Cellular respiration: glycolysis and CAC

Intuition [Beginner]

Cells need energy to survive, grow, and divide. The primary energy currency is ATP. Cellular respiration is the process by which cells extract energy from glucose (and other fuel molecules) to regenerate ATP from ADP and inorganic phosphate. This process occurs in three main stages: glycolysis (in the cytoplasm), the citric acid cycle (in the mitochondrial matrix), and oxidative phosphorylation (on the inner mitochondrial membrane). This unit covers the first two.

Glycolysis (Greek: glykys = sweet, lysis = splitting) breaks one molecule of glucose (6 carbons) into two molecules of pyruvate (3 carbons each). It happens in ten enzyme-catalysed steps in the cytoplasm. The net yield per glucose is 2 ATP (used directly by the cell) and 2 NADH (electron carriers that feed into oxidative phosphorylation). Glycolysis does not require oxygen; it is an ancient pathway that evolved before atmospheric oxygen existed.

Glycolysis has two phases. The investment phase (steps 1-5) uses 2 ATP to phosphorylate glucose and split it into two three-carbon molecules. The payoff phase (steps 6-10) produces 4 ATP (for a net of 2) and 2 NADH. The key regulated enzyme is phosphofructokinase (PFK), which commits glucose to glycolysis and is inhibited by ATP and citrate (signals of abundant energy).

After glycolysis, pyruvate enters the mitochondrion (in eukaryotes) and is oxidised to acetyl-CoA by the pyruvate dehydrogenase complex (PDC). This reaction produces one NADH per pyruvate (2 per glucose) and releases one CO2 per pyruvate. Acetyl-CoA is a two-carbon molecule that feeds into the citric acid cycle.

The citric acid cycle (also called the Krebs cycle or TCA cycle) takes place in the mitochondrial matrix. Each turn of the cycle oxidises one acetyl-CoA (2 carbons) to 2 CO2, producing 3 NADH, 1 FADH2, and 1 GTP (equivalent to 1 ATP). Because each glucose produces 2 acetyl-CoA, the cycle turns twice per glucose, yielding 6 NADH, 2 FADH2, and 2 GTP.

The NADH and FADH2 from glycolysis, pyruvate oxidation, and the citric acid cycle carry high-energy electrons to the electron transport chain (unit 17.04.02), where the majority of ATP is produced.

Visual [Beginner]

The carbon accounting for one glucose: glucose (6C) becomes 2 pyruvate (2 x 3C = 6C), which release 2 CO2 during pyruvate oxidation, leaving 2 x 2C = 4C in two acetyl-CoA molecules. The citric acid cycle releases 4 CO2 (2 per turn), accounting for all 6 carbons from the original glucose.

Worked example [Beginner]

Calculate the total ATP yield from one glucose molecule through glycolysis and the citric acid cycle (before oxidative phosphorylation).

Step 1. Glycolysis: 1 glucose yields 2 pyruvate. Net: 2 ATP + 2 NADH.

Step 2. Pyruvate oxidation: 2 pyruvate yield 2 acetyl-CoA. Net: 2 NADH + 2 CO2.

Step 3. Citric acid cycle (2 turns): 2 acetyl-CoA yield 6 NADH + 2 FADH2 + 2 GTP + 4 CO2.

Step 4. Total electron carriers:

Stage	NADH	FADH2	Substrate-level ATP/GTP
Glycolysis	2	0	2 ATP
Pyruvate oxidation	2	0	0
Citric acid cycle	6	2	2 GTP
Total	10	2	4

Step 5. ATP equivalent estimate. Using the approximate ATP yield per electron carrier (from established P/O ratios): each NADH produces ~2.5 ATP, each FADH2 produces ~1.5 ATP through oxidative phosphorylation. Therefore:

• NADH: $10\times 2.5=25$ ATP
• FADH2: $2\times 1.5=3$ ATP
• Substrate-level: 4 ATP (2 from glycolysis + 2 GTP)
• Total: ~32 ATP per glucose

However, the cytoplasmic NADH from glycolysis (2 NADH) must be shuttled into the mitochondrion, and the shuttle mechanism affects the yield. The malate-aspartate shuttle preserves the full 2.5 ATP/NADH; the glycerol-3-phosphate shuttle transfers electrons to FAD, yielding only ~1.5 ATP per NADH. With the glycerol-3-phosphate shuttle: $8\times 2.5+(2+2)\times 1.5+4=20+6+4=30$ ATP. This unit focuses on the substrate-level yield (4 ATP equivalent); the full accounting is completed in unit 17.04.02.

Check your understanding [Beginner]

Formal definition [Intermediate+]

Glycolysis: ten-step pathway

The ten reactions of glycolysis, catalysed by soluble cytoplasmic enzymes:

Investment phase (steps 1-5, consume 2 ATP):

Step	Enzyme	Reaction	ATP/NADH
1	Hexokinase	Glucose + ATP -> Glucose-6-phosphate + ADP	-1 ATP
2	Phosphoglucose isomerase	Glucose-6-P <-> Fructose-6-P	-
3	Phosphofructokinase-1 (PFK-1)	Fructose-6-P + ATP -> Fructose-1,6-bisP + ADP	-1 ATP
4	Aldolase	Fructose-1,6-bisP -> DHAP + G3P	-
5	Triose phosphate isomerase	DHAP <-> G3P	-

Payoff phase (steps 6-10, produce 4 ATP + 2 NADH per glucose):

Step	Enzyme	Reaction	ATP/NADH
6	G3P dehydrogenase	2 G3P + 2 NAD+ + 2 Pi -> 2 1,3-BPG + 2 NADH	+2 NADH
7	Phosphoglycerate kinase	2 1,3-BPG + 2 ADP -> 2 3-PG + 2 ATP	+2 ATP
8	Phosphoglycerate mutase	2 3-PG -> 2 2-PG	-
9	Enolase	2 2-PG -> 2 PEP + 2 H2O	-
10	Pyruvate kinase	2 PEP + 2 ADP -> 2 Pyruvate + 2 ATP	+2 ATP

Net per glucose: 2 ATP + 2 NADH + 2 pyruvate.

The three irreversible steps (1, 3, 10) are sites of regulation. Steps 1 and 3 are regulated by feedback inhibition; step 10 (pyruvate kinase) is regulated by allosteric effectors (activated by fructose-1,6-bisphosphate via feedforward, inhibited by ATP and alanine).

Pyruvate dehydrogenase complex (PDC)

The PDC is a multi-enzyme complex that converts pyruvate to acetyl-CoA:

$$\text{Pyruvate}+\text{CoA}+{\text{NAD}}^{+}\to \text{Acetyl-CoA}+{\text{CO}}_2+\text{NADH}$$

The complex contains three enzymes (E1: pyruvate dehydrogenase, E2: dihydrolipoyl transacetylase, E3: dihydrolipoyl dehydrogenase) and five cofactors (TPP, lipoamide, CoA, FAD, NAD+). The reaction is irreversible and commits pyruvate to acetyl-CoA rather than other fates (lactate, oxaloacetate, alanine).

PDC regulation: phosphorylation of E1 by PDK (pyruvate dehydrogenase kinase, activated by ATP, NADH, acetyl-CoA) inactivates the complex. Dephosphorylation by PDP (pyruvate dehydrogenase phosphatase, activated by Ca2+) reactivates it.

Citric acid cycle: eight reactions

The eight steps of the cycle (per acetyl-CoA):

Step	Enzyme	Reaction	Product
1	Citrate synthase	Acetyl-CoA + Oxaloacetate -> Citrate	Citrate
2	Aconitase	Citrate <-> Isocitrate	Isocitrate
3	Isocitrate dehydrogenase	Isocitrate + NAD+ -> alpha-KG + CO2 + NADH	alpha-KG + NADH
4	alpha-KG dehydrogenase	alpha-KG + NAD+ + CoA -> Succinyl-CoA + CO2 + NADH	Succinyl-CoA + NADH
5	Succinyl-CoA synthetase	Succinyl-CoA + GDP + Pi -> Succinate + GTP + CoA	GTP
6	Succinate dehydrogenase	Succinate + FAD -> Fumarate + FADH2	FADH2
7	Fumarase	Fumarate + H2O -> Malate	Malate
8	Malate dehydrogenase	Malate + NAD+ -> Oxaloacetate + NADH	NADH

Net per acetyl-CoA: 3 NADH + 1 FADH2 + 1 GTP + 2 CO2.

Per glucose (2 acetyl-CoA): 6 NADH + 2 FADH2 + 2 GTP + 4 CO2.

Key theorem with proof [Intermediate+]

Theorem (Complete carbon and energy accounting for glucose oxidation through glycolysis and the citric acid cycle). One molecule of glucose (C6H12O6) oxidised through glycolysis, pyruvate oxidation, and two turns of the citric acid cycle yields 6 CO2, 10 NADH, 2 FADH2, and 4 ATP equivalents (2 ATP + 2 GTP). All 6 carbons are accounted for as CO2. The 24 electrons removed from glucose are carried by NADH and FADH2 to the electron transport chain.

Proof (carbon and electron bookkeeping).

