Structural basis of unisite catalysis of bacterial F0F1-ATPase

Atsuki Nakano; Jun-ichi Kishikawa; Atsuko Nakanishi; Kaoru Mitsuoka; Ken Yokoyama

doi:10.1093/pnasnexus/pgac116

. 2022 Jul 11;1(3):pgac116. doi: 10.1093/pnasnexus/pgac116

Structural basis of unisite catalysis of bacterial F₀F₁-ATPase

Atsuki Nakano ¹, Jun-ichi Kishikawa ^2,³, Atsuko Nakanishi ^4,⁵, Kaoru Mitsuoka ⁶, Ken Yokoyama ^7,^✉

Editor: Shibu Yooseph

PMCID: PMC9896953 PMID: 36741449

Abstract

Adenosine triphosphate (ATP) synthases (F₀F₁-ATPases) are crucial for all aerobic organisms. F₁, a water-soluble domain, can catalyze both the synthesis and hydrolysis of ATP with the rotation of the central γε rotor inside a cylinder made of α₃β₃ in three different conformations (referred to as β_E, β_TP, and β_DP). In this study, we determined multiple cryo-electron microscopy structures of bacterial F₀F₁ exposed to different reaction conditions. The structures of nucleotide-depleted F₀F₁ indicate that the ε subunit directly forces β_TP to adopt a closed form independent of the nucleotide binding to β_TP. The structure of F₀F₁ under conditions that permit only a single catalytic β subunit per enzyme to bind ATP is referred to as unisite catalysis and reveals that ATP hydrolysis unexpectedly occurs on β_TP instead of β_DP, where ATP hydrolysis proceeds in the steady-state catalysis of F₀F₁. This indicates that the unisite catalysis of bacterial F₀F₁ significantly differs from the kinetics of steady-state turnover with continuous rotation of the shaft.

Significance Statement.

The F₀F₁-ATPase rotates its central axis by continuously changing the structure of the three β subunits upon ATP hydrolysis. Here, we reconstructed cryo-electron microscopy structures under unisite conditions that permit only a single catalytic β subunit per enzyme to bind ATP. The structures indicated that hydrolysis of the first ATP occurs at β_TP instead of β_DP, where ATP hydrolysis proceeds in the steady-state catalysis of F₀F₁. This indicates that unisite catalysis is an initial reaction that is distinguished from steady-state rotary catalysis in F₀F₁.

Introduction

Adenosine triphosphate (ATP) synthases (F₀F₁) are crucial for aerobic organisms and reside in the inner membranes of mitochondria, plasma membranes of bacteria, and thylakoid membranes of chloroplasts in plants (1–3). F₀F₁ consists of a hydrophilic F₁ domain responsible for ATP hydrolysis or synthesis and a hydrophobic F₀ domain housing proton translocation across the membranes. ATP hydrolysis/synthesis in F₁ is coupled to proton flow in F₀ through the rotation of a common shaft.

F₀F₁ from the thermophilic bacteria Geobacillus stearothermophilus is one of the best characterized ATP synthases because of its structural stability and simple subunit structure (α₃β₃γ₁ε₁δ₁a₁b₂c₁₀; Fig. 1A). In particular, single-molecule rotation experiments using this enzyme enabled direct observation of the rotation of ATP synthase, which considerably improved our understanding of the mechanochemical cycle of F₀F₁ (4–11).

Fig. 1. — Structure of nucleotide depleted wild-type F₀F₁ (*ND-wt-*F₀F₁) of *G. stearothermophilus*. Cryo-electron microscopy density map of ND-wt-F₀F₁ (state 1). (A) Side view of the overall structure of F₀F₁ as a ribbon model. All α subunits are represented as gray, and other subunits are colored in F₁ domain. (B) View of the F₀ side of F₁ domain at C terminal region. Superscript indicates the closed or open structure of the β subunit. (C) Ribbon representation of view from the F₀ side. No nucleotide was found at noncatalytic or catalytic nucleotide-binding sites.

The γε rotor in the F₁ domain (α₃β₃γ₁δ₁ε₁) is surrounded by a cylinder composed of three noncatalytic α and three catalytic β subunits arranged alternately (Fig. 1). The ε subunit modulates ATP hydrolysis activity by the structural change from a contracted to an extended form with a C-terminus helix towards α₃β₃ (12). Several experimental studies have demonstrated that the extended C-terminus of the ε subunit (ε-CT up form) strongly inhibits ATPase activity of the F₁ domain; however, the inhibition mechanism of the ε-CT up form remains elusive.

In F₁ domain, six nucleotide-binding sites are located at different interfaces between the α and β subunits; three catalytic sites are located mainly in the β subunit, while the other three sites, called noncatalytic nucleotide-binding sites, are located mainly in the α subunit. According to the binding change mechanism of ATP synthesis, the three catalytic β subunits in F₁ are in different conformations (open, loose, and tight); However, they interconvert sequentially between three different conformations as catalysis proceeds (2). Thus, at a given time, all three catalytic β subunits are in different conformations. The crystal structure of F₁ from bovine heart mitochondria demonstrated the following asymmetry of catalytic sites: β_E, which adopts an open structure with no nucleotide; β_TP, which adopts a structure containing ATP analog (AMPPNP); and β_DP, which adopts a structure containing ADP (13, 14). The C-terminal region of each β subunit interacts with the γ subunit, and the differences are a consequence of the asymmetric association of the γ subunit with the α₃β₃ cylinder. Thus, sequential interconversion between three different β subunits drives the rotation of the γ subunit with ATP hydrolysis (4).

The asymmetric structures of α₃β₃ were found in the F₁ structure of other species (15–17) and V-ATPases (16), which is another rotary ATPase that is evolutionarily and structurally related to F₀F₁. In this study, we refer to the asymmetric architecture as a “Walker structure.”

Assuming that F₀F₁ adopts the Walker structure during ATP hydrolysis, the alternating participation of β subunits within ATP hydrolysis does not require positive cooperation. Early studies showed that F₁-ATPases exhibited multiple K_m, likely due to the activation of F₁-ATPase by the binding of ATP to the α subunits (18). However, using both bulk and single-molecule experiments, one rotary mechanism was found to govern the entire range of nanomolar to millimolar ATP (19). This indicates that thermophilic F₁-ATPase obeys simple Michaelis–Menten kinetics with a single K_m value. However, previous experiments using a mitochondrial F₁ indicated strong positive cooperation between the catalytic sites. F₁-ATPase was compared under unisite and multisite catalytic conditions in which the ATP/enzyme ratio was adjusted to facilitate operation of either one or three sites (20). The multisite/unisite rate enhancement ratio of 10⁶ fold was interpreted as a reflection of the strong positive cooperation among the three catalytic sites. The unisite catalysis experiment also suggests a high affinity of unisite to ATP with a binding constant K_d of ∼10⁻¹² M, which is markedly higher than the K_m of ∼10^–5 M in bacterial F₁ (19, 21). For thermophilic F₁-ATPase, similar ATP hydrolysis in a single catalytic site has been reported (21–23). However, the structural basis of the high affinity for ATP observed in unisite catalysis and whether unisite catalysis reflects steady-state activity (multisite catalysis) remains elusive.

