Module 3: Oxidative Phosphorylation

Before 1961, bioenergetics was dominated by the search for a “high-energy intermediate” X∼I that would link oxidation to phosphorylation. This mythical molecule was never found. Peter Mitchell, working almost alone at Glynn House in Cornwall, proposed that the intermediate was not a chemical bond but a vectorial proton gradient maintained across a topologically closed membrane (Mitchell 1961, Nature, 191, 144–148).

The central claims of the chemiosmotic hypothesis are:

• Electron-transport complexes are vectorial pumps that translocate protons from the matrix to the intermembrane space (IMS).
• The inner mitochondrial membrane is impermeable to protons, so the gradient persists.
• ATP synthase (F₀F₁) is a reversible molecular motor that uses the gradient to synthesize ATP.
• Coupling is obligatory but indirect: any process that discharges the gradient (uncoupling, leak, ionophores) blocks ATP synthesis without blocking electron transport.

The hypothesis was met with intense skepticism. Paul Boyer and others argued for a conformational coupling model. Experimental vindication came gradually through the 1960s and 1970s: reconstitution of purified bacteriorhodopsin into vesicles with F₀F₁(Racker & Stoeckenius 1974) showed that a light-driven proton pump alone was sufficient to drive ATP synthesis—proof that the “intermediate” was indeed just a proton gradient. Mitchell received the Nobel Prize in Chemistry in 1978.

For a typical mammalian mitochondrion with \(\Delta\psi \approx 140\) mV (negative inside the matrix) and \(\Delta\text{pH} \approx 1.0\)(alkaline inside), the total PMF is\(\Delta p \approx 140 + 61.5 \times 1.0 \approx 200\) mV. In plants and some bacteria the balance is different: chloroplasts at pH 8 matrix vs. pH 5 stroma have almost all of their PMF in the \(\Delta\text{pH}\) term.

The free-energy change for moving one proton from the IMS (high [H⁺], low pH) to the matrix (low [H⁺], high pH) is

With \(F = 96{,}485\) C/mol and \(\Delta p = 0.2\) V, each proton stores \(\sim 19.3\) kJ/mol of free energy. If ATP synthase couples ~3 H⁺ to one ATP (Module 4 will refine to the c-ring stoichiometry), the input is \(\sim 58\) kJ/mol—comfortably exceeding the\(\Delta G_{\text{ATP}} \approx +50\) kJ/mol required under cellular conditions ([ATP]/[ADP][Pi] far from equilibrium).

Full oxidation of one NADH by O₂ releases\(\Delta G^\circ = -220\) kJ/mol. The ETC pumps approximately 10 H⁺ per NADH (4 at C1, 2 at C3, 4 at C4), capturing\(10 \times 19.3 \approx 193\) kJ/mol in the gradient, for a thermodynamic efficiency of \(\sim 88\%\). Downstream, ATP synthase converts the gradient to ATP at \(\sim 65\%\) efficiency, yielding an overall oxidation-to-ATP efficiency of ~40%.

Complex I is the entry point for electrons from NADH into the chain and the largest respiratory enzyme. Mammalian C1 contains 45 subunits in an L-shape with a hydrophilic arm protruding into the matrix (which hosts FMN and 8 iron-sulfur clusters) and a membrane arm embedded in the IMM (which contains the proton translocation machinery).

The reaction:

The electron path: NADH donates a hydride to FMN (flavin mononucleotide), producing FMNH₂. Electrons then tunnel through eight Fe-S clusters (N1a, N3, N1b, N4, N5, N6a, N6b, N2) spaced <14 Å apart (Moser-Dutton criterion for biological tunneling)—a physiological electron wire ~90 Å long from FMN to the Q-binding site. The final cluster N2 transfers electrons to bound quinone, which after two-electron reduction and protonation leaves as QH₂.

The proton pumping mechanism is uniquely long-range: four antiporter-like subunits (ND2, ND4, ND5, and the swapped-symmetry fourth channel in ND6/ND4L) are coupled to the Q-binding site by a ~100 Å lateral helix (HL) on the matrix side of ND5. This helix acts as a mechanical piston: Q reduction at the hydrophilic arm triggers conformational waves that propagate the length of the membrane arm, driving sequential protonation/deprotonation in each of the four half-channels. The mechanism was first proposed from the bacterial cryo-EM structure by Baradaran et al. (2013, Nature), refined in ovine C1 by Zhu et al. (2016, Nature), and in active/deactive conformations by Agip et al. (2020, Nat. Struct. Mol. Biol.).