Carbon accounting:

• Glycolysis: C6 (glucose) -> 2 x C3 (pyruvate). No carbon lost.
• Pyruvate oxidation: 2 x C3 (pyruvate) -> 2 x C2 (acetyl-CoA) + 2 x CO2. Lost: 2 CO2.
• CAC turn 1: C2 (acetyl-CoA) + C4 (oxaloacetate) -> C6 -> ... -> C4 (oxaloacetate) + 2 CO2.
• CAC turn 2: C2 (acetyl-CoA) + C4 (oxaloacetate) -> C6 -> ... -> C4 (oxaloacetate) + 2 CO2.
• Total CO2: 2 + 2 + 2 = 6 CO2. Matches the 6 carbons in glucose.

Electron accounting:

Glucose: ${\text{C}}_6{\text{H}}_{12}{\text{O}}_6$. Fully oxidised: $6{\text{CO}}_2$.

Oxidation state of C in glucose: each C is bonded to H, O, and other C. Average oxidation state: 0 (since ${\text{C}}_6{\text{H}}_{12}{\text{O}}_6$: $6x+12(+1)+6(-2)=0$, so $6x-0=0$, $x=0$).

Oxidation state of C in CO2: +4. Change per C: +4 electrons removed. For 6 carbons: 24 electrons removed.

Each NADH carries 2 electrons. Each FADH2 carries 2 electrons.

Electrons carried: $10\times 2+2\times 2=24$ electrons. Matches.

Energy accounting:

10 NADH x ~2.5 ATP = 25 ATP 2 FADH2 x ~1.5 ATP = 3 ATP 4 substrate-level ATP/GTP Total: ~32 ATP per glucose (using modern P/O ratios and the malate-aspartate shuttle). $\square$

Worked example: energetic efficiency

The standard free energy change for complete glucose oxidation is $\Delta G^{{\circ }^{\prime }}=-2840\text{kJ/mol}$. The free energy of ATP hydrolysis under cellular conditions is approximately $\Delta G_{\text{ATP}}\approx -50\text{kJ/mol}$.

If 32 ATP are produced: total energy captured = $32\times 50=1600\text{kJ/mol}$.

Efficiency = $1600/2840\times 100\%=56\%$.

This is remarkably efficient compared to human-made engines (typically 25-40%). The remaining 44% is released as heat, which maintains body temperature in endothermic animals.

Exercises [Intermediate+]

Glycolysis flux control and metabolic control analysis [Master]

The textbook story names phosphofructokinase-1 (PFK-1) as the rate-limiting step of glycolysis and stops there. The systems-biology story is more honest and more useful: there is no single rate-limiting enzyme, and the question "which enzyme controls the flux" has a precise quantitative answer formulated by Henrik Kacser and Jim Burns at the 1973 Society for Experimental Biology symposium ^{[Kacser and Burns 1973]} and independently by Reinhart Heinrich and Tom Rapoport in 1974 ^{[Heinrich and Rapoport 1974]}. Their framework — metabolic control analysis (MCA) — replaces the qualitative notion of a bottleneck with a distribution of fractional sensitivities.

For a pathway carrying steady-state flux $J$ through enzymes $E_1,E_2,\dots ,E_n$, the flux control coefficient of enzyme $E_i$ is

$$C_J^{E_i}=\frac{\partial \ln J}{\partial \ln [E_i]}=\frac{[E_i]}{J}\frac{\partial J}{\partial [E_i]},$$

a dimensionless number measuring the fractional change in pathway flux when the enzyme concentration is perturbed by a small fractional amount. A coefficient of 1 means the enzyme is the unique bottleneck; a coefficient of 0 means changing that enzyme's level has no effect on flux. Kacser and Burns proved the summation theorem

$$\sum _{i=1}^n{C}_J^{E_i}=1$$

by differentiating the steady-state condition with respect to a global scaling of all enzyme concentrations. The theorem is structural: control is a property of the whole pathway, and the total amount of control is conserved across enzymes.

Experimental measurements in yeast, hepatocytes, and erythrocytes consistently show that the largest flux control coefficient in glycolysis under physiological conditions belongs to glucose transport across the plasma membrane (often $C_J\approx 0.3$ to $0.5$ in skeletal muscle and liver), followed by PFK-1 ($C_J\approx 0.2$ to $0.3$) and pyruvate kinase ($C_J\approx 0.1$ to $0.2$). Hexokinase has substantial control in glucose-limited conditions but is product-inhibited by glucose-6-phosphate, which buffers its contribution. The remaining $\approx 0.2$ to $0.3$ of total control is distributed among the seven near-equilibrium glycolytic enzymes, each contributing a small fraction. The summation theorem is honored to within experimental error.

The mathematical mechanism behind the distribution of control is elasticity. The elasticity ${\epsilon }_S^{E_i}=\partial \ln v_i/\partial \ln [S]$ of enzyme $E_i$ to metabolite $S$ measures how strongly the enzyme's velocity responds to a fractional change in metabolite concentration. The connectivity theorem of MCA states that for each shared metabolite, the elasticities of upstream and downstream enzymes are coupled by

$$\sum _i{C}_J^{E_i}{\epsilon }_S^{E_i}=0,$$

reflecting the steady-state mass balance on $S$. Enzymes operating far from equilibrium (saturated, $K_m$-buffered) have small elasticities to their substrates and large flux control coefficients; enzymes near equilibrium have large elasticities and small control coefficients. The intuition is that a near-equilibrium step is forced by the surrounding metabolite pools and cannot itself dictate flux, whereas a far-from-equilibrium step is the gatekeeper.

PFK-1 is the textbook regulatory enzyme not because it carries the most control in absolute terms but because its allosteric architecture makes it tunable. The mammalian PFK-1 tetramer integrates at least four allosteric inputs into a multi-state cooperative response. ATP at high concentration binds an inhibitory regulatory site distinct from the catalytic ATP-binding pocket, dropping the enzyme into a low-activity tense (T) state. AMP at micromolar levels displaces ATP from the regulatory site and stabilises the relaxed (R) state, restoring activity. Citrate (signaling that the citric acid cycle is saturated with carbon) reinforces ATP inhibition. Fructose-2,6-bisphosphate — a separately regulated allosteric activator synthesized and degraded by the bifunctional enzyme PFK-2/FBPase-2 under hormonal control by insulin and glucagon — is the most potent activator known, dropping the apparent $K_m$ for fructose-6-phosphate by an order of magnitude when present at submicromolar concentrations. The Michaelis-Menten formalism developed at 14.08.01 is the kinetic substrate for these regulatory mechanisms; the hormone-driven changes in fructose-2,6-bisphosphate level are an output of the receptor-tyrosine-kinase and G-protein-coupled-receptor cascades treated at 17.07.02 and 17.07.01 pending.

What MCA reveals, beyond the level of any single enzyme, is that control is a configuration, not a location. The same pathway can have different control distributions in different physiological states — feeding versus fasting, rest versus exercise, normal versus tumor cell. In a hepatocyte during fasting, PFK-1 control is high because gluconeogenesis is dominant and fructose-2,6-bisphosphate is low, keeping PFK-1 partially inhibited and sensitive to fractional changes. In the same hepatocyte after a meal, with high fructose-2,6-bisphosphate, PFK-1 saturates and its control coefficient drops, while hexokinase (glucokinase isoform in liver) takes a larger share. The Kacser-Burns formalism captures this redistribution rigorously through the elasticity-coefficient algebra.

The crossover theorem of Chance and Williams (1955) gives MCA its diagnostic teeth. When an inhibitor blocks a pathway at a specific enzyme, metabolites upstream of the block accumulate and metabolites downstream are depleted; the crossover from accumulation to depletion locates the site of inhibition. Pre-MCA, the crossover technique was the principal way of identifying control points; post-MCA, it becomes a way of measuring elasticities indirectly. The diagnostic logic generalises to genetic perturbations: a yeast strain with a hexokinase deletion shows glucose accumulation outside the cell and depletion of all downstream glycolytic intermediates, locating the lesion at the transport-or-hexokinase border. This is the same crossover signature seen with the fluorocitrate-aconitase inhibition discussed in Exercise 9, illustrating that the diagnostic principle applies symmetrically to glycolysis and the citric acid cycle.

The MCA formalism extends from linear pathways to branched networks with a matrix-algebra generalisation. For a network with $n$ enzymes and $m$ internal metabolites, the steady-state condition is $\mathbf{Nv}(\mathbf{x})=0$, where $\mathbf{N}$ is the $m\times n$ stoichiometric matrix, $\mathbf{v}$ is the vector of enzyme velocities, and $\mathbf{x}$ is the vector of metabolite concentrations. Implicit differentiation of this condition with respect to a global enzyme-scaling parameter yields the control coefficients as solutions to a linear system involving the elasticity matrix $\mathbf{\epsilon }$ and a basis of the null space of $\mathbf{N}$. The summation and connectivity theorems become, respectively, statements about the rank of the elasticity matrix and about the orthogonality of control vectors to elasticity rows. For glycolysis with eleven enzymes and ten internal metabolites the system has one free flux (the net glucose-to-pyruvate throughput) and the control coefficients are the components of a unique flux-control vector once elasticities are measured. Branched pathways such as the glycolysis-to-pentose-phosphate-pathway split add additional null-space dimensions and additional summation theorems, one per independent flux.