In this study, we constructed a mutant F₀F₁ with a C-terminus-truncated ε subunit and showed that it does not undergo ε inhibition. The structure of this mutant F₀F₁ was determined using cryo-electron microscopy (cryo-EM) and compared with that of wild-type (wt) F₀F₁ to understand the structural basis of ε inhibition. Furthermore, we determined the structure of the mutant F₀F₁ under unisite catalytic conditions to capture the structure after ATP hydrolysis of the β subunit.

Results

Cryo grid preparation of wt-F₀F₁ and ATPase-active ΔεCT-F₀F₁

In this study, we used purified F₀F₁ from G. stearothermophilus expressed in Escherichia coli (24). The ATPase activity of wt-F₀F₁ is ∼10 s⁻¹ (Supplementary Fig. S1 and Supplementary Table S1), which is considerably lower than that of F₁ (α₃β₃γ, ∼70 s⁻¹) (25). Several studies have indicated that ATP hydrolysis by G. stearothermophilus F₀F₁ is significantly inhibited when the extended C-terminal region of the ε subunit penetrates the α₃β₃ cavity (12, 26–28). The initial ATPase activity of wt-F₀F₁ was very low because of the initial lag (Supplementary Fig. S1). To obtain ATPase-active F₀F₁, we constructed a mutant F₀F₁ with a C-terminal-truncated ε subunit (ΔεCT-F₀F₁). The prepared ΔεCT-F₀F₁ was subjected to dialysis in phosphate-buffered saline to deplete the bound nucleotide, as described in the Materials and Methods section. For the nucleotide-depleted thermophilic F₀F₁ (ND-ΔεCT-F₀F₁), the lag time was shorter than that of wt-F₀F₁, and the ATPase activity at 1,000 s after the start of reaction was 130 s⁻¹ (Supplementary Table S1), which was comparable to the ATPase activity of wt-F₁ without the ε subunit (5, 25).

First, we prepared a cryo grid of nucleotide-depleted wt-F₀F₁ (ND-wt-F₀F₁) in the absence or presence of 4 mM ATPγS. In addition, the cryo grid of ΔεCT-F₀F₁ under unisite conditions (molar ratio of enzyme to ATP is approximately 1:4.) was used for structural analysis. These cryo grids were subjected to cryo-EM image acquisition using a Titan Krios (Thermo Fisher Scientific) equipped with a K3 direct electron detector.

Structure of ND-wt-F₀F₁

Flowcharts showing the image acquisition and reconstitution of the 3D structure of nucleotide-depleted wt-F₀F₁ (ND-wt-F₀F₁) are summarized in Supplementary Fig. S2. We reconstructed three rotational states (state 1: 3.1 Å, state 2: 3.0 Å, and state 3: 3.7 Å resolutions) from the single-particle images of ND-wt-F₀F₁. The structure of ND-wt-F₁ domain is similar to the structure of F₁ domain in wt-F₀F₁ reported in a study by Guo et al.(6N2Y(12)), which adopted the following structures: “open in β_E without nucleotide,” “closed in β_TP containing ADP,” and “open in β_DP without nucleotide.” In fact, the three β subunits in ND-wt-F₁ are similar to their counterparts in wt-F₀F₁ (Supplementary Fig. S3). In the structure of ND-wt-F₀F₁, the C-terminal helix of the ε subunit (ε-CT) also adopts the up form and penetrates the cavity of the α₃β₃ cylinder (Fig. 1B). In contrast to the previous structure of wt-F₀F₁ (12), no nucleotide density was observed at the noncatalytic nucleotide-binding site in the three α subunits or the catalytic site in β_TP (Figs. 1C and 3A). Instead, a density likely corresponding to a phosphate was observed at the catalytic site of β_TP (Fig. 3A and Supplementary Fig. S4A). This phosphate was possibly derived from the phosphate buffer used for nucleotide depletion.

Fig. 3. — Structure of catalytic sites of wild-type and mutated F₀F₁ under different condition. Magnified views of the three catalytic sites (β_E, β_TP, and β_DP) in each structure are shown as follows: (A) *ND-wt*-F₀F₁, (B) wt-F₀F₁ε up form with ATPγS, (C) wt-F₀F₁ε down form with ATPγS, (D) *ND-ΔCT*-F₀F₁, and (E) *US-ΔCT*-F₀F₁. Bound nucleotides and Mg ions are highlighted as stick and spherical representations, respectively. Scale bar is 5 Å.

Therefore, the structure indicates that the closed structure of β_TP is stabilized by the penetration of ε-CT into the cavity between α_DP and β_TP, which is independent of nucleotide binding to β_TP.

Structure of wt-F₀F₁ exposed to 4 mM ATPγS

Previous studies have indicated that the binding of ATP to an isolated ε subunit induces a conformational change in ε-CT from the “up” to “down” form (25, 29). The ATPase activity of wt-F₀F₁ gradually accelerated in the presence of 4 mM ATP (Supplementary Fig. S1 and Supplementary Table S1). In this study, we determined the cryo-EM structure of F₀F₁ exposed to 4 mM ATPγS, which is a slow hydrolyzable ATP analog (30), to prove the conformational change of the ε subunit in the complex. After focused 3D classification using a mask covering the F₁ domain (Supplementary Fig. S5B), we obtained the following two structures: F₁ with the up form of ε subunit from 526,524 particles at 2.6 Å resolution and F₁ with the down form of ε subunit from 39,991 particles at 3.3 Å resolution (Fig. 2 and Supplementary Fig. S5B). In the down form of ε subunit in the complex, the density of bound ATPγS was identified with similar coordination of surrounding amino acid residues to that in the monomeric ε subunit (Supplementary Fig. S6) (31). F₀F₁ with the down form of ε subunit adopted the following canonical Walker structures: “open in β_E with ATPγS,” “closed in β_TP with ATPγS,” and “closed in β_DP with ADP ” (Figs. 2B and 3C). ATPγS molecules were also identified at the noncatalytic sites of the three α subunits (Fig. 2, lower). In contrast, the structure of F₀F₁ with the extended ε-CT was similar to the ND-wt-F₀F₁ structure, except the nucleotide occupancy of “open in β_E with ATPγS,” “closed in β_TP with ADP,” and “open in β_DP with ADP” (Fig. 3B). These results indicate that the extended ε-CT hampers the conformational change of the F₁ domain from the inhibitory conformation with “open in β_E,” “closed in β_TP,” and “open in β_DP” to the Walker structure with “open in β_E,” “closed in β_TP,” and “closed in β_DP” during ATP hydrolysis.

Fig. 2. — Structure of wild-type F₀F₁ (*wt-*F₀F₁) exposed to 4 mM ATPγS. Structures of wt-F₀F₁ exposed to 4 mM ATPγS with the retracted (A) or extended (B) ε-subunit viewed from vertical side (upper) and from F₁ side (lower). The bound nucleotides are represented as color spheres. All α subunits are represented in gray, and all β subunits are colored.

Structure of ΔεCT-F₀F₁ under unisite catalysis conditions

To capture the structure of the unisite catalysis of F₀F₁, 15 µM of ND-ΔεCT-F₀F₁ was mixed with 4 µM of the ATP-containing regeneration system described in the Materials and Methods section. The mixture was incubated for 120 s at 25°C, then loaded onto a holey grid, and subjected to flash freezing.

