The Citric Acid Cycle
The Pyruvate Dehydrogenase Complex: The One-Way Gate to the Cycle
The pyruvate dehydrogenase complex is the one-way gate between glycolysis and the TCA cycle: it converts pyruvate to acetyl-CoA, releasing one CO2 and banking one NADH, and once a cell walks through it there is no walking back. One physical machine, three enzymes, five coenzymes, and a 14-Angstrom swinging arm hand the carbon down an assembly line without ever letting it escape into solution. Because the step is irreversible and sits at a metabolic crossroads, it is also one of the most heavily regulated reactions in the cell.
Glycolysis ends with pyruvate, and pyruvate is standing at a one-way gate. Once it walks through, there is no walking back. The reaction on the other side, run by the pyruvate dehydrogenase complex, converts pyruvate into acetyl-CoA, releases a carbon as CO2, and banks one NADH. It is the bridge between two terminals: glycolysis in the cytosol and the citric acid cycle in the mitochondrial matrix. And it is the reason animals cannot turn fat into sugar.
Picture an airport. Glycolysis is Terminal 1. The TCA cycle is Terminal 2. Between them sits a security gate that only opens one way. Pyruvate clears the gate, becomes acetyl-CoA, and commits to oxidation. Traffic does not flow backward. That single fact, written in thermodynamics, organizes the entire chapter.
The reaction and the one-way gate
The net reaction is worth memorizing exactly:
Pyruvate + CoA + NAD+ -> Acetyl-CoA + CO2 + NADH + H+
It is an oxidative decarboxylation. Pyruvate (3 carbons) loses its carboxyl carbon as CO2 and emerges as a 2-carbon acetyl group carried by coenzyme A. The oxidation that strips that carbon also reduces NAD+ to NADH. It happens in the matrix, so pyruvate must first cross the inner mitochondrial membrane through the mitochondrial pyruvate carrier.
Here is the load-bearing point: in animals this reaction is irreversible, with a large negative free-energy change. There is no enzyme to run it backward. Acetyl-CoA cannot be turned back into pyruvate, which is why gluconeogenesis cannot simply reverse glycolysis through this step; it must detour around PDC entirely using pyruvate carboxylase and PEPCK. The same wall explains a dinner-table fact: you cannot make a net amount of glucose from fatty acids, because fat enters metabolism as acetyl-CoA, on the far side of the gate.
Acetyl-CoA: the energy is in the bond
Students often imagine coenzyme A as the valuable part. It is not. CoA is the handle; the energy lives in the thioester bond that joins the acetyl group to CoA’s sulfur. Hydrolyzing that bond releases about -31 kJ/mol, roughly the same as ATP’s -30.5. An ordinary oxygen ester releases only about -21.
Why the difference? Sulfur is a larger atom than oxygen, so its electron orbitals overlap poorly with the carbonyl carbon. In an oxygen ester, the oxygen lone pairs stabilize the C=O by resonance, making the bond comfortable and low-energy. Sulfur cannot lend that stabilization. The thioester is left strained, like a spring wound tight rather than relaxed, or a boulder perched on a cliff edge. Release it and a lot of energy comes out. That stored potential is exactly what lets acetyl-CoA drive the first step of the TCA cycle and seed fatty-acid synthesis. Acetyl-CoA has two fates: complete oxidation for ATP, or storage as fat.
One machine, three enzymes, five coenzymes
PDC is one of the largest enzyme assemblies in the cell, roughly 10 megadaltons in eukaryotes, larger than a ribosome. E2 forms a central core scaffold (classically 60 copies, the pentagonal dodecahedron; recent mammalian cryo-EM suggests the functional core may be smaller), and E1 and E3 dock onto that scaffold through dedicated peripheral subunit-binding domains, close enough that the swinging lipoamide arm can reach every active site without the acetyl intermediate ever diffusing free.
That architecture buys substrate channeling: the intermediate never escapes into solution. Think of a robot arm on an assembly line. A lipoic acid molecule is amide-bonded to a lysine on E2, forming lipoamide, and this ~14-Angstrom arm physically swings the growing acetyl unit from one active site to the next by Brownian motion. Nothing is dropped, nothing diffuses away, nothing is wasted.
