Cellular Respiration
Oxidative Phosphorylation: The Electron-Transport Chain
The whole point of oxidative phosphorylation is that electron flow and ATP synthesis are two separate machines linked by one thing: a proton gradient across the inner mitochondrial membrane. Electrons fall from NADH down a five-story redox ladder to oxygen, and the three big drops pump protons; ATP synthase then spends that gradient to mint ~30-32 ATP per glucose. Get the gradient as the currency and the famous misconceptions dissolve: why FADH2 yields less ATP, why uncouplers make heat instead of ATP, and why cyanide kills in minutes.
Cyanide kills in minutes, and the reason is one enzyme at the very bottom of the electron-transport chain. Block that last step and electrons have nowhere to fall, the whole chain backs up, the proton gradient collapses, and a cell that runs on oxygen suffocates from the inside while surrounded by air. To understand why a single blocked step is lethal that fast, you have to understand what the chain is actually doing, and the central idea is simpler and stranger than most students expect.
Oxidative phosphorylation is two machines, not one. The first machine drops electrons down a ladder and uses the released energy to pump protons across a membrane. The second machine lets those protons flow back and uses that to make ATP. The two are linked by exactly one thing: the proton gradient. Hold onto that, because nearly every famous misconception in this chapter dissolves the moment you treat the gradient as the currency.
The five-story building
Electrons fall the way water falls: from high potential energy to low. The measure of that potential is the reduction potential, E0’. A more negative E0’ means a molecule holds its electrons loosely and donates them eagerly; a more positive E0’ means it grabs electrons hard. Electrons flow spontaneously from negative toward positive, and the free energy released at each drop is ΔG°’ = −nFΔE°’.
Picture the chain as a five-story building. NADH sits in the penthouse at −0.32 V. Oxygen waits in the basement at +0.82 V. In between are the landmarks the field measured one by one: FMN at −0.30, the iron-sulfur clusters near 0, coenzyme Q around +0.04, cytochrome c at +0.25, cytochrome a at +0.29. The total fall from NADH to O2 is large, and the cell does not waste it as a single crash. It breaks the descent into stages, and at three of those stages the drop is steep enough to do work: pump a proton uphill, against its own gradient, out of the matrix.
The four complexes, in order
Everything happens on the inner mitochondrial membrane, folded into cristae and impermeable to ions. That impermeability is the whole foundation, because a gradient is only useful if it cannot leak away. The outer membrane is porous (VDAC lets through anything under ~5000 Da); the matrix inside runs the citric acid cycle and fatty-acid oxidation.
Complex I (NADH-Q oxidoreductase) is an L-shaped giant of ~46 subunits. NADH hands two electrons to FMN, which passes them through a relay of about seven Fe-S clusters to coenzyme Q. The drop is steep, and Complex I pumps about 4 protons.
Complex II (succinate dehydrogenase) is the quiet one, and it carries the chapter’s most-tested lesson. It is also a citric acid cycle enzyme, oxidizing succinate to fumarate, and it feeds the resulting FADH2 electrons to Q. But Complex II pumps zero protons. It is small, only four subunits, repurposed from the cycle without ever evolving the pumping hardware. This is the entire reason FADH2 yields less ATP than NADH. The electrons are real, but they enter below the first pump and skip it.
Complex III (cytochrome bc1) runs the Q cycle, which solves a packaging problem. Reduced QH2 carries two electrons, but cytochrome c can only take one at a time. So the complex splits them: one electron goes through the Rieske Fe-S cluster to cyt c1 and out to cyt c, while the other loops back through the b hemes to reduce another Q at a second site. Two turns reduce two cytochrome c and move 4 protons to the intermembrane space. The cost is a semiquinone radical (Q•⁻) that lingers at the Qp site, and if it meets O2 it makes superoxide. Complex III is both elegant and dangerous, which is why antimycin A blocks it as a classic experimental tool.
Complex IV (cytochrome c oxidase) is the terminal step, and it handles the most dangerous molecule in metabolism without ever fumbling it. Electrons pass from cyt c to CuA, to heme a, to the heme a3/CuB binuclear center, where O2 binds. Think of that binuclear center as a vise grip: it clamps O2 and refuses to release it until all four electrons have arrived, so no half-reduced, radical-spitting intermediate ever escapes. Four electrons plus four protons plus O2 give two water molecules. Per O2, Complex IV removes 8 protons from the matrix (4 become water, 4 are pumped). Cyanide and CO jam this site, and the chain dies here.
Cryo-EM now shows these complexes assembling into supercomplexes (the respirasome, I + III2 + IV), which may shorten the diffusion path for the mobile carriers and reduce leak. Whether that matters for normal function is still debated.
Here is the whole chapter in one sentence: the respiratory chain and ATP synthase are two different machines that never touch. The chain is a pump that charges a battery (the H⁺ gradient); ATP synthase is a motor that spends it. Cut the wire between them — add an uncoupler like 2,4-dinitrophenol — and electrons still flow, O₂ is still burned, but the protons leak back as heat and almost no ATP gets made. That is exactly how brown fat keeps a newborn warm and why a fever spikes your metabolic rate. It is also why DNP, sold as a 1930s diet pill, melted fat by literally turning the battery into a space heater — and killed people by cooking them from the inside. Same physics, two destinies, decided entirely by whether the protons go through the motor.
