Sun Runguang, Zhang Jing ^*
(College of Physics and Information Technology, Department of Food Engineering, Shaanxi Normal University, Xi'an 710062, China)

Abstract Many scientists have observed that the F₁F₀-ATPase complexes are comprised of 8,9, and 16 different subunits respectively in Escherichia coli, chloroplasts and mitochondria. All F-type ATPase have a similar structure. The globular F₁ domain and the intrinsic membrane F₀ domain linked by a central and peripheral stalk. One of the stalks belongs to the rotor and is built of the g and e subunits. The other stalk belongs to the stator and consists of subunits b₂ (I, II) and d. The central stalk in ATP synthase, made of the g, d and e subunits in the mitochondrial enzyme, is the key rotary element in the enzyme's catalytic mechanism. The g subunit penetrates the catalytic (a₃b₃) domain and protrudes beneath it, interacting with a ring of the c subunits in the membrane that drives rotation of the stalk during ATP synthesis. The d and e subunits interact with a Rossmann fold in the g subunit, forming a foot. In ATP synthase, this foot interacts with the c-ring and couples the transmembrane proton motive force to catalysis in the (a₃b₃) domain. A fuzzy cap and a mysterious collar structure have also been visualized by electron microscopy. The F₁F₀-ATPase from Escherichia coli, chloroplasts, and bovine heart mitochondria, catalyses the formation of ATP (adenosine triphosphate) from ADP (adenosine diphosphate) and Pi (inorganic phosphate) by using the energy of electrochemical proton gradient derived from electron transport.
Keywords F₁F₀-ATPase; structure of subcomplex; rotational catalysis; central stalk; second stalk;

1. INTRODUCTION
ATP synthase, also known as F₁F₀-ATPase from Escherichia coli, chloroplasts, and bovine heart mitochondria, catalyses the formation of ATP (adenosine triphosphate) from ADP (adenosine diphosphate) and Pi (inorganic phosphate) by using the energy of electrochemical proton gradient derived from electron transport ^[1]. The F₁F₀-ATPase complexes are comprised of 8,9, and 16 different subunits respectively in Escherichia coli (E. coli) ^[2,3], chloroplasts ^[4] and mitochondria ^[5].
    Many scientists have observed that all F-type ATPase have a similar structure. The ATP synthase consists of two portions, globular F₁ domain and the intrinsic membrane F_0, connected by two stalks: a central rotor stalk containing g and e subunits and a peripheral, the second stalk formed by d and two copies of the b subunits (F₀). The second stalk is expected to keep the stator subunits from spinning along with the rotor ^[6-9]. The simplest F₁F₀-type enzymes, e.g. as in Escherichia coli, are composed of eight types of subunits in an unusual stoichiometry of a₃b₃g d e (F₁) and ab₂c₁₂ (F₀). F₁ extends from the membrane, with the a , and b subunits alternating around a central subunit g . ATP synthesis occurs alternately in different b subunits, the cooperative tight binding of ADP + Pi at one catalytic sites being coupled to ATP release at a second, the differences in binding affinities appear to be caused by rotation of the g subunit in the center of the a₃b₃ hexamer. The g subunit traverses a 4.5 nm stalk connecting the catalytic subunits to the membrane-traversing F₀ sector. Subunit c is the H⁺-translocating subunit of F_0. Protonation/deprotonation of Asp61 in the center of the membrane is coupled to structural changes in an extramembranous loop of subunit c which interacts with both g and e subunits. Subunits g and e appear to move from one subunit c to another as ATP is synthesized. The torque of such movement is proposed to cause the rotation of g within the a₃b₃ complex. Four protons are translocated for each ATP synthesized. The movement of g and e therefore probably involves a unit of four c subunits ^[2] (Fig.1). In chloroplasts, Fromme et al. identified that the F₁ part has five different subunits with the stoichiometry a₃b₃g d e , the F₀ part has four subunits with the likely stoichiometry b(I), b'(II), c₁₂(III₁₂), a (VI)^[4]. J. Walker's group established that bovine heart F₁F₀-ATPase consists of 16 different subunits ^[10,11]. Five of them (a₃b₃g₁d₁e₁) form the F₁ portion with OSCP. The mitochondrial oligomycin sensitivity conferring protein (OSCP) is considered equivalent to the d subunit in E. coli. The mitochondrial subunit d is considered equivalent to the e subunit in E. coli ^[12]. The other subunits, (a(25 kDa), b(24 kDa), c_9-12(8 kDa), d(19 kDa), e(8 kDa), f(10 kDa), g(11 kDa). A6L(8 kDa), and F₆(9 kDa), form the membrane domain F₀, which transports protons through the membranes ^[13]. The inhibitor protein IF₁ (8 kDa) is present also in apparently stoichiometric amounts (Fig.2). Subunits g , d , e , b, d and F₆ together with subunit OSCP were thought in the past to contribute to the central connecting stalk ^[14]. Subunits b, d, OSCP and F₆ are known to be present in the complex in one copy each ^[15,16]. This sub-complex (referred to as stalk) formed a stoichiometric complex with F₁ ^[17,18]. The crystal structure of bovine F₁ clearly shows that the g subunit is a major component of the central stalk ^[19]. The rotation of the g subunit has been visualized by optical microscopy ^[20]. Therefore, this subunit could represent part of the peripheral feature, together with subunits d and e ^[8]. The roles of subunits d and e and their locations in the bovine enzyme remain obscure. Bovine ATP synthase preparations also contain three small polypeptides, designated subunits e, f, and g respectively. Subunits e, f, and g could be immunoprecipitated with anti-OSCP IgG from a fraction of bovine submitochondrial particles enriched in oligomycin-sensitive ATPase. The NH₂ termini of subunits f and g are exposed on the matrix side of the mitochondrial inner membrane and can be curtailed by proteolysis. The COOH termini of all three polypeptides are exposed on the cytosolic side of the inner membrane. Subunit f cross-links to A6L and to subunit g, and subunit e cross-links to subunit g and appears to form an e-e dimer. Thus, the bovine ATP synthase complex appears to have 16 unlike subunits, twice as many as its counterpart in Escherichia coli ^[21].