Seven of the 45 subunits are mitochondrially encoded (ND1–ND6, ND4L) and constitute the proton-pumping core. This explains why Complex I is the most common site of inherited mitochondrial disease (LHON, MELAS-associated mutations, Leigh syndrome). The inhibitor rotenone binds the Q-binding pocket and is used experimentally to dissect electron flow; it was historically a fish poison used by indigenous peoples.

Complex I also has an active/deactive (A/D) transition: when oxygen is absent, the enzyme adopts an inactive “D” state that is slow to reactivate. On reperfusion, a large fraction of C1 is blocked in D-form, producing a lag that contributes to ischemia-reperfusion injury (Chouchani et al. 2013). During the reactivation phase, C1 runs in reverse as described below.

Complex II is the respiratory chain’s simplest enzyme and the only one that participates dually in the TCA cycle. It catalyzes the oxidation of succinate to fumarate and donates the resulting electrons directly to the ubiquinone pool. Unlike C1, C3, and C4, C2 does not pump protons—a thermodynamic consequence of the small redox drop from FADH₂to ubiquinone (\(\Delta E \approx 30\) mV).

The mammalian enzyme has four subunits:

• SDHA: catalytic flavoprotein, carries FAD covalently linked to histidine
• SDHB: iron-sulfur subunit with three Fe-S clusters [2Fe-2S], [4Fe-4S], [3Fe-4S]
• SDHC, SDHD: membrane anchors, cytochrome b₅₆₀ heme, Q-binding pocket

All four subunits are nuclear-encoded. Mutations in SDHB, SDHC, SDHD cause hereditary paraganglioma and pheochromocytoma (Baysal et al. 2000, Science). Loss of SDH activity causes succinate accumulation, which stabilizes HIF-1α by inhibiting prolyl hydroxylases—linking mitochondrial dysfunction to oncometabolism.

Malonate is the canonical competitive inhibitor of succinate dehydrogenase (structural analog of succinate). It has re-emerged as a therapeutic candidate in ischemia-reperfusion: preemptive malonate blocks the succinate buildup that would otherwise drive RET-dependent ROS (Valls-Lacalle et al. 2016).

Complex III (cytochrome bc₁) transfers electrons from QH₂ to cytochrome c and translocates 2 H⁺/2 e^-. It is a dimer of 11 subunits (mammalian) with three redox-active centers per monomer: cytochrome b (two hemes: low-potential b_L and high-potential b_H), cytochrome c₁ (one heme c type), and the Rieske iron-sulfur protein(ISP, [2Fe-2S] cluster with coordinating histidines).

The mechanism is the famous Q-cycle proposed by Mitchell himself (1975, FEBS Lett.) and later refined with cryo-EM. The electrons from QH₂ bifurcate at the outer Q-binding pocket (Q_o site, near the IMS):

• First electron: from QH₂ at Q_o to Rieske ISP, then to cytochrome c₁, then to soluble cytochrome c. Two protons released into IMS.
• Second electron: from the semiquinone Q•^- intermediate at Q_o to heme b_L, tunneling through the membrane to b_H, reducing a second Q at the inner Q_i site to semiquinone.
• A second round repeats and fully reduces the Q at Q_i to QH₂. Two protons taken up from matrix.

The Rieske ISP executes a dramatic ~60° mechanical rotation during catalysis, carrying its Fe-S cluster between contact with QH₂ at Q_o and delivery to cytochrome c₁. This “headpiece motion” was first revealed in yeast by Iwata et al. (1998, Science) and is the slowest step in the cycle.

Two classes of C3 inhibitor illustrate the bifurcation:

• Myxothiazol, stigmatellin: block Q_o site, prevent first-electron transfer
• Antimycin A: blocks Q_i site, traps semiquinone at Q_o → ROS generation

The antimycin-induced semiquinone at Q_o is the major physiological source of superoxide from C3 (covered in Module 5). The site is exposed to both leaflets of the IMM, so ROS appear on both sides—an important distinction from C1 which releases superoxide only into the matrix.

Complex IV (cytochrome c oxidase, COX) is the terminal electron acceptor of the chain. It accepts electrons from cytochrome c and uses them to reduce molecular oxygen to water. The mammalian enzyme has 13 subunits; the three catalytic subunits (COX1, COX2, COX3) are mitochondrially encoded, while the remaining ten are nuclear.

The catalytic center is a binuclear heme a₃-Cu_B site buried in the COX1 subunit. Heme a₃and Cu_B are only ~4.5 Å apart and coordinate O₂ as a peroxide bridge during turnover. Electrons enter from cytochrome c at a second copper site, Cu_A (dinuclear, in COX2), pass to heme a (in COX1), and then to the binuclear center for O₂ reduction.

