The Krebs cycle takes place in the mitochondrial matrix, the gel-like compartment enclosed by the inner mitochondrial membrane. In eukaryotic cells, all eight enzymes of the cycle except one are soluble proteins in that matrix; the exception, succinate dehydrogenase, is embedded in the inner membrane. Prokaryotes run the same reactions in the cytosol, because they have no mitochondria.
This localization is not trivia. It explains why the cycle is physically coupled to the electron transport chain, why oxygen deprivation halts it within seconds to minutes, and why inherited defects in one matrix enzyme produce tissue-specific disease. The following sections trace the mechanism, the numbers, the exceptions, and the practical implications.
Matrix architecture and enzyme geography
The mitochondrion has two membranes and two compartments. The outer membrane is porous; the inner membrane is folded into cristae and is selectively permeable. Between them sits the intermembrane space. Inside the inner membrane is the matrix, which holds the cycle's enzymes, mitochondrial DNA, ribosomes, and a high concentration of soluble metabolites.
The cycle's eight steps are catalyzed by citrate synthase, aconitase, isocitrate dehydrogenase, alpha-ketoglutarate dehydrogenase, succinyl-CoA synthetase, succinate dehydrogenase, fumarase, and malate dehydrogenase. Seven are matrix-soluble. Succinate dehydrogenase sits on the matrix face of the inner membrane as Complex II of the electron transport chain, which is why the cycle and oxidative phosphorylation share a physical junction. This arrangement lets NADH and FADH2 produced in the matrix hand electrons directly to the membrane-embedded chain.
A useful mental model: the matrix is a reaction chamber bolted to a power plant. The chamber makes reduced carriers; the plant converts them to a proton gradient. The bolt is succinate dehydrogenase.
Why location changes what the cycle can do
Compartmentalization is the mechanism, not just the address. The inner membrane maintains a proton gradient of roughly 150 to 180 millivolts, with matrix pH near 7.8 and intermembrane space near 7.0. That gradient powers ATP synthase. Because the cycle runs in the matrix, its NADH is generated where Complex I can accept it, and its succinate is generated where Complex II can oxidize it.
Location also gates the cycle on oxygen. When oxygen is absent, Complexes III and IV cannot pass electrons, the chain backs up, NADH accumulates, and NAD+ regeneration stalls. Because isocitrate dehydrogenase and alpha-ketoglutarate dehydrogenase require NAD+, the cycle slows within seconds to minutes. The cell then shifts toward anaerobic glycolysis. This is why sprinting muscle accumulates lactate: the cycle's substrate flow is throttled by the matrix's dependence on the membrane's electron sink.
Compartmentalization also separates the cycle from cytosolic glycolysis. Pyruvate made in the cytosol must be imported into the matrix and converted to acetyl-CoA by pyruvate dehydrogenase before it can enter. That import step is a control point, regulated by acetyl-CoA and NADH feedback.
Quantified context: yields, rates, and turnover
Per turn, the cycle produces three NADH, one FADH2, one GTP (readily converted to ATP), and two CO2. Each acetyl-CoA entering the cycle yields roughly 10 ATP equivalents when fully oxidized through the chain. A single glucose molecule, after glycolysis and pyruvate oxidation, yields about 30 to 32 ATP in most human tissues, with the exact figure depending on shuttle efficiency.
The cycle also turns over fast. In working skeletal muscle, flux through citrate synthase can rise several-fold within seconds of contraction onset. In the liver, the cycle supplies precursors for gluconeogenesis and heme synthesis, so flux is tuned to biosynthetic demand as well as energy demand. Cardiac muscle runs the highest oxidative flux per gram of tissue of any human organ, sustained by dense cristae and high mitochondrial volume.
Carbon accounting matters. Two carbons enter as acetyl-CoA, two leave as CO2, but the specific carbons are not the same atoms. The cycle is catalytic: oxaloacetate is regenerated each turn. A common exam error is assuming the two CO2 molecules come directly from the acetyl carbons; in fact, they are released at the isocitrate and alpha-ketoglutarate steps from carbons that entered earlier.
Edge cases: prokaryotes, red blood cells, and inherited defects
Prokaryotes such as Escherichia coli run the cycle's reactions in the cytosol, using membrane-bound dehydrogenases for the analogous electron transfer steps. The chemistry is conserved; the geography is not.
Mature human red blood cells have no mitochondria and therefore no Krebs cycle. They rely on glycolysis and the pentose phosphate pathway. This is an edge case worth remembering because it shows the cycle is not universal to all human cells.
Inherited defects in matrix enzymes cause disease with tissue-specific patterns. Mutations in fumarase cause hereditary leiomyomatosis and renal cell cancer, with skin and uterine tumors. Mutations in succinate dehydrogenase subunits cause hereditary paraganglioma and pheochromocytoma. These are not abstract biochemistry: the matrix location means defects in one step can alter metabolite levels that act as oncogenic signals.