Fig. 1 Model of features and structure of ATP synthase in Escherichia coli

2. MOLECULAR MODELS OF MECHANISMS OF ROTATIONAL CATALYSIS
A model for proton translocation was presented by Vik and Antonio ^[22] in 1994. In the model, protonated Arg210 of subunit a interacts with unprotonated Asp61 from one subunit c, at site 1. A second subunit c, adjacent to the first at site 2, is also unprotonated, and interacts with subunit a. The remaining c subunits have a protonated Asp61, and interact with the membrane lipids. The Asp61 interacting with Arg210 at site 1 can be protonated from protons originating from the periplasm through an access channel controlled by His245 and Glu219. When this protonation occurs, the oligomer of the c subunits is free to rotate such that the newly protonated Asp61 can enter the lipid environment. This rotation will be electrostatically driven, as the Arg210 will be strongly attracted to the remaining unprotonated Asp61 at site 2. This arrangement assures that rotation is unidirectional. When rotation occurs, a protonated Asp61 of subunit c will move from the lipid environment into contact with subunit a. In this environment the pK_a of the Asp61 will be shifted down due to the more polar environment, and the proton will be released. The role of Glu96, and perhaps-other residues, is likely to be to contribute indirectly to the polar environment of the Asp61 of the subunit c at site 2. This ensures that the proton is released and that the cycle will continue. The negative charge would help to draw the released proton the cytoplasmic surface.
    After that, in 1996, Junge et al has proposed a model of how proton transport through F₀ might generate a rotary motion ^[23]. The essence of this hypothetical rotary motor is that the essential carboxyl circumference of the c annulus. Part of the external surface of the annulus interacts with subunit a, and in this region the carboxyl groups are negatively charged. It is envisaged that subunit a has an inlet port on the external surface of the membrane which allows a proton to neutralize one of the negatively charged carboxylate group. The resulting un-ionized carboxyl group will find its way by thermal vibrations to its preferred environment in contact with the phospholipid bilayer. The neutralization of this carboxylate group at one point of the circumference is accompanied by reionization of another group further around the circumference of the c annulus, release of the proton on the opposite side of the membrane through an exit port in subunit a and regeneration of a negative charge in the c annulus-subunit a interface. These protonation/ deprotonation events result in a rotary movement of the c annulus. The rotation brings the next negatively charged carboxyl group to the inlet port, where in turn it is neutralized by another proton. The accompanying release of another proton through the exit port generates further rotation. It is envisaged that this rotary device is directly coupled to the g subunit. The synthesis of each ATP molecule requires a rotation of the g subunit by 120° ; thus, each complete rotation of the g subunit in F₁ produces three ATP molecules. In a hypothetical proton motor with a c annulus consisting of twelve c subunits, this corresponds to the sequential neutralization of four carboxylate groups by proton binding. In other word, a H⁺: ATP ratio of four is compatible with a molar ratio of c subunits in F₀ of twelve. Likewise, a H⁺: ATP ratio of three requires nine c subunits in each F₀.