We obtained multiple F₀F₁ structures with and without nucleotides (ND-ΔεCT-F₀F₁) in the β_TP using 418,497 selected particle images of ΔεCT-F₀F₁ (Supplementary Fig. S7B). For ND-ΔεCT-F₀F₁, the structures of two rotational states without nucleotides were obtained at the following resolutions: state 1 at 3.6 Å and state 2 at 3.4 Å, with the γ subunit positions differing by 120° in each state. The structure of state 3 without nucleotides was not identified because of the small number of particles. Further, we obtained the F₁ parts of states 1 and 2 at resolutions of 3.4 Å and 3.3 Å, respectively, using focused refinement of the F₁ domain (Supplementary Fig. S7B).

The structure of ND-ΔεCT-F₀F₁ significantly differs from the structure of ND-wt-F₀F₁ in which three catalytic β subunits adopt “open in β_E,” “closed in β_TP,” and “open in β_DP” (Fig. 1C). In contrast, all three β subunits in ND-ΔεCT-F₀F₁ adopt an almost identical structure of open conformations as follows: “open in β_E,” “open in β_TP,” and “open in β_DP” (Fig. 4A). There was no nucleotide bound to the three open β subunits of ND-ΔεCT-F₀F₁, but a density corresponding to phosphate was observed in the β_DP (Fig. 3D and Supplementary Fig. S4D). This phosphate may be derived from the phosphate buffer used for nucleotide depletion. Our findings suggest that nucleotide depletion from the F₁ domain causes all three β subunits to adopt an open conformation. The interaction of the γ subunit with the three open β subunits are shown in Supplementary Fig. S8. The C termini region of closed β_TP is in close proximity to the coiled coil of the γ subunit, while the open β_TP is in close proximity to the globular domain of the γ subunit (Supplementary Fig. S8C).

Fig. 4. — Comparison of the structure of nucleotide-depleted *ΔεCT*-F₀F₁ with that of US-*ΔεCT*-F₀F₁. Cryo-electron microscopy structures of *ND-ΔεCT*-F₀F₁ (A) and *ΔεCT*-F₀F₁ under unisite conditions (B). The slice view of the F₁ domain is represented in the left panel, and the main chain of each β subunit is lined. *Right panels*, each β subunits (β_TP, β_DP, and β_E) of ND-*ΔεCT*-F₀F₁ or US-*ΔεCT*-F₀F₁ is represented as a colored chain. (C) Comparison of the β subunits between ND-*ΔεCT*-F₀F₁ and US-*ΔεCT*-F₀F₁. Each β subunit (β_TP, β_DP, and β_E) of ND-*ΔεCT*-F₀F₁(colored chain) or *US-ΔεCT*-F₀F₁ (black chain) is superimposed on the β-barrel domain (1 to 80 amino acids).

In other words, the penetration of extended ε-CT into the α₃β₃ cavity does not result in the open conformation of β_DP but rather forces open β_TP without bound nucleotides to the closed conformation (cf. Figs. 1C and 4A).

For ND-ΔεCT-F₀F₁ containing a nucleotide in β_TP, we determined the structures of the following three states: state 1 at 3.2 Å, state 2 at 3.4 Å, and state 3 at 4.0 Å. For state 1, another subclass with nearly identical structure (state 1) was isolated (Supplementary Fig. S6C). Upon focused refinement using an F₁ mask, the structure of the F₁ domain for each state was obtained (Supplementary Fig. S6B, lower). We refer to the ΔεCT-F₀F₁ structure containing a nucleotide at the β_TP as UniSite-ΔεCT-F₀F₁ (US-ΔεCT-F₀F₁). The structure of US-ΔεCT-F₀F₁ was different from that of ND-ΔεCT-F₀F₁. The structures of two β subunits, β_E and β_DP, adopted an open form without nucleotides, whereas β_TP adopted a closed form containing nucleotide density (Fig. 3E). These findings indicate that the structure of US-ΔεCT-F₀F₁ is very similar to that of ND-wt-F₀F₁. Specifically, F₀F₁ in both structures adopt “open in β_E,” “closed in β_TP,” and “open in β_DP.” In addition, the relative position of the γ subunit to α₃β₃ was completely analogous in the two structures (Supplementary Fig. S9). The structure of US-ΔεCT-F₀F₁ revealed a nucleotide density due to ADP at the catalytic site in β_TP, indicating that ATP bound to β_TP was already hydrolyzed (Figs. 3E and 4B).

Furthermore, the position of the γ subunit of US-ΔεCT-F₀F₁ relative to α₃β₃ is slightly different from that of ND-ΔεCT-F₀F₁. Superimposition of the two structures with α₃β₃ shows ∼ 7° rotation of the γ subunit in the hydrolysis direction (Supplementary Fig. S8A). Assuming that the first ATP binds to β_TP, the rotation of the γ subunit driven by unisite catalysis is faint compared to the 120° rotation of the γ subunit upon binding and hydrolysis of one ATP molecule under ATP-saturated conditions (multisite conditions).

Discussion

In this study, we reveal that substoichiometric ATP is hydrolyzed at the β_TP of ND-ΔεCT-F₀F₁ under unisite catalysis conditions, where a single catalytic site per enzyme molecule binds ATP.

All three β subunits in ND-ΔεCT-F₀F₁ adopted an open form, implying that the first ATP can bind to any of the β subunits. Assuming that the first ATP binds to β_TP, the conformation change from the open to the closed form of β_TP occurs without the 120° rotation of γ subunit (Fig. 5A), which is consistent with previous studies suggesting unisite catalysis without rotation of γ subunit in E. coli F₁ (32). In the wt-F₀F₁ structure, ADP is bound only to β_TP, and both β_DP and β_E are in the open form without nucleotides, which suggests that β_TP can easily change to the closed form upon ATP binding, particularly when compared to β_DP and β_E. β_TP bound with ATP immediately changes to the closed form due to the zippering motion by the bound ATP, which likely increases the affinity of β_TP for ATP. This is consistent with the high affinity for ATP reported in unisite catalysis experiments (20, 22, 33). Alternatively, it is also possible that the first ATP bound to β_E and the γ subunit rotated 120° with a structural change of the ATP-bound β_E to β_TP, resulting in the structure of US-ΔεCT-F₀F₁. Although both catalytic pathways (Fig. 5A and B) exhibit high possibilities for unisite catalysis, it is more likely that the first ATP binds to β_TP to explain the high affinity of the unisite for ATP.

Fig. 5. — Schematic representation of possible catalytic pathways for unisite catalysis at β_TP and promotion of ATP hydrolysis by chase ATP. Three β subunits (β_TP, β_DP, and β_E) and γ subunits are colored in blue, blue–purple, magenta and yellow, respectively. All α subunits are represented as gray. (A) First ATP binds to β_TP in the nucleotide-depleted (ND) structure where the three β subunits are open. Then, changes to with hydrolysis of bound ATP. The conformation change of β_TP induces a 7° rotation of the γ subunit. The phosphate is released spontaneously, resulting in the unisite structure (US). (B) The first ATP binds to β_E in the ND structure, followed by structural changes of three β subunits with 120° rotation of the γ subunit. (C) The bound ATP on β_TP is slowly hydrolyzed by coordination of catalytic amino acid residues not suitable for hydrolysis of ATP. Binding of the second ATP to β_E causes 120° rotation of γ subunit and structural transition of the each of the following β subunits: β_E to β_TP, β_TP to β_DP, and β_DP to β_E. The β subunit to which the first ATP binds becomes β_DP, and the hydrolysis of ATP is accelerated.

