Contents of this page:
H+ flux linked to ATP synthesis or hydrolysis
Composition & roles of major domains of the ATP Synthase
Binding change mechanism
Structure of F1 & central stalk
Evidence for rotation
Fo & peripheral stalk subunits
F1Fo ATP Synthase of mitochondria, chloroplasts, and bacteria is represented schematically at right. When the electrochemical H+ gradient is favorable, F1Fo catalyzes ATP synthesis coupled to spontaneous H+ flux toward the side of the membrane where F1 protrudes. E.g., in mitochondria, the pH and electrical gradients drive H+ transport from the intermembrane space to the matrix compartment. The chemiosmotic theory is discussed elsewhere.
If no membrane potential or pH gradient exists to drive the forward reaction, the Keq favors the reverse reaction, ATP hydrolysis (ATPase activity).
In some bacteria, the reverse reaction has a physiological role, providing a mechanism for ATP-dependent creation of a proton gradient that drives other reactions.
Viewed by electron microscopy with negative staining, the ATP synthase appeared as "lollipops" on the inner mitochondrial membrane, facing the matrix (V & V Fig. 22-36 p. 827). Higher resolution cryo-electron microscopy later showed each lollipop to have two stalks. E.g., see movie on a website of J. Rubinstein.
Roles of major subunits were determined in studies of submitochondrial particles (SMP). If mitochondria are treated with ultrasound, the inner membrane breaks and reseals as vesicles, with F1 on the outer surface. Since F1 of intact mitochondria faces the interior matrix space, these SMP are said to be inside out.
F1, the lollipop head, when extracted from SMP, catalyzes ATP hydrolysis (the spontaneous reaction in the absence of an energy input). Thus F1 contains the catalytic domain(s).
After removal of F1, the SMP membrane containing Fo is leaky to H+. Adding back F1 restores the normal low permeability to H+. Thus it was established that Fo includes a "proton channel."
Either oligomycin or DCCD blocks the H+ leak in membranes depleted of F1. Thus oligomycin and DCCD inhibit the ATP Synthase by interacting with Fo.
|ATP synthase complexes of
bacteria, mitochondria and chloroplasts are all very similar, with only
minor differences. Mitochondria are
believed to have evolved from symbiotic aerobic
bacteria ingested by an anaerobic host cell. The limiting
membrane of the bacterium became the inner mitochondrial membrane.
Mitochondria contain a small DNA chromosome, but genes that encode most
mitochondrial proteins are located in the nucleus, consistent with
transfer of some DNA to the nucleus during evolution.
The subunit composition of the ATP Synthase was first established for E. coli, which has an operon that encodes genes for all subunits. Stalk subunits were classified initially as being part of either F1 or Fo, based on whether they co-purified with extracted F1.
Mammalian mitochondrial F1Fo is slightly more complex than the bacterial enzyme, with a few additional subunits. Also, since names were assigned based on apparent molecular weights, some subunits were given different names in different organisms.
There is evidence that the ATP Synthase (F1Fo) may form a complex with the adenine nucleotide translocase (ADP/ATP antiporter) and the phosphate carrier (Pi/H+ symporter). This complex has been designated the ATP Synthasome.
|The binding change mechanism of energy coupling
was proposed by Paul Boyer. He shared the Nobel
prize for this model, which accounts for the existence of 3
sites in F1.
For simplicity, only the catalytic b subunits are shown in the diagram at right.
It is proposed that an irregularly shaped "shaft" linked to Fo rotates relative to the 3b subunits, which are arranged in a ring. The rotation is driven by flow of H+ through Fo. The conformation of each b subunit changes sequentially, as it interacts with the rotating shaft. Each of the 3 b subunits is in a different stage of the catalytic cycle at any time. For example, the green subunit shown above sequentially changes to:
This model is supported by two major lines of evidence:
1. The crystal structure of F1 with the central stalk was determined by John Walker, who shared the Nobel prize for that achievement. The g (gamma) subunit was found to include a bent helical loop that constitutes a "shaft" within the ring of a and b subunits.
|Shown at right is bovine F1,
treated with DCCD to yield crystals in
which more of the central stalk is ordered, allowing structure determination. (Structure solved by C. Gibbons,
M. G. Montgomery, A. G. W. Leslie, & J. E. Walker, 2000, PDB 1E79).
Subunit colors: a yellow, b green, g red, d blue, and e magenta.
Note the wide base of the rotary shaft, including part of g as well as d and e subunits.
Recall that the bovine d subunit, which is located at the base of the shaft, is equivalent to the e subunit of bacterial F1.
In crystals of F1 not treated with DCCD (PDB file 1COW), less of the shaft structure is elucidated, but ligand binding may be observed under more natural conditions.
The 3 b subunits are found to differ in conformation and bound ligand:
Bound to one b subunit is a non-hydrolyzable analog of ATP (assumed to be the tight conformation).
Bound to another b subunit is a molecule of ADP (assumed to be the loose conformation).
The third b subunit has an empty active site (assumed to be the open conformation).
These findings are consistent with the binding change mechanism, which predicts that each of the three b subunits, being differently affected by the irregularly shaped rotating shaft, will be in a different stage of the catalytic cycle.