    Wang & Oster assumed that the entire F₁F₀-ATPase structure is arranged as a counter-rotating "rotor" and "stator" assembly ^[24,25]. The stator portion consists of the catalytic sites contained in the hexameter, together with subunits a, b₂, and d . The rotor consists of 9-12 c-subunits arranged in a ring and connected to the g-, and d-subunits that form the central "shaft". ATP synthase is conceived as a rotary enzyme, the proton-transporting membrane portion, F₀, and the catalytic peripheral portion, F₁. They are mounted on a central shaft (subunit g ) and held together by an eccentric bearing. Proton flow drives the rotor (namely, subunits c₁₂eg ) relative to stator (namely, subunits ab₂da₃b₃) and extrudes spontaneously formed ATP from three symmetrically arranged binding sites on a₃b₃ into the solution. The binding of subunit d to a₃b₃ is of sufficient strength to hold against the elastic strain, which is generated during the operation of this enzyme. A single binding site for d on the hexagon of a₃b₃ is sufficient for the assumed hold-function of d in the stator ^[26].
    The rotor-stator concept requires two stalks connecting F₁ to F₀. One of the stalks belongs to the rotor and is built of g and e subunits. The other stalk belongs to the stator and might consist of subunits b₂ (I, II) and d ^[27].
    Recently, using electron micrographs of solubilised and negatively stained F₁F₀-ATPase respectively from Escherichia coli^[7], chloroplasts^[27] and bovine heart mitochondria^[8], three new features in the F₁F₀-ATPase are seen, one of which is a smaller peripheral stalk. The second feature is a collar structure like an elliptical disk just above the very compact density of the membrane domain, and the third feature is some additional density on top of the F₁ domain. The peripheral stalk seems to connect the catalytic F₁ domain to the collar structure above F₀ domain. These are very interesting features.

3. HYDROPHILIC GLOBULAR F₁ PORTION  
The structure of F₁ portion from various sources is very clearly. It is made up of three a subunits, three b subunits and one each of the g, d, and e subunits. The three catalytic sites are found mainly on the b subunit. The three catalytic sites are known to pass sequentially through three different conformations associated with substrate binding, formation of tightly bound ATP, and release of the ATP. These changes are thought to occur through a rotational catalysis where rotation of the g subunit causes the requisite sequential changes in the b subunit ^[1,28]. In 1994, J. Walker's group reported the X-ray structure of the major portion of F₁ with three different conformations of the b subunits that are likely to be interconverted by rotation of the g subunit ^[19]. Duncan and Aggeler showed positional interchange of the b subunits as catalysis proceeds ^[29,30]. In 1997, Noji et al. attached a fluorescently labeled actin filament to the g subunit of the F₁ portion fixed on a slide then watched the filament spin as the enzyme cleaved ATP. They dramatically showed the rotation directly and provided strong evidence for the rotation by a specialized fluorescence technique ^[20]. In 2002, Rubinstein and Walker found that the bound avidin has been localised close to the F₁ domain by electron microscopy of negatively stained particles of the ATP synthase-avidin complex. The images were subjected to multi-reference alignment and classification. Because of the presence of a flexible linker between the OSCP and the biotinylation signal, the class-averages differ in the position of the avidin relative to the F₁ domain. These positions lie on an arc, and its center indicates the position of the C terminus of the OSCP on the surface of the F₁ domain. Since the N-terminal region of the OSCP is known to interact with the N-terminal regions of the a -subunits, which are on top of the F₁ domain distal from the F₀ membrane domain, the OSCP extends almost 10 nm along the surface of F₁ down towards F₀ where it interacts with the C terminus of the b subunit, which extends up from F₀^[31].