Inline graphic — Schematic representation of possible catalytic pathways for unisite catalysis at β_TP and promotion of ATP hydrolysis by chase ATP. Three β subunits (β_TP, β_DP, and β_E) and γ subunits are colored in blue, blue–purple, magenta and yellow, respectively. All α subunits are represented as gray. (A) First ATP binds to β_TP in the nucleotide-depleted (ND) structure where the three β subunits are open. Then, changes to with hydrolysis of bound ATP. The conformation change of β_TP induces a 7° rotation of the γ subunit. The phosphate is released spontaneously, resulting in the unisite structure (US). (B) The first ATP binds to β_E in the ND structure, followed by structural changes of three β subunits with 120° rotation of the γ subunit. (C) The bound ATP on β_TP is slowly hydrolyzed by coordination of catalytic amino acid residues not suitable for hydrolysis of ATP. Binding of the second ATP to β_E causes 120° rotation of γ subunit and structural transition of the each of the following β subunits: β_E to β_TP, β_TP to β_DP, and β_DP to β_E. The β subunit to which the first ATP binds becomes β_DP, and the hydrolysis of ATP is accelerated.

Both β_DP and β_E retain open structures after unisite catalysis at β_TP; therefore, the next catalytic event is ATP binding to either β_DP or β_E. During ATP hydrolysis by F₀F₁, each conformational change from β_E to β_TP, β_TP to β_DP, and β_DP to β_E occurs with 120° rotation of the γ subunit. Assuming that ATP then binds to β_E and induces a conformational change of each β subunit with 120° rotation of the γ subunit, F₀F₁ adopts the following canonical Walker structure observed in most F₁ structures: “open in β_E,” “closed in β_TP with the second ATP,” and “closed in β_DP with the ADP from first ATP” (Fig. 5C). Several studies using various F₁-ATPases have reported that hydrolysis of ATP bound to the first catalytic site is markedly accelerated by the addition of an excess ATP (cold chase experiment) (20, 21, 35). In Walker structures, β_TP and β_DP adopt very similar closed conformations; however, their catalytic sites are not equivalent (Supplementary Fig. S10). Although these differences are relatively small, they have a marked effect on catalysis. ATP in β_DP is immediately hydrolyzed, whereas hydrolysis of ATP in β_TP proceeds slowly (14, 33). Therefore, most crystal structures of F₁ contain ATP analogs in β_TP and ADP in β_DP. Upon binding of ATP to β_E by the addition of excess ATP, the conformational change of β_TP to β_DP promoted the hydrolysis of ATP at the catalytic site (Fig. 5C).

The F₀F₁ structures exposed to 4 mM ATPγS provided direct evidence of conformational changes in the ε subunit by ATP binding in the holo-F₀F₁ complex. The retracted ε subunit in the complex contains ATP with the similar coordination as the monomeric ε subunit (Supplementary Fig. S5) (31). In another wt-F₀F₁ structure exposed to 4 mM ATPγS concentration, the ε-inhibited structure, which adopted “open in β_E,” “closed in β_TP,” and “open in β_DP,” was maintained, although all nucleotide binding sites were occupied with ADP or ATPγS (Figs. 2 and 3B). This indicates that the inhibited wt-F₀F₁ structure by the extended ε subunit is not activated by ATP binding to β_E or β_DP (Fig. 2A), and that the conformational change of the ε subunit by ATP binding is crucial for the activation of wt-F₀F₁ for ATP hydrolysis activity (Fig. 2B).

In this study, we determined the multiple structures of F₀F₁ ATPase during catalysis by structural analysis using cryo-EM, which allows the capture of states inaccessible to crystallization. The techniques and approaches used in this study can potentially assist the elucidation of the detailed reaction and regulatory mechanisms of other enzymes.

Materials and methods

Protein purification

Wt-F₀F₁-ATP synthase from G. stearothermophilus was purified from E. coli DK 8 strain containing an expression vector (pTR19-ASDS) for F₀F₁, as described previously (24). The expression vector for ΔεCT-F₀F₁, which lacks the C-terminal of ε subunit (83 to 133 amino acids), was constructed from pTR19-ASDS. Transformed E. coli cells were grown in 2 × YT medium at 37°C for 16 h before harvesting by centrifugation at 5000 × g. The cell pellet was suspended in lysis buffer (50 mM Tris–Cl pH 8.0, 5 mM MgCl₂, and 10% [w/v] glycerol), and the cell membranes were collected by ultracentrifugation at 35,000 rpm for 20 min and solubilized by mixing in solubilization buffer (50 mM Tris–Cl, 5 mM MgCl₂, 10% [w/v] glycerol, and 2% n-dodecyl-d-maltoside [DDM]) at 4°C for 3 h. The supernatant was then applied to a Ni-NTA column. For bound nucleotide removal, the eluted fractions containing F₀F₁ were dialyzed against 200 mM sodium phosphate (pH 8.0), 10 mM EDTA, and 0.03% DDM overnight at 25°C and the dialysis buffer was changed thrice. The dialyzed F₀F₁ fractions were concentrated using ultrafiltration with an Amicon filter (100 KDa cut-off, Amicon corp.) and loaded onto a Superose 6 Increase 10/300 column (Cytiva) equilibrated with gel permeation buffer (20 mM Tris–HCl pH 8.0, 150 mM NaCl, and 0.03% DDM). The peak fractions (6 to 9 mg/mL) were used for the cryo grid preparation or ATPase assay.

Grid preparation

For cryo grid preparation, Quanfifoil R1.2/1.3 Mo grids were glow-discharged for 1 min using an Ion Bombarder (Vacuum Device). Prior to blotting, 3 µL of the samples were placed on a cryo grid and incubated for 15 min. Further, 2.6 to 3.5 µL F₀F₁ was loaded onto the grid and blotted for 3 to 10 s with a blot force of 10, drain time of 0.5 s, and 100% humidity using a FEI Vitrobot (ThermoFisher). The blotted grid was then plunged into a liquid ethane. For the unisite catalysis condition, 1 µL reaction buffer (0.2 M Tris–Cl pH 8.0, 40 µM ATP, 40 mM PEP, 1 M KCl, and 5 mg/mL pyruvate kinase) was added to 9 µL of the sample. The mixtures were incubated for 60 s at 25°C, followed by blotting and vitrification.

Cryo-EM imaging under ATPγS and US conditions was performed using a Titan Krios (FEI/Thermo Fisher Scientific) operating at 300 kV acceleration voltage and equipped with an electron detector K3 (Gatan) in electron counting mode (CDS). Cryo-EM imaging under wt-ND conditions was performed in CDS using CRYOARM 300 (JEOL) operating at 300 kV and an electron detector K3 (Gatan).