Additional data are consistent with there being an intermediate conformation between the major transitions discussed above. This intermediate conformation may have nucleotide bound at all three sites. By one model, considering the left-most image in the diagram above: ATP synthesis (on the green subunit) is associated with transition to an intermediate conformation that allows binding of ADP + Pi to the adjacent, previously empty site (magenta subunit). A further conformational change then occurs as ATP formed in the previous step is released (from the cyan subunit).
See also recent articles, especially the paper by Kagawa et al.
Explore at right the structure of bovine F1 with bound ADP and
The non-hydrolyzable AMPPNP is used as a substitute for ATP, which would hydrolyze during crystallization.
2. Rotation of the g shaft relative to the ring of a and b subunits was demonstrated by H. Noji, R. Yasuda, M.Yoshida & K. Kinoshita.
b subunits of a bacterial F1 were tethered to a glass surface, as represented at right. A fluorescent-labeled actin filament (shown in yellow) was attached to the protruding end of the g subunit.
Video recording showed the fluorescent actin filament rotating like a propeller. The rotation was found to be ATP-dependent.
Studies using varied techniques have shown ATP-induced rotation to occur in discrete 120o steps, with intervening pauses. Some observations indicate that each 120o step consist of 80-90o and 30-40o substeps, with a brief intervening pause. Such substeps are consistent with evidence for an intermediate conformation between the major transitions, discussed above.
Although the binding change mechanism is widely accepted, some details of the reaction cycle are still debated.
View videos showing F1 rotation, at a website that includes details of the experimental approach used.
Then view at right an animation based on observed variation in conformation of F1 subunits attributed to rotation of the g shaft.
The c subunit of Fo has a hairpin structure, with 2 transmembrane a-helices and a short connecting loop. (Structure at right determined via NMR by M. E. Girvin, V. K. Rastogi, F. Abildgaard, J. L. Markley, & R. H. Fillingame, 1998).
The small c subunit (79 amino acid residues in E. coli), is also referred to as proteolipid, because of its hydrophobicity.
One a-helix of the c subunit includes an aspartate or glutamate residue whose carboxyl group reacts with DCCD (Asp61 in E. coli). Mutation studies have shown this DCCD-reactive carboxyl group, which is located in the middle of the bilayer, to be essential for H+ transport through Fo.
View at right a low resolution, partial structure of yeast F1 with the central stalk and attached Fo c subunits (D. Stock, A. Leslie, & J. Walker, 1999, PDB file 1Q01).
Display as backbone and color chain.
Question: How many c subunits, are in the Fo c-ring?
Visualize the aspartate residue near the middle of one transmembrane segment of each c subunit.
An atomic resolution structure of the complete ATP Synthase, including F1 and Fo with peripheral as well as central stalks, has not yet been achieved. However partial or complete structures of individual protein constituents, mutational studies, and evidence for inter-subunit interactions, have defined the roles of most subunits.
The image at right, depicting models of mitochondrial and bacterial ATP Synthase subunit structure, was provided by Dr. John Walker. Keep in mind that some equivalent subunits from different organisms are assigned different names.
The proposed "rotor" consists of the ring 10 c subunits, plus the central stalk (subunits g, d, & e in the mitochondrial enzyme; or g & e in E. coli).
The proposed "stator" consists of the 3a and 3b F1 subunits, the a subunit of Fo, and a peripheral stalk that connects these. The peripheral stalk consists of 2b, and d in E. coli, or subunits b, d, F6, and OSCP in bovine mitochondria..
Mitochondrial ATP Synthase E. coli ATP Synthase
The b subunit includes a membrane anchor, one transmembrane a-helix in E. coli and two in mammalian F1Fo, that interacts with the intramembrane a subunit. A polar, a-helical domain of b extends out from the membrane.
OSCP, which is homologous to the E. coli d subunit, interacts with the protruding end of the b subunit and with the distal end of an F1 a subunit. This linkage, along with interactions of the b subunit with residues on the surface of F1, are postulated to hold back the ring of a and b subunits, keeping it from rotating along with the central stalk.
The a subunit of Fo (271 amino acid residues in E. coli) is predicted, e.g., from hydropathy plots, to include several transmembrane a-helices.
It has been proposed that the intramembrane a subunit contains two half-channels or proton wires (each a series of protonatable groups or embedded water molecules), that allow passage of protons between the two membrane surfaces and the interior of the bilayer.
Protons may be relayed from one half-channel or proton wire to the other only via the DCCD-sensitive carboxyl group of a c-subunit. Recall that the essential carboxyl group of each c-subunit (Asp61 in E. coli) is located half way through the membrane (see above). An essential arginine residue on one of the transmembrane a-subunit a-helices has been identified as the group that accepts a proton from Asp61 and passes it to the exit channel.
As the ring of 10 c subunits rotates, the c-subunit carboxyls relay protons between the 2 a-subunit half-channels. This allows H+ gradient-driven H+ flux across the membrane to drive the rotation.
It has been proposed that rotation of the ring of c subunits may result from concerted swiveling movements of the c-subunit helix that includes the DCCD-sensitive Asp61 and transmembrane a-subunit helices having the residues that transfer H+ to or from Asp61, as protons are passed from or to each half-channel. See also Fig. 22-43 p. 832.
Copyright © 1998-2007 by Joyce J. Diwan. All rights reserved.
Additional material on F1Fo