Rotary catalysis of F_oF₁-ATP synthase

Abstract

The synthesis of ATP, the key reaction of biological energy metabolism, is accomplished by the rotary motor protein; F_oF₁-ATP synthase (F_oF₁). In vivo, F_oF_1, located on the cell membrane, carries out ATP synthesis by using the proton motive force. This heterologous energy conversion is supposed to be mediated by the mechanical rotation of F_oF₁; however, it still remained unclear. Recently, we developed the novel experimental setup to reproduce the proton motive force in vitro and succeeded in directly observing the proton-driven rotation of F_oF₁. In this review, we describe the interesting working principles determined so far for F_oF₁ and then introduce results from our recent study.

F_oF₁ ATP synthase (F_oF₁) is a molecular energy-converter which catalyzes physiologically important synthesis of ATP from ADP and inorganic phosphate (P_i) by using the electrochemical energy in proton gradient; the proton motive force (pmf) across bio-membranes1–3. The prominent feature of F_oF₁ is that mechanical rotation of the inner rotor complex mediates the aforementioned heterologous energy conversion with high efficiency and reversibility4,5, which is not found in other biological systems, and therefore, F_oF₁ attracts great interest from many researchers across a wide range of research fields. F_oF₁ comprises 2 rotary motors, F₁ and F_o (Fig. 1a). F₁ (α₃β₃γδɛ), a water-soluble part of F_oF₁, is an ATP-driven rotary motor, which couples ATP hydrolysis or synthesis at α₃β₃δ stator ring to the mechanical rotation of γɛ complex in counterclockwise direction (Fig. 1b)6. Each αβ interface possesses a catalytic site for ATP hydrolysis7; F₁ hydrolyzes 3 ATP molecules per turn8. Extensive studies have been done to understand the chemo-mechanical coupling mechanism of F₁9–18, and therefore, currently, F₁ is one of the best characterized molecular motor proteins.

Figure 1.

Structural model of bacterial F_oF₁-ATP synthase

(a) The space-filling model of F_oF₁ (side view) is assembled from several partial structures obtained by X-ray crystallography and NMR (PDB codes 1C17, 1B9U, 2KHK, 1L2P, 3UDO, 3OAA, and 1ABV). F_oF₁ has subunit stoichiometry of α₃β₃ γδεab₂c_8–15. F_oF₁ is composed of 2 rotary motors; a membrane-embedded F_o (ab₂c_8–15) and a water-soluble F₁ (α₃β₃γδε). Proton flows through the channels located at the a–c interface of F_o, while ATP synthesis is carried out at the catalytic sites located at the α–β interface of F₁. (b) Top view of F_o (left) and F₁ (right).

F_o, a membrane-embedded part of F_oF₁, is the proton-driven rotary motor, which couples proton translocation to mechanical rotation of the oligomer ring of the c-subunit against the ab₂ complex (Fig. 1b)5,19,20. Each c-subunit has a proton-binding site and mediates proton translocation: 1 proton per c-subunit per rotation. Thus, the number of c-subunits in the c-ring is thought to determine the total number of protons translocated per rotation. While the number of c-subunits varies among species from 8 to 1521,22, the bacterial F_o, such as E-coli and thermophilic Bacillus PS3, has 10 c-subunits that form a c₁₀-ring 23,24. It is difficult to handle F_o since it is embedded in the membrane, and therefore, the rotary mechanism of F_o remains elusive.

In a cell, F₁ and F_o are connected via central and peripheral stalks (Fig. 1a), which allow the torque transmission between the 2 motors. Under physiological conditions, F_o generates a larger torque than F₁ and reverses the rotary direction of F₁, thereby inducing the reverse reaction of ATP hydrolysis, i.e. ATP synthesis (Fig. 2a). In contrast, when pmf diminishes, F₁ hydrolyzes ATP and reverses the rotary direction of F_o, thereby enforcing pumping of protons by F_o in order to generate pmf (Fig. 2b). Thus, F_oF₁ manifests reversibility with regard to the process of energy conversion, and the rotation of inner rotor complex plays an important role in this reversibility.

Figure 2.

Heterologous energy conversion by F_oF₁

Energy conversion mechanism of F_oF₁ in ATP synthesis (a) or hydrolysis (b) conditions. Orange and light green represent the rotor and stator subcomplex, respectively.