The proton budget, and the FADH2 tax
Now the accounting that earns the central idea. Pumping protons charges a battery, the proton-motive force. ATP synthase discharges it. The exchange rate, including the cost of importing ADP and exporting ATP across the membrane, runs about 4 protons per ATP.
NADH drives all three pumps, so it banks roughly 10 protons, which buys about 2.5 ATP. FADH2 enters at Complex II and skips the first pump, banking roughly 6 protons, which buys about 1.5 ATP. The difference traces to one structural fact: Complex II, alone among the four, was never built with proton-pumping subunits, so its electrons bank four fewer protons before they ever reach ATP synthase. Multiply across a glucose and you reach the headline number: ETC plus chemiosmotic gradient plus ATP synthase yields roughly 30 to 32 ATP per glucose, an order of magnitude beyond what glycolysis alone delivers.
| Stage | Direct ATP/GTP | NADH | FADH₂ | ATP from this stage |
|---|---|---|---|---|
| Glycolysis (cytosol) | 2 | 2* | 0 | — |
| Pyruvate → acetyl-CoA (×2) | 0 | 2 | 0 | — |
| Citric acid cycle (×2 turns) | 2 | 6 | 2 | — |
| Grand total per glucose | 4 | 10 | 2 | — |
— · —. * The 2 cytosolic NADH from glycolysis can't cross the mitochondrial membrane directly — a shuttle moves their electrons in, and which shuttle decides whether they're cashed as NADH or as FADH₂.
The famous "~38 ATP per glucose" is the old textbook number, built on round P/O ratios of 3 and 2. Measured proton stoichiometry gives the modern estimate of roughly 30–32, and the exact figure isn't a constant of nature — it depends on which shuttle a tissue uses and on the leakiness of the membrane. The lesson of this ledger isn't a number to memorize; it's that the cell does accounting, and the answer shifts with the assumptions you bring to the books.
When the chain breaks
The same logic that builds the chain explains how it fails. Inhibitors freeze it at a single point, and the freeze is diagnostic. Rotenone blocks Complex I, antimycin A blocks Complex III, cyanide blocks Complex IV. Apply any one and every carrier upstream piles up reduced while everything downstream goes oxidized. That crossover pattern, read by spectroscopy, is literally how the order of the chain was worked out before anyone had a structure.
Uncouplers punch a proton leak in the membrane. Protons rush back without passing through ATP synthase, so the energy of the fall comes out as heat instead of ATP. Your brown fat does this on purpose to keep you warm; the diet drug DNP did it lethally.
Reactive oxygen species are the standing tax of running on oxygen. About 1 to 2 percent of the O2 handled leaks out as superoxide, mostly at the Complex I flavin and the Complex III semiquinone (Complex IV almost never leaks). Superoxide dismutase converts two superoxides to H2O2 and O2, then catalase and peroxidases finish the cleanup.
These mechanisms are clinical. In Friedreich’s ataxia, a GAA-repeat expansion silences frataxin, Fe-S cluster assembly fails, and the ETC collapses first in the neurons and heart that demand the most ATP. In ischemia-reperfusion injury, electrons queue up in an oxygen-starved chain during a heart attack or stroke, then dump onto O2 in a ROS tsunami the instant blood flow returns, which is why the first 24 hours after an MI are so dangerous. Even the Gulf of Mexico dead zone is this chemistry at ecosystem scale: nutrient runoff fuels a bloom, the bloom’s respiration strips the water of oxygen, and the terminal electron acceptor everything depends on simply runs out.
Next, in CH21, we follow the gradient into the second machine: ATP synthase, the rotary motor that spends the battery this chapter charged.
How we measure it
Difference spectroscopy of cytochromes
Keilin's microspectroscope read respiration in real time: oxidized and reduced cytochromes absorb light at different wavelengths, so watching absorption bands appear and vanish as oxygen is added or consumed tells you which carrier is reduced and which is oxidized. Add an inhibitor and the chain freezes in a 'crossover' pattern (reduced upstream, oxidized downstream), and that pattern is how the order of the carriers was deduced before any structure existed.
Inhibitor crossover analysis
Specific blockers cut the chain at known points: rotenone at Complex I, antimycin A at the Complex III Qi site, cyanide and CO at Complex IV. Apply one, and every carrier upstream of the block piles up reduced while everything downstream goes oxidized. Mapping that boundary for each inhibitor pins down where each component sits in the sequence and is the classic logic for ordering an electron-transfer pathway.
Submitochondrial particles and the pH-electrode pulse
Tear mitochondria apart and the inner membrane reseals into closed vesicles that still pump and still make ATP, proving OxPhos lives on the membrane while the matrix runs the soluble reactions. Mitchell and Moyle put a fast glass pH electrode in a mitochondrial suspension, pulsed in oxygen, and watched the medium acidify, measuring close to two protons ejected per ATP and turning chemiosmosis from hypothesis into number.
Cryo-EM of supercomplexes
Single-particle cryo-electron microscopy froze whole respiratory chains and revealed the 'respirasome' (Complex I + III2 + IV) assembling into supercomplexes, which may shorten the diffusion path for CoQ and cytochrome c and reduce electron leak. The same technique solved the individual complexes (Complex I's L-shaped 46-subunit machine, the bc1 Q-cycle, cytochrome c oxidase) at near-atomic resolution.