trace 1). When IF1-treated SMP were diluted in FCCP-free medium containing succinate, ATPase activity was initially negligible and was recovered mainly after FCCP addition (Fig. 5A, trace 3). Fig. 5B (black squares) shows the extent of Fig. 3. IF1-dependent inhibition and FCCP-induced recovery of GTPase activity in SMPs. The reaction was monitored as in Figs 1 and 2. The reaction mixture contained 2 m M GTP. (A) GTP hydrolysis initiated by addition of SMP, inhibited by IF1 at two different con- centrations, and further reactivated with FCCP. Additions (SMP 18 lg, IF1, FCCP 5 l M ) are indicated by arrows. Curves 1 and 2: addition of IF1 1.25 l M and 0.125 l M , respectively, 2.5 min after FCCP. Curve 3, control with IF1 0.125 l M added after FCCP. Only one sample of the first part of the curves is shown, because it does not change with subsequent additions. (B) Time course of GTP hydrolysis just before and just after FCCP addition (extended scale), for three differentIF1concentrations.Curves1,2,3:IF10.125 l M ,0.25 l M and 1 l M , respectively. Straight lines were obtained from linear regression using all the displayed data before FCCP additions, and the data of the first 30 s, 20 s and 15 s, respectively, after FCCP addition. Their slopes are proportional to GTPase activities. Fig. 4. Rate of GTP hydrolysis and GTPase inhibition as a function of IF1 concentration. Conditions as in Fig. 3. (A) Rates of GTP hydro- lysis just before (s, curve 1) and just after (d, curve 2) FCCP addition to coupled SMP, as a function of IF1 concentration. Rates were cal- culated as shown in Fig. 3B. Curve 3 (dashed) was obtained by mul- tiplying ordinates of curve 1 by the ratio between uncoupled and coupled GTPase rates in the absence of IF1. It indicates expected uncoupled activity if catalysed by the same enzyme form as coupled activity. Curve 4 (dashed) was obtained by subtracting curve 1 to the uncoupled activity without IF1. It gives expected activity due to a hypothetical tight-coupled SMP subpopulation resistant to IF1 before FCCP addition. Neither curve 3, nor curve 4 correctly fits the data (see text for details). (B) Apparent rate constant of inhibition after FCCP addition, vs. IF1 concentration. 1966 M. Galkin et al. (Eur. J. Biochem. 271) Ó FEBS 2004 recovery of ATPase activity as a function of the time separating dilution of pretreated SMP and FCCP addition. It is also shown that some recovery of ATPase activity was actually observed before FCCP addition (white squares), but it was weak, even though one multiplies its value by a factor two (Fig. 5B, dashed curve) to take into account the back pressure effect of D ~ l H þ , which is about 50%. This comparison confirms that ATPase activities measured before and after FCCP are different by nature. Lastly, black triangles in Fig. 5B show the ATPase activity recovered in the presence of FCCP, present from the beginning. It is about twice the activity recovered without FCCP, as expected if the only difference between these two modes of slow activation is the back-pressure effect exerted by D ~ l H þ . Anyway, in both cases, this background reacti- vation, due to the high pH of the reaction medium [39] and not related to membrane energization, remains well below that triggered by D ~ l H þ and further revealed by FCCP addition. Discussion It is thought that unidirectional IF 1 binding to ATP synthase is independent of the presence of D ~ l H þ [17], whereas its release depends on D ~ l H þ [10–17]. Here, inhibi- tion of coupled NTP hydrolysis by IF 1 , and the unexpected reactivation of IF 1 -inhibited enzyme on uncoupler addition, were measured in the same assay, in conditions where consumption of NTP had little effect (ATP case) or no effect (GTP case) in itself. This uncoupler-induced recovery of ATPase activity is hardly compatible with the assumption of only