4. HYDROPHOBIC F₀ PORTION
There is one a subunit, two b subunits and 9-12 c subunits in the F₀ portion from the bacterium Escherichia coli. The F₀ portion from various plants and animals is more complex, but it still contains the multiple copies of the c-type subunit ^[32]. Three types of subunits make up of the F₀ portion that indicated an annular arrangement of twelve c subunits, with subunit a and b₂ on the periphery of the cylinder ^[33,34]. The long helical subunit-b₂ dimer together with the d form the peripheral stalk, linking subunit a of the F₀ portion with one of the a-subunits of the F₁ portion ^[35,36]. The ring of the c subunits makes contact with subunit a in the F₀ portion ^[34]. The subunit c ring is also linked to the g-subunit of the F₁ core both directly and indirectly through mutual contacts with the e-subunit. Asp61 in subunit c (an essential carboxylate of Escherichia coli) translates protons at the interface between subunit a and c ^[32,37].
    Current models postulate that protonation and subsequent ionization of the Asp61 side chain in subunit c lead to rotation of the c₁₂ oligomer with respect to the a and b₂ subunits ^[32,38,39]. Rotation of the c₁₂ oligomer in turn cause rotation of the e- and g -subunits with respect to the catalytic sites on the b-subunits in the F₁ portion ^[32]. The a and b subunits constitute the stator elements in the F₀ sector of F₁F₀-ATP synthase. The a subunit in association with the ring of the c subunits houses the proton channel through F₁F₀-ATP synthase. The two b subunits dimerize, forming an extended flexibly unit in the peripheral stalk linking the F₁ and F₀ sectors ^[40].
    Subunit c in the F₀ portion of the H+ transporting ATP synthase is an essential part of its membrane domain that participates in transmembrane proton conduction. The annular architecture of the subunit c from different species has been previously reported. However, little is known about the type of interactions that affect the formation of c-rings in the ATPase complex. In 2002, Arechaga et al. reported that subunit c over-expressed in Escherichia coli and purified in non-ionic detergent solutions self-assembles into annular structures in the absence of other subunits of the complex. The results suggested that the ability of subunit c to form rings were determined by its primary structure ^[41].

