An F1Fo-type ATP synthase is found in the inner mitochondrial membrane, the inner membrane of bacteria and the thylakoid membrane of chloroplasts, where it functions to convert the free energy of the proton-motive force into the chemical energy source ATP (for reviews, see Boyer, 1993; Junge et al., 1997; Weber and Senior, 1997). This large enzyme complex is composed of two major parts, a water-soluble F1 sector made up of three α subunits, three β subunits and one copy of each of the γ, δ and ε subunits, and a membrane-embedded Fo sector consisting of one a, two b and 12 c subunits. The overall molecular mass of the complex is 520 kDa. There are three catalytic nucleotide-binding sites located on the β subunits of the F1 and one proton channel formed by the a and c subunits in the membrane-embedded Fo sector.
The available evidence suggests that the coupling of catalytic sites to proton translocation in both the ATP synthesis and hydrolysis modes is conformational and that this enzyme works as a rotary molecular motor. The key to its functioning is in the nature of the interaction between the F1and Fosectors. These interactions have been worked out largely by a combination of electron microscopy and cross-linking studies.
ECF1is linked to the Fosector by two stalks The bipartite nature of the F1Focomplex had been observed in early electron micrographs using negative staining to optimize the contrast between protein and solvent, and such pictures appeared to show a stalk linking the two parts (Fernandez-Moran, 1962; Soper et al., 1979; Telford et al., 1984). However, it was uncertain whether this stalk was an
artifact of the staining procedure until our cryoelectron microscopy images confirmed its presence in buffered solutions without heavy atoms added (Gogol et al., 1987; Lücken et al., 1990). In these first cryoelectron microscopy pictures of ECF1Fo, F1appeared to extend to Foat one side in some of the images, as though there were two linkages between the two parts.
More recently, we have obtained electron micrographs of ECF1Fo in a monodisperse, detergent-solubilized form (Wilkens and Capaldi, 1998a,b). Under these conditions, the overlap of molecules seen in membranous material is absent, simplifying the analysis of images. These new data establish conclusively that there are two stalks in ECF1Fo (Fig. 1) and, in concert with recent data for chloroplast F1Fo(Böttcher et al., 1998) and mitochondrial F1Fo(J. E. Walker, personal communication), indicate that this is a common feature of F1Fo-type proton-pumping ATPases and possibly of V1Vo-type ATPases as well (Boekema et al., 1997).
The central stalk contains the γand εsubunits Key evidence that the γ subunit contributes to the central stalk came from the high-resolution structure of mitochondrial F1α3β3γsubcomplex of Abrahams et al. (1994). This structure showed the γsubunit passing through a cavity within the α3β3 hexagon and extending from the bottom of the structure as expected of a stalk component. Our finding that the γsubunit can be cross-linked in high yield to the c subunit ring (Watts et al., 1996) established that the γ subunit extends the full length of the stalk. Also present in the central stalk is the ε subunit. This polypeptide has a two-domain structure (Wilkens Printed in Great Britain © The Company of Biologists Limited 2000
JEB2341
ATP synthase, also called F1Fo-ATPase, catalyzes the synthesis of ATP during oxidative phosphorylation. The enzyme is reversible and is able to use ATP to drive a proton gradient for transport purposes. Our work has focused on the enzyme from Escherichia coli (ECF1Fo). We have used a combination of methods to study this enzyme, including electron microscopy and chemical cross-linking.
The utility of these two approaches in particular, and the important insights they give into the structure and mechanism of the ATP synthase, are reviewed.
Key words: F1Fo-type ATPase, ATP synthase, electron microscopy,
cross-linking, rotation.
Summary
Introduction
CROSS-LINKING AND ELECTRON MICROSCOPY STUDIES OF THE STRUCTURE
AND FUNCTIONING OF THE ESCHERICHIA COLI ATP SYNTHASE
RODERICK A. CAPALDI*, BIRTE SCHULENBERG, JAMES MURRAY ANDROBERT AGGELER Institute of Molecular Biology, University of Oregon, Eugene, OR 97403-1229, USA
*e-mail: [email protected]
et al., 1995; Uhlin et al., 1997). There is a C-terminal pair of α-helices that fits below F1and interacts with either the α or β subunit, depending on nucleotide binding in catalytic sites (Aggeler et al., 1995; Aggeler and Capaldi, 1996; Wilkens and Capaldi, 1998c). The N-terminal domain is a 10-stranded β sandwich that binds on one face to the γ subunit (Tang and Capaldi, 1996; Watts et al., 1996) and by the bottom of the sandwich to the c subunit ring (Zhang and Fillingame, 1995).