The practical impact of MCA on metabolic engineering has been substantial. Industrial fermentation processes — production of lysine in Corynebacterium glutamicum, ethanol in yeast, succinate in engineered E. coli, artemisinin precursors in engineered yeast — all face the same conceptual problem: the engineer wants to redirect carbon flux from biomass to product, but pathway flux is distributed across many enzymes with small control coefficients. Overexpressing the textbook "rate-limiting" enzyme typically yields modest gains, because the new bottleneck reveals itself once the first one is relieved. Modern strain-engineering campaigns use MCA-guided global rebalancing — measuring elasticities and control coefficients in the wild-type strain, predicting the redistribution of control under proposed perturbations, and engineering enzyme levels to redistribute control toward the desired bottleneck (typically the irreversible step before the product node). This has industrialised what was previously a trial-and-error practice.

A second extension is time-dependent MCA, formulated by Heinrich and Reder in the late 1980s. The steady-state MCA above analyses small perturbations around a steady operating point. Dynamic perturbations — a step change in glucose supply, a hormonal transient — are described by frequency-dependent control coefficients that interpolate between the steady-state value at low frequency and the elasticity-only value at high frequency. The crossover from steady-state to high-frequency response occurs at the inverse of the metabolite relaxation time. For glycolysis the relaxation times are seconds for the most mobile metabolites and tens of seconds for the more buffered intermediates, setting a kilohertz-range frequency cutoff that is biologically irrelevant — the cell never sees such fast transients — but is theoretically informative about how the network filters input fluctuations.

A worked numerical illustration: in human erythrocytes, Schuster, Holzhütter, and Jacobasch (1988) measured the flux control coefficients of all ten glycolytic enzymes and reported PFK at $C_J\approx 0.32$, hexokinase at $0.25$, pyruvate kinase at $0.18$, glucose transporter at $0.12$, and the remaining seven enzymes contributing $0.13$ combined. The summation theorem is recovered to within $5\%$ experimental uncertainty. Doubling PFK expression in this system increases pathway flux by only $\sim 25\%$, not the doubling that the rate-limiting-step picture would predict, because the flux-control coefficient is $0.32$ and not $1$. The same overexpression also redistributes control: with PFK relaxed, hexokinase's coefficient rises and becomes the new principal control point. The redistribution is a direct prediction of the connectivity theorem applied to the elasticity matrix and is verified empirically.

Synthesis. The foundational reason MCA gave systems biology its quantitative spine is that it identifies pathway control with a measurable real number summing to 1. The central insight is that no single enzyme owns the rate; the summation theorem ${\sum }_i{C}_J^{E_i}=1$ distributes control across the entire chain, and putting these together with the connectivity theorem ${\sum }_i{C}_J^{E_i}{\epsilon }_S^{E_i}=0$ produces a closed algebraic system from which both elasticities and control coefficients can be inferred from finite perturbation experiments. This is exactly the bridge from the qualitative "rate-limiting step" intuition to the quantitative algebra of pathway regulation, and the pattern generalises through Heinrich-Schuster (1996) to non-linear and branched networks, builds toward the dynamic control structure of signaling cascades at 17.07.02, and appears again in cancer-metabolism control at the IDH and PKM2 nodes treated below. The bridge from Michaelis-Menten kinetics at 14.08.01 to whole-pathway behavior runs through the elasticity definition.

Citric acid cycle thermodynamics and the energy budget [Master]

The citric acid cycle as written in textbooks looks like a closed loop of eight reactions returning oxaloacetate to itself. Thermodynamically the loop is anything but closed: it is a one-way ratchet driven by three large free-energy drops, harvesting redox equivalents into NADH and FADH2 while emitting CO2 and a small amount of substrate-level GTP. Reading the cycle as a thermodynamic ratchet rather than a chemical loop is the master-tier reframe.

The standard free energies of the eight CAC reactions at physiological pH (pH 7) and 1 M reactant concentrations are

Step	Enzyme	$\Delta G^{\circ \prime }$ (kJ/mol)	$\Delta G$ (kJ/mol, in vivo)
1	Citrate synthase	$-31.5$	$\approx -32$
2	Aconitase	$+5.0$	$\approx 0$ (near-equilibrium)
3	Isocitrate dehydrogenase	$-8.4$	$\approx -21$
4	$\alpha$-KG dehydrogenase	$-30.1$	$\approx -34$
5	Succinyl-CoA synthetase	$-3.3$	$\approx -3$
6	Succinate dehydrogenase	$0$	$\approx 0$ (near-equilibrium)
7	Fumarase	$-3.4$	$\approx 0$ (near-equilibrium)
8	Malate dehydrogenase	$+29.7$	$\approx 0$ (driven forward by step 1 pull)

Three reactions are decisively exergonic in vivo: citrate synthase (step 1), isocitrate dehydrogenase (step 3), and $\alpha$-ketoglutarate dehydrogenase (step 4). Their cumulative free-energy drop is approximately $87$ kJ/mol per acetyl-CoA processed, which exceeds the standard free energy of ATP hydrolysis ($\approx 50$ kJ/mol in vivo) by a large margin and which makes the cycle operationally irreversible despite containing several near-equilibrium steps. The thermodynamics of 14.06.01 — equilibrium constants, $\Delta G=\Delta G^{\circ \prime }+RT\ln Q$, coupling of exergonic and endergonic steps — is the framework that converts these numbers into a working model of metabolic direction.

Step 8 is the famous puzzle. Malate dehydrogenase has $\Delta G^{\circ \prime }=+29.7$ kJ/mol; in isolation the reaction would run backward, with malate accumulating and oxaloacetate scarce. Yet in vivo the reaction proceeds forward at the cycle's steady-state flux. The resolution is that the product oxaloacetate is removed at vanishingly low concentration (typically $\approx 10\mu$M in mitochondria) by the very fast citrate synthase, which has $K_m$ for oxaloacetate near the in-vivo concentration. The mass-action ratio $Q=[\text{OAA}][\text{NADH}]/([\text{malate}][{\text{NAD}}^{+}])$ is therefore driven far below the equilibrium constant, and the actual $\Delta G$ for step 8 approaches zero or becomes slightly negative. The cycle achieves directional flux through the step 1 pull, not through a step 8 push.

The same coupling logic explains the otherwise paradoxical aconitase step (step 2). The standard equilibrium of the citrate-to-isocitrate isomerization is $\approx 93\%$ citrate to $\approx 7\%$ isocitrate. If aconitase were the only consumer or producer, isocitrate would accumulate to only a tiny fraction of citrate. But isocitrate is consumed rapidly by isocitrate dehydrogenase at step 3, dragging the aconitase equilibrium toward isocitrate by mass action. This is the same near-equilibrium-coupled-to-far-from-equilibrium pattern that organises whole metabolism: a few exergonic gates do the thermodynamic work; the rest of the pathway flows through near-equilibrium channels whose direction is set by the gates' pull.

The cycle's redox accounting is the second piece of its thermodynamic logic. Per acetyl-CoA the cycle produces 3 NADH at steps 3, 4, and 8, plus 1 FADH2 at step 6, plus 1 GTP (the substrate-level phosphorylation at step 5). Per glucose (two acetyl-CoA from one pyruvate-oxidation pair) the totals double: 6 NADH and 2 FADH2 from the cycle, plus 2 NADH from pyruvate oxidation, plus 2 NADH from glycolysis — 10 NADH and 2 FADH2 in total. Each NADH carries two high-energy electrons at a redox potential of $E^{\circ \prime }({\text{NAD}}^{+}/\text{NADH})=-320$ mV, and each FADH2 carries two electrons at $E^{\circ \prime }(\text{FAD}/{\text{FADH}}_2)\approx -220$ mV (enzyme-bound). Donating these electrons to molecular oxygen at $E^{\circ \prime }({\text{O}}_2/{\text{H}}_2\text{O})=+815$ mV releases approximately $220$ kJ/mol per NADH and $150$ kJ/mol per FADH2 via the electron transport chain.

The free-energy bookkeeping for one glucose oxidation closes as follows. The Gibbs free energy of complete glucose combustion under cellular conditions is $\Delta G\approx -2870$ kJ/mol. Of that, the cycle and upstream stages capture approximately $90\%$ as electron-carrier potential ($10\times 220+2\times 150=2500$ kJ/mol of stored redox potential) plus $4$ substrate-level ATP equivalents ($4\times 50=200$ kJ/mol). The substrate-level contribution is approximately $7\%$ of the total energy captured. The remaining $93\%$ flows through the electron transport chain and oxidative phosphorylation — the next unit, 17.04.02 pending, picks up the redox carriers and converts them to the bulk of the cell's ATP through the chemiosmotic proton circuit.