two states of the enzyme (active, without IF1 and inactive, with IF 1 bound), which suggests that the proton- motive force converts the IF1-inhibited ATPase into a form which has a lower affinity for IF1, but which remains inactive even after IF1 release, unless D ~ l H þ is collapsed. Before developing this idea, it is necessary to examine other explanations. A possible interpretation of our data could be that SMP preparation would be a heterogeneous mixture of energized vesicles, always insensitive to IF1 unless FCCP is added, and permanently deenergized vesicles, sensitive to IF1. The recovery of ATPase activity after FCCP addition should only be due to well-coupled vesicles. This does not seem realistic for quantitative reasons. In Fig. 1 indeed, the maximal uncoupled ATPase activity, representing in the context of heterogeneity the uncoupled activity of the whole preparation, is 3 lmolÆmin )1 Æmg protein )1 (curve a), whereas the FCCP-induced ATPase recovery, believed to be only due to well-coupled vesicles, is 2.2 lmolÆmin )1 Æmg protein )1 (curve c; this value, calculated some seconds after FCCP addition, is probably underesti- mated). The difference between these two activities (0.8 lmolÆmin )1 Æmg protein )1 ) would be the activity of permanently uncoupled vesicles, the only ones to be inhibited by IF1 added before the uncoupler. This activity should be subtracted from the apparent coupled activity (1.4 lmolÆmin )1 Æmg protein )1 , curve c) to obtain the true coupled ATPase activity of the subpopulation of energized vesicles. The resulting activity is expected to be resistant to IF1, so ATPase activity after IF1 and before FCCP addition should remain as high as 0.6 lmolÆmin )1 Æmg protein )1 .This is not the case, the measured activity (curve c) is less than 0.05 lmolÆmin )1 Æmg protein )1 (in fact practically zero). In other words, if a special SMP subpopulation is responsible for the FCCP-induced recovery of activity after IF1 treatment, one expects IF1 to inhibit only the activity of Fig. 5. ATP hydrolysis in SMPs preincubated with concentrated IF1. SMP (0.9 mgÆmL )1 ) were preincubated in standard succinate-free mixture in presence of 50 l M IF 1 ,2.3m M ATP and 8.3 m M MgCl 2 (SMP IF1 ). Additions in a 1-mL spectrophotometric cuvette (1.4 m M ATP, 5 l M FCCP, 9 lgSMP,100l M ADP) are indicated by arrows. (A) Time courses of ATPase activity of pretreated SMP after 100-fold dilution in the standard reaction medium containing succinate. Curve 1, control; curves 2–4, SMP preincubated with IF 1 (SMP IF1 ); curve 2, FCCP added before SMP IF1 ; curve 3, FCCP added 2 min after SMP IF1 . Labels near curves 1 and 3 give rates of hydrolysis in lmolÆmin )1 Æmg protein )1 . (B) Dependence of the uncoupler-induced ATPase activity (j) on the time separating dilution of SMP IF1 in succinate-containing medium and addition of uncoupler; 100% corresponds to the control SMP (A, trace 1). (h), Coupled ATPase activity (no FCCP) as a function of the time following SMP IF1 addition, calculated from slopes of pH recordings, under conditions similar to those of the part of curve 3 (A) preceding FCCP addition; dashed curve (- - -), double of the coupled ATPase activity, representing the expected activity after FCCP addition; (m), ATPase activity as a function of the time following SMP IF1 addition with FCCP initially present, calculated from the slope of curves like curve 2 in (A). Ó FEBS 2004 IF1-mediated transitions of F1-ATPase in yeast SMP (Eur. J. Biochem. 