The peripheral stalk contains the δand b subunits
Obvious candidates for the second stalk in the bacterial enzyme are the δ and b subunits. Our nuclear magnetic resonance (NMR) structure determination of the δsubunit shows a largely globular α-helical protein (Wilkens et al., 1997) which cross-linking studies have established is attached near the top of F1by interactions with an αsubunit (Lill et al., 1996; Ogilvie et al., 1997). Along with N-terminal segments of the three α subunits, it probably generates the cap seen at the top of ECF1Fo in recent electron micrographs (Wilkens and Capaldi, 1998a,b). Cross-linking studies indicate that the C-terminal part of the
b subunits interacts with the C-terminal region of the δsubunit and with the α subunit (Rodgers et al., 1997; Rodgers and Capaldi, 1998) and that these interactions are close to the top of the F1sector. The two b subunits extend as paired αhelices through the length of F1Fo, with interactions inside the lipid bilayer (Dmitriev et al., 1999).
Two stalks, a central rotor and a peripheral stator F1is unusual in being a trimeric enzyme (three α–βpairs).
During functioning, three catalytic sites must be coupled alternately to one proton channel. To explain how this coupling might occur, both Cox and colleagues (e.g. Cox et al., 1984) and Boyer (1993) independently proposed that, during ATP synthesis, a central element in the enzyme rotated in one direction between catalytic sites, driven by a proton gradient. During ATP hydrolysis in the three catalytic sites, the central element rotates in the opposite direction. Electron microscopy provided the first indication that this central element of the F1 sector was the γ subunit (Gogol et al., 1989, 1990). Furthermore, we were able to show that the γsubunit was not fixed in the cavity within the α3β3hexagon, but could be found at each of the three α–βpairs (Gogol et al., 1990). Additional evidence that the γsubunit could rotate within the F1sector was obtained from cross-linking studies (Duncan et al., 1995) and by polarized absorption relaxation after photobleaching (PARAP) measurements (Sabbert et al., 1996). Conclusive evidence of ATP-hydrolysis-driven rotation of the γ subunit came with the video microscopy of single particles by Yoshida, Kinosita and their colleagues (Noji et al., 1997). In our earlier cryoelectron microscopy studies, we had observed movements of the γ subunit relative to the εsubunit. We now know that the εsubunit moves in concert with the γsubunit, as shown by cross-linking experiments in which εcan be covalently linked to γwithout alteration of function (Aggeler et al., 1992; Tang and Capaldi, 1996; Schulenberg et al., 1997). More recently, rotation of the εsubunit has been observed directly by video microscopy (Kato-Yamada et al., 1998). In our earlier experiments, the interaction of εwith γhad been perturbed by the monoclonal antibody used to tag it, resulting in this subunit remaining fixed at one βsubunit.
The rotation of the γ–εcentral stalk, two subunits that react with the c subunit ring, suggests a model in which γ–εand the
c subunits form a rotor, which moves relative to the rest of the
complex, α3β3δb2a. In such a scheme, the second stalk of δb2 should act as a stator positioning the α3β3subunits in a fixed disposition relative to the a subunit. Rotation of the c subunit ring relative to the a subunit then provides the proton-translocation mechanism. The evidence that the γ–ε–c12 subcomplex is moving relative to the rest of the complex is so far based only on cross-linking studies. The cross-linking of γ to ε, or of ε to c, does not block ATP-driven proton translocation in ECF1Fo, whereas the cross-linking of either γ or εto αor βdoes (Aggeler and Capaldi, 1992, 1996; Aggeler et al., 1992, 1993, 1995).
Cross-linking of δ or b to α also has no effect on the functioning of the ATP synthase (Ogilvie et al., 1997; Rodgers and Capaldi, 1998), as expected if these two subunits are the stator element. A summary of our cross-linking experiments to date in ECF1Fois shown in Fig. 2. Additional evidence for the rotation of γand εduring ATP hydrolysis in ECF1Fo(Aggeler et al., 1997; Zhou et al., 1997) and of γduring ATP synthesis (Bulygin et al., 1998) has been provided by cross-linking experiments, and rotation of the c subunit ring relative to subunit
a is implied by recent cross-linking studies of Fillingame and
[image:2.609.42.291.72.275.2]coworkers (Jones et al., 1998). However, convincing evidence
Fig. 1. Electron microscopy images of detergent-dispersed F1Fo
-ATPase of Escherichia coli (ECF1Fo). Parts A–C show three
different classes of images resolved. A and C are mirror images of one another and are side views that clearly separate a central and peripheral stalk. Part B is a view more end-on, in which the Fosector
of c subunit rotation from visualizing it in real time, as in the video microscopy experiments, has not yet been provided.
