The Proton-Motive Force and ATP Synthase

In the last chapter we built a dam. Complexes I through IV spent the energy of electrons falling toward oxygen to pump protons out of the mitochondrial matrix, stacking them in the intermembrane space like water rising behind a wall. That dam stored real energy. Today we open the floodgates.

The Itaipu Dam holds back a reservoir for one reason: height stores energy, water or not — the same mgh physics whether you are damming a river or stacking protons. Let it fall through a turbine and the height becomes electricity. A mitochondrion does the same trick with protons. The reservoir is the proton gradient; the turbine is ATP synthase, an actual rotary motor a few nanometers wide; and the electricity is ATP. Everything in this chapter is one idea wearing different clothes: electron transport and ATP synthesis are a single coupled machine, joined by protons.

The seventeen-year argument

Here is the part most textbooks rush past. The person who figured this out was ridiculed for nearly two decades.

In 1961 Peter Mitchell proposed that the energy linking electron flow to ATP synthesis is not stored in some high-energy chemical intermediate — the molecule everyone was hunting and no one could find — but in a gradient of protons across a sealed membrane. He called it the chemiosmotic hypothesis. The field’s reaction ranged from skeptical to scornful. One contemporary recalled that some treated Mitchell as a court jester. Forced out of Edinburgh by illness, he restored a half-ruined Cornish mansion, founded a private institute partly funded by a prize-winning dairy herd, and kept measuring protons with his collaborator Jennifer Moyle.

The vindication came from experiments you can hold in your head. Jagendorf and Uribe soaked chloroplast membranes in acid, then jumped them into base — a hand-built proton gradient, no light, no electron transport — and watched them make ATP in the dark. Racker and Stoeckenius went further: they rebuilt artificial vesicles containing only purified lipids, bacteriorhodopsin (a light-driven proton pump borrowed from a salt-loving microbe), and ATP synthase. Shine light, and the pump builds a gradient; the gradient drives the synthase; ATP appears. No respiratory chain anywhere in the vesicle. That experiment ended the argument. In 1978 the Nobel phone call reached Mitchell at lunch. Asked his reaction, he gave one word: “astonishment.”

What the dam actually stores: Δp

The proton-motive force, written Δp, has two parts, and forgetting one is the single most common mistake here.

The first is chemical: protons are simply more concentrated outside the matrix, a difference of roughly 1.4 pH units. The second is electrical: pumping positive charge outward leaves the matrix negative, a membrane potential ΔΨ of about 0.14 volts. You need both. Together they come to roughly 21 kJ per mole of protons returning — and a student who tracks only the pH difference throws away about half the energy. Note the asymmetry the cell maintains: the matrix is both alkaline and negative, which is exactly what pulls a positive proton back in.

A motor, not a metaphor

ATP synthase has two parts named, unhelpfully, F_o and F₁.

F_o sits in the membrane and is the proton channel: a ring of c-subunits (the rotor) pressed against a stationary a-subunit that holds two offset half-channels. A proton enters the half-channel from the intermembrane space and protonates a conserved acidic residue on the c-ring — a glutamate in mitochondria, an aspartate in E. coli. Neutralizing that charge lets the ring rotate one notch into the oily membrane interior; one full turn later, the proton is released into the matrix through the second half-channel. Protons walk the ring around, one step at a time.

F₁ is the catalytic head poking into the matrix: three α and three β subunits arranged like orange segments around a single bent, asymmetric γ shaft. The c-ring and γ are bolted together. So the proton flow that spins the rotor spins the shaft, and the shaft’s own rotation is what drives chemistry: γ’s off-center bend touches only one β subunit at full force at any instant, forcing that subunit through a shape change while holding the other two in different states — the mechanical seed of the three-state cycle below.

Boyer’s inside-out insight

Here is the question Paul Boyer posed that reorganized the whole field. What would you say if I told you it takes no energy to make ATP at the catalytic site — but it takes energy to get the ATP off?