The overall reaction (per 4 electrons, 1 O₂):

The reaction cycle passes through four spectroscopically distinct intermediates (R, A, P_R, F, O_H) that cleave the O=O bond without releasing harmful intermediates—a triumph of enzyme design, since free hydroxyl or superoxide released here would wreak oxidative havoc. The structure of bovine COX was solved at atomic resolution by Tsukihara et al. (1996, Science).

Cyanide, azide, and carbon monoxide are classic COX inhibitors. Cyanide binds heme a₃ as an axial ligand and brings respiration to a near-immediate halt, causing rapid death by energy starvation. COX has extraordinarily high O₂ affinity (\(K_m \approx 0.1\) µM), which is why tissue p_O2 can fall to extremely low values before respiration limits.

Electrons cross from complex to complex via two small mobile carriers:

• Ubiquinone (Q): a hydrophobic isoprenoid quinone (coenzyme Q₁₀ in humans) that diffuses within the lipid bilayer. Q/QH₂ redox potential \(\approx +60\) mV; 50–100 Q per complex in the pool. As a two-electron carrier, it shuttles 2 e^- + 2 H⁺per turnover, translating reducing equivalents efficiently across the membrane.
• Cytochrome c: a small (12 kDa, 104 residues) water-soluble heme protein that diffuses in the IMS along the outer surface of the IMM. Redox potential \(\approx +250\) mV; carries one electron per turnover. Also plays a second role: when released to the cytosol during apoptosis, it triggers caspase-9 activation and programmed cell death.

Cytochrome c is covalently tethered to its heme via two thioether bonds to cysteines in a conserved CXXCH motif—a rare modification found in few other proteins. The bovine holo-cytochrome c structure (Bushnell, Louie & Brayer 1990) shows a compact fold that presents a hydrophobic patch for docking to cytochrome c₁ and COX via surface electrostatic complementarity.

For decades the “fluid mosaic” view held that complexes I–IV diffuse independently in the IMM and encounter substrates by random Q/c collision. Blue-native PAGE (Schägger & Pfeiffer 2000, EMBO J.) changed this: mild solubilization preserves intact supercomplexeswith defined stoichiometries:

• I₁III₂: minimal respirasome core
• I₁III₂IV₁: the canonical mammalian respirasome
• I₁III₂IV₂: megacomplex in some tissues
• III₂IV_1-2: supercomplex without C1 (bacteria, plants)
• ATP synthase dimers: V-shape rows along cristae ridges

Cryo-EM structures of the mammalian respirasome by Letts et al. (2016, Nature), Sousa et al. (2016), and Gu et al. (2016) revealed the architecture: C1’s membrane arm docks against the C3 dimer, which in turn contacts C4. The assembly creates a protected micro-environment where Q and cyt c can diffuse short distances without being “lost” to the bulk membrane.

The adaptive value of supercomplex assembly is debated. Substrate channeling (Lapuente-Brun et al. 2013, Science) would speed electron transfer and reduce ROS escape. Alternative views emphasize structural stabilization without kinetic channeling (Blaza et al. 2014). A compromise view: supercomplexes partition the Q pool into accessible and sequestered fractions that respond differently to substrate and inhibitor challenges.

The supercomplex assembly factor SCAFI(COX7A2L) has been implicated in respirasome biogenesis; its expression varies by mouse strain and tissue, and correlates with the fraction of C1 in supercomplex form. This introduces an intriguing form of respiratory-capacity regulation at the organelle level.

Accounting for every proton in and out yields the P/O ratio: moles of ATP synthesized per gram-atom of oxygen consumed. Classical biochemistry texts used 3 for NADH and 2 for FADH₂, but careful modern measurements give smaller values:

Now the ATP synthesis side (details in Module 4): the c-ring of mammalian F₀has 8 c-subunits, each carrying one H⁺; three ATPs are made per full revolution (120° per ATP, three beta catalytic sites). This gives\(8/3 \approx 2.67\) H⁺/ATP. Adding the cost of P_i/H⁺ symport into the matrix (+1 H⁺ per ATP) and ATP/ADP antiport (neutral but draws on \(\Delta\psi\)), the physiological total is ~4 H⁺ per cytosolic ATP.

Over a human lifetime at basal metabolic rate, ATP synthase spins at roughly 100 revolutions/s per enzyme, and the body synthesizes roughly its own weight in ATP each day (Module 4 will quantify this).

The respiratory chain is the target of many natural products and drugs:

Uncouplers illustrate Mitchell’s argument beautifully: in their presence, the ETC keeps running (electrons still flow to O₂), the PMF collapses, and ATP synthesis stops. Oxygen consumption actually increases (“state 3u”) because the brake of thermodynamic back-pressure is released.