Elasticity of the F1Focomplex during rotation
Most studies of the rotary motor model have focused on the
[image:3.609.351.533.73.224.2]torque generated in relation to energy conservation during coupling within the enzyme complex. However, as pointed out by Cherepanov et al. (1999), there is likely to be an elasticity of the transmission between Foand F1when four small steps of rotation of c relative to a are coupled to one 120 ° rotation of γ–ε. We have been looking for conformational changes in ECF1Fo that could correspond to this storage and release of elastic energy. We have recently found that there are significant nucleotide-dependent alterations in the relationship between F1 and Fothat should be part of the functioning of the enzyme. Thus, single particles of F1Fo show the now characteristic pattern of F1separated by two stalks when Mg2+ADP is bound, but the picture is altered with ATP in catalytic sites (in the presence of Mg2+AMP.PNP). There is a contraction of the complex in the ATP state such that the F1and Fosectors are closer together and the two stalks are only partly discerned. As
Fig. 2. A summary of the cross-links generated so far between subunits of F1Fo-ATPase of Escherichia coli (ECF1Fo). Red bow-ties
[image:3.609.90.260.74.306.2]show cross links that inhibit both ATP hydrolysis and ATP synthesis. The blue bow-tie differentiates the cross link between γand c from others because, in this case, ATP synthesis is affected but not ATP hydrolysis. Yellow bow-ties show cross links that have no effect on functioning of the enzyme.
Fig. 3. Size distribution of F1Fo-ATPase of Escherichia coli
(ECF1Fo) single particles when examined in side view.
Cross-hatched bars are molecules in Mg2+ADP. Open bars are for enzyme
in Mg2+AMP.PNP.
β β
α
a b2
δ α
α
c10-12
ε γ
4.5 5 5.5 6 6.5 60
50 40 30 20 10 0
Number of particles
Particle length (mm) 7
b2
β
γ
β
α
δ
ε
a c12
b2
β
γ
a
β δ α
ε Mg2+ADP
c12 H+
Mg2+ATP
1
3 2
Fig. 4. A diagram of the motions of the F1Fo-type
[image:3.609.257.564.493.736.2]shown in Fig. 3, the average particle length of side views is 10 % shorter in the presence of ATP than in the presence of ADP, a contraction of 1.5–2.0 nm. Such a conformational change would require alterations in both stalks. If the conformational changes described above represent energy storage by elastic strain, the uncoupling of catalytic site and proton-translocation events seen in many different mutants of ECF1Fo may be due to perturbation of rotation or of the up–down movements as well as to direct disruption of the proton channel (Fig. 4).
Epilogue
As reviewed here, electron microscopy and cross-linking studies of the E. coli ATP synthase have contributed significantly to our understanding of the structure and functioning of F1Fo-type ATPases. The dynamic nature of the functioning of the enzyme means that a full understanding of its mechanism will require high-resolution structural data for the intact complex trapped in different conformations. Cross-linking to fix subunits in position and, thereby, trap specific conformations can continue to provide new insights into the workings of the ATP synthase, particularly if ECF1 and ECF1Focan be crystallized in a variety of trapped states. Our efforts to obtain such crystals continue.
Work in this laboratory is supported by National Institutes of Health grants HL 58671 and HL 24526. We thank our many colleagues who have contributed to the studies described here.
References
Abrahams, J. P., Leslie, A. G. W., Lutter, R. and Walker, J. E.
(1994). Structure at 2.8 Å resolution of F1-ATPase from bovine
heart mitochondria. Nature 370, 621–628.
Aggeler, R., Cai, S. X., Keana, J. F. W., Koike, T. and Capaldi, R. A. (1993). The γsubunit of the E. coli F1ATPase can be cross
linked near the glycine-rich loop region of a β subunit when ADP+Mg2+occupies catalytic sites but not when ATP+Mg2+is
bound. J. Biol. Chem. 268, 20831–20837.
Aggeler, R. and Capaldi, R. A. (1992). Cross linking of the γsubunit of the E. coli ATPase (ECF1) via cysteines introduced by
site-directed mutagenesis. J. Biol. Chem. 267, 21355–21359.
Aggeler, R. and Capaldi, R. A. (1996). Nucleotide dependent
movement of the ε subunit between α and β subunits in the
Escherichia coli F1Fo type ATPase. J. Biol. Chem. 271,
13888–13891.