Data collection was performed using SerialEM software with a calibrated magnification of 0.88 Å pixel^–1 for the ATPγS and ΔεCT-US conditions and a calibrated magnification of 1.01 Å pixel^–1 for the wt-ND condition. Under wt-ND and ATPγS conditions, data were collected at an electron dose of 50.0 e^–/Å² for an exposure time of 5 s, and under unisite conditions, data were collected at an electron dose of 60.0 e^–/Å² for an exposure time of 6 s. The defocus range was 0.8 to 2.0 µm, and data were collected at 50 frames for the wt-ND condition, 55 frames for the ATPγS condition, and 59 frames for the unisite condition.

Image processing

The details of the image-processing procedure for each condition are described in Supplementary Figs. 2, 4, and 6. Image analysis was performed using RELION 4.0β (36) and cryoSPARC v3.2 (37). The file format conversion between RELION and cryoSPARC was performed using the script csparc2star.py in Pyem. The beam-induced drift was corrected using MotionCor2 (38), and the CTF was estimated using CTFFIND 4.1 (39). We analyzed 7,329 movies for the unisite condition, 9,625 movies for the wt-ND condition, and 17,261 movies for the ATPγS condition. Particle picking was performed using Topaz (40). Good particles were selected by 2D classification and trained using 4,000 particles. Autopicking using the trained topaz model yielded 499,788 particles for the unisite condition, 1,381,269 particles for the wt-ND condition, and 1,020,321 particles for the ATPγS condition. Particles picked by Topaz were subjected to 2D classification for further selection of good particles. Then heterogeneous refinements using cryoSPARC were performed to eliminate junk particles. Further, we selected 418,497 particles for unisite conditions, 622,109 particles for wt-ND conditions, and 912,931 particles for ATPγS conditions. Heterogeneous refinement was then used to classify F₀F₁ into multiple conformational states. These particles were re-extracted at full pixel size and subjected to repeated 3D auto-refinement, CTF refinement, and Bayesian polishing (36). The structures of the three rotational states, including their subclasses, were obtained at 3 to 4.5 Å resolution. All obtained classes of F₀F₁ were subjected to focused refinement on the F₁ part to reduce the resolution loss due to the relative motion of the F₀ part with F₁. In the analysis of the ATPγS dataset to determine a small percentage of ε-retracted structures, the class of F₀F₁ of all states were added together to increase the number of particles, and focused refinement was performed on the F₁ part. The obtained F₁ structure was subjected to focused 3D classification by masking with γε-β_DP to detect structural changes between the ε and β subunits. The resolution was estimated using the gold standard Fourier shell correlation (FSC) = 0.143 criterion.

Model building and refinement

We used Phenix real-space refinement (41), ISOLDE (42), and COOT (41) for the atomic model building. The epsilon-extended F₁ model was built using PDB 6N2Y as the initial model, and the epsilon-retracted subunit was built using PDB 2E5Y. The initial model, which was a rigid body fitted to the density map by UCSF ChimeraX (43), was first refined by Phenix real-space refinement. Residues that did not fit correctly into the map were manually placed using the COOT and ISOLDE. Refinement and manual modification were repeated until the model parameters were improved. Lastly, the refinement model was evaluated using MolProbity (44) and EMRinger (45).

Supplementary Material

pgac116_Supplemental_File

Click here for additional data file.^{(5.9MB, pdf)}

ACKNOWLEDGEMENTS

We are grateful to all the members of the Yokoyama Lab for their continuous support and technical assistance. We also thank Prof. Allison for his enthusiastic discussion on unisite catalysis long time ago.

Notes

Competing Interest: The authors declare no competing interests.

Contributor Information

Atsuki Nakano, Department of Molecular Biosciences, Kyoto Sangyo University, Kamigamo-Motoyama, Kita-ku, Kyoto 603-8555, Japan.

Jun-ichi Kishikawa, Department of Molecular Biosciences, Kyoto Sangyo University, Kamigamo-Motoyama, Kita-ku, Kyoto 603-8555, Japan; Institute for Protein Research, Osaka University, 3-2 Yamadaoka, Suita, Osaka 565-0871, Japan.

Atsuko Nakanishi, Department of Molecular Biosciences, Kyoto Sangyo University, Kamigamo-Motoyama, Kita-ku, Kyoto 603-8555, Japan; Research Center for Ultra-High Voltage Electron Microscopy, Osaka University, 7-1 Mihogaoka, Ibaraki, Osaka 567-0047, Japan.

Kaoru Mitsuoka, Research Center for Ultra-High Voltage Electron Microscopy, Osaka University, 7-1 Mihogaoka, Ibaraki, Osaka 567-0047, Japan.

Ken Yokoyama, Department of Molecular Biosciences, Kyoto Sangyo University, Kamigamo-Motoyama, Kita-ku, Kyoto 603-8555, Japan.

Funding

Our research was supported by Grant-in-Aid for Scientific Research (JSPS KAKENHI) Grant Number 20H03231 to K.Y., 20K06514 to J.K., and Grant-in-Aid for JSPS Fellows Grant Number 20J00162 to A.N., and Takeda Science foundation to K.Y. Our research was also supported by Platform Project for Supporting Drug Discovery and Life Science Research (Basis for Supporting Innovative Drug Discovery and Life Science Research [BINDS]) from AMED under Grant Number JP17am0101001 (support number 1312), Grants-in-Aid from “Nanotechnology Platform” of the Ministry of Education, Culture, Sports, Science and Technology (MEXT) to K.M. (Project Number. 12024046), and the Research Program for Next Generation Young Scientists of “Five-star Alliance” in “NJRC Mater. & Dev.” under Grant Number 20215008 to A.N..

Authors’ contributions

K.Y., J.K., A. Nakanishi, and A. Nakano designed, performed, and analyzed the experiments; A. Nakano analyzed the data and contributed to the preparation of the samples; J.K. and K.M. provided technical support and conceptual advice; and K.Y. designed and supervised the experiments and wrote the manuscript. All authors discussed the results and commented on the manuscript.

Data availability

Data is available in the manuscript and supplementary materials. Cryo-EM density maps (.mrc files) and atomic models (.pdb files) obtained in this study were deposited to EMDB and PDB. The accession codes (PDBID and EMDBID) 7XKH, 7XKQ, 7XKR, 7XKO, 7XKP, 33,251, 33,252, 33,253, 33,264, 33,265, 33,266, 33,267, 33,268, 33,277, 33,258, 33,278, 33,259, 33,279, 33,260, 33,280, 33,261, 33,281, 33,269, 33,282, 33,262, 33,283, and 33,263 are summarized in Supplementary Tables S2–S5. The data that support the findings of this study are available from PDB and EMDB.