271) 1967 permanently deenergized vesicles, and then to cause the same drop of ATPase activity before and after FCCP addition. This is far from being observed in Fig. 1, as previously discussed, and also in Fig. 4A, which shows the dependency of GTPase activity on IF1 concentration before and after FCCP addition and clearly indicates that the two activities do not follow parallel curves (curves 2 and 4 are clearly distinct). More, if we consider a heterogeneous preparation, it cannot reasonably consist of two discrete populations, noncoupled and tightly coupled, where ATP- ases are, respectively, inhibited and 100% resistant. It should contain of course partially coupled vesicles, where some inhibition by IF1 is expected. To get ATPase inhibition by IF1, complete collapse of the protonmotive force is not necessary provided that ATPase works, even slowly, in the direction of ATP (GTP) hydrolysis. As a consequence, if it was due to functional heterogeneity, FCCP-induced GTPase activity should actually follow not curve 4, but a curve located between curves 3 and 4, still more remote from the experimental data. Therefore the results cannot be explained simply by functional heterogen- eity of SMP. Of course, this does not mean that SMP are fully homogeneous, and functional homogeneity of SMP is actually impossible to check. But heterogeneity is generally associated with poor reproducibility, and the different SMP preparations used in this work had a reproducible ATPase activity in the presence of succinate, the stimulation factor of ATPase activity by FCCP being practically constant (100 ± 10%). Finally, our data are consistent with the existence of the following states of ATPase (or GTPase): where E is the only active form of ATPase. During ATP hydrolysis, IF1 can bind to the ATPase both in the absence and in the presence of D ~ l H þ . Only in the presence of D ~ l H þ , does the E*IF1 complex undergo energy- dependent conversion to a latent form of the enzyme still inactive in ATP hydrolysis (E ¼ IF1) but presenting a lower affinity for IF1. Upon uncoupling (indicated by a vertical arrow) the latent ATPase which is free of IF1 (E ¼) immediately converts, within the time resolution limit, into the active ATPase (E) (uncoupler-induced activity), which again undergoes inhibition by IF1. Due to the presence of the uncoupler, this second inhibition cannot be reversed anymore. In the case of GTP hydrolysis, the dependency of the uncoupler-induced ATPase on IF1 concentration allows one to determine the affinity of the latent state (E ¼ in the above scheme) of the ATPase for IF 1 :K d is %1 l M .This value is much higher than the K d for the IF 1 –ATPase interaction in the absence of D ~ l H þ ,whichis% 40 n M at pH 8 under conditions similar to the present ones [39] and is consistent with the classical energy-dependent release of IF 1 . A plausible mechanism for the IF1-transition to the ATPase latent form can be proposed by focusing on the release of ADP from one catalytic site during net ATP hydrolysis (we will not consider the other sites; a complete description of the enzyme should take into account the increase of nucleotide occupancy induced by IF1 [43,44]). The statements are the following: The ADP-loaded site is successively closed, half-closed and open [45]. D ~ l H þ tends to reverse the closed to half- closed transition (back pressure effect). When the ADP- loaded site is closed, the enzyme has some probability to be converted into the latent form. Under normal conditions, this conversion is negligible because the steady state concentration of the closed state is low. When IF1 is bound, it is assumed to block, or to slow down, the transition from half-closed to open conformation. In the latter case, it fully blocks the next step. IF1 acts on the considered catalytic site either directly, or indirectly, by blocking ATP hydrolysis on another site [24]. In the absence of D ~ l H þ , the final state is a dead-end complex where IF1, initially loosely bound, is now locked [31]. When IF1 is bound and D ~ l H þ present, the ADP-loaded catalytic site remains essentially closed and is then converted into its latent form, not represented in the above scheme. As in