5. CENTRAL AND PERIPHERAL STALK
That two stalks link the F₁ and F₀ sectors of ATP synthase respectively in Escherichia coli, chloroplasts, and mitochondria have been evidenced by electron microscopy after negative staining of specimens ^[7-9]. The two b subunits and the d subunit associate to form b₂d , a second, peripheral stalk extending from the membrane up to the side of a₃b₃ and binding to the N-terminal regions of the a subunits, which are approx. 12.5 nm from the membrane ^[42]. The E. coli. d-subunit interacts with the N-terminal part of the a subunits that is located on top of the F₁ portion ^[43].
    ATP synthase contains a rotary motor involved in biological energy conversion. Its membrane-embedded F₀ sector has a rotation generator fueled by proton-motive force, which provides the energy required for the synthesis of ATP by the F₁ domain. A ring of 10 c sub-subunits was obtained from crystals of a F₁c₁₀ complex of yeast mitochondrial ATP synthase by Stock et al in 1999. This is probably the most novel of features in the mitochondrial ATP synthase. Each c subunit forms an α-helical hairpin. The interhelical loops of six to seven of the c subunits are in close contact with the g and d subunits of the central stalk. The extensive contact between the c ring and the stalk suggests that they may rotate as an ensemble during catalysis ^[44].
    The recent structure of bovine a₃b₃g e d by Gibbons et al. has opened up new discussions in 2000. The central stalk in ATP synthase, made of the g, d and e subunits in the mitochondrial enzyme, is the key rotary element in the enzyme's catalytic mechanism. The g subunit penetrates the catalytic (a₃b₃) domain and protrudes beneath it, interacting with a ring of the c subunits in the membrane that drives rotation of the stalk during ATP synthesis. In other crystals of the F₁-ATPase, the protrusion was disordered, but with crystals of the F₁-ATPase inhibited with dicyclohexylcarbodiimide, the complete structure was revealed. The d and e subunits interact with a Rossmann fold in the g subunit, forming a foot. In ATP synthase, this foot interacts with the c-ring and couples the transmembrane proton motive force to catalysis in the (a₃b₃) domain ^[45].
    In 2002, Meier et al. found that The Na⁺-translocating ATP synthases from Ilyobacter tartaricus and Propionigenium modestum contain undecameric c subunit rings of unusual stability. These c11 rings were isolated from both ATP synthases and crystallized in two dimensions. Cryo-transmission electron microscopy projection maps of the c-rings from both organisms were identical at 7Å resolution. Different crystal contacts were induced after treatment of the crystals with dicyclohexylcarbodiimide (DCCD), which is consistent with the binding of the inhibitor to glutamate 65 in the C-terminal helix on the outside of the ring. The c subunits of the isolated c11 ring of I. tartaricus were modified specifically by incubation with DCCD with kinetics that was indistinguishable from those of the F1Fo holoenzyme ^[46].

6. FUZZY CAP
A fuzzy cap at the top of the F₁ part has been found ^[7-9]. In 1988, Stephan Wilkens & Roderick A. Capaldi found that there is a cap at the top of the F₁ part at which the second stalk may bind from Escherichia coli by electron microscopy after negative staining of specimens. After them, Simone Karrasch and John E. Walker found there is some additional density, a fuzzy cap, on top of the F₁ domain from bovine heart mitochondria ^[8]. This is likely includes N-terminal stretches of the three copies of the a subunit and a part of the d subunit. The second stalk of the d and b subunits is the stator which makes this rotation possible ^[9]. It is known that the bacterial d subunit and the equivalent bovine OSCP is required for binding F₁ to F₀, implying that it interacts with subunits in the membrane domain. Therefore, the inescapable conclusion is that OSCP (and possibly also subunit d in the bacterial and chloroplast enzymes) extends from the top of the F₁, down its external surface, to a region associated with the membrane domain ^[14]. In the bovine enzyme, the N-terminal part of OSCP interacts with the N-terminal part of the a subunit ^[47].
    Wilkens et al. have determined the location of the d subunit directly by imaging the intact E.coli ATP-synthase labeled with a monoclonal antibody specific to d . The part of the d subunit is bound in the "dimple" formed by the N terminal of the large subunits a and b on the very top of the F₁, which provides an explanation for why there is only one binding site for d on the F₁ despite the existence of three identical a b pairs. The fuzzy cap is clear on the very top of the F₁ part, too ^[48].