The regulation of the cycle follows the thermodynamic structure: the three far-from-equilibrium steps are the regulatory nodes. Citrate synthase is allosterically inhibited by ATP, NADH, and succinyl-CoA — signals that the cycle is energetically saturated. Isocitrate dehydrogenase is activated by ADP and Ca${}^{2+}$ (signals of energy demand) and inhibited by ATP and NADH. $\alpha$-Ketoglutarate dehydrogenase is product-inhibited by succinyl-CoA and NADH and activated by Ca${}^{2+}$. The Ca${}^{2+}$ activation of two of three regulatory enzymes is biologically significant: cytosolic Ca${}^{2+}$ signals from hormone or neurotransmitter action enter mitochondria through the MCU calcium uniporter and tune the cycle's flux to match demand — a fast metabolic response to electrical and chemical signaling that links the cycle to the receptor-cascade machinery of 17.07.02 without going through transcription. The redox-state signals (NAD${}^{+}$/NADH ratio, ATP/ADP ratio) tune the cycle's flux on a slower timescale through allosteric inhibition.

The pyruvate dehydrogenase complex (PDC) is a separate but tightly coupled regulatory node. PDC is not itself part of the cycle but the gateway feeding acetyl-CoA into it. PDC is inhibited by phosphorylation of its E1 subunit by PDH kinase (PDK, activated by ATP, NADH, and acetyl-CoA — feedback signals of energy and substrate abundance) and reactivated by dephosphorylation by PDH phosphatase (activated by Ca${}^{2+}$ and Mg${}^{2+}$). In Type 2 diabetes and metabolic syndrome, PDK isoforms are upregulated, locking PDC in its phosphorylated inactive state and forcing the cell toward glycolytic and fatty-acid oxidation rather than glucose oxidation through the cycle. This is a regulatory leverage point with substantial therapeutic interest: dichloroacetate (DCA) inhibits PDK and forces PDC into its active state, increasing pyruvate flux into the cycle, with reported effects on lactic acidosis and on the Warburg metabolism of certain cancer cells discussed in the next section.

The cycle's chemistry is also a worked example of biochemical efficiency in carbon-skeleton processing. Each step's reaction has been optimised by hundreds of millions of years of stabilising selection to use the chemically simplest mechanism that does the job. Citrate synthase performs a Claisen-aldol-like condensation through a single enzyme active site whose chemistry is essentially the same in the bacterial and the mammalian enzymes. Aconitase uses an iron-sulfur cluster to dehydrate citrate and rehydrate the cis-aconitate intermediate to isocitrate, achieving the required net stereochemistry through a single ferric-coordinated water molecule. Isocitrate dehydrogenase and the two $\alpha$-keto acid dehydrogenase complexes all use NAD${}^{+}$-coupled oxidative decarboxylation through the same conserved mechanism — a $\beta$-keto carboxylic acid first oxidised to a $\beta$-keto aldehyde-equivalent, then decarboxylated through electrostatic stabilisation of the developing carbanion. Succinate dehydrogenase is the cycle's only membrane-embedded enzyme and is Complex II of the electron transport chain, illustrating that the cycle is not topologically separate from the respiratory machinery.

The Ca${}^{2+}$ signal that activates three cycle enzymes (isocitrate dehydrogenase, $\alpha$-KG dehydrogenase, pyruvate dehydrogenase phosphatase) deserves more attention. Cytosolic Ca${}^{2+}$ enters mitochondria primarily through the mitochondrial calcium uniporter (MCU), a low-affinity, high-capacity Ca${}^{2+}$ channel in the inner membrane. The MCU has a $K_d$ for cytosolic Ca${}^{2+}$ of $\sim 1\mu$M, well above the resting cytosolic concentration of $\sim 100$ nM, so Ca${}^{2+}$ entry is silent under basal conditions and triggered only when cytosolic Ca${}^{2+}$ rises above a threshold (typically during synaptic activity in neurons, during contraction in cardiomyocytes, or during hormonal stimulation in other cell types). Mitochondrial matrix Ca${}^{2+}$ then rises into the micromolar range and binds allosteric sites on the three cycle enzymes, increasing their $V_{\max }$ and decreasing their $K_m$ for substrates. The cycle's flux rises rapidly to match the increased ATP demand that accompanies the same Ca${}^{2+}$ signal. This Ca${}^{2+}$-flux coupling is the cell's primary mechanism for matching mitochondrial output to acute energy demand, on a timescale of seconds, with no need for slower transcriptional or translational adjustment.

The redox state of the cell itself feeds back into the cycle's flux. The free ${\text{NAD}}^{+}/\text{NADH}$ ratio in the mitochondrial matrix is typically $\sim 8$ under aerobic conditions (NADH is fast-consumed by the electron transport chain) and can drop to $\sim 1$ or below under hypoxia or respiratory-chain inhibition. The free ${\text{NAD}}^{+}/\text{NADH}$ ratio is a direct allosteric modulator of isocitrate dehydrogenase, $\alpha$-KG dehydrogenase, and malate dehydrogenase: high NADH inhibits all three, slowing the cycle. The same ratio appears in the mass-action quotient $Q$ for steps 3, 4, and 8, raising the actual $\Delta G$ of each step toward zero as NADH accumulates. Both mechanisms — direct allosteric inhibition and mass-action thermodynamic equilibration — act in the same direction, jointly arresting the cycle when its downstream consumer (the electron transport chain) cannot keep up. This is respiratory control, the master-tier integration of cycle kinetics with electron-transport-chain kinetics, and it is the primary way the cell prevents wasteful ATP overproduction.

Synthesis. The foundational reason the citric acid cycle is biochemically robust is that three exergonic gates carry the thermodynamic load while five near-equilibrium reactions handle the carbon skeleton chemistry without contributing to directionality. The central insight is that the cycle's irreversibility is not a per-step property but a system property — putting these together with the malate-dehydrogenase step pulled forward by oxaloacetate depletion, the cycle becomes a one-way ratchet whose direction is set by step 1's free-energy drop. This is exactly the structure that identifies thermodynamic feasibility with kinetic structure: the bridge is from $\Delta G^{\circ \prime }$ tables to in-vivo $\Delta G$ via the mass-action correction at 14.06.01, and the same logic generalises to all coupled enzymatic pathways. The pattern recurs in the calcium-driven cycle activation that builds toward 17.04.02 pending respiratory control, and appears again in the cancer-metabolism reorganisation of the next section.

Anaplerotic and cataplerotic reactions: the cycle as a metabolic hub and cancer rewiring [Master]

The citric acid cycle is described in undergraduate biochemistry as a catabolic ring oxidising acetyl-CoA to CO2. In real cells the cycle is also the central hub for biosynthesis. Cycle intermediates are continuously drawn off as carbon skeletons for amino acids, nucleotides, heme, lipids, and gluconeogenic precursors. The cell replenishes the cycle through dedicated anaplerotic (Greek: filling up) reactions that import carbon at specific nodes; the withdrawals are called cataplerotic. The amphibolic character of the cycle — simultaneously catabolic and anabolic — is the master-tier reframe of its biological function.

The most quantitatively important anaplerotic reaction is catalysed by pyruvate carboxylase, a biotin-dependent enzyme located in the mitochondrial matrix:

$$\text{Pyruvate}+{\text{CO}}_2+\text{ATP}⟶\text{Oxaloacetate}+\text{ADP}+{\text{P}}_i.$$

Pyruvate carboxylase is allosterically activated by acetyl-CoA, which is a powerful regulatory signal: when acetyl-CoA accumulates, the enzyme infers that the cycle is starved of oxaloacetate relative to acetyl-CoA supply, and it produces more oxaloacetate to balance the stoichiometry. This is a feedforward regulation distinct from the feedback-inhibition logic governing the cycle's exergonic gates. In liver and kidney, pyruvate carboxylase is the entry point for gluconeogenesis: pyruvate is carboxylated to oxaloacetate, which is reduced to malate, exported from the mitochondrion, reoxidised to cytosolic oxaloacetate, and decarboxylated-phosphorylated to phosphoenolpyruvate by PEP carboxykinase (PEPCK). The PEPCK step reverses the irreversible pyruvate kinase step of glycolysis, allowing the cell to traverse the glycolytic pathway backward from PEP to glucose.

Several other anaplerotic entries operate alongside pyruvate carboxylase. Glutamate dehydrogenase catalyses the reversible reductive amination of $\alpha$-ketoglutarate:

$$\alpha \text{-ketoglutarate}+{\text{NH}}_4^{+}+\text{NADH}\rightleftharpoons \text{Glutamate}+{\text{NAD}}^{+}+{\text{H}}_2\text{O}.$$

In the forward (anabolic) direction, glutamate dehydrogenase synthesises glutamate from cycle carbon and free ammonia, providing the universal amino-group donor for transamination. In the reverse (catabolic) direction, the enzyme deaminates glutamate, feeding the cycle with $\alpha$-ketoglutarate from amino acid catabolism. The reaction direction is set by mass action: the cellular ammonia concentration, the redox state, and the glutamate-to-$\alpha$-ketoglutarate ratio together determine whether the enzyme is filling or draining the cycle. Other anaplerotic transaminases (aspartate aminotransferase, alanine aminotransferase) feed cycle intermediates from amino-acid pools through similar redox-coupled equilibria.