the latent form the catalytic reaction is stopped upstream of ADP release, IF1 remains loosely bound and can be released if its concentration is low, which gives the IF1-free latent form. This outline of mechanism can explain the synergetic effect of IF1 and D ~ l H þ on the formation of the latent form in the presence of MgATP. These three effectors must be simultaneously present to stabilize the closed state leading to the latent form. The proposed mechanism can also explain the recovery of the active form after D ~ l H þ collapse. The latent form indeed can be considered so sensitive to the D ~ l H þ back-pressure that ADP stays on the catalytic site and the rate of ATP hydrolysis is practically zero. However, when D ~ l H þ is collapsed, ADP release occurs at a rate which may be low with respect to the time scale of the catalytic cycle (milliseconds), but which remains faster than the experimental time response (seconds). Once ADP is released, one gets again the active form of ATPase. The simplest model of D ~ l H þ -induced activation of mitochondrial ATPase involves two functional forms: an inactive and an active IF1-free form. Early studies have suggested the existence of an additional IF1-bearing form active in ATP synthesis but practically inactive in ATP hydrolysis [10,19]. The present data suggest that in the presence of D ~ l H þ , IF1 could act as a catalyst inducing the transformation of ATP synthase from a fully active enzyme to a latent form, unable to hydrolyse ATP, even after IF1 release. Further investigations of the mechanism of ATPase regulation by D ~ l H þ and IF1 should now take into account 1968 M. Galkin et al. (Eur. J. Biochem. 271) Ó FEBS 2004 this possible additional form of ATP synthase. This IF1- dependent conversion of active ATPase into latent ATPase resembles that induced by MgADP and D ~ l H þ in bovine heart SMP [32,34] and in Pseudomonas denitrificans [46]. Whether the presently described latent state of ATPase could synthesize ATP under appropriate conditions or not has not been established so far. Acknowledgements We thank Gwe ´ nae ¨ lle Moal-Raisin for her excellent technical help, and Dr Sigalat for HPLC analysis of nucleotides and helpful suggestions. References 1. Boyer, P.D. (1997) The ATP synthase – a splendid molecular machine. Annu.Rev.Biochem.66, 717–749. 2. Abrahams, J.P., Leslie, A.G.W., Lutter, R. & Walker, J.E. (1994) Structure at 2.8 A ˚ resolution of F 1 -ATPasefrombovineheart mitochondria. Nature 370, 21–28. 3. Weber, J. & Senior, A.E. (1997) Catalytic mechanism of F 1 - ATPase. Biochim. Biophys. Acta 1319, 19–58. 4. Berden, J.A. (2003) Rotary movements within the ATP synthase do not constitute an obligatory element of the catalytic mechan- ism. IUBMB Life 55, 473–481. 5. Noji, H., Yasuda, R., Yoshida, M. & Kinosita, K. Jr (1997) Direct observation of the rotation of F 1 -ATPase. Nature 386, 299–302. 6. Yasuda, R., Noji, H., Yoshida, M., Kinosita, K. Jr & Itoh, H. (2001) Resolution of distinct rotational substeps by submillisecond kinetic analysis of F 1 -ATPase. Nature 410, 898–904. 7. Stock, D., Leslie, A.G.W. & Walker, J.E. (1999) Molecular architecture of the rotary motor in ATP synthase. Science 26, 1700–1705. 8. Pullman, M.E. & Monroy, G.C. (1963) A naturally occurring inhibitor of mitochondrial adenosine triphosphatase. J. Biol. Chem. 238, 3762–3768. 9. Green, D.W. & Grover, G.J. (2000) The IF 1 inhibitor protein of the mitochondrial F 1 F 0 -ATPase. Biochim. Biophys. Acta 1458, 343–355. 10. Schwerzmann, K. & Pedersen, P.L. (1986) Regulation of the mitochondrial ATP synthase/ATPase complex. Arch. Biochem. Biophys. 250, 1–18. 11. Van de Stadt, R.J., de Boer, B.L. & Van Dam, K. (1973) The interaction between the mitochondrial ATPase (F 1 )andthe ATPase inhibitor. Biochim. Biophys. Acta 292, 338–349. 