To understand the precise role of the rotation on the energy conversion of F_oF₁, some single-molecule studies have been carried out5,19,20,25–27. Due to ease of handling, ATP-driven rotation of detergent-solubilized F_oF₁, which is no longer part of a bio-membrane, has been studied; however, pmf-driven rotation has not been studied sufficiently. In this year, we developed a novel experimental setup that for the first time allowed direct observation of the proton-driven rotation of F_oF₁ 20. In this review, we focus on the role of rotary motion in the energy conversion of F_oF₁, and introduce the fundamental working principles determined so far, including the results from our recent study20.

Energy conversion mechanism

As mentioned above, ATP synthesis/hydrolysis reaction is reversibly coupled with proton translocation across the membrane. To understand the reversible energy conversion mechanism of F_oF₁, extensive biochemical studies have been performed. Since it is difficult to measure the time course of pmf generated by F_oF₁, most studies pertained to the measurement of ATP synthesis activity under a given pmf. In general, pmf mainly comprises 2 components: the trans-membrane proton gradient (ΔpH) and the potential difference (Δψ). At first, biochemical studies examined ATP synthesis activity by changing the amplitude of each of these components of pmf, and examined the component-dependence of ATP synthesis activity of F_oF₁ 28–34. In recent studies, F_oF₁ was purified and reconstituted into liposomes, and the ATP synthesis rate was quantitatively measured by the acid–base transition or valinomycin-mediated K⁺ diffusion potential for producing ΔpH or Δψ. It was found that ΔpH and Δψ contribute equally to ATP synthesis rate of F_oF₁, and, moreover, either ΔpH or Δψ alone can drive synthesis of ATP29. Kinetic equivalence implies that F_oF₁ can use ΔpH or Δψ to drive rotation at an equal efficiency.

Next, to determine the coupling efficiency in F_oF₁, biochemical studies examined the number of protons translocated per synthesis of 1 molecule of ATP; the H⁺/ATP ratio, by enzymes from various organisms, e.g., Escherichia coli, yeast, and chloroplasts4,35,36. In view of the rotary catalysis model of F_oF₁, the H⁺/ATP ratio should coincide with the ratio of the number of proton-binding c-subunits to the 3 catalytic nucleotide-binding β-subunits when proton-translocation and ATP synthesis are highly coupled. Analyses of the equilibrium point, where the free energy of ATP synthesis is balanced with that of proton translocation, allowed determination of the H⁺/ATP ratio. The determined H⁺/ATP ratios were dependent on the stoichiometry of the c and β subunits, although they were not identical to the c/β ratios. In particular, H⁺/ATP ratios in a recent study were smaller than the c/β ratios36, which implies that proton translocation is stochastically coupled to the synthesis of ATP; however, the coupling efficiency is extremely high.

ATP-driven rotation

To understand the precise role of rotation in the energy conversion, the rotation of F_oF₁ was observed at single-molecule level. At first, due to the ease of handling, ATP-driven rotation of solubilized F_oF₁ from E. coli and thermophilic Bacillus PS3 was observed (Fig. 3)25,26,37. In this condition where F_oF₁ was not a part of membrane, pmf was not imposed on F_oF₁, and F_o did not generate the rotary torque. Therefore, as same as isolated F₁, F_oF₁ rotated in a counter-clockwise direction, and showed a 120°-stepping rotation at low ATP concentrations, in which ATP binding was the rate-limiting step of rotation. This 120°-stepping rotation reflects the structural symmetry of F₁; 3 catalytic sites for ATP hydrolysis/synthesis are located on a single molecule of F₁7.

Figure 3.

ATP-driven rotation of F_oF₁

Time course of the ATP-driven rotation of solubilized F_oF₁ in the presence of 1 mM ATP (left) and 50 nM ATP (right). The insets show the centroid trajectories of the rotating particles. At low ATP concentration, F_oF₁ showed a 120° stepping rotation.