A third major anaplerotic entry is the propionyl-CoA carboxylase pathway. Odd-chain fatty acid oxidation and the catabolism of valine, isoleucine, methionine, and threonine all generate propionyl-CoA, which is carboxylated to D-methylmalonyl-CoA, racemized to L-methylmalonyl-CoA, and isomerized by the vitamin-B12-dependent methylmalonyl-CoA mutase to succinyl-CoA. The pathway feeds carbon directly into the cycle at the succinyl-CoA node. Inherited deficiencies in methylmalonyl-CoA mutase or in B12 metabolism cause methylmalonic acidemia, with toxic accumulation of propionyl-CoA and methylmalonate. The clinical presentation — metabolic acidosis, failure to thrive, neurological damage — illustrates how essential the anaplerotic pathway is: when the cycle cannot be replenished from this source, the cell cannot maintain steady cycle flux under conditions of high amino acid or odd-chain fatty acid catabolism.

The matching cataplerotic withdrawals route cycle carbon into biosynthetic pathways. Citrate exported from the mitochondrion through the citrate carrier is cleaved in the cytosol by ATP-citrate lyase to acetyl-CoA and oxaloacetate, providing acetyl-CoA for fatty-acid and cholesterol synthesis. The same export delivers oxaloacetate to the cytosol where it is reduced to malate and recycled. $\alpha$-Ketoglutarate is withdrawn for glutamate synthesis as above. Succinyl-CoA is withdrawn for heme biosynthesis (the rate-limiting step of porphyrin synthesis, $\delta$-aminolevulinate synthase, uses succinyl-CoA and glycine). Aspartate (the transamination product of oxaloacetate) is withdrawn for nucleotide biosynthesis and urea-cycle entry. Each withdrawal must be balanced by an anaplerotic input to maintain cycle flux, and the cell maintains this balance through the allosteric and hormonal regulation of the anaplerotic enzymes.

The malate-aspartate shuttle is a system of cytosolic and mitochondrial isoenzymes that imports NADH reducing equivalents into the mitochondrion without violating the impermeability of the inner mitochondrial membrane to NADH itself. Cytosolic NADH reduces oxaloacetate to malate via cytosolic malate dehydrogenase. Malate enters the mitochondrion via the malate-$\alpha$-ketoglutarate antiporter and is reoxidised by mitochondrial malate dehydrogenase, regenerating intramitochondrial NADH. Oxaloacetate cannot exit the mitochondrion directly, so it is transaminated to aspartate, which exits via the aspartate-glutamate antiporter and is transaminated back to oxaloacetate in the cytosol. The net effect is to transfer two reducing equivalents from cytosolic NAD(H) to mitochondrial NAD(H) at the cost of consuming one molecule of glutamate and exporting one molecule of $\alpha$-ketoglutarate, all of which are recycled. The shuttle exists because the inner membrane has no NADH transporter, and without it the glycolytic NADH would build up in the cytosol and shut down glycolysis. The alternative shuttle, the glycerol-3-phosphate shuttle, transfers electrons to mitochondrial FAD rather than NAD, giving a lower ATP yield per cytosolic NADH (approximately 1.5 versus 2.5).

Cancer metabolism demonstrates how rewiring the cycle reorganises whole-cell physiology. The Warburg effect — Otto Warburg's 1956 observation that proliferating tumor cells convert most of their imported glucose to lactate even in the presence of oxygen ^{[Warburg 1956]} — was originally interpreted as a defect in mitochondrial function. Modern interpretation reframes it as adaptive: aerobic glycolysis sustains a high biosynthetic flux through diversion of glycolytic intermediates into the pentose phosphate pathway (for nucleotide synthesis) and into serine biosynthesis (via 3-phosphoglycerate). The lactate output is not pathology but pathway design — it regenerates NAD${}^{+}$ for the high-flux glycolytic machinery without committing the cell to high oxidative phosphorylation, which would generate damaging reactive oxygen species at the rate of proliferative growth. PKM2, a pyruvate kinase isoform expressed in proliferating cells, is allosterically regulated to operate at sub-maximal velocity, allowing upstream glycolytic intermediates to accumulate and be siphoned into biosynthesis. The Warburg phenotype is the rewiring of glycolytic flux from energy capture toward carbon-skeleton supply for growth.

IDH mutations in glioma, acute myeloid leukemia, and chondrosarcoma illustrate the cycle as a chemical engineer's plaything. Recurrent point mutations at R132 in cytosolic IDH1 or at R140 and R172 in mitochondrial IDH2 produce a neomorphic enzyme that reduces $\alpha$-ketoglutarate to D-2-hydroxyglutarate (D-2HG) using NADPH as reductant ^{[Dang 2009]}. D-2HG is a structural analog of $\alpha$-ketoglutarate that competitively inhibits the broad family of $\alpha$-ketoglutarate-dependent dioxygenases, including the TET DNA demethylases and the JmjC histone demethylases. The downstream consequence is widespread hypermethylation of CpG islands and histones, locking the cell in a state of impaired differentiation. The IDH mutation is thus an oncometabolite-generating lesion: a single amino-acid substitution in a cycle enzyme produces a small-molecule signal that reprograms the epigenome. Clinically targeted inhibitors of mutant IDH (ivosidenib for IDH1, enasidenib for IDH2) restore $\alpha$-ketoglutarate levels and have shown clinical activity in IDH-mutant AML.

Glutaminolysis is the third major cancer-metabolic theme. Many proliferating tumor cells become glutamine-addicted: they import glutamine at high rate and convert it via glutaminase to glutamate and via glutamate dehydrogenase or aminotransferases to $\alpha$-ketoglutarate. The $\alpha$-ketoglutarate feeds the cycle from the side, providing a route to cycle intermediates that bypasses pyruvate input from glycolysis. In MYC-driven tumors, glutamine catabolism supplies both the biosynthetic carbon and the redox cofactors (via NADH from glutamate dehydrogenase and via reductive carboxylation of $\alpha$-ketoglutarate back to citrate for fatty-acid synthesis) needed for rapid proliferation. Glutaminase inhibitors such as CB-839 (telaglenastat) target this dependence and have entered clinical trials.

The succinate dehydrogenase (SDH) and fumarate hydratase (FH) loss-of-function mutations identified in hereditary paragangliomas, pheochromocytomas, and hereditary leiomyomatosis and renal-cell-cancer syndrome are the cycle's third class of "metabolic oncogene." Inactivating mutations in SDH cause accumulation of succinate; mutations in FH cause accumulation of fumarate. Both succinate and fumarate are competitive inhibitors of the same family of $\alpha$-ketoglutarate-dependent dioxygenases inhibited by D-2HG in IDH-mutant tumors. The downstream consequence is the same: hypermethylation of CpG islands, stabilisation of hypoxia-inducible factor (HIF) through inhibition of the prolyl hydroxylases that normally tag HIF for degradation, and a pseudohypoxic transcriptional response that promotes angiogenesis and glycolytic adaptation. The convergence of three independent loss-of-function mutations (in SDH, FH, and IDH) on the same family of epigenetic-modifier targets through three different oncometabolites is one of the cleanest demonstrations in modern oncology that metabolism and epigenetics are intertwined regulatory layers.

The cycle's anaplerotic-cataplerotic balance is also tissue-specific. In hepatocytes, the dominant cataplerotic flux is oxaloacetate-to-glucose via PEPCK during gluconeogenesis; pyruvate carboxylase is the dominant anaplerotic input. In adipocytes, the dominant cataplerotic flux is citrate-to-cytosolic-acetyl-CoA via ATP-citrate lyase for fatty-acid synthesis; pyruvate carboxylase again anaplerotic. In skeletal muscle at rest, anaplerotic input from branched-chain amino acid catabolism (leucine, isoleucine, valine) replenishes succinyl-CoA via methylmalonyl-CoA pathway; during exercise, glutamate-to-$\alpha$-ketoglutarate transamination contributes additional anaplerotic flux. In cardiac muscle, $\beta$-oxidation of fatty acids delivers most of the acetyl-CoA, with anaplerotic input from propionyl-CoA (odd-chain fatty acid catabolism) and aspartate transamination. Each tissue's anaplerotic-cataplerotic profile reflects its biological role: liver as glucose factory, adipose as lipid factory, muscle as ATP consumer.

The intersection of anaplerotic regulation with whole-body metabolism runs through the Cori cycle, the glucose-alanine cycle, and the ketone body cycle. In the Cori cycle, muscle lactate is exported to the liver, reconverted to glucose (consuming 6 ATP in the liver per Cori turn), and recycled to muscle as glucose. The Cori cycle has whole-body energetic cost but redistributes the metabolic burden between tissues — muscle gets fast ATP via anaerobic glycolysis, liver pays the gluconeogenic ATP debt during recovery. The glucose-alanine cycle is analogous but transports nitrogen as well as carbon: muscle releases alanine from protein catabolism plus pyruvate transamination, liver removes the amino group via the urea cycle and synthesises glucose from the carbon skeleton. The ketone body cycle (liver acetyl-CoA to $\beta$-hydroxybutyrate and acetoacetate, transported to brain and muscle, reconverted to acetyl-CoA and burned through the cycle) is a glucose-sparing route used during fasting and ketogenic states. All three cycles are anaplerotic-cataplerotic balanced at the tissue level but unbalanced at the whole-body level — the liver runs a net cataplerotic deficit (citric acid cycle intermediates exported as glucose, fatty acids, or ketones), and peripheral tissues run a net anaplerotic surplus (importing these same molecules and feeding them back into the cycle). Whole-body metabolic homeostasis is the integral of these tissue-level imbalances.

