The Glyoxylate Cycle

Seed biology and agriculture. Oilseed crops (castor, cucumber, sunflower, cotton) rely entirely on glyoxysomal glyoxylate-cycle flux during the heterotrophic phase of germination, before photosynthesis begins. Understanding and engineering this pathway's regulation matters for seedling vigor and stress tolerance in crop breeding programs.

Antitubercular drug discovery. Mycobacterium tuberculosis upregulates isocitrate lyase (icl1/icl2) dramatically during the dormant, non-replicating persistent state inside granulomas, where host lipids (cholesterol, fatty acids) are essentially the only available carbon source. Because humans lack isocitrate lyase entirely, it is an attractive selective drug target; several small-molecule inhibitors (3-nitropropionate derivatives, itaconate and its analogs) have been explored specifically because they can attack latent TB without an equivalent human off-target.

Itaconate and immunometabolism. Itaconate, produced by activated macrophages via decarboxylation of cis-aconitate (a citric acid cycle intermediate, via the enzyme IRG1/ACOD1), is itself a potent inhibitor of isocitrate lyase in engulfed bacteria and fungi — meaning the mammalian immune system has evolved its own small molecule to specifically disable this pathway in pathogens, a striking case of host-pathogen metabolic warfare built directly around glyoxylate-cycle biochemistry.

Industrial and synthetic biology. Engineered E. coli and yeast strains with amplified or rebalanced glyoxylate-shunt flux are used industrially to improve yields of succinate, malate, and other dicarboxylic-acid platform chemicals from cheap feedstocks (acetate, glycerol, fatty-acid waste streams), since the shunt offers a carbon-conserving route unavailable to the standard oxidative cycle.

Fungal pathogenesis. Candida albicans and other pathogenic fungi upregulate glyoxylate-cycle enzymes when growing within host tissue on lipid-rich, glucose-poor niches (such as inside macrophage phagosomes), making fungal isocitrate lyase and malate synthase active areas of antifungal target research as well.

Conceptual bridge to gluconeogenesis. The succinate (or downstream malate/oxaloacetate) exported from the glyoxylate cycle is exactly the substrate pool that phosphoenolpyruvate carboxykinase acts on to generate PEP, making this pathway the essential upstream carbon-supply step whenever Gluconeogenesis: Reversing Glycolysis must build sugar from fat or acetate rather than from lactate or amino acids.

Ask the tutor: How does phosphoenolpyruvate carboxykinase couple the oxaloacetate exiting the glyoxylate cycle to the first committed step of gluconeogenesis?

The essential difficulty with using acetyl-CoA to make glucose is arithmetic, not chemistry. Pyruvate carboxylase can add CO₂ to pyruvate to make oxaloacetate, and phosphoenolpyruvate carboxykinase can turn oxaloacetate into PEP, the entry point to reversed glycolysis. But acetyl-CoA has no route to pyruvate in animals — the pyruvate dehydrogenase reaction is physiologically irreversible. So the only way fat-derived acetyl-CoA could yield net sugar is via the citric acid cycle: acetyl-CoA condenses with oxaloacetate to give citrate (6C), and if oxaloacetate could be regenerated in greater quantity than it started, the surplus could peel off toward gluconeogenesis. Running the ordinary citric acid cycle does not achieve this. Trace the carbons: citrate (6C) → isocitrate (6C) → α-ketoglutarate (5C, one CO₂ lost) → succinyl-CoA (4C, one CO₂ lost) → succinate → fumarate → malate → oxaloacetate (4C). Two carbons went in as acetyl-CoA; two carbons left as CO₂. The oxaloacetate at the end is numerically regenerated — same one that started — but not net increased. Every carbon atom from acetyl-CoA that entered was oxidized to CO₂ within that turn (label the entering acetyl-CoA carbons and they turn out, after scrambling through the symmetric succinate intermediate, to be distributed between the released CO₂ and the retained skeleton, but the total carbon count is conserved with two atoms leaving as gas). Net new carbon skeleton: zero. The glyoxylate cycle changes the topology of this loop. After isocitrate forms, instead of oxidative decarboxylation by isocitrate dehydrogenase, isocitrate lyase cleaves the C2–C3 bond of isocitrate in a retro-aldol reaction, releasing succinate (C1–C4 of the original citrate skeleton) and glyoxylate (OHC–COO⁻, a two-carbon fragment). No CO₂ is released here — all six carbons of isocitrate are conserved, split as 4 + 2. The succinate (4C) can leave the cycle entirely, headed for gluconeogenesis. The glyoxylate (2C) is then condensed, by malate synthase, with a second incoming acetyl-CoA to form malate (4C), which oxidizes to oxaloacetate and closes the loop, ready to accept a third acetyl-CoA.