12. Klein, G., Satre, M. & Vignais, P. (1977) Natural protein ATPase inhibitor from Candida utilis mitochondria. FEBS Lett. 84, 129– 134. 13. Husain, I. & Harris, D.A. (1983) ATP synthesis and hydrolysis in submitochondrial particles subjected to an acid-base transition. FEBS Lett. 160, 110–114. 14. Power, J., Cross, R.L. & Harris, D.A. (1983) Interaction of F 1 - ATPase from ox heart mitochondria with its naturally occurring inhibitor protein. Studies using radio-iodinated inhibitor protein. Biochim. Biophys. Acta 724, 128–141. 15. Husain, I., Jackson, P.J. & Harris, D.A. (1985) Interaction be- tween F 1 -ATPase and its naturally occurring inhibitor protein. Studies using a specific anti-inhibitor antibody. Biochim. Biophys. Acta 806, 64–74. 16. Lippe, G., Sorgato, M.C. & Harris, D.A. (1988) Kinetics of the release of the mitochondrial inhibitor protein. Correlation with synthesis and hydrolysis of ATP. Biochim. Biophys. Acta 933, 1–11. 17. Lippe, G., Sorgato, M.C. & Harris, D.A. (1988) The binding and release of the inhibitor protein are governed independently by ATP and membrane potential in ox-heart submitochondrial vesicles. Biochim. Biophys. Acta 933, 12–21. 18. Beltran, C., Tuena de Go ´ mez-Puyou, M., Go ´ mez-Puyou, A. & Darszon, A. (1984) Release of the inhibitory action of the natural ATPase inhibitor protein on the mitochondrial ATPase. Eur. J. Biochem. 144, 151–157. 19. Schwerzmann, K. & Pedersen, P.L. (1981) Proton-adenosine- triphosphatase complex of rat liver mitochondria: effect of energy state on its interaction with the adenosinetriphosphatase inhibitory peptide. Biochemistry 20, 6305–6311. 20. Jackson, P.J. & Harris, D.A. (1988) The mitochondrial ATP synthase inhibitor protein binds near the C-terminus of the F 1 b-subunit. FEBS Lett. 229, 224–228. 21. Ichikawa, N., Yoshida, Y., Hashimoto, T. & Tagawa, K. (1996) An intrinsic ATPase inhibitor binds near the active site of yeast mitochondrial F 1 -ATPase. J. Biochem. (Tokyo) 119, 193–199. 22. Mimura, H., Hashimoto, T., Yoshida, Y., Ichikawa, N. & Tagawa, K. (1993) Binding of an intrinsic ATPase inhibitor to the interface between a-andb-subunits of F 1 F 0 ATPase upon de-energization of mitochondria. J. Biochem. (Tokyo) 113, 350– 354. 23. Minauro-Sanmiguel, F., Bravo, F. & Garcia, J.J. (2002) Cross- linking of the endogenous inhibitor protein (IF 1 )withrotor(c,e) and stator (a) subunits of the mitochondrial ATP synthase. J. Bioenerg. Biomembr. 34, 433–443. 24. Cabezo ´ n, E., Montgomery, M.G., Leslie, A.G.W. & Walker, J.E. (2003) The structure of bovine F1-ATPase in complex with its regulatory protein IF1. Nat. Struct. Biol. 10, 744–750. 25.Horstman,L.L.&Racker,E.(1970)Partialresolutionofthe enzymes catalyzing oxidative phosphorylation. XXII. Interaction between mitochondrial adenosine triphosphatase inhibitor and mitochondrial adenosine triphosphatase. J. Biol. Chem. 245, 1336–1344. 26. Panchenko, M.V. & Vinogradov, A.D. (1985) Interaction between the mitochondrial ATP synthetase and ATPase inhibitor protein. Active/inactive slow pH-dependent transitions of the inhibitor protein. FEBS Lett. 184, 226–230. 27. Khodjaev, E Yu, Komarnitsky, F.B., Capozza, G., Dukhovich, V.F., Chernyak, B.V. & Papa, S. (1990) Activation of a complex of ATPase with the natural protein inhibitor in submitochondrial particles. FEBS Lett. 272, 145–148. 28. Fujii, S., Hashimoto, T., Yoshida, Y., Miura, R., Yamano, T. & Tagawa, K. (1983) pH-induced conformational change of ATPase inhibitor from yeast mitochondria. A proton magnetic resonance study. J. Biochem. (Tokyo) 93, 189–196. 29. Rouslin, W. & Broge, C.W. (1993) Factors affecting the species- homologous and species-heterologous binding of mitochondrial ATPase inhibitor, IF1, to the mitochondrial ATPase of slow and fast heart-rate hearts. Arch. Biochem. Biophys. 303, 443–450. 