References

1. Kühlbrandt W. 2019. Structure and mechanisms of F-type ATP synthases. Annu Rev Biochem. 88:515–549. [DOI] [PubMed] [Google Scholar]
2. Boyer PD. 1997. The ATP synthase—a splendid molecular machine. Annu Rev Biochem. 66:717–749. [DOI] [PubMed] [Google Scholar]
3. Yoshida M, Muneyuki E, Hisabori T. 2001. ATP synthase—a marvellous rotary engine of the cell. Nat Rev Mol Cell Biol. 2:669–677. [DOI] [PubMed] [Google Scholar]
4. Noji H, Yasuda R, Yoshida M, Kinosita K Jr. 1997. Direct observation of the rotation of F₁-ATPase. Nature. 386:299–302. [DOI] [PubMed] [Google Scholar]
5. Yasuda R, Noji H, Yoshida M, Kinosita K Jr., Itoh H. 2001. Resolution of distinct rotational substeps by submillisecond kinetic analysis of F₁-ATPase. Nature. 410:898–904. [DOI] [PubMed] [Google Scholar]
6. Imamura H, et al. 2003. Evidence for rotation of V1-ATPase. Proc Natl Acad Sci U S A. 100:2312–2315. [DOI] [PMC free article] [PubMed] [Google Scholar]
7. Yanagisawa S, Frasch WD. 2021. pH-dependent 11°F₁F₀ ATP synthase sub-steps reveal insight into the F₀ torque generating mechanism. eLife. 10:e70016. [DOI] [PMC free article] [PubMed] [Google Scholar]
8. Spetzler D, et al. 2006. Microsecond time scale rotation measurements of single F₁-ATPase molecules. Biochemistry. 45:3117–3124. [DOI] [PMC free article] [PubMed] [Google Scholar]
9. Adachi K, et al. 2007. Coupling of rotation and catalysis in F(1)-ATPase revealed by single-molecule imaging and manipulation. Cell. 130:309–321. [DOI] [PubMed] [Google Scholar]
10. Adachi K, Oiwa K, Yoshida M, Nishizaka T, Kinosita K Jr. 2012. Controlled rotation of the F₁-ATPase reveals differential and continuous binding changes for ATP synthesis. Nat Commun. 3:1022. [DOI] [PMC free article] [PubMed] [Google Scholar]
11. Diez M, et al. 2004. Proton-powered subunit rotation in single membrane-bound F₀F₁-ATP synthase. Nat Struct Mol Biol. 11:135–141. [DOI] [PubMed] [Google Scholar]
12. Guo H, Suzuki T, Rubinstein JL. 2019. Structure of a bacterial ATP synthase. eLife. 8:e43128. [DOI] [PMC free article] [PubMed] [Google Scholar]
13. Abrahams JP, Leslie AG, Lutter R, Walker JE. 1994. Structure at 2.8-A resolution of F₁-ATPase from bovine heart mitochondria. Nature. 370:621–628. [DOI] [PubMed] [Google Scholar]
14. Menz RI, Leslie AG, Walker JE. 2001. Structure of bovine mitochondrial F(1)-ATPase with nucleotide bound to all three catalytic sites: implications for the mechanism of rotary catalysis. Cell. 106(3):331–341. [DOI] [PubMed] [Google Scholar]
15. Suzuki K, et al. 2016. Crystal structures of the ATP-binding and ADP-release dwells of the V(1) rotary motor. Nat Commun. 7:13235. [DOI] [PMC free article] [PubMed] [Google Scholar]
16. Arai S, et al. 2013. Rotation mechanism of Enterococcus hirae V1-ATPase based on asymmetric crystal structures. Nature. 493:703–707. [DOI] [PubMed] [Google Scholar]
17. Nakanishi A, Kishikawa JI, Tamakoshi M, Mitsuoka K, Yokoyama K. 2018. Cryo EM structure of intact rotary H(+)-ATPase/synthase from Thermus thermophilus. Nat Commun. 9:89. [DOI] [PMC free article] [PubMed] [Google Scholar]
18. Jault JM, Allison WS. 1994. Hysteretic inhibition of the bovine heart mitochondrial F₁-ATPase is due to saturation of noncatalytic sites with ADP which blocks activation of the enzyme by ATP. J Biol Chem. 269:319–325. [PubMed] [Google Scholar]
19. Sakaki N, et al. 2005. One rotary mechanism for F₁-ATPase over ATP concentrations from millimolar down to nanomolar. Biophys J. 88:2047–2056. [DOI] [PMC free article] [PubMed] [Google Scholar]
20. Grubmeyer C, Cross RL, Penefsky HS. 1982. Mechanism of ATP hydrolysis by beef heart mitochondrial ATPase. Rate constants for elementary steps in catalysis at a single site. J Biol Chem. 257:12092–12100. [PubMed] [Google Scholar]
21. Yokoyama K, Hisabori T, Yoshida M. 1989. The reconstituted α₃β₃δ complex of the thermostable F₁-ATPase. J Biol Chem. 264:21837–21841. [PubMed] [Google Scholar]
22. Konishi J, Yohda M, Hashimoto T, Yoshida M. 1987. Single site catalysis of the F₁-ATPase from Saccharomyces cerevisiae and the effect of inorganic phosphate on it. J Biochem. 102:273–279. [DOI] [PubMed] [Google Scholar]
23. Hisabori T, et al. 1992. Single site hydrolysis of 2′,3′-O-(2,4,6-trinitrophenyl)-ATP by the F₁-ATPase from thermophilic bacterium PS₃ is accelerated by the chase-addition of excess ATP. J Biol Chem. 267:4551–4556. [PubMed] [Google Scholar]
24. Suzuki T, Ueno H, Mitome N, Suzuki J, Yoshida M. 2002. F(0) of ATP synthase is a rotary proton channel. Obligatory coupling of proton translocation with rotation of c-subunit ring. J Biol Chem. 277:13281–13285. [DOI] [PubMed] [Google Scholar]
25. Kato S, Yoshida M, Kato-Yamada Y. 2007. Role of the epsilon subunit of thermophilic F₁-ATPase as a sensor for ATP. J Biol Chem. 282:37618–37623. [DOI] [PubMed] [Google Scholar]
26. Shirakihara Y, et al. 2015. Structure of a thermophilic F₁-ATPase inhibited by an ε-subunit: deeper insight into the ε-inhibition mechanism. FEBS J. 282:2895–2913. [DOI] [PubMed] [Google Scholar]
27. Suzuki T, et al. 2003. F₀F₁-ATPase/synthase is geared to the synthesis mode by conformational rearrangement of epsilon subunit in response to proton motive force and ADP/ATP balance. J Biol Chem. 278:46840–46846. [DOI] [PubMed] [Google Scholar]
28. Tsunoda SP, et al. 2001. Large conformational changes of the epsilon subunit in the bacterial F₁F₀-ATP synthase provide a ratchet action to regulate this rotary motor enzyme. Proc Natl Acad Sci U S A. 98:6560–6564. [DOI] [PMC free article] [PubMed] [Google Scholar]
29. Imamura H, et al. 2009. Visualization of ATP levels inside single living cells with fluorescence resonance energy transfer-based genetically encoded indicators. Proc Natl Acad Sci U S A. 106:15651–15656. [DOI] [PMC free article] [PubMed] [Google Scholar]
30. Imamura H, et al. 2005. Rotation scheme of V₁-motor is different from that of F₁-motor. Proc Natl Acad Sci U S A. 102:17929–17933. [DOI] [PMC free article] [PubMed] [Google Scholar]
31. Yagi H, et al. 2007. Structures of the thermophilic F₁-ATPase epsilon subunit suggesting ATP-regulated arm motion of its C-terminal domain in F₁. Proc Natl Acad Sci U S A. 104:11233–11238. [DOI] [PMC free article] [PubMed] [Google Scholar]
32. García JJ, Capaldi RA. 1998. Unisite catalysis without rotation of the γ-ε domain inEscherichia coli F₁-ATPase. J Biol Chem. 273:15940–15945. [DOI] [PubMed] [Google Scholar]
33. Al-Shawi MK, Senior AE. 1988. Complete kinetic and thermodynamic characterization of the unisite catalytic pathway of Escherichia coli F₁-ATPase. Comparison with mitochondrial F₁-ATPase and application to the study of mutant enzymes. J Biol Chem. 263:19640–19648. [PubMed] [Google Scholar]
34. Senior AE. 1992. Catalytic sites of Escherichia coli F₁-ATPase. J Bioenerg Biomembr. 24:479–484. [DOI] [PubMed] [Google Scholar]
35. Amano T, Tozawa K, Yoshida M, Murakami H. 1994. Spatial precision of a catalytic carboxylate of F₁-ATPase beta subunit probed by introducing different carboxylate-containing side chains. FEBS Lett. 348:93–98. [DOI] [PubMed] [Google Scholar]
36. Scheres SH. 2012. RELION: implementation of a Bayesian approach to cryo-EM structure determination. J Struct Biol. 180:519–530. [DOI] [PMC free article] [PubMed] [Google Scholar]
37. Punjani A, Rubinstein JL, Fleet DJ, Brubaker MA. 2017. cryoSPARC: algorithms for rapid unsupervised cryo-EM structure determination. Nat Methods. 14:290–296. [DOI] [PubMed] [Google Scholar]
38. Zheng SQ, et al. 2017. MotionCor2: anisotropic correction of beam-induced motion for improved cryo-electron microscopy. Nat Methods. 14:331–332. [DOI] [PMC free article] [PubMed] [Google Scholar]
39. Rohou A, Grigorieff N. 2015. CTFFIND4: fast and accurate defocus estimation from electron micrographs. J Struct Biol. 192:216–221. [DOI] [PMC free article] [PubMed] [Google Scholar]
40. Bepler T, et al. 2019. Positive-unlabeled convolutional neural networks for particle picking in cryo-electron micrographs. Nat Methods. 16:1153–1160. [DOI] [PMC free article] [PubMed] [Google Scholar]
41. Liebschner D, et al. 2019. Macromolecular structure determination using X-rays, neutrons and electrons: recent developments in Phenix. Acta Crystallogr D Struct Biol. 75:861–877. [DOI] [PMC free article] [PubMed] [Google Scholar]
42. Croll TI. 2018. Isolde: a physically realistic environment for model building into low-resolution electron-density maps. Acta Crystallogr D Struct Biol. 74:519–530. [DOI] [PMC free article] [PubMed] [Google Scholar]
43. Pettersen EF, et al. 2021. UCSF ChimeraX: Structure visualization for researchers, educators, and developers. Protein Sci. 30:70–82. [DOI] [PMC free article] [PubMed] [Google Scholar]
44. Barad BA, et al. 2015. EMRinger: side chain-directed model and map validation for 3D cryo-electron microscopy. Nat Methods. 12:943–946. [DOI] [PMC free article] [PubMed] [Google Scholar]
45. Chen VB, et al. 2010. MolProbity: all-atom structure validation for macromolecular crystallography. Acta Crystallogr D Biol Crystallogr. 66:12–21. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