The two enzymes unique to this pathway — isocitrate lyase and malate synthase — are absent from vertebrate mitochondria, which is precisely why mammals cannot convert net fatty-acid carbon into glucose. The four other steps (citrate synthase, aconitase, malate dehydrogenase, plus the final oxidation of succinate to malate via succinate dehydrogenase and fumarase) are shared machinery borrowed from the ordinary citric acid cycle, run either in the same mitochondrion (some organisms) or, in plants, in a dedicated peroxisome-derived organelle called the glyoxysome, with succinate shuttled out to mitochondria for oxidation to malate/OAA and onward to the cytosol.

The net reaction of one full turn of the glyoxylate cycle, from two acetyl-CoA to one succinate, is:

where:

• the two acetyl-CoA molecules supply all four carbons that end up in succinate,
• NAD⁺ is reduced once, at the malate dehydrogenase step regenerating oxaloacetate from malate,
• no CO₂ is released — this is the entire point of the bypass.

Reading: compare this to one full turn of the ordinary citric acid cycle oxidizing one acetyl-CoA completely,

In the citric acid cycle, carbon is destroyed for energy; in the glyoxylate cycle, carbon is conserved for biosynthesis, at the cost of extracting far less reducing power per acetyl unit (one NADH per two acetyl-CoA here, versus three NADH + one FADH₂ + one GTP per single acetyl-CoA fully oxidized). The organism trades energy yield for carbon skeleton. The overall carbon economy, run twice (to make a full four-carbon unit ready to double into hexose via gluconeogenesis and eventually condense two triose units), is best seen by combining the glyoxylate cycle with gluconeogenesis:

where:

• four acetyl-CoA (from complete β-oxidation of one fatty-acid chain) feed two turns of the glyoxylate cycle,
• yielding two succinate, each convertible via fumarate/malate/OAA to one PEP,
• two PEP molecules then combine, via reversed glycolysis, into one glucose (C₆) — the net conversion of fat to sugar.

Goal: show explicitly how the C2–C3 bond of isocitrate breaks to give succinate and glyoxylate without loss of any carbon as CO₂, and why this is chemically an aldol cleavage in reverse. Isocitrate is HOOC–CH(OH)–CH(COOH)–CH₂–CH₂–COOH numbered as C1 (carboxyl attached to the CHOH), C2 (the CH bearing the extra carboxyl branch), C3 (CH₂ attached to C2), and continuing C4–C5–C6 to the terminal carboxylate. The key structural feature is the β-hydroxy acid arrangement: C2 carries a carboxylate branch alpha to the C3–C4 chain, and this whole unit is analogous to a citrate-derived aldol adduct.

Step 1. The enzyme active site (a divalent Mg²⁺ or Mn²⁺-independent enzyme in most organisms, though some isocitrate lyases show metal dependence) positions isocitrate so that a general base (an active-site histidine or a deprotonated carboxylate, depending on organism) abstracts the C3 proton adjacent to the C2 carboxylate branch. Removal of this proton generates a resonance-stabilized carbanion at C3.

Step 2. This carbanion is stabilized as it collapses: the electron pair flows from C3 into the forming C2–C3 π bond, and simultaneously the C2–C1(carboxylate) bond that connects the branch carbon (C1 of the isocitrate numbering, bearing the hydroxyl) breaks heterolytically. This is the retro-aldol step: an aldol addition product (isocitrate can be viewed as the aldol adduct of glyoxylate's enol equivalent and the succinate-derived carbanion) reverts to its two carbonyl/enolate fragments.

