10 February 2004
We are probably all aware of what a dynamo is, if not precisely how it works. For the enthusiasts, a dynamo is a machine that converts mechanical energy to electrical energy via the motion of a conducting material through a magnetic field. We may well think this is fairly smart science, but it appears as though nature has been manufacturing its own dynamos for at least 2 or 3 billion years. Writing in Nature, Hiroyasu Itoh and colleagues describe an ingenious approach to investigate the function of the F1-ATPase (1). It appears as though the essential energy-generating component of all plants, animals, fungi and many bacteria works in a similar way to an electrical generator. Where we use water or steam, nature uses protons and where we make electricity, nature manufactures ATP.
ATP-synthase is a ubiquitous enzyme found throughout nature and it is the principal means by which ATP is synthesised in intermediary metabolism. By contrast ATP production through direct synthesis (e.g. the coupled conversion of phosphoenoylpyruvate and ADP to pyruvate and ATP via the glycolytic pathway) produces only 6% that made by the conversion of glucose to CO2 and water through oxidative phosphorylation.
At some point between 2 and 3 billion years ago nature stumbled across the use of chemical gradients to generate usable energy. The key to this strategy is the conversion of the energy locked within chemical gradient (and there are many routes to creating this gradient) to a usable form: the cell has to be able to generate ATP using this gradient. This is not an easy task. Ions will naturally tend to diffuse down concentration gradients or move under the influence of applied voltages. However, getting something as small as a proton to make something as large as a molecule of ATP seems a tall order. How does the cell do it?
ATP-synthase consists of two linked complexes, F0 in which protons translocates through the membrane and the F1 complex in which ATP synthesis occurs. These two complexes, which are made up of a number of different subunits, are joined to one another through an ‘axle protein’ - the g-subunit (2). A number of different mechanisms for ATP-synthase had been suggested based on this structure (2,3,4). The most promising route involved the rotation of the entire F0 complex (made up of 12 ‘c’-subunits) coupled to the synthesis of ATP through conformational changes in the associated trimeric F1 component (5). Itoh and colleagues elegantly demonstrate that this is true (1). Protons enter compartments in one of the ‘c’-subunits via a partial channel in the tightly associated ‘a’ subunit. However once bound here, these protons are unable to flow directly through the protein. Instead channelling through to the other side of the membrane (down concentration and pH gradients) is achieved by the rotation of the whole F0-complex of ‘c’-subunits and the g-subunit through 120o. Once rotation has occurred, the compartment in which the proton is held, encounters the second partial channel in the ‘a’ subunit that connects with the opposite side of the membrane. The proton can now escape.
But what of ATP synthesis? Itoh and co-workers did some lateral thinking. It had been known from previous work that the motor protein myosin would synthesise ATP spontaneously from ADP and phosphate. However, this ATP remained irreversibly bound to the protein unless an external force was applied (3,6). They then surmised that if ATP-synthase could be rotated manually, the ATP that was generated spontaneously, from ADP and inorganic phosphate, might then be released. Itoh and colleagues took the ‘axle’ and the F1 subunit, link the axle to a magnetic bead and the F1 head to a surface via histidine tags. Using an externally applied magnetic field the bead was rotated, dragging the g-subunit with it. In a medium enriched in ADP and inorganic phosphate, ATP synthesis was detected through luciferase activity. For every rotation through 120o, at a frequency of 3Hz, 5 molecules of ATP were synthesised – roughly half of the expected 9 ATP molecules observed in vitro (5). (By contrast, the reverse ATP-hydrolysis reaction drove the bead at a maximum rotation frequency of 130Hz when it was allowed to turn freely in the presence of excess ATP.) The authors attribute the slightly reduced rate of ATP-synthesis to the slow 3Hz rotation that would allow near-equilibrium conditions to operate throughout ATP-synthesis (1).
Through comparisons with myosin and other molecular motors the mechanism of ATP-synthesis appears as follows. In one of the three F1 subunits, molecules of phosphate and ADP bind. Coupled rotation of the F0 and g-subunits alters the shape of the ADP/phosphate-binding pocket forcing the condensation reaction to occur. Normally the product ATP molecule would remain bound. However, further rotation of the g-subunit, driven by proton translocation, forces the expulsion of the ATP-product molecule through further conformational change. As this happens in one F1 subunit, synthesis of ATP is occurring at differing stages in the two other subunits. For every 120o turn one molecule of ATP is made; 3 for every 360o full turn.
Although the authors are still working on improving their experimental set-up, this is the first demonstration of mechanically driven ATP-synthesis. It is also an elegant reminder that however fancy we may think our technology is, you can bet nature got there first.
1) Mechanically driven ATP synthesis by F1-ATPase. Hiroyasu Itoh, Akira, Takahashi, Kenge Adachi, Hiroyuki Noji, Ryohei Yasuda, Masasuko Yoshida and Kazuhiko Kinosita Jr. (2004). Nature, 427, 465-468.
2) Structure at 2.8A resolution of F1-ATP synthase from bovine heart mitochondria Abrahams, J.P., Leslie, A.G.W., Lutter, R. and
3) The ATP synthase – a splendid molecular machine. Boyer, P.D. (1997). Ann Rev. Biochemistry, 66, 717-749.
4) A rotary molecular motor that can work at near 100% efficiency. Yoshida, R., Noji, H. and Kinosita, M. (2000). Phil. Trans. Act. Royal. Soc. Lond. B. 355, 473-489.
5) F1 ATP synthase is a highly efficient molecular motor that rotates with discreet 120o steps. Yoshida, R., Noji, H. and Kinosita, M. (1998). Cell, 93, 1117-1124.
6) The reversal of the myosin and actinomyosin ATPase reactions and the free-energy of ATP binding to myosin. Wolcott, R.G. and Boyer, P.D. Eur. J. Biochem., 48, 287-295.