Next, the rotation of F_oF₁ reconstituted in a membrane was observed. Ishmukhametov et al. developed an experimental setup with a gold nanorod and phospholipid bilayer nanodisc, which has been shown to provide a good model for a lipid bilayer membrane, and attempted to visualize the rotary motion of membrane-constituted E. coli F_oF₁ driven by ATP hydrolysis27. In their setup, the gold nanorod was attached to the c-subunit as a rotation probe. The intensity of scattered red light from a nanorod changes in a sinusoidal manner as a function of the rotary position, and therefore, the rotary motion of F_oF₁ could be visualized from the analysis of red light scattered from the nanorod. On the other hand, the F_o module was buried within the phospholipid bilayer nanodisc, which is large enough to allow incorporation of F_o, but which is on the same scale as the F_oF₁ complex, and thus, it was difficult to generate a pmf across the nanodisc. Using this experimental setup, the 36° stepping rotation of F_oF₁ in the presence of a high concentration of polyethylene glycol was observed for the first time. This reflects the structural symmetry of the F_o module; 10 proton binding sites are located on a single molecule of F_o24. In this setup, as mentioned above, pmf was not imposed on F_oF₁, and F_o did not generate rotary torque; and therefore, it could not be confirmed whether the 36° step was coupled to the translocation of protons.

pmf-driven rotation

Diez et al. developed a method to indirectly visualize the rotation of E. coli F_oF₁ that had been reconstituted in liposomes by using single molecule Förster resonance energy transfer (sm-FRET)5. In their method, they introduced a pair of FRET probes at a stator and rotor subunit of F_oF₁ for visualization of the rotary motion of F_oF₁, and generated the pmf by using the acid–base transition or valinomycin-mediated K⁺ diffusion potential method, as mentioned above. By using this method, they observed the 120° stepping rotation driven by ATP hydrolysis5, and moreover, for the first time observed the obscure 36º stepping rotation driven by pmf 19. However, due to the low signal-to-noise ratio and fast photobleaching of the fluorescent dyes used in sm-FRET, high resolution tracking and long-term recording of the rotational dynamics of F_oF₁ has not yet been achieved (recording time < approximately 300 ms). Therefore, the fundamental features of the pmf-driven rotation of F_oF₁, such as the exact step size, unidirectionality of the rotation, and stochasticity of the steps, has remained elusive to date.

To solve this problem, we recently developed a novel experimental setup that allows long-term direct observation of the pmf-driven rotation of E. coli F_oF₁ with a high spa-tiotemporal resolution (Fig. 4a)20. In this setup, the F_oF₁-reconstituted, supported membrane was expanded on a cover-slip covered with Ni-NTA-modified agarose, where F_oF₁ molecules were anchored via His-tags that had been introduced to the periplasmic side of the c-subunits. The 80-nm gold colloid was attached as a rotation probe onto the β subunits of F₁ to allow visualization using a total internal reflection dark-field illumination system, which facilitated the long-term recording of rotation (approximately 10 s) with a high spatiotemporal resolution of about 5 nm and <0.5 ms38. In addition, pmf across the supported lipid bilayer was generated by photolysis of caged protons [1-(2-Nitrophenyl) ethyl sulfate] with a total internal reflection illumination of UV light (λ=404 nm) that selectively acidified the space between the coverslip and the lipid bilayer (the interspace). This novel setup can stably generate ΔpH of 1.8–3.7 for several tens of seconds, while the conventional method, i.e. acid–base transition, can generate ΔpH for only a few seconds. The magnitude of ΔpH upon photolysis of the caged protons was measured using a pH-sensitive fluorescent dye, pHrodo-Red (pHrodo)39, which increased the fluorescent signal upon acidification.

Figure 4.

Proton-driven rotation of F_oF₁

(a) Schematic model of the novel experimental setup used to directly visualize the proton-driven rotation of F_oF₁. Evanescent illumination by 532-nm and 404-nm lasers was used. (b) Time course of the proton-driven rotation of F_oF₁ in the presence of 50 μM ADP, 1 μM ATP, and 200 μM P_i before (blue) and after (red) UV irradiation. The left inset shows an enlarged view of the time course. Orange and light blue points represent the pauses before clockwise and counterclockwise steps, respectively. The right inset shows the centroid trajectory of the rotating particle during UV irradiation. Under these conditions, F_oF₁ showed the stochastic 120° stepping rotation. (c) Rotational speed plotted against the trans-membrane proton gradient (ΔpH).