30. Klein, G. & Vignais, P.V. (1983) Effect of the protonmotive force on ATP-linked processes and mobilization of the bound natural ATPase inhibitor in beef heart submitochondrial particles. J. Bioenerg. Biomembr. 15, 347–362. 31. Milgrom, Ya. M. (1989) An ATP-dependence of mitochondrial F 1 -ATPase inactivation by the natural inhibitor protein agrees with the alternating-site binding-change mechanism. FEBS Lett. 246, 202–206. 32. Galkin, M.A. & Vinogradov, A.D. (1999) Energy-dependent transformation of the catalytic activities of the mitochondrial F 0 F 1 -ATP synthase. FEBS Lett. 448, 123–126. 33. Vinogradov, A.D. (2000) Steady-state and pre-steady-state kinet- ics of the mitochondrial F 1 -F 0 ATPase: is ATP synthase a reversible molecular machine? J. Exp. Biol. 203, 41–49. Ó FEBS 2004 IF1-mediated transitions of F1-ATPase in yeast SMP (Eur. J. Biochem. 271) 1969 34. Syroeshkin, A.V., Vasilyeva, E.A. & Vinogradov, A.D. (1995) ATP synthesis catalyzed by the mitochondrial F 1 -F 0 ATP synthase is not a reversal of its ATPase activity. FEBS Lett. 36, 29–32. 35. Baubichon, H., Godinot, C., Di Pietro, A. & Gautheron, D.C. (1981) Competition between ADP and nucleotide analogues to occupy regulatory sites(s) related to hysteretic inhibition of mitochondrial F 1 -ATPase. Biochem. Biophys. Res. Commun. 100, 1032–1038. 36. Bullough, D.A., Brown, E.L., Saario, J.D. & Allison, W.S. (1988) On the location and function of the noncatalytic sites on the bovine heart mitochondrial F 1 -ATPase. J. Biol. Chem. 263, 4053–4060. 37. Berden, J.A., & Hartog, A.F. (2000) Analysis of the nucleotide binding sites of mitochondrial ATP synthase provides evidence for a two-site catalytic mechanism. Biochim. Biophys. Acta 1458, 234–251. 38. Gue ´ rin, B., Labbe, P. & Somlo, M. (1979) Preparation of yeast mitochondria (Saccharomyces cerevisiae)withgoodP/Oand respiratory control ratios. Methods Enzymol. 55, 149–159. 39. Venard, R., Bre ` thes, D., Giraud, M F., Vaillier, J., Velours, J. & Haraux, F. (2003) Investigation of the role and mechanism of IF1 and STF1 proteins, twin inhibitory peptides which interact with the yeast mitochondrial ATP synthase. Biochemistry 42, 7626– 7636. 40. Lowry, O.H., Rosebrough, N.J., Farr, A.L. & Randall, R.J. (1951) Protein measurement with the Folin phenol reagent. J. Biol. Chem. 193, 265–275. 41. Kotlyar, A.V. & Vinogradov, A.D. (1984) Interaction of the membrane-bound succinate dehydrogenase with substrate and competitive inhibitors. Biochim. Biophys. Acta 784, 24–34. 42. Nishimura, M., Ito, T. & Chance, B. (1962) Studies on bacterial photophosphorylation. III. A sensitive and rapid method of determination of photophosphorylation. Biochim. Biophys. Acta 59, 177–182. 43. Di Pietro, A., Penin, F., Julliard, J H., Godinot, C. & Gautheron, D.C. (1988) IF1 inhibition of mitochondrial F 1 -ATPase is corre- lated to entrapment of four adenine- or guanine-nucleotides including at least one triphosphate. Biochem. Biophys. Res. Commm. 152, 1319–1325. 44. Milgrom, Ya. M. (1991) When beef-heart mitochondrial F 1 -ATPase is inhibited by inhibitor protein a nucleotide is trapped in one of the catalytic sites. Eur. J. Biochem. 200, 789–795. 45. Menz, R.I., Walker, J.E. & Leslie, A.G.W. (2001) Structure of bovine mitochondrial F 1 -ATPase with nucleotide bound to all three catalytic sites: implications for the mechanism of rotary catalysis. Cell 10, 331–341. 46. Zharova, T.V. & Vinogradov, A.D. (2004) Energy-dependent transformation of F 0 F 1 -ATPase in Paracoccus denitrificans plasma membranes. J.Biol. Chem. 279, 12319–12324. 1970 M. Galkin et al. (Eur. J. Biochem. 271) Ó FEBS 2004 . Functional transitions of F 0 F 1 -ATPase mediated by the inhibitory peptide IF1 in yeast coupled submitochondrial particles Mikhail Galkin 1, *,. its naturally occurring protein inhibitor (IF1) was investi- gated in submitochondrial particles by studying the IF1- mediated ATPase inhibition in the presence