A counterintuitive caveat for professionals: some cancer cells rewire the cycle so that citrate is exported from the matrix to the cytosol for lipid synthesis, reversing part of the normal flow. The matrix is a hub, not a fixed one-way street.
Common misconceptions and red flags
Misconception: the cycle happens in the cytoplasm. It does not, in eukaryotic cells. The cytoplasm runs glycolysis; the matrix runs the cycle.
Misconception: the cycle directly consumes oxygen. The cycle itself uses no O2. Oxygen is consumed by the electron transport chain, which recycles the NADH and FADH2 the cycle produces. Oxygen deprivation stalls the cycle indirectly.
Misconception: the cycle only makes energy. It is amphibolic, supplying precursors for amino acids, heme, and fatty acid synthesis.
Red flags when reasoning about the cycle:
Assuming CO2 released in one turn comes from the acetyl-CoA just added. Forgetting that succinate dehydrogenase is both a cycle enzyme and Complex II. Treating the cycle as identical in all cells despite red blood cell and prokaryotic exceptions. Ignoring the NAD+/NADH ratio as the real gate on flux. Confusing mitochondrial matrix with the intermembrane space when interpreting lab results.
Comparison table: compartments and their roles
| Compartment | Main carbon pathway | Oxygen needed directly | Key carriers produced |
|---|---|---|---|
| Cytosol | Glycolysis | No | NADH, pyruvate |
| Mitochondrial matrix | Krebs cycle | No (indirectly) | NADH, FADH2, GTP |
| Inner mitochondrial membrane | Electron transport chain | Yes | Proton gradient |
| Intermembrane space | Proton reservoir | No | none (gradient) |
| Mitochondrial matrix (biosynthesis) | Precursor export | No | citrate, succinyl-CoA |
The table makes the division of labor explicit. Two compartments run carbon chemistry, one membrane runs electron chemistry, and the gradient lives in the space between.
Timeline table: what happens after oxygen drops
| Time after oxygen loss | Matrix event | Consequence |
|---|---|---|
| 0-10 seconds | Electron chain backs up | NADH accumulates |
| 10-60 seconds | NAD+ falls | Isocitrate and alpha-ketoglutarate dehydrogenase slow |
| 1-3 minutes | Cycle flux drops | Acetyl-CoA pools rise |
| 3-10 minutes | Pyruvate dehydrogenase inhibited | Pyruvate diverted to lactate |
| Minutes to hours | Glycolysis dominates | ATP from substrate-level phosphorylation |
| Hours | Matrix metabolite shifts | Biosynthetic output falls |
The timeline reconciles two thresholds that are often conflated. The cycle's enzymes do not require oxygen, so the cycle does not stop at time zero. It stops when the NAD+/NADH ratio falls below what the dehydrogenases need, which typically occurs within the first minute or two of anoxia in working tissue, and sooner in high-flux organs such as heart.
Practical actions for students and clinicians
For students: draw the matrix, place all eight enzymes, and mark which are soluble versus membrane-bound. Quiz yourself on the two CO2 release steps and the one GTP-producing step. When solving yield problems, track NADH and FADH2 separately before converting to ATP.
For clinicians interpreting metabolic labs: a rising lactate with a falling bicarbonate suggests impaired mitochondrial oxidation, not necessarily a cycle enzyme defect. Matrix defects often show elevated specific metabolites, such as fumarate or succinate, rather than a simple lactate rise. Order targeted metabolite testing when a matrix enzyme disorder is suspected.
For researchers: when measuring cycle flux, account for compartmentation. Cytosolic and mitochondrial pools of citrate and malate are not interchangeable, so whole-cell metabolite measurements can mislead. Use isotope tracing with compartment-aware modeling.
Threshold reconciliation: the common claim that the cycle "stops without oxygen" and the accurate claim that it "slows as NAD+ falls" describe the same event at different timescales. The first is a simplification for exams; the second is the mechanistic reality. Neither should be used to claim a precise universal cutoff, because tissue-specific buffering changes the timing.
Limitations and what remains uncertain
The core localization is settled: matrix in eukaryotes, cytosol in prokaryotes. What remains active research is how matrix metabolites are channeled between enzymes, whether the cycle operates as a transient multienzyme assembly called a metabolon, and how compartment-specific NAD+ pools are maintained. These questions affect drug targeting and biomarker interpretation but not the basic answer.
Quantitative yields vary with shuttle systems and tissue, so treat 30 to 32 ATP per glucose as a range, not a fixed value. Do not present any single figure as universal. For any clinical decision about a suspected mitochondrial disorder, the general information here is not a diagnosis or treatment plan; consult a qualified metabolic specialist or genetic counselor.