Synthesis. The foundational reason the cycle is biological rather than merely chemical is that it is amphibolic, identifying carbon catabolism with carbon supply for biosynthesis. The central insight is that anaplerotic and cataplerotic flows are balanced by allosteric and hormonal regulation: putting these together with the redox accounting of the previous section produces the cell's metabolic budget. This is exactly the structure that allows tumor cells to rewire the cycle for proliferation rather than energy capture, and the bridge is from balanced flux through the cycle to imbalanced flux through specific anaplerotic and cataplerotic nodes. The pattern recurs in tissue-specific metabolic specialisation — hepatocytes prioritise gluconeogenesis, adipocytes prioritise citrate export to cytosolic acetyl-CoA, neurons prioritise glutamate-glutamine cycling — and builds toward whole-organism metabolic integration at 17.04.02 pending and downstream signaling control at 17.07.02. The pyruvate-carboxylase node generalises the same allosteric-feedforward logic that drove PFK-1 in glycolysis.

Comparative and evolutionary metabolism: glycolysis across taxa and the metabolic-rate scaling [Master]

Glycolysis is the oldest known metabolic pathway. Its enzymes are conserved across all three domains of life — bacteria, archaea, and eukarya — and the canonical Embden-Meyerhof-Parnas (EMP) pathway is functionally identical from E. coli to human hepatocytes. This conservation is biological evidence that glycolysis evolved before the divergence of cellular life, before atmospheric oxygenation, and before the endosymbiotic acquisition of mitochondria. The cycle is younger, requires an oxidising environment to be useful, and is essentially absent from obligate anaerobes — but its core chemistry of two-carbon condensation onto a four-carbon acceptor, sequential decarboxylation, and redox-driven regeneration of the acceptor is also broadly conserved across aerobic life. The master-tier reframe is that the universality of these pathways is not coincidence but the signature of metabolism's deep evolutionary roots.

The EMP pathway is one of three known routes from glucose to pyruvate. The alternative Entner-Doudoroff (ED) pathway, found primarily in Gram-negative bacteria (Pseudomonas, Azotobacter) and many archaea, also converts glucose to two pyruvate but with different stoichiometry and a different chemical logic. ED replaces the EMP investment phase (phosphorylation, isomerization, second phosphorylation, aldol cleavage to two trioses) with a parallel route: glucose-6-phosphate is oxidised to 6-phosphogluconate (the entry to the pentose phosphate pathway), then dehydrated to 2-keto-3-deoxy-6-phosphogluconate (KDPG), and aldol-cleaved to one pyruvate and one glyceraldehyde-3-phosphate. The G3P then runs through the EMP payoff phase, producing one ATP and one NADH (rather than the EMP's $2$ and $2$ per glucose). The total ED yield is $1$ ATP $+1$ NADH $+1$ NADPH per glucose, versus EMP's $2$ ATP $+2$ NADH per glucose. ED is thermodynamically less efficient but generates NADPH (useful for biosynthesis and for redox defense) at the entry to the pathway, and the ED enzymes are fewer in number, requiring less protein investment for the same throughput. The phylogenetic prevalence of ED in oligotrophic environments and in chemolithotrophs suggests it is the more ancient route — Kleiber and others noted that ED enzymes can be reconstructed in primitive thermophiles, hinting that EMP may be a later refinement that traded protein investment for higher per-glucose ATP yield.

A third route, the phosphoketolase pathway found in heterofermentative lactic acid bacteria and bifidobacteria, splits pentoses or hexoses through a phosphoketolase that produces acetyl-phosphate and glyceraldehyde-3-phosphate from xylulose-5-phosphate. The pathway is the metabolic basis of yogurt, sourdough, and a substantial fraction of the human gut microbiome's fermentative output. The phosphoketolase route generates less ATP per substrate than EMP but provides flexibility in carbon-skeleton output, including direct production of the two-carbon fragment used in propionate fermentation.

The fermentative endpoints of pyruvate diverge by organism. In mammalian skeletal muscle and in many bacteria, pyruvate is reduced to lactate by lactate dehydrogenase, regenerating NAD${}^{+}$ for continued glycolysis. The lactate output is acidic and limits the duration of intense exercise through cellular pH drop; in mammals the lactate is exported via the Cori cycle to the liver for gluconeogenic reconversion to glucose. In yeasts and some bacteria, pyruvate is decarboxylated and then reduced to ethanol via pyruvate decarboxylase and alcohol dehydrogenase, regenerating NAD${}^{+}$ without acidic byproduct. The ethanol output is volatile and self-limiting (toxicity to the producer) but is the chemical basis of brewing, baking, and biofuel production. In propionic acid bacteria the fermentation runs through methylmalonyl-CoA and the cycle's succinyl-CoA node to propionate, the carboxylic-acid responsible for Swiss-cheese flavor and acetate fermentation in the gut. In Clostridia the fermentation diverges to acetate, butyrate, acetone, and butanol — the latter is the basis of the World War I-era acetone-butanol-ethanol (ABE) industrial fermentation.

The evolutionary timing of the cycle is more recent and less universally distributed. Strict anaerobes generally lack a complete oxidative cycle. Many anaerobes run a reductive cycle — the cycle in reverse, fixing CO2 onto succinyl-CoA and running back to pyruvate, providing carbon skeletons for biosynthesis from CO2 alone (the reductive TCA cycle is one of the six known autotrophic CO2 fixation pathways, found in green sulfur bacteria and in some thermophiles). The reductive cycle is plausibly very ancient and may predate the oxidative cycle; the modern oxidative cycle is thought to have evolved by reversal of the reductive variant after atmospheric oxygenation made aerobic energetics favorable. The molecular phylogeny of citrate synthase and aconitase places their divergence from a common ancestor near the Archaean-Proterozoic boundary, consistent with this scenario. Some extant organisms run both forward (oxidative) and reverse (reductive) cycle segments simultaneously in different cellular compartments, supplying biosynthesis from one direction while running catabolism from the other.

The comparative biology of muscle illustrates how the choice between glycolytic and oxidative metabolism is itself tuned across species. Fast-twitch glycolytic muscle (type IIb fibers, predominating in sprinters, in the white meat of chicken breast, in the burst-swimming muscle of fish) generates ATP almost entirely through glycolysis with lactate output, sustains very high power output for seconds, and fatigues rapidly from lactate accumulation. Slow-twitch oxidative muscle (type I fibers, predominating in marathon runners, in the red meat of poultry leg, in cardiac muscle) generates ATP almost entirely through oxidative phosphorylation, sustains moderate power output for hours, and is fatigue-resistant. The mitochondrial density, capillary density, and myoglobin content of the two fiber types differ by an order of magnitude. The same actin-myosin contractile machinery is powered by two entirely different metabolic strategies. The genetic and developmental program selecting one fiber type over the other is itself a target of training, hormonal regulation, and disease — illustrating that metabolic strategy is a tunable phenotype, not a fixed property of the cell.

Kleiber's law ^{[Kleiber 1932]} provides the most striking comparative-physiological scaling result for metabolism. Across animals from a 30-gram mouse to a 5000-kilogram elephant, basal metabolic rate (BMR) scales with body mass $M$ approximately as

$$\text{BMR}\propto M^{3/4}.$$

The $3/4$ exponent is not the $2/3$ that would follow from naive surface-to-volume thinking (heat loss across a surface of area $\propto M^{2/3}$ would limit metabolism to that scaling). The deviation from $2/3$ has been the subject of an extensive theoretical literature. The West-Brown-Enquist (1997) hypothesis derives the $3/4$ exponent from the fractal architecture of vascular networks delivering oxygen and substrates to the metabolic tissues: the network must be space-filling, terminate in capillaries whose size is invariant across species, and minimize total energy dissipation. The resulting allometric scaling of nutrient delivery enforces a $3/4$ exponent on whole-organism metabolic rate. An alternative derivation through Banavar-Damuth-Maritan emphasises the network's capacity-to-volume ratio and arrives at the same exponent through a slightly different optimisation. The empirical result — robust across nine orders of magnitude in body mass and across mammals, birds, fish, reptiles, and plants — anchors metabolism to network geometry.

The per-cell consequence of Kleiber scaling is that smaller organisms have higher metabolic rates per cell. A mouse hepatocyte respires roughly an order of magnitude faster per cell than an elephant hepatocyte. Mitochondrial density, cristae surface area per mitochondrion, and enzyme concentrations per cell all scale with the $-1/4$ power of body mass. This puts hard constraints on the per-cell flux through glycolysis and through the cycle, observable directly as differences in basal $\dot{V}{\text{O}}_2$ per gram of tissue across species. The metabolic machinery described mechanistically in this unit operates within a quantitative scaling envelope set by the organism's size — and the same enzymes, the same allosteric regulation, the same redox cofactors run faster in a shrew and slower in a whale by an amount predicted by the Kleiber exponent.