Step 3. The bond cleaved is specifically the C2–C3 bond in the older citrate-based numbering (or equivalently the bond between the alkoxide-bearing carbon and its neighbor), releasing two fragments: succinate (HOOC–CH₂–CH₂–COOH, symmetric and now bearing the double bond electrons quenched by protonation) and the enol of glyoxylate, which tautomerizes to the aldehyde form OHC–COO⁻.

Step 4. Protonation of the succinate carbanion at C3 (from solvent or an active-site acid) and tautomerization of the glyoxylate enolate to the stable aldehyde complete the reaction. Every one of isocitrate's six carbons is accounted for: four in succinate, two in glyoxylate. No CO₂ leaves.

This is mechanistically the mirror image of citrate synthase's aldol-type condensation (acetyl-CoA enolate attacking the oxaloacetate carbonyl), except run on isocitrate rather than oxaloacetate, and cleaving rather than forming the bond. The enzyme is a member of the same broad structural superfamily as PEP mutase and other divalent-metal-independent lyases in some species, but its catalytic strategy — general base-mediated retro-aldol cleavage stabilized by an active-site electrophile that stabilizes developing negative charge — is common across isocitrate lyases regardless of metal dependence.

Goal: show how malate synthase performs a Claisen-type condensation analogous to citrate synthase, joining a second acetyl-CoA to glyoxylate to regenerate a four-carbon dicarboxylic acid, thereby closing the cycle. Malate synthase catalyzes essentially the same class of reaction as citrate synthase — condensation of the acetyl-CoA methyl group (as a stabilized enolate) with an aldehyde/ketone carbonyl — but the electrophile here is glyoxylate's aldehyde carbon rather than oxaloacetate's ketone carbon, and the product is a hydrolyzed free acid (malate) rather than a thioester-linked citryl-CoA that is subsequently hydrolyzed.

Step 1. An active-site general base (often an aspartate or a coordinated water positioned by the enzyme) abstracts a proton from the methyl group of acetyl-CoA. Because this proton is alpha to the thioester carbonyl, its removal is stabilized by delocalization of the resulting carbanion's electron density into the C=O π system, forming an enolate of the thioester — the same activation strategy used by citrate synthase.

Step 2. The enolate carbon (nucleophilic CH₂⁻ adjacent to the thioester) attacks the electrophilic aldehyde carbon of glyoxylate (OHC–COO⁻). Electrons flow from the enolate carbon into a new C–C bond, while the glyoxylate carbonyl π electrons shift onto its oxygen, generating an alkoxide at what becomes the malate C2 hydroxyl-bearing carbon.

Step 3. This new C–C bond formation produces malyl-CoA, an enzyme-bound intermediate: CoA–S–CO–CH₂–CH(OH)–COO⁻, analogous to citryl-CoA in the citrate synthase mechanism.

Step 4. The alkoxide (or an active-site base) then activates a water molecule for nucleophilic attack on the thioester carbonyl carbon of malyl-CoA. Water's oxygen attacks the carbonyl carbon; the tetrahedral intermediate collapses, expelling CoA-thiolate (a good leaving group, stabilized as CoASH after protonation) and leaving behind the free carboxylate of malate.

This hydrolysis step is strongly exergonic (thioester bond cleavage, ΔG°′ ≈ −31 kJ/mol for a typical acyl-CoA thioester hydrolysis) and is what renders the whole condensation essentially irreversible, exactly as it does for citrate synthase. This irreversibility is thermodynamically important: it is the "ratchet" step that pulls the whole glyoxylate cycle forward, just as citrate synthase's irreversibility pulls the ordinary citric acid cycle forward.