By using this experimental setup, we for the first time observed the clockwise rotary motion of F_oF₁ upon UV irradiation (averaged velocity was 2.7 rps.), while no rotating particles were observed prior to UV irradiation (Fig. 4b). To investigate the correlation between the rotational rate and ΔpH, we measured ΔpH by using the pHrodo located at the region where F_oF₁ showed rotations. Although F_oF₁ showed a large variation in velocity, it was still evident that faster rotation occurred at higher ΔpH (Fig. 4c). In addition, the data points of higher velocity at a given ΔpH qualitatively agreed with the aforementioned biochemical measurement of ATP synthesis activity28, showing that the rotation observed in this study was coupled to the proton translocation and ATP synthesis, and vice versa, the rotation of F_oF₁ can mediate the energy conversion with high efficiency.

We also observed a 120° stepping rotation of F_oF₁ driven by pmf (Fig. 4b). The step size of rotation, viz., 120°, implies that the rotary potential of F₁ with 3-fold symmetry dominated the overall rotary potential of F_oF₁. In other words, the kinetic bottleneck of the pmf-driven rotation was not proton-translocation in F_o, but a catalytic event(s) on F₁, such as ATP release or P_i binding14,40. This result is consistent with a previous sm-FRET measurement of rotation of F_oF₁ in ATP synthesis condition where a pair of FRET probes were introduced at a stator of F_o and a rotor subunit of F₁ 5, while small steps that was estimated to be 36° were recorded when FRET probes were introduced into F_o19. In the present study, we immobilized the rotor part of F_o on a coverslip and attached the rotation probe at the stator part of F₁. Therefore, the observed rotation reflects the stepping behavior both of F₁ and F_o as shown in the other works, in which 36°-steps of ATP-driven rotation of F_oF₁ were observed in a similar experimental setup27. However, we do not exclude the possibility that the difference in the probe position caused the apparently different step size of the rotation. Another possible reason for the inconsistency in step size is the difference in the components of the pmf; while pmf in the sm-FRET measurements was composed of both ΔpH and Δψ, pmf was essentially composed only of ΔpH in our study. To confirm this, a method for direct observation of the proton-driven rotation of F_oF₁ by applying Δψ is crucial.

On the other hand, noted that the stepping rotation of F_oF₁ was highly stochastic; F_oF₁ showed forward-and-backward (clockwise-and-counterclockwise) steps during rotation (Fig. 4b). This is a prominent feature of the pmf-driven rotation of F_oF₁ that is not seen in the ATP-driven rotation of F₁ or F_oF₁. Surprisingly, the stepping was also observed in the absence of pmf, suggesting that pmf biased the rotary diffusion of F_oF₁ to the clockwise direction. To confirm this, we analyzed the pause durations between the 120°-steps. In the absence of pmf, the histograms of the pause duration before clockwise or counterclockwise steps showed single exponential decay. The rate constants of the clockwise and counterclockwise steps were determined to be 65 and 61 s⁻¹, respectively. The equilibrium constant of clockwise rotation was, thus, almost 1. In the presence of pmf, the rate constant of clockwise stepping markedly increased about two-fold, while that of counterclockwise step decreased slightly. Thus, the equilibrium constant in the presence of pmf increased to 2, showing that pmf actually biased step direction. This result also suggests that chemical equilibrium was slightly biased toward ATP synthesis by pmf. The stochastic rotation of F_oF₁ would represent rotation under physiological conditions, where free energy of ATP synthesis almost balances pmf.

Future prospects

Owing to the progress of single molecule observation techniques, we can directly observe the rotary motion of F_oF₁ both in ATP hydrolysis and synthesis conditions. In particular, the introduced novel experimental setup that allows us to stably apply, under an optical microscope, pmf to the membrane will push forward to understand the rotary catalysis mechanism of F_oF₁ in ATP synthesis condition, which had been unclear for long time. This experimental strategy is fundamentally applicable to the study on the dynamics of other membrane proteins driven by electrochemical potential. The most promising experiment is the application of this protocol to transporters and ion channels since there are several caged compounds that release specific ions or chemicals. Such studies would reveal the generality and uniqueness of the finding in single-molecule studies on F_oF₁.