Glycolysis and the cycle therefore sit at an interesting intersection of biological time scales. They are universal at the molecular level — every cell that uses glucose runs essentially the same enzymes — yet their flux is tuned across a $10^9$-fold range from the slowest hibernating animal to the fastest hummingbird in flight. The mechanistic biochemistry is invariant; the regulation, the flux distribution, the anaplerotic balance, and the absolute flux level are all evolved variables.

Archaeal glycolysis illustrates how dramatically regulation can vary while keeping the chemistry. Thermophilic and hyperthermophilic archaea (Pyrococcus, Thermococcus, Sulfolobus) live at temperatures up to $113^{\circ }$C and run modified versions of the EMP and ED pathways with archaea-specific enzyme variants. The hyperthermophile Pyrococcus furiosus uses an ADP-dependent glucokinase and an ADP-dependent phosphofructokinase rather than the standard ATP-dependent enzymes, generating ADP-AMP exchange currencies rather than ATP-ADP. The pathway's overall ATP yield is similar, but the regulatory architecture is different — ADP-dependent kinases are insensitive to the AMP-ATP allosteric signaling that controls mesophilic PFK-1. Several archaeal organisms run a non-phosphorylated variant of ED in which glucose is oxidised through gluconate to KDG (2-keto-3-deoxygluconate) without phosphorylation, splitting into pyruvate and glyceraldehyde, with phosphorylation occurring only downstream. The chemistry is the same as bacterial ED in form but rearranged in timing, suggesting that the phosphorylation steps in canonical glycolysis are an evolutionary refinement rather than a chemical necessity.

The temperature dependence of cycle flux follows the Arrhenius framework developed at 14.08.01. Enzymatic rate constants scale as $k=A\exp (-E_a/RT)$ with activation energies $E_a$ typically in the range $40-70$ kJ/mol, giving Q${}_{10}$ values (the fold-change in rate per $10^{\circ }$C) of approximately $2-3$. Mammalian cycle enzymes have evolved $E_a$ values that match the cell's operating temperature ($37^{\circ }$C) — too low and the enzyme would be too active at room temperature; too high and the enzyme would be too sluggish at body temperature. Hibernating mammals exploit this Arrhenius dependence: during torpor, body temperature drops to near-freezing and metabolic rate drops by an order of magnitude, allowing extended survival on stored fat. Migratory birds and Arctic fish have evolved enzyme variants with cold-shifted activation energies, maintaining cycle flux at temperatures that would shut down the mammalian enzymes. The cycle's enzymes are not chemically frozen but evolutionarily tunable around the same basic mechanisms.

The regulatory architecture has also evolved. Bacterial PFK-1 has a different allosteric pattern from mammalian PFK-1: it is allosterically activated by ADP and GDP (signals of energy demand) rather than AMP, and inhibited by phosphoenolpyruvate (a downstream signal of high-glycolytic-intermediate concentration) rather than citrate. The two regulatory schemes use different small-molecule signals but achieve the same functional outcome — slowing glycolysis when energy is abundant and accelerating it when energy is needed. The convergence on a common regulatory function from different allosteric mechanisms illustrates how the cell can implement the same logic with different molecular vocabularies. The architecture of regulation is more conserved than the specific allosteric ligands.

Comparative analysis of microbial fermentation diversity reveals the breadth of solutions to the NAD${}^{+}$ regeneration problem. Lactic acid bacteria reduce pyruvate to lactate; yeasts reduce acetaldehyde to ethanol; propionic acid bacteria run pyruvate through the methylmalonyl-CoA pathway to propionate; Clostridium acetobutylicum runs pyruvate through butyryl-CoA to butyrate, butanol, and acetone; many anaerobes can switch between several products depending on the carbon source and the redox state. Each solution regenerates NAD${}^{+}$ from NADH without molecular oxygen but produces different metabolic by-products that have shaped the chemistry of foods, fuels, and the gut microbiome. The diversity is constrained by the requirement to balance NADH oxidation against ATP generation: solutions that produce more reduced products (ethanol, butanol) regenerate less NAD${}^{+}$ per substrate than solutions that produce more oxidised products (acetate, succinate), and the cell must balance the two to keep the redox state homeostatic. The constraints map directly onto the stoichiometry of cycle and glycolysis intermediates.

Synthesis. The foundational reason metabolism is biology's deepest invariant is that glycolysis and (in aerobic descendants) the cycle were evolved before the modern domain divergence and have been conserved by stabilising selection ever since. The central insight is that variation occurs at the regulatory and flux-distribution layer, not at the enzymatic-chemistry layer: putting these together with the comparative-physiology scaling of Kleiber's law, the same molecular machinery runs at vastly different rates across taxa under network-geometric constraints. This is exactly the structure that identifies cellular biochemistry with whole-organism physiology: the bridge is the allometric scaling of capillary delivery to mitochondrial demand, and the pattern recurs from skeletal muscle fiber-type specialisation to bacterial pathway-choice (EMP versus ED versus phosphoketolase). The generalises both downward through quantitative ecology and upward through evolutionary biology, and appears again in the metabolic-rate-bounded growth rates of cancer cells discussed in the previous section. The cycle's flexibility builds toward the integrated metabolic physiology of 17.04.02 pending.

Connections [Master]

• Chemical kinetics: rate laws and Arrhenius 14.08.01. The Michaelis-Menten formalism that quantifies every enzymatic step in glycolysis and the cycle is developed at this peer unit, including the steady-state derivation of $v=V_{\max }[S]/(K_m+[S])$ and the temperature dependence of rate constants. Allosteric regulation of PFK-1, citrate synthase, isocitrate dehydrogenase, and $\alpha$-ketoglutarate dehydrogenase is the biological context in which the cooperative-binding extensions of Michaelis-Menten kinetics become essential.

• Chemical thermodynamics and equilibrium 14.06.01. The free-energy framework $\Delta G=\Delta G^{\circ \prime }+RT\ln Q$, the relation between equilibrium constants and standard free energies, and the coupling of exergonic and endergonic reactions are the foundation of the cycle's directional flux analysis above. The cycle's irreversibility-via-three-exergonic-gates argument depends entirely on the in-vivo mass-action correction to standard free energies.

• Oxidative phosphorylation 17.04.02 pending. The 10 NADH and 2 FADH2 per glucose produced through the pathways analysed here are the input to the electron transport chain and the chemiosmotic proton circuit. The next unit picks up the redox carriers and converts them to the bulk of cellular ATP via the F1F0 ATP synthase, completing the energy-capture story begun in glycolysis.

• Receptor tyrosine kinases and the MAPK cascade 17.07.02. Hormonal control of glycolytic flux through insulin and glucagon signaling acts on the bifunctional PFK-2/FBPase-2 enzyme via the same RTK and PKA cascades developed in this peer unit. Cancer rewiring of glycolysis and the cycle (Warburg effect, glutaminolysis, IDH mutations) couples upstream growth-factor signaling to metabolic pathway choice.

• Cell signaling: receptors and GPCRs 17.07.01 pending. Glucagon-driven elevation of cAMP through the glucagon receptor (a GPCR) is the canonical hormonal trigger for hepatic gluconeogenesis and glycolysis suppression. The Ca${}^{2+}$ signals that activate isocitrate dehydrogenase, $\alpha$-ketoglutarate dehydrogenase, and pyruvate dehydrogenase phosphatase come from GPCR and ion-channel pathways treated in this peer unit.

• Biomolecules in cells 17.01.01. Glucose, ATP, NAD${}^{+}$/NADH, FAD/FADH2, acetyl-CoA, and the citric-acid-cycle intermediates are the biomolecular building blocks introduced in this peer unit. The molecular structures and redox chemistry described there are the substrate-level prerequisites for the pathway analysis above.

• Cellular organisation: organelles 17.03.01 pending. The mitochondrion — matrix where the cycle and pyruvate oxidation occur, inner membrane where the electron transport chain lives, cristae geometry that organises ATP synthesis — is the structural setting for the biochemistry above. The malate-aspartate shuttle depends on the inner-membrane impermeability and the specific transporters analysed in this peer unit.

• Photosynthesis 17.10.01. Photosynthesis is the metabolic complement: it captures light energy to fix CO2 into glucose, which is then oxidised through the pathways developed here. The Calvin cycle and the citric acid cycle share enzymatic and evolutionary connections, and the global carbon cycle balances photosynthetic fixation against respiratory release.

• Enzyme mechanism 15.14.01 pending. Every enzyme-catalysed step in glycolysis and the TCA cycle deploys one or more of the catalytic strategies formalised in the enzyme mechanism unit: covalent catalysis (aldolase Schiff base with Lys), metal-ion catalysis (hexokinase Mg${}^{2+}$), general acid-base catalysis (citrate synthase His/Asp), and Lewis-acid cofactors (aconitase Fe/S cluster). The mechanistic vocabulary for each enzymatic transformation traces back to the principles established there.

• Photosynthesis: light and dark reactions 17.04.03 pending. The triose phosphates exported from the chloroplast Calvin cycle enter the carbohydrate metabolism pathways described here. G3P is converted to glucose, starch, or sucrose and then feeds glycolysis and the citric acid cycle during respiration. Plants respire continuously, consuming roughly 50% of the carbon they fix during daylight — photosynthesis and respiration are complementary halves of the photosynthetic cell's carbon-energy economy.