Aggeler, R., Chicas-Cruz, K., Cai, S-X., Keana, J. F. W. and Capaldi, R. A. (1992). Introduction of reactive cysteine residues
in the εsubunit of Escherichia coli F1ATPase. Modification of
these sites with tetrafluorophenlyazide-maleimides and examination of changes in the binding of the ε subunit when different nucleotides are in catalytic sites. Biochemistry 31, 2956–2961.
Aggeler, R., Haughton, M. A. and Capaldi, R. A. (1995). Disulfide
bond formation between the COOH terminal domain of the β subunits and the γand εsubunits of the E. coli F1ATPase. J. Biol. Chem. 270, 9185–9191.
Aggeler, R., Ogilvie, I. and Capaldi, R. A. (1997). Rotation of a γ–ε
subunit domain in the E. coli F1FoATP synthase complex. The γ–ε
subunits are essentially randomly distributed relative to the α3β3δ
domain in the intact complex. J. Biol. Chem. 272, 19621–19624.
Boekema, E. J., Ubbink-Kok, T., Lolkema, J. S., Brisson, A. and Konings, W. N. (1997). Visualization of the peripheral stalk in
V-type ATPase: evidence for a stator structure essential to rotational catalysis. Proc. Natl. Acad. Sci. USA 94, 14291–14293.
Böttcher, B., Schwarz, L. and Graber, P. (1998). Direct indication
for the existence of a double stalk in C FoF1. J. Mol. Biol. 281,
757–762.
Boyer, P. D. (1993). The binding change mechanism for ATP
synthase – some probabilities and possibilities. Biochim. Biophys.
Acta 1140, 215–250.
Bulygin, V. V., Duncan, T. M. and Cross, R. L. (1998). Rotation
of the εsubunit during catalysis by E. coli FoF1ATPase synthase. J. Biol. Chem. 273, 31765–31769.
Cherepanov, D. A., Mulkidjanian, A. Y. and Junge, W. (1999).
Transient accumulation of elastic energy in proton translocating ATP synthase. FEBS Lett. 449, 1–6.
Cox, G. B., Jans, D. A., Fimmel, A. L., Gibson, F. and Hatch, L.
(1984). The mechanism of ATP synthase: conformational changes by rotation of the b subunit. Biochim. Biophys. Acta 768, 201–208.
Dmitriev, O., Jones, P. C., Jiang, W. and Fillingame, R. H. (1999).
Structure of the membrane domain of subunit b of the E. coli FoF1
ATP synthase. J. Biol. Chem. 274, 15598–15604.
Duncan, T. M., Bulygin, V. V., Zhou, Y., Hutcheon, M. L. and Cross, R. L. (1995). Rotation of subunits during catalysis of
Escherichia coli F1-ATPase. Proc. Natl. Acad. Sci. USA 92,
10964–10968.
Fernandez-Moran, H. (1962). Cell membrane ultrastructure. Low
temperature electron microscopy and X-ray diffraction of lipoprotein components in lamellar systems. Circulation 26, 1039–1065.
Gogol, E. P., Aggeler, R., Sagermann, M. and Capaldi, R. A.
(1989). Cryoelectron microscopy of Escherichia coli F1
adenosinetriphosphatase decorated with monoclonal antibodies to individual subunits of the complex. Biochemistry 28, 4717–4724.
Gogol, E. P., Johnson, E., Aggeler, R. and Capaldi, R. A. (1990).
Ligand-dependent structural variations in Escherichia coli F1
ATPase revealed by cryoelectron microscopy. Proc. Natl. Acad.
Sci. USA 87, 9585–9589.
Gogol, E. P., Lücken, U. and Capaldi, R. A. (1987). The stalk
connecting the F1and Fodomains of ATP synthase visualized by
electron microscopy of unstained specimens. FEBS Lett. 219, 274–278.
Jones, P. C., Jiang, W. and Fillingame, R. H. (1998). Arrangement
of the multicopy H+translocating subunit c in the membrane sector
of the E. coli F1FoATP synthase. J. Biol. Chem. 273, 17178–17185.
Junge, W., Lill, H. and Engelbrecht, S. (1997). ATP synthase: an
electrochemical transducer with rotary mechanics. Trends
Biochem. Sci. 22, 420–423.
Kato-Yamada, Y., Noji, H., Yasuda, R., Kinosita, K. and Yoshida, M. (1998). Direct observation of the rotation of εsubunit in F1
ATPase. J. Biol. Chem. 273, 19375–19377.
Lill, H., Hensel, F., Junge, W. and Engelbrecht, S. (1996). Cross
linking of engineered δto (αβ)3in chloroplast F1ATPase. J. Biol. Chem. 271, 32737–32742.