Example 1 — Foundational: tracking carbons through one full turn with explicit atom labels. Take citrate synthase's product, citrate, formed from oxaloacetate (label its four carbons Oa–Ob–Oc–Od, where Oa and Ob are the two carbons nearer the carbonyl/keto end) plus acetyl-CoA (label its two carbons Aa–methyl and Ab–carbonyl). Citrate synthase joins Aa to Ob (the keto carbon of OAA), giving citrate with six carbons in the chain Oa–Ob(Aa)(Ab)... — for bookkeeping purposes here we only need the total count, not the exact stereochemical assignment (which is handled rigorously in citrate synthase's own mechanism). Citrate (6C) → isocitrate (6C, aconitase just moves the OH group, no atoms lost) → isocitrate lyase cleaves to give succinate (4C, made of Oa-Ob-Oc-Od, the original OAA-derived carbons) and glyoxylate (2C, made of Aa-Ab, the first acetyl-CoA's carbons). Now malate synthase combines this glyoxylate (Aa-Ab, 2C) with a second, fresh acetyl-CoA (call its carbons Ba-methyl, Bb-carbonyl) to give malate: 4 carbons, Aa-Ab-Ba-Bb in some connectivity. Malate dehydrogenase oxidizes this to a new oxaloacetate (still Aa-Ab-Ba-Bb, 4C), ready to accept a third acetyl-CoA next turn. Tally: two acetyl-CoA consumed (four carbons: Aa, Ab from the first; Ba, Bb from the second). Output: one succinate (Oa-Ob-Oc-Od, four carbons — note these are the original oxaloacetate's carbons, not the incoming acetyl-CoA's!) exported to gluconeogenesis, and one new oxaloacetate (Aa-Ab-Ba-Bb) retained to run the cycle again. Total carbon in: 4 (two acetyl groups). Total useful product carbon out: 4 (succinate). The catalytic oxaloacetate is a different four carbons than it started with, but its quantity (one molecule) is unchanged — exactly the invariant identified in the second derivation.

Example 2 — Applied: acetate as sole carbon source for bacterial growth, quantitative carbon and NADH yield. E. coli grown on acetate as sole carbon source activates it to acetyl-CoA (via acetate kinase/phosphotransacetylase or acetyl-CoA synthetase, costing one ATP-equivalent per acetate). Suppose a cell needs to synthesize one glucose-6-phosphate equivalent (for cell wall/glycogen precursors) starting purely from acetate. Two turns of the glyoxylate cycle are needed to generate two succinate molecules (using 4 acetyl-CoA total, i.e., 4 acetate consumed), since gluconeogenesis needs two triose-phosphate-equivalent units (from two PEP) to build one hexose.

Each succinate is then oxidized (succinate dehydrogenase, generating 1 FADH₂ each) to fumarate, hydrated to malate, oxidized (malate dehydrogenase, 1 more NADH each) to oxaloacetate, and decarboxylated by PEP carboxykinase (costing 1 GTP each, releasing 1 CO₂ each — this CO₂ loss is a necessary cost of the OAA→PEP step, distinct from the glyoxylate-cycle-avoided decarboxylations) to PEP:

Two PEP then run reversed glycolysis (costing additional ATP/GTP for the 1,3-BPG and PEP-forming steps run backward, well known from gluconeogenesis) to yield one glucose. Counting carbons: 4 acetate × 2 C = 8 C in; 2 CO₂ released in the OAA→PEP step = 2 C out as gas; 6 C remain, matching 2 PEP (2 × 3C = 6C) exactly, which combine into 1 glucose (6C). The arithmetic closes perfectly: 8 carbons in, 2 lost as CO₂ in the necessary decarboxylative step of gluconeogenesis (not the glyoxylate cycle itself), 6 carbons make it into glucose — a 75% carbon-retention efficiency from acetate to glucose, a number directly checkable against the mass balance above (6/8 = 0.75).

