ATP Synthase

A power plant the size of a protein

Every contraction of your heart, every spark across a neuron, every molecule your liver detoxifies runs on ATP — adenosine triphosphate. When a cell needs energy, it snaps the terminal phosphate bond of ATP, releasing about 30.5 kJ/mol under standard conditions (and closer to 50 kJ/mol inside a real cell). The problem is that ATP is spent almost as fast as it is made. Your body holds only about 250 grams of ATP at any instant, yet it consumes roughly 40 to 80 kilograms of it per day. That accounting works only because each ATP molecule is recharged hundreds of times a day — and the device that recharges it, again and again, is ATP synthase.

ATP synthase is the terminal enzyme of oxidative phosphorylation. It is not a furnace that burns fuel directly. Instead it is a turbine: it harvests energy that has already been stored as a proton gradient across a membrane and converts the flow of those protons into the synthesis of a chemical bond. In your cells it lives in the inner mitochondrial membrane. In plants and algae it lives in the thylakoid membrane of the chloroplast, where light, not food, charges the gradient. In bacteria and archaea it lives in the plasma membrane. The architecture is so deeply conserved that the bacterial and human versions are unmistakably the same machine — a sign that this motor was running in the last universal common ancestor billions of years ago.

Chemiosmosis: charging the battery

To understand the motor you first have to understand its fuel. In the 1960s Peter Mitchell proposed an idea so heretical it took fifteen years to win acceptance: the energy from food is not passed directly, hand-to-hand, from one molecule to ATP. Instead the electron transport chain uses the energy from electrons falling from NADH and FADH₂ down to oxygen to pump protons across the inner mitochondrial membrane, from the matrix into the intermembrane space. This builds up a difference in proton concentration (a pH gradient of roughly one unit) and an electrical charge difference across the membrane. Together these form the proton-motive force, an electrochemical battery of about 150 to 200 millivolts — enormous for a membrane only about 4 nanometers thick, equivalent to a field strength of tens of millions of volts per meter.

This is chemiosmosis, and it earned Mitchell the 1978 Nobel Prize in Chemistry. The proton-motive force is a form of stored potential energy, like water held behind a dam. ATP synthase is the sluice gate and the turbine combined: it provides the only easy path for protons to flow back down the gradient, and it makes them pay a toll. Every proton that flows back through the enzyme does a small amount of work, and the enzyme banks that work as ATP.

Two motors on one shaft: Fo and F1

The enzyme's name, F1Fo-ATP synthase, encodes its two-part design. The "o" in Fo originally stood for oligomycin, an antibiotic that jams the membrane portion (it is the letter O, not a zero). Think of the whole thing as two rotary motors mounted on a common axle, one inside the membrane and one outside it, connected so that whatever one does, the other must follow.

• Fo — the proton turbine. Embedded in the membrane, Fo is built around a ring of identical c-subunits (the c-ring) plus a stationary a-subunit. The a-subunit has two offset half-channels that don't quite meet. A proton enters from the high-concentration side, binds a charged amino acid (a conserved aspartate or glutamate) on one c-subunit, rides the ring almost all the way around, then exits into the low-concentration side. Because protons can only get on and off at specific points, their flow forces the c-ring to rotate in one direction — a Brownian ratchet biased by the gradient.
• F1 — the catalytic head. Sitting in the watery interior (the matrix), F1 is a roughly spherical assembly of three α-subunits and three β-subunits arranged like the segments of an orange, with a central γ-subunit stalk running up its core. The three β-subunits hold the catalytic sites where ADP and phosphate are joined. F1 alone, detached from the membrane, will happily run backwards and hydrolyze ATP — which is how its rotary nature was first proven.
• The rotor and the stator. The c-ring and the γ-stalk are glued together as one rotating unit. To keep the α₃β₃ head from simply spinning along uselessly with the rotor, an external peripheral stalk (the b-subunits and others) clamps the head to the stationary a-subunit — like the casing of a drill holding still while the bit turns. The rotor turns; the stator holds.

The binding-change mechanism

How does a spinning shaft make a chemical bond? The answer is Paul Boyer's binding-change mechanism, for which he and structural biologist John Walker shared the 1997 Nobel Prize (Walker solved the atomic structure of F1 by X-ray crystallography, confirming Boyer's prediction in beautiful detail). The key insight is that the γ-stalk is bent and asymmetric, like a crankshaft. As it spins inside the α₃β₃ head, it pushes against each β-subunit in turn, forcing each one to cycle through three distinct shapes:

• Open (O): the site is empty and relaxed — finished ATP has just left, and the site can now grab fresh ADP and phosphate.
• Loose (L): the site has loosely bound ADP and inorganic phosphate but cannot yet react.
• Tight (T): the γ-stalk squeezes the site shut, clamping ADP and phosphate so close together that they spontaneously condense into ATP. Remarkably, forming the bond at this stage needs almost no energy input — the energy of the proton gradient is spent not on making the bond but on prying the finished, tightly bound ATP off the enzyme and on resetting the cycle.