Lücken, U., Gogol, E. P. and Capaldi, R. A. (1990). Structure of
the ATP synthase complex (ECF1Fo) of Escherichia coli from
Noji, H., Yasuda, R., Yoshida, M. and Kinosita, J., Jr (1997).
Direct observation of the rotation of F1 ATPase. Nature 386,
299–302.
Ogilvie, I., Aggeler, R. and Capaldi, R. A. (1997). Cross linking
of the δ to one of the three α subunits has no effect on functioning, as expected if δis a part of the stator that links F1
and Fo parts of the E. coli ATP synthase. J. Biol. Chem. 272,
16652–16656.
Rodgers, A. J. W. and Capaldi, R. A. (1998). The second stalk
composed of the b and δ subunits connects Foto F1 via the α
subunit in E. coli ATP synthase. J. Biol. Chem. 273, 29406–29410.
Rodgers, A. J. W., Wilkens, S., Aggeler, R., Morris, M. B., Howitt, S. M. and Capaldi, R. A. (1997). The subunit δ-subunit
b domain of the E. coli F1FoATPase: the b subunits interact with
F1 as a dimer and through the δ subunit. J. Biol. Chem. 272,
31058–31064.
Sabbert, D., Engelbrecht, S. and Junge, W. (1996). Intersubunit
rotation in active F-ATPase. Nature 381, 623–625.
Schulenberg, B., Wellmer, F., Lill, H., Junge, W. and Engelbrecht, S. (1997). Cross-linking of chloroplast FoF1-ATPase subunit εto γ
without effect on activity. ε and γare parts of the rotor. Eur. J.
Biochem. 249, 131–141.
Soper, J. W., Decker, G. L. and Pederson, P. L. (1979).
Mitochondrial ATPase complex. A dispersed, cytochrome deficient, oligomycin-sensitive preparation from rat liver mitochondria containing molecules with a tripartite structural arrangement. J. Biol. Chem. 254, 11170–11176.
Tang, C. and Capaldi, R. A. (1996). Characterization of the interface
between γand εsubunits of Escherichia coli F1-ATPase. J. Biol. Chem. 271, 3018–3024.
Telford, J. N., Langworthy, T. A. and Racker, E. (1984). Three
proton pumps, morphology and movements. J. Bioenerg.
Biomembr. 16, 335–351.
Uhlin, U., Cox, G. B. and Goss, J. M. (1997). Crystal structure of
the εsubunit of the proton translocating ATP synthase from E. coli.
Structure 5, 1219–1230.
Watts, S. D., Tang, C. and Capaldi, R. A. (1996). The stalk region
of E. coli ATP synthase: Tyr 205 of the γsubunit is in the interface between F1and Foparts and can interact with both the εand the c
oligomer. J. Biol. Chem. 271, 28341–28347.
Weber, J. and Senior, A. E. (1997). Catalytic mechanism of F1
ATPase. Biochim. Biophys. Acta 1319, 19–58.
Wilkens, S. and Capaldi, R. A. (1998a). Two stalks link the F1to
the Fopart in the E. coli ATP synthase. A rotor and a stator? Nature
393, 29.
Wilkens, S. and Capaldi, R. A. (1998b). Electron microscopic
evidence of two stalks linking the F1and Foparts of the Escherichia coli ATP synthase. Biochim. Biophys. Acta 1365, 93–97.
Wilkens, S. and Capaldi, R. A. (1998c). Solution structure of the ε
subunit of the F1 ATPase from E. coli and interactions of this
subunit with β subunits in the complex. J. Biol. Chem. 273, 26645–26651.
Wilkens, S., Dahlquist, F. W., Mcintosh, L. P., Donaldson, L. W. and Capaldi, R. A. (1995). Structural features of the εsubunit of the Escherichia coli ATP synthase (ECF1Fo) from
NMR-spectroscopy. Nature Struct. Biol. 2, 961–967.
Wilkens, S., Dunn, S. D., Chandler, J., Dahlquist, F. W. and Capaldi, R. A. (1997). Solution structure of the N terminal domain
(residues 1–134) of the δ subunit of the E. coli ATP synthase (ECF1Fo). Nature Struct. Biol. 4, 198–201.
Zhang, Y. and Fillingame, R. H. (1995). Subunits coupling H+
transport and ATP synthesis in the Escherichia coli ATP synthase: Cys–Cys cross linking of F1subunit εto the polar loop of Fosubunit c. J. Biol. Chem.270, 24609–24614.
Zhou, Y., Duncan, T. M. and Cross, R. L. (1997). Subunit rotation