pgac116_Supplemental_File

Click here for additional data file.^{(5.9MB, pdf)}

Data Availability Statement

[bib1] 1. Kühlbrandt W. 2019. Structure and mechanisms of F-type ATP synthases. Annu Rev Biochem. 88:515–549. [DOI] [PubMed] [Google Scholar]

[bib2] 2. Boyer PD. 1997. The ATP synthase—a splendid molecular machine. Annu Rev Biochem. 66:717–749. [DOI] [PubMed] [Google Scholar]

[bib3] 3. Yoshida M, Muneyuki E, Hisabori T. 2001. ATP synthase—a marvellous rotary engine of the cell. Nat Rev Mol Cell Biol. 2:669–677. [DOI] [PubMed] [Google Scholar]

[bib4] 4. Noji H, Yasuda R, Yoshida M, Kinosita K Jr. 1997. Direct observation of the rotation of F₁-ATPase. Nature. 386:299–302. [DOI] [PubMed] [Google Scholar]

[bib5] 5. Yasuda R, Noji H, Yoshida M, Kinosita K Jr., Itoh H. 2001. Resolution of distinct rotational substeps by submillisecond kinetic analysis of F₁-ATPase. Nature. 410:898–904. [DOI] [PubMed] [Google Scholar]

[bib6] 6. Imamura H, et al. 2003. Evidence for rotation of V1-ATPase. Proc Natl Acad Sci U S A. 100:2312–2315. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib7] 7. Yanagisawa S, Frasch WD. 2021. pH-dependent 11°F₁F₀ ATP synthase sub-steps reveal insight into the F₀ torque generating mechanism. eLife. 10:e70016. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib8] 8. Spetzler D, et al. 2006. Microsecond time scale rotation measurements of single F₁-ATPase molecules. Biochemistry. 45:3117–3124. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib9] 9. Adachi K, et al. 2007. Coupling of rotation and catalysis in F(1)-ATPase revealed by single-molecule imaging and manipulation. Cell. 130:309–321. [DOI] [PubMed] [Google Scholar]

[bib10] 10. Adachi K, Oiwa K, Yoshida M, Nishizaka T, Kinosita K Jr. 2012. Controlled rotation of the F₁-ATPase reveals differential and continuous binding changes for ATP synthesis. Nat Commun. 3:1022. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib11] 11. Diez M, et al. 2004. Proton-powered subunit rotation in single membrane-bound F₀F₁-ATP synthase. Nat Struct Mol Biol. 11:135–141. [DOI] [PubMed] [Google Scholar]

[bib12] 12. Guo H, Suzuki T, Rubinstein JL. 2019. Structure of a bacterial ATP synthase. eLife. 8:e43128. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib13] 13. Abrahams JP, Leslie AG, Lutter R, Walker JE. 1994. Structure at 2.8-A resolution of F₁-ATPase from bovine heart mitochondria. Nature. 370:621–628. [DOI] [PubMed] [Google Scholar]

[bib14] 14. Menz RI, Leslie AG, Walker JE. 2001. Structure of bovine mitochondrial F(1)-ATPase with nucleotide bound to all three catalytic sites: implications for the mechanism of rotary catalysis. Cell. 106(3):331–341. [DOI] [PubMed] [Google Scholar]

[bib15] 15. Suzuki K, et al. 2016. Crystal structures of the ATP-binding and ADP-release dwells of the V(1) rotary motor. Nat Commun. 7:13235. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib16] 16. Arai S, et al. 2013. Rotation mechanism of Enterococcus hirae V1-ATPase based on asymmetric crystal structures. Nature. 493:703–707. [DOI] [PubMed] [Google Scholar]

[bib17] 17. Nakanishi A, Kishikawa JI, Tamakoshi M, Mitsuoka K, Yokoyama K. 2018. Cryo EM structure of intact rotary H(+)-ATPase/synthase from Thermus thermophilus. Nat Commun. 9:89. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib18] 18. Jault JM, Allison WS. 1994. Hysteretic inhibition of the bovine heart mitochondrial F₁-ATPase is due to saturation of noncatalytic sites with ADP which blocks activation of the enzyme by ATP. J Biol Chem. 269:319–325. [PubMed] [Google Scholar]