7. THE POSITION OF THE IF₁ SUBUNIT
Pullman and Monroy discovered a soluble, heat-stable protein in mitochondria from bovine heart that inhibits the ATP hydrolytic activity of the F₁F₀-ATPase ^[49]. Further in vitro characterization of this "IF₁" protein was consistent with it being active under the physiological conditions where ATP hydrolysis rather that ATP synthesis would occur. The binding of the IF₁ to the F₁F₀-ATPase in submitochondrial particles requires the hydrolysis of ATP in the absence of an electrochemical gradient across the inner membrane ^[50-52]. Similar conditions might be found in tumor cells, where oxygen levels vary from the perivascular regions to the anoxic necrotic centers. Thus, the IF₁ could play an important role in the pathology of tissue ischemia and tumor growth by helping conserve ATP under conditions of oxygen deprivation ^[53].
    The position of the IF₁ subunit in the ATP synthase has been mapped by cross-linking experiments. Although yeast IF₁ interacts at the interface of the a and b subunits, cross-linking of bovine IF₁ to the bovine F₁ domain has only identified residues 394-359 on the C-terminus of the b subunit as the binding site ^[54]. According to the binding change mechanism, only one of the three b subunits can be in the open conformation and the 1:1 binding stoichiometry observed between IF₁ and F₁ is consistent with this ^[50]. The ATPase inhibitor protein (IF₁) binds at one side of the F₁F₀ connection. The carboxyl-terminal segment of the IF₁ apparently binds to OSCP. The 42L-58K segment of the IF₁ binds at the surface of one of the three a b pairs of the F₁, thus preventing the cyclic interconversion of the catalytic sites required for ATP hydrolysis ^[55]. Gaballo et al. suggested that the IF₁ has a relevant physiopathological role for the conservation of the cellular ATP pool in ischemic tissues. Under these conditions IF₁, which appears to be over expressed, prevents dissipation of the glycolytic ATP ^[56].

8. MYSTERIOUS COLLAR STRUCTURE
Recently, Simone Karrasch and John E. Walker, using electron micrographs of solubilised and negatively stained F₁F₀-ATPase from bovine heart mitochondria, showed that the new features in bovine F₁F₀-ATPase is a collar structure like an elliptical disk just above the very compact density of the membrane domain ^[11]. The mysterious collar above the membrane domain F₀ may be a circular structure and could be connected to the subunit c complex, which is believed to form a ring structure in the membrane ^[57-59]. It is possible that this ring structure is extended to the area outside the membrane by the help of hydrophilic parts of other subunit of the F₀ ^[11]. The transmembrane sector of the F₀F₁ rotary ATP synthase is proposed to organize with an oligomeric ring of the c subunits, which function as a rotor, interacting with two b subunits at the periphery of the ring, the b subunits functioning as a stator. The b subunit cross-linked to subunit c, which suggests that, either only one of the two b subunits lie adjacent to the c-ring or that both b subunits interact with a single c subunit. The c subunit lying adjacent to subunit b was shown to be mobile and to exchange with the c subunits that initially occupied non-neighboring positions. The oligomeric c-ring can move with respect to the b-stator and provide further support for a rotary catalytic mechanism in the ATP synthase ^[39].
    Simone Karrasch and John E. Walker suggest that subunits g, d, e, b, d, F₆, and OSCP are likely to contribute to the central as well as to the peripheral stalk, the collar and the additional density on top of the F₁ particle ^[8]. The question that now arises is, which subunits of the F₁F₀-ATPase contribute to the peripheral stalk, the collar, and which the features and structure of subunits e, f, g, A6L are not known in the membrane domain.

9. NOVEL STRUCTURAL MODEL OF ATP SYNTHASE     
Subunit b appears to have two transmembrane α-helical spans and an extensive hydrophilic domain that is highly charged and lies outside the membrane where it interacts with the F₁ via to OSCP. OSCP could form a stable binary complex with the F₁-ATPase. A stable sub-complex was also assembled from one copy each of the subunit b, d, F₆, and OSCP ^[16]. Subunit b and d are components of the membrane domain of the ATP synthase ^[17]. The F₆ subunit is also associated with the F₀ portion ^[60]. As subunits F₆ and d have no extensive hydrophobic regions in their sequence, they are likely to be attached to the matrix surface of the inner mitochondrial membrane ^[61].
    It is very clearly that subunit b and OSCP may in one copy each form the peripheral stalk. The subunit d and F₆ in one copy each may form a collar structure; namely, a double helical, ring structure around the central stalk. This circular collar structure may be composed of parts of the c subunits. The image of the bovine enzyme contains the novel feature, seen as disc-shaped or collar evidently sitting between the central stalk and the F₀ domains. A similar collar was observed in the Clostridial enzyme. The role of the collar is an even bigger mystery ^[14]. The mysterious collar structure remains to be confirmed.