That is the binding-change mechanism. Each β subunit cycles through three states as γ turns: Open (loosely binds ADP and Pi, or releases product), Loose (substrates trapped), and Tight (ATP forms spontaneously, held so firmly it won’t leave). The proton gradient’s job is to wrench the tight site open and let the finished ATP go. Each 120° turn of γ advances one β through one state, so a full 360° rotation releases three ATP. John Walker later caught all three β subunits in different nucleotide states in a single crystal — a frozen snapshot of Boyer’s cycle, drawn one structure at a time.

You can watch it turn. Noji and Yoshida glued isolated α₃β₃ heads to glass, attached a fluorescent actin filament to the γ shaft, added ATP, and saw the filament rotate in discrete 120° steps — roughly 130 revolutions per second, about 400 ATP per second. The world’s smallest motor, running under a microscope. And run it backward — isolated F₁ with no gradient will hydrolyze ATP and spin the other way. The gradient does not make the chemistry possible; it sets the direction.

Counting protons and ATP

The vertebrate c-ring has 8 subunits, the most efficient stoichiometry known. Eight protons per turn, three ATP per turn, gives about 2.7 H⁺/ATP. But the ATP must still leave the matrix: the ATP–ADP translocase (the most abundant protein in the inner membrane) swaps ATP⁴⁻ out for ADP³⁻ in, an electrogenic 1:1 exchange that lets ΔΨ pull one net negative charge outward — a proton tax on every ATP shipped. Add the phosphate carrier importing Pi with a proton, and the real cost is closer to ~3.7 H⁺ per exported ATP.

Two more wrinkles set the final glucose tally. Cytoplasmic NADH from glycolysis can’t cross the inner membrane, so it hands its electrons across by a shuttle. The glycerol-3-phosphate shuttle (muscle, brain) delivers to FAD and CoQ, bypassing Complex I — fast, but worth only ~1.5 ATP. The malate-aspartate shuttle (heart, liver, kidney) regenerates matrix NADH — slower, but worth ~2.5 ATP. Speed versus efficiency, and the tissue chooses. This is why the honest answer is ~30–32 ATP per glucose, with NADH ≈ 2.5 and FADH₂ ≈ 1.5. The word “approximately” is doing real work, and it is fair game on an exam.

Setting the rate, and breaking it on purpose

The cell throttles all of this with acceptor (respiratory) control: ADP availability sets the pace. Plenty of ADP and respiration races (State 3); ADP runs low and the gradient backs up and the chain idles (State 4). The master gauge is energy charge, ([ATP] + ½[ADP]) / ([ATP] + [ADP] + [AMP]), held near 0.85–0.90 — a fuel gauge that simultaneously tunes PFK-1, isocitrate dehydrogenase, and pyruvate dehydrogenase.

Poisons let you dissect the coupling, and the three-way table they make is heavily tested. ETC inhibitors (rotenone at CI, antimycin A at CIII, cyanide/CO at CIV) stop pumping: O₂ down, ATP down, gradient down. Oligomycin plugs F_o: protons can’t fall, the gradient rises as back-pressure, and the stalled pumps stop consuming O₂ — so O₂ down, ATP down, gradient up. Uncouplers (DNP, FCCP) punch a leak in the membrane: the gradient drains without making ATP, and the chain races to refill it — so O₂ up, ATP down, gradient down. The biology runs its own uncoupler: UCP1 (thermogenin) in brown fat deliberately leaks protons to make heat instead of ATP, the basis of nonshivering thermogenesis you can light up on a PET-CT in a cold adult. (Pigs lack UCP1, which is why piglets shiver.) And the IF1 inhibitor protein locks the rotor at low pH during ischemia, stopping the synthase from running backward and burning ATP when the gradient collapses — a switch some cancers exploit.

Mutations bring it home. mtDNA is maternally inherited and present in many copies, so disease depends on heteroplasmy — the fraction of mutant genomes — and strikes high-demand tissues first. LHON (the m.11778G>A mutation in ND4, a Complex I subunit) blinds otherwise healthy young men over weeks. The proton gradient is older than mitochondria: chloroplasts, bacterial flagella, and bacteriorhodopsin all run on it.

We have now followed one glucose from the gut to the last drop of ATP. Next we turn the engine around and ask where the fuel itself comes from when the diet runs short — gluconeogenesis and the management of blood glucose.