[bib19] 19. Sakaki N, et al. 2005. One rotary mechanism for F₁-ATPase over ATP concentrations from millimolar down to nanomolar. Biophys J. 88:2047–2056. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib20] 20. Grubmeyer C, Cross RL, Penefsky HS. 1982. Mechanism of ATP hydrolysis by beef heart mitochondrial ATPase. Rate constants for elementary steps in catalysis at a single site. J Biol Chem. 257:12092–12100. [PubMed] [Google Scholar]

[bib21] 21. Yokoyama K, Hisabori T, Yoshida M. 1989. The reconstituted α₃β₃δ complex of the thermostable F₁-ATPase. J Biol Chem. 264:21837–21841. [PubMed] [Google Scholar]

[bib22] 22. Konishi J, Yohda M, Hashimoto T, Yoshida M. 1987. Single site catalysis of the F₁-ATPase from Saccharomyces cerevisiae and the effect of inorganic phosphate on it. J Biochem. 102:273–279. [DOI] [PubMed] [Google Scholar]

[bib23] 23. Hisabori T, et al. 1992. Single site hydrolysis of 2′,3′-O-(2,4,6-trinitrophenyl)-ATP by the F₁-ATPase from thermophilic bacterium PS₃ is accelerated by the chase-addition of excess ATP. J Biol Chem. 267:4551–4556. [PubMed] [Google Scholar]

[bib24] 24. Suzuki T, Ueno H, Mitome N, Suzuki J, Yoshida M. 2002. F(0) of ATP synthase is a rotary proton channel. Obligatory coupling of proton translocation with rotation of c-subunit ring. J Biol Chem. 277:13281–13285. [DOI] [PubMed] [Google Scholar]

[bib25] 25. Kato S, Yoshida M, Kato-Yamada Y. 2007. Role of the epsilon subunit of thermophilic F₁-ATPase as a sensor for ATP. J Biol Chem. 282:37618–37623. [DOI] [PubMed] [Google Scholar]

[bib26] 26. Shirakihara Y, et al. 2015. Structure of a thermophilic F₁-ATPase inhibited by an ε-subunit: deeper insight into the ε-inhibition mechanism. FEBS J. 282:2895–2913. [DOI] [PubMed] [Google Scholar]

[bib27] 27. Suzuki T, et al. 2003. F₀F₁-ATPase/synthase is geared to the synthesis mode by conformational rearrangement of epsilon subunit in response to proton motive force and ADP/ATP balance. J Biol Chem. 278:46840–46846. [DOI] [PubMed] [Google Scholar]

[bib28] 28. Tsunoda SP, et al. 2001. Large conformational changes of the epsilon subunit in the bacterial F₁F₀-ATP synthase provide a ratchet action to regulate this rotary motor enzyme. Proc Natl Acad Sci U S A. 98:6560–6564. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib29] 29. Imamura H, et al. 2009. Visualization of ATP levels inside single living cells with fluorescence resonance energy transfer-based genetically encoded indicators. Proc Natl Acad Sci U S A. 106:15651–15656. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib30] 30. Imamura H, et al. 2005. Rotation scheme of V₁-motor is different from that of F₁-motor. Proc Natl Acad Sci U S A. 102:17929–17933. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib31] 31. Yagi H, et al. 2007. Structures of the thermophilic F₁-ATPase epsilon subunit suggesting ATP-regulated arm motion of its C-terminal domain in F₁. Proc Natl Acad Sci U S A. 104:11233–11238. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib32] 32. García JJ, Capaldi RA. 1998. Unisite catalysis without rotation of the γ-ε domain inEscherichia coli F₁-ATPase. J Biol Chem. 273:15940–15945. [DOI] [PubMed] [Google Scholar]

[bib33] 33. Al-Shawi MK, Senior AE. 1988. Complete kinetic and thermodynamic characterization of the unisite catalytic pathway of Escherichia coli F₁-ATPase. Comparison with mitochondrial F₁-ATPase and application to the study of mutant enzymes. J Biol Chem. 263:19640–19648. [PubMed] [Google Scholar]

[bib34] 34. Senior AE. 1992. Catalytic sites of Escherichia coli F₁-ATPase. J Bioenerg Biomembr. 24:479–484. [DOI] [PubMed] [Google Scholar]

[bib35] 35. Amano T, Tozawa K, Yoshida M, Murakami H. 1994. Spatial precision of a catalytic carboxylate of F₁-ATPase beta subunit probed by introducing different carboxylate-containing side chains. FEBS Lett. 348:93–98. [DOI] [PubMed] [Google Scholar]

[bib36] 36. Scheres SH. 2012. RELION: implementation of a Bayesian approach to cryo-EM structure determination. J Struct Biol. 180:519–530. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib37] 37. Punjani A, Rubinstein JL, Fleet DJ, Brubaker MA. 2017. cryoSPARC: algorithms for rapid unsupervised cryo-EM structure determination. Nat Methods. 14:290–296. [DOI] [PubMed] [Google Scholar]

[bib38] 38. Zheng SQ, et al. 2017. MotionCor2: anisotropic correction of beam-induced motion for improved cryo-electron microscopy. Nat Methods. 14:331–332. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib39] 39. Rohou A, Grigorieff N. 2015. CTFFIND4: fast and accurate defocus estimation from electron micrographs. J Struct Biol. 192:216–221. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib40] 40. Bepler T, et al. 2019. Positive-unlabeled convolutional neural networks for particle picking in cryo-electron micrographs. Nat Methods. 16:1153–1160. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib41] 41. Liebschner D, et al. 2019. Macromolecular structure determination using X-rays, neutrons and electrons: recent developments in Phenix. Acta Crystallogr D Struct Biol. 75:861–877. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib42] 42. Croll TI. 2018. Isolde: a physically realistic environment for model building into low-resolution electron-density maps. Acta Crystallogr D Struct Biol. 74:519–530. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib43] 43. Pettersen EF, et al. 2021. UCSF ChimeraX: Structure visualization for researchers, educators, and developers. Protein Sci. 30:70–82. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib44] 44. Barad BA, et al. 2015. EMRinger: side chain-directed model and map validation for 3D cryo-electron microscopy. Nat Methods. 12:943–946. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib45] 45. Chen VB, et al. 2010. MolProbity: all-atom structure validation for macromolecular crystallography. Acta Crystallogr D Biol Crystallogr. 66:12–21. [DOI] [PMC free article] [PubMed] [Google Scholar]

PERMALINK

Structural basis of unisite catalysis of bacterial F0F1-ATPase

Atsuki Nakano

Jun-ichi Kishikawa

Atsuko Nakanishi

Kaoru Mitsuoka

Ken Yokoyama

Roles

Abstract

Significance Statement.

Introduction

Fig. 1.

Results

Cryo grid preparation of wt-F0F1 and ATPase-active ΔεCT-F0F1

Structure of ND-wt-F0F1

Fig. 3.

Structure of wt-F0F1 exposed to 4 mM ATPγS

Fig. 2.

Structure of ΔεCT-F0F1 under unisite catalysis conditions

Fig. 4.