Fig. 2 Model of features and structure of ATP synthase in bovine heart mitochondria^[64]

    Recent research indicates that ATP synthases (F₀F₁) contain two distinct nanomotors, one an electrochemically driven proton motor contained within F₀ that drives an ATP hydrolysis-driven motor (F₁) in reverse during ATP synthesis. This is depicted in recent models as involving a series of events in which each of the three ab pairs comprising F₁ is induced via a centrally rotating subunit (g) to undergo the sequential binding changes necessary to synthesize ATP (binding change mechanism). Stabilization of this rotary process is provided in current models by a peripheral stalk or "stator" that has recently been shown to extend from near the bottom of the ATP synthase molecule to the very top of the F₁. Although quite elegant, these models envision the stator as fixed during ATP synthesis, i.e., bound to only a single a b pair. This is despite the fact that the binding change mechanism views each a b pair as going through the same sequential order of conformational changes which demonstrate a chemical equivalency among them. For this reason, Blum et al. proposed two different dynamic models for stator function during ATP synthesis ^[62]. Up to the present moment, three protein motors have been unambiguously identified as rotary engines: the bacterial flagellar motor and the two motors that constitute ATP synthase (F₀F₁ ATPase)^[63]. The F₁ motor is driven by ATP hydrolysis.
    According to the results of previous experiments, we suggest a model of the structure of F₁F₀-ATPase in shown as Fig.2. This enzyme structure consists of a fuzzy cap, a hydrophilic globular F₁ portion, a central stalk, a peripheral stalk, a collar structure, and a hydrophobic F₀ portion that is embedded in a phospholipid bilayer membrane. The well characterized subunits of the bovine ATP synthase complex are the subunits a₃b₃gde and OSCP of the catalytic sector, F₁; the ATPase inhibitor protein (IF₁); and subunits a, b, c_9-12, d, e, f, g, F_6, and A6L, which are present in the membrane sector, F₀, and the two stalk (a central and a peripheral stalk) that connects F₁ to F₀. A central rotor stalk containing g-, d- and e-subunits. The g subunit penetrates the catalytic (a₃b₃) domain and connects the catalytic subunits to the membrane-traversing F₀ sector, interacting with a ring of the c subunits in the membrane that drives rotation of the stalk during ATP synthesis. The d and e subunits interact with a Rossmann fold in the g subunit, forming a foot. This foot interacts with the c-ring and couples the transmembrane proton motive force to catalysis in the (a₃b₃) domain. Thus, rotor portion consists of g d e c_9-12 sub-coplex, which is the key rotary element in the mitochondrial enzyme's catalytic mechanism. The peripheral stalk consists of one b and OSCP subunits, which is the stator that makes the rotation possible. Only one b subunit interacts with a single c subunit. The c subunit lying adjacent to subunit b was shown to be mobile and to exchange with the c subunits that initially occupied non-neighboring positions. The oligomeric c-ring can move with respect to the b-stator and provide further support for a rotary catalytic mechanism in the ATP synthase. Sub-complex aba₃b₃ together with OSCP subunits form a stator portion in the mitochondrial ATPsynthase. Subunit f cross-links to A6L and to subunit g, and subunit e cross-links to subunit g and appears to form an e-e dimer. The circular collar structure may be composed of the subunits d and F₆ together with parts of the c subunits, which sits between the central stalk and F₀ domains. In the mitochondrial enzyme, the N-terminal domain part of interacts with the N-terminal part of the a subunits form a fuzzy cap at the very top of the F₁ part.
    Although the research progress of the ATP synthase is commendable, questions and uncertainties remain. A further to the structure of ATP synthase will probably uncover yet more surprising features.

[ Back ] [ Home ] [ Up ] [ Next ]Mirror Site in USA  Europe  China  GBNet