Five coenzymes do the chemistry. A useful mnemonic is “Tender Loving Care For Nancy”: TPP, Lipoamide, CoA, FAD, NAD+. Four trace back to vitamins you must eat: TPP from B1 (thiamine, at E1), CoA from B5 (pantothenate, at E2), FAD from B2 (riboflavin, at E3), and NAD+ from B3 (niacin, at E3). Lipoic acid is the exception; the body synthesizes it, so it is not a true vitamin.
Walking the mechanism
Trace the carbon through the complex below. Open each step and watch where the CO2 leaves and, critically, where the NADH is actually made.
Pyruvate made by glycolysis in the cytosol is carried across the inner mitochondrial membrane by the mitochondrial pyruvate carrier. Nothing is consumed or produced here; this is the geographic move from glycolysis into the matrix, where the complex waits at the one-way gate.
The C2 carbanion (ylide) of thiamine pyrophosphate attacks pyruvate's carbonyl, and the carboxyl carbon leaves as CO₂. This decarboxylation is the irreversible, committed event: a large negative ΔG, the carbon gone as gas, and animals have no enzyme to put it back. E1 is also the principal control point.
Regulation. E1 is switched off by PDH kinase (phosphorylation) when acetyl-CoA and NADH are high, and on by PDH phosphatase when ADP/pyruvate/Ca²⁺ rise.
Clinical. Thiamine (B1) deficiency starves E1 of TPP; the complex stalls, pyruvate and lactate accumulate, and beriberi or Wernicke's follows. Give thiamine BEFORE glucose in malnourished patients, or a glucose load drives lactic acidosis.
Still on E1: the two-carbon hydroxyethyl unit is oxidized to an acetyl group and transferred onto lipoamide's disulfide in the same motion. The S–S of lipoamide is reduced to a dithiol and TPP is regenerated. The electrons are now held on the lipoamide arm, NOT yet on NAD⁺ — so no NADH is made here, a common trap.
The swinging lipoamide arm carries the acetyl group to the E2 core, which transfers it to coenzyme A, forming the high-energy thioester acetyl-CoA (~−31 kJ/mol). Acetyl-CoA is released; the lipoamide is left reduced and must be reset before the cycle runs again.
Clinical. Arsenite (As³⁺) and mercury (Hg²⁺) bridge the two –SH groups of the dithiol lipoamide, locking the arm. The complex freezes, and the brain and heart, most dependent on its flux, are hit hardest.
E3 reoxidizes the dithiol back to the lipoamide disulfide, resetting the arm. The electrons pass to E3's tightly bound FAD, making FADH₂ transiently. This FADH₂ stays on the enzyme — an internal relay, not a released product — which is why PDC's net carriers list NADH and not FADH₂.
The electrons move from FADH₂ to NAD⁺, producing NADH and regenerating FAD. This is the only NADH-producing step of the whole complex. No molecular O₂ is used directly; the complex needs aerobic conditions only because the electron-transport chain must regenerate NAD⁺ downstream.
Clinical. In hyperglycemia, excess glucose floods the TCA cycle with electron donors, the mitochondrial proton gradient climbs past a threshold, and electrons leak from the chain as superoxide; that oxidative stress contributes to diabetic neuropathy.
The five mechanistic moves, in order: E1 uses TPP to attack pyruvate and lop off the carboxyl carbon as CO2, the irreversible step, leaving a hydroxyethyl group on TPP. Still on E1, that group is oxidized to an acetyl group and handed to lipoamide, reducing its disulfide to a dithiol. E2 transfers the acetyl from lipoamide to CoA, releasing acetyl-CoA and leaving lipoamide reduced. E3 reoxidizes that dithiol, passing the electrons to its bound FAD to make FADH2. Finally E3 moves those electrons from FADH2 to NAD+, producing NADH and resetting FAD.
The exam trap lives here, so flag it hard: NADH is produced only at E3, at the very end, not at the oxidation step on E1. The FADH2 formed along the way never leaves the enzyme; it is an internal relay, which is why the net carrier is NADH, not FADH2. And note that no molecular O2 is consumed anywhere in PDC. The reaction needs aerobic conditions only indirectly, because the electron transport chain must regenerate NAD+ for the complex to keep running.