• Mutation and repair 17.06.01 pending. The electron transport chain leaks approximately 1–2% of electrons to oxygen, generating superoxide and downstream reactive oxygen species (hydrogen peroxide, hydroxyl radical) that are the primary endogenous source of oxidative DNA damage, including 8-oxoG. The mutation rate is therefore coupled to metabolic rate: tissues with high oxidative phosphorylation activity face a larger endogenous damage burden, as quantified in the mutation and repair unit.

Historical & philosophical context [Master]

Glycolysis was the first metabolic pathway to be reconstructed in its entirety. The investigation began with Eduard Buchner's 1897 demonstration that cell-free yeast extracts can ferment sugar ^{[Buchner 1897]}, a result that disposed of the vitalist claim that fermentation required a living cell. Buchner won the 1907 Nobel Prize in Chemistry and the field of biochemistry was launched. Over the next four decades, Arthur Harden and William Young identified the role of phosphate and what they called "cozymase" (later NAD${}^{+}$); Otto Meyerhof and Karl Lohmann linked muscle glycolysis to lactate production and to ATP; Gustav Embden and Jakub Karol Parnas worked out the central reactions; and Otto Warburg developed the spectrophotometric methods that allowed enzyme-by-enzyme dissection. The pathway as taught today bears the names of Embden, Meyerhof, and Parnas (EMP) for the central glucose-to-pyruvate sequence.

Hans Krebs proposed the citric acid cycle in 1937 ^{[Krebs and Johnson 1937]} on the basis of careful tracer experiments in minced pigeon breast muscle. Krebs observed that adding any of the catalytic intermediates — citrate, isocitrate, $\alpha$-ketoglutarate, succinate, fumarate, malate, or oxaloacetate — to the tissue increased oxygen consumption in a substoichiometric fashion (a little intermediate, a lot of oxygen consumption), implying that the intermediates were catalytic rather than consumed. This was the experimental signature of a cycle. The two-carbon unit entering the cycle was identified as acetyl-CoA by Fritz Lipmann in 1945 ^{[Krebs and Johnson 1937]}; CoA itself had been characterised earlier by Lipmann in studies of acetylation reactions. Krebs and Lipmann shared the 1953 Nobel Prize in Physiology or Medicine for these discoveries. The pyruvate dehydrogenase complex was characterised by Lester Reed in the late 1950s and 1960s, revealing the substrate-channelling architecture with its swinging-lipoyl arm — one of the most beautiful pieces of molecular machinery in biology.

Metabolic control analysis grew out of two independent papers in the early 1970s. Henrik Kacser and Jim Burns presented their formulation at the Society for Experimental Biology symposium in 1973 ^{[Kacser and Burns 1973]}, building on engineering control theory and on Kacser's earlier interest in genetic dominance. Reinhart Heinrich and Tom Rapoport developed an equivalent formalism in 1974 ^{[Heinrich and Rapoport 1974]}, motivated by quantitative modelling of erythrocyte metabolism. The Kacser-Burns and Heinrich-Rapoport formulations were initially viewed as competing but were later shown to be mathematically equivalent. The summation and connectivity theorems are now standard, and MCA software (METATOOL, COPASI, JWS Online) has industrialised the analysis. The framework's broader impact is that it gave systems biology its first rigorous quantitative methodology — pathway flux is a measurable scalar, and control over flux is a measurable vector of fractional sensitivities. This anticipated by two decades the systems-biology programs of the 1990s and 2000s.

The cancer-metabolism field traces to Otto Warburg's 1924 observation that tumor cells produce large amounts of lactate from glucose even in the presence of oxygen, an effect he later named aerobic glycolysis or the Warburg effect ^{[Warburg 1956]}. Warburg interpreted the effect as evidence of damaged mitochondria, a view that fell out of favor when it was shown that tumor mitochondria are largely functional. The field re-emerged after the 2008 discovery of IDH1 mutations in glioma and AML ^{[Dang 2009]}, which provided the first clear-cut "metabolic oncogene" — a mutation in a single cycle enzyme that drives malignant transformation through a small-molecule epigenetic signal (D-2-hydroxyglutarate). The decade since has seen targeted IDH inhibitors enter clinical practice, glutaminase inhibitors enter trials, and a broad reframing of tumor metabolism as an adaptive biosynthetic response to proliferation rather than a defect in energetics. The Warburg effect is now read not as broken metabolism but as a deliberate flux redirection toward growth, illustrating that the same enzymes can be wired for different biological purposes.

The conceptual significance of the metabolic-pathway program is that it established biochemistry as a quantitative science. Before glycolysis was solved, biological catalysis was a black box. After it was solved, every enzyme had a name, a structure (eventually), kinetic parameters, and a place in a wiring diagram. The same Michaelis-Menten formalism that quantifies invertase in a test tube quantifies hexokinase in a liver cell. This unification is the foundation on which modern molecular biology, pharmacology, and metabolic engineering rest. The cycle is also a worked example of the principle that biological systems are organised at multiple levels — the chemistry is local to each enzyme; the flux is a property of the network; the regulation is layered with allosteric, hormonal, transcriptional, and epigenetic mechanisms — and the layers must be analysed together to understand how the cell works.

Bibliography [Master]

@article{KrebsJohnson1937,
  author = {Krebs, Hans A. and Johnson, William A.},
  title = {The role of citric acid in intermediate metabolism in animal tissues},
  journal = {Enzymologia},
  volume = {4},
  year = {1937},
  pages = {148-156},
}

@article{Buchner1897,
  author = {Buchner, Eduard},
  title = {Alkoholische {G}{\"a}rung ohne {H}efezellen},
  journal = {Berichte der deutschen chemischen Gesellschaft},
  volume = {30},
  year = {1897},
  pages = {117-124},
}

@article{KacserBurns1973,
  author = {Kacser, Henrik and Burns, James A.},
  title = {The Control of Flux},
  journal = {Symposia of the Society for Experimental Biology},
  volume = {27},
  year = {1973},
  pages = {65-104},
}

@article{HeinrichRapoport1974,
  author = {Heinrich, Reinhart and Rapoport, Tom A.},
  title = {A linear steady-state treatment of enzymatic chains: general properties, control and effector strength},
  journal = {European Journal of Biochemistry},
  volume = {42},
  year = {1974},
  pages = {89-95},
}

@article{Warburg1956,
  author = {Warburg, Otto},
  title = {On the origin of cancer cells},
  journal = {Science},
  volume = {123},
  year = {1956},
  pages = {309-314},
}

@article{Dang2009,
  author = {Dang, Lenny and White, David W. and Gross, Stefan and Bennett, Bryson D. and Bittinger, Mark A. and Driggers, Edward M. and Fantin, Valeria R. and Jang, Hyun Gyung and Jin, Shengfang and Keenan, Marie C. and Marks, Kevin M. and Prins, Robert M. and Ward, Patrick S. and Yen, Katharine E. and Liau, Linda M. and Rabinowitz, Joshua D. and Cantley, Lewis C. and Thompson, Craig B. and Vander Heiden, Matthew G. and Su, Shinsan M.},
  title = {Cancer-associated {IDH1} mutations produce 2-hydroxyglutarate},
  journal = {Nature},
  volume = {462},
  year = {2009},
  pages = {739-744},
}

@article{Kleiber1932,
  author = {Kleiber, Max},
  title = {Body size and metabolism},
  journal = {Hilgardia},
  volume = {6},
  year = {1932},
  pages = {315-353},
}

@article{WestBrownEnquist1997,
  author = {West, Geoffrey B. and Brown, James H. and Enquist, Brian J.},
  title = {A general model for the origin of allometric scaling laws in biology},
  journal = {Science},
  volume = {276},
  year = {1997},
  pages = {122-126},
}

@book{Alberts2022,
  author = {Alberts, Bruce and Heald, Rebecca and Johnson, Alexander and Morgan, David and Raff, Martin and Roberts, Keith and Walter, Peter},
  title = {Molecular Biology of the Cell},
  edition = {7th},
  publisher = {Garland Science},
  year = {2022},
}

@book{BergTymoczkoStryer2019,
  author = {Berg, Jeremy M. and Tymoczko, John L. and Stryer, Lubert},
  title = {Biochemistry},
  edition = {9th},
  publisher = {W. H. Freeman},
  year = {2019},
}

@book{NichollsFerguson2013,
  author = {Nicholls, David G. and Ferguson, Stuart J.},
  title = {Bioenergetics},
  edition = {4th},
  publisher = {Academic Press},
  year = {2013},
}

@book{Fell1997,
  author = {Fell, David},
  title = {Understanding the Control of Metabolism},
  publisher = {Portland Press},
  year = {1997},
}

@book{HeinrichSchuster1996,
  author = {Heinrich, Reinhart and Schuster, Stefan},
  title = {The Regulation of Cellular Systems},
  publisher = {Chapman \& Hall},
  year = {1996},
}

---

Cycle 6 Track B deepening of the Wave 3 biology unit. Deepened 2026-05-20 to math-style four-sub Master tier per the Cycle 4 style-parity contract. All hooks_out targets are proposed. Pending Tyler review and external biology reviewer per BIOLOGY_PLAN.