Example 3 — Exam/competition level: isotopic labeling to distinguish the two cycles experimentally. A classic experiment (conceptually following Kornberg and Krebs, 1957, who first proposed the glyoxylate cycle from labeling data in E. coli grown on acetate) uses acetate labeled with ¹⁴C specifically at the carboxyl carbon (i.e., ¹CH₃–¹⁴COOH). Predict where the label ends up in succinate after one glyoxylate-cycle turn, and contrast with the ordinary citric acid cycle prediction. Set up the labeling using citrate synthase stereochemistry: acetyl-CoA's carbonyl carbon (labeled, "Ab" using Example 1's notation) becomes bonded to oxaloacetate's C2 (the keto carbon) in forming citrate; the methyl carbon (unlabeled, "Aa") becomes a CH₂ group of citrate. Citrate synthase is known to act with strict stereospecificity — it uses only the pro-S arm of citrate for subsequent aconitase chemistry, meaning the "new" acetyl-derived carbons and the "old" OAA-derived carbons are not symmetrically scrambled at the isocitrate stage, unlike a naive symmetric-molecule argument might suggest. Following the established stereochemical mapping: in isocitrate, the carbons derived from the incoming acetyl-CoA (Aa, Ab) end up specifically in the segment that isocitrate lyase releases as glyoxylate, while the four carbons derived from the original oxaloacetate (Oa–Od) are released as succinate. This is a nontrivial, experimentally established result (not guessable from molecular symmetry alone) that confirms citrate synthase's known "prochiral" stereospecificity — if citrate reacted symmetrically, the label would scramble 50/50 between succinate and glyoxylate, but it does not. Therefore: label the incoming acetyl-CoA's carboxyl carbon (Ab position). Prediction: this label appears in glyoxylate, not in the succinate released by isocitrate lyase in that same turn. The glyoxylate then condenses (malate synthase) with a second unlabeled or differently-labeled acetyl-CoA to form malate, and eventually a new oxaloacetate — so the original label re-enters the cycle's catalytic OAA pool rather than exiting as succinate immediately. Only after a second turn, when this newly-formed (labeled) oxaloacetate reacts with a fresh acetyl-CoA and its isocitrate is cleaved, does the ¹⁴C label finally appear in an exported succinate. This distinguishes the glyoxylate cycle from the ordinary citric acid cycle experimentally: in the ordinary cycle, label from acetyl-CoA's carboxyl carbon is lost as ¹⁴CO₂ within the first turn (specifically at the isocitrate dehydrogenase step, following established labeling patterns), whereas in the glyoxylate cycle that same carbon survives, gets temporarily retained in the OAA pool, and eventually surfaces in exported succinate one turn later — with zero loss to CO₂ at any point. Kornberg and Krebs's actual experiments, tracking labeled acetate incorporation into cell-wall carbohydrate in acetate-grown bacteria, found exactly this pattern of full carbon retention, which is what first revealed the glyoxylate bypass's existence.

Hans Adolf Krebs, already the discoverer of the citric acid cycle and the urea cycle, turned in the 1950s (with his student Hans Kornberg, no relation to Arthur Kornberg) to a puzzle that his own cycle could not explain: how do bacteria and germinating plants convert two-carbon compounds like acetate into net cell carbohydrate? Krebs had spent the 1930s and 1940s establishing that the citric acid cycle was a closed, catalytic combustion loop — elegant precisely because oxaloacetate is always regenerated, never net consumed or produced. That same elegance was the obstacle: a perfectly catalytic cycle, by definition, cannot manufacture new carbon skeletons from its substrate. Kornberg and Krebs approached the problem the way Krebs always did — by careful manometric and isotopic tracer experiments on intact bacterial and plant tissue, watching where labeled carbon actually went rather than reasoning from presumed pathways. Working with E. coli grown on acetate as sole carbon source, and with germinating castor bean endosperm (which Kornberg showed contained a specialized particulate fraction — later named the glyoxysome by others), they identified two enzyme activities never before characterized: an isocitrate-cleaving enzyme (isocitrate lyase) and a glyoxylate-condensing enzyme (malate synthase). Their 1957 paper in Nature, "Synthesis of Cell Constituents from C2-Units by a Modified Tricarboxylic Acid Cycle," laid out essentially the pathway taught today, remarkable both for its speed of discovery and for the fact that it required Krebs to revise, in public, the very cycle that had made him a Nobel laureate a few years earlier — a mark of exactly the kind of self-correcting rigor that defines good biochemistry.