Two layers of regulation
A gate this important is heavily controlled, and PDC reads the cell’s energy charge with the same logic as PFK-1 from the glycolysis chapter.
The fast layer is allosteric, responding in seconds. The products themselves are inhibitors: ATP, NADH, and acetyl-CoA all say “we have plenty, slow down.” ADP and pyruvate are activators: “we are hungry, speed up.”
The slower, sustained layer is covalent, acting over minutes. A dedicated kinase, PDK (pyruvate dehydrogenase kinase), phosphorylates a serine on E1 and switches the complex OFF. A phosphatase, PDP, removes that phosphate and switches it back ON. Note the direction: here phosphorylation inactivates, the same convention as glycogen synthase. PDK is itself stimulated by the high-energy signals ATP, NADH, and acetyl-CoA, reinforcing the allosteric brake.
Turns PDH OFF
(activates PDK)
Turns PDH ON
(activates PDP / blocks PDK)
With no effectors set, the phosphatase keeps the complex dephosphorylated and active by default.
Phosphorylation is the master switch: PDK adds a phosphate to shut PDH off when the cell is fuel-rich; PDP removes it to turn PDH on when the cell needs to burn pyruvate. The products of the reaction (acetyl-CoA, NADH) feed back to switch it off — classic product inhibition through a kinase.
The cleanest illustration is muscle. When a muscle contracts, the cytosol fills with Ca2+. Calcium activates PDP, PDP dephosphorylates E1, PDC switches on, and acetyl-CoA surges into the TCA cycle exactly when the muscle needs ATP. The same calcium that triggers contraction also unlocks the fuel.
When the gate fails
These mechanisms turn into medicine the moment they break. Beriberi is B1 deficiency: no thiamine, no TPP, E1 cannot work, pyruvate and lactate pile up, and the nervous and cardiovascular systems suffer. This is why a malnourished or alcohol-dependent patient must get thiamine before a glucose drip, or the sugar load drives a dangerous lactic acidosis (a cause of refeeding syndrome). Arsenite and mercury bind the dithiol form of lipoamide and clamp the swinging arm shut, freezing the whole complex; brain and heart, most dependent on PDC, are hit first. And the Warburg effect in cancer is PDC regulation hijacked: HIF-1a drives up PDK, E1 stays phosphorylated and off, pyruvate is shunted to lactate even when oxygen is abundant, and the tumor gets fast ATP plus biosynthetic carbon for growth.
With acetyl-CoA now in hand, we are standing at the door of Terminal 2. The next chapter runs that acetyl group through the citric acid cycle, where each turn squeezes out the CO2, NADH, FADH2, and GTP that feed the electron transport chain.
How we measure it
Substrate channeling in multienzyme complexes
When several active sites are tethered into one physical complex, an intermediate made at one site is handed directly to the next without diffusing into the cytosol. PDC is the textbook case: the lipoamide arm of E2 swings the growing acetyl unit between E1, E2, and E3, so reactive intermediates never escape, side reactions are suppressed, and local concentrations stay high. The same principle is read out experimentally by cryo-EM of the assembled core and by kinetic comparison of the intact complex versus the dissociated enzymes.
Reductive carboxylation / decarboxylation tracing with labeled carbon
Feed cells pyruvate labeled at C1 (the carboxyl) versus C2/C3, and the released CO2 tells you which carbon left. This is how the field established that PDC strips the carboxyl carbon (C1) as CO2 and that the surviving two carbons (C2, C3) become the acetyl group. Scaled up, the same logic underlies clinical 13C-pyruvate breath and MRI hyperpolarization studies of metabolism in tumors and failing heart.
Phospho-state assays for covalent regulation
To see whether PDC is on or off you measure the phosphorylation state of the E1 serine, using phospho-specific antibodies or activity assays before and after adding PDK, PDP, ATP, or Ca2+. Watching activity fall when PDK phosphorylates E1 and rise when PDP (Ca2+-activated) removes the phosphate is how the kinase/phosphatase switch was mapped, and it is the same readout used to test PDK inhibitors like dichloroacetate.