Because the three β-subunits are 120° apart and the γ-stalk is asymmetric, at any instant the three sites are in three different states — one open, one loose, one tight. Each 120° step of rotation advances all three sites to the next state, and one full 360° revolution therefore releases three ATP. Single-molecule experiments by Masasuke Yoshida and Kazuhiko Kinosita, who glued a fluorescent actin filament to the γ-stalk and literally watched it spin under a microscope, confirmed the discrete 120° steps — and even resolved them into sub-steps tied to proton binding and ATP release.

The gearing: protons, c-rings, and yield

Here is where the engineering gets elegant. The F1 head always makes three ATP per turn — that is fixed by its three-fold symmetry. But the number of protons per turn is set by the number of c-subunits in the Fo ring, because each c-subunit carries exactly one proton across per revolution. Different organisms have evolved different ring sizes, and that single number sets the proton cost of ATP and the minimum gradient the enzyme can run on.

Organism / source	c-subunits in ring	Protons per turn	Protons per ATP (≈ ring ÷ 3)
Mammalian mitochondria (e.g. bovine, human)	8	8	~2.7
Yeast mitochondria	10	10	~3.3
E. coli	10	10	~3.3
Chloroplast (spinach)	14	14	~4.7
Alkaliphilic bacteria (Bacillus sp.)	13–15	13–15	~4.3–5.0

The pattern reveals a trade-off. A small ring (mammals, 8 subunits) is a low-gear, high-efficiency design: few protons spent per ATP, but it needs a strong, steady gradient. A large ring (chloroplasts and alkaliphiles, 14–15 subunits) is a high-torque, low-gear design: it burns more protons per ATP but can keep turning even when the gradient is weak — exactly the situation an alkaliphile faces, living in an environment where its external pH fights against building a strong proton store. Evolution tuned the gear ratio to the habitat.

The performance of this motor is staggering for its size — the whole assembly is about 10 nanometers across and around 500–600 kilodaltons. A single ATP synthase spins faster than 100 revolutions per second and turns out several hundred ATP molecules per second. Its mechanical efficiency, the fraction of proton-gradient energy it converts into ATP rather than wasting as heat, approaches 90–100% under ideal conditions — better than any engine humans have built. Multiply by the tens of thousands of copies packed into the folded cristae of a single mitochondrion, and by the hundreds to thousands of mitochondria in an active cell, and the daily kilograms of recycled ATP stop sounding absurd.

A motor that runs both ways

ATP synthase is fully reversible — its very name as both "synthase" and "ATPase" admits this. Run protons through it and it makes ATP. Take the gradient away and feed it ATP, and it will hydrolyze that ATP and use the energy to pump protons the other way, spinning in reverse. This is not a flaw; it is a feature many cells exploit. Some bacteria deliberately run the enzyme backwards to maintain their membrane potential when respiration stalls. But for a cell with a fragile gradient — say a heart cell during a heart attack, when oxygen is cut off and the electron transport chain stops pumping — reverse running would be catastrophic, draining the cell's last ATP. Mammals guard against this with a dedicated inhibitor protein, IF1, that swings in and clamps the rotor the instant the gradient drops below a threshold, preventing the machine from cannibalizing the cell's reserves.

Evolutionary and clinical significance

Because ATP synthase is so ancient and so universal, it is a window into the deep history of life. Its homology with bacterial enzymes is one of the cornerstones of endosymbiotic theory: mitochondria and chloroplasts carry their own ATP synthases that resemble those of free-living bacteria, evidence that these organelles were once independent microbes. The enzyme's reversibility hints that the very first cells may have used proton gradients across mineral membranes — gradients that existed for free in alkaline hydrothermal vents — before they ever learned to pump their own.

Clinically, the enzyme is both a target and a vulnerability. Oligomycin blocks the Fo channel and is a classic research tool (and the source of the "o"). The antibiotic class that includes bedaquiline, used against drug-resistant tuberculosis, works by jamming the c-ring of Mycobacterium tuberculosis's ATP synthase — starving the pathogen of energy while sparing the differently shaped human enzyme. Inherited mutations in ATP synthase subunits cause severe mitochondrial diseases such as NARP and Leigh syndrome, which strike the most energy-hungry tissues — brain, heart, and muscle — first. And because cancer cells rewire their energy metabolism, ATP synthase and its inhibitor IF1 are active targets in oncology research. A motor this central to life is, inevitably, central to disease.