Eicosanoids and Lipid Signaling Molecules
A single aspirin tablet dissolves in the bloodstream within minutes. It silently acetylates a serine residue buried in a cyclooxygenase channel. By that one covalent stroke, it shuts down platelet aggregation, fever, and pain signaling all at once. The reason one small modification has such sweeping consequences is simple. The target enzyme sits at the source of an entire signaling economy. That economy is built from a twenty-carbon fatty acid. That fatty acid is arachidonic acid. The messengers carved from it — prostaglandins, thromboxanes, leukotrienes, collectively the eicosanoids — are the local hormones of inflammation, platelet function, smooth muscle tone, and immune defense. Eicosanoids matter because they are not stored, and not transported over long distances. Nor are they translated from a gene when needed, unlike a peptide hormone. They are synthesized transiently. They act within seconds to minutes on cells adjacent to their site of production. They are degraded almost as fast as they appear. This "autocrine/paracrine" logic — release and self-destruct locally — makes them like a fire alarm rather than a broadcast signal. It is exactly this logic that makes cyclooxygenase (COX) and lipoxygenase (LOX) such powerful drug targets. Block the enzyme, and the local alarm never sounds. There is no need to intercept a receptor everywhere in the body. The big picture is a single branching pathway. A membrane phospholipid releases arachidonic acid via phospholipase A2 activation. The free fatty acid, still tethered to the endoplasmic reticulum and nuclear envelope, is then seized by one of two competing enzyme families. Cyclooxygenase converts it into the cyclic endoperoxide PGH2. This is the shared precursor of prostaglandins, prostacyclin, and thromboxane A2. Lipoxygenase instead abstracts a hydrogen and inserts molecular oxygen to build the leukotrienes. This culminates in the potent bronchoconstrictor cysteinyl leukotrienes of asthma. Both branches begin from the same fatty acid backbone described under Fatty Acid Structure and Nomenclature. Both illustrate how oxygen chemistry — radical intermediates, peroxide rearrangements, epoxide openings — can be harnessed with exquisite regio- and stereocontrol by a protein active site. This article works through the full arachidonate cascade mechanistically. It covers the free-radical cyclization chemistry of COX, and the reaction cycle of 5-lipoxygenase. It explains the enzymatic logic distinguishing PGI2 from TXA2 in the vasculature. It also covers the pharmacology of aspirin, NSAIDs, and leukotriene modifiers. Every mechanism is drawn as electron flow. Every number — carbon counts, double bond positions, kinetic parameters — is independently checkable against the structures shown.
- Arachidonic acid (20:4, all-cis-Δ5,8,11,14) is released from membrane phospholipids by phospholipase A2 and is the universal eicosanoid precursor.
- Cyclooxygenase (COX-1/COX-2) uses a tyrosyl radical to abstract the 13-pro-S hydrogen, triggering bis-oxygenation and cyclization to PGG2, then peroxidase reduction to PGH2.
- PGH2 is partitioned by tissue-specific synthases into PGE2, PGF2α, PGD2, prostacyclin (PGI2, endothelium) and thromboxane A2 (TXA2, platelets) — a physiological push–pull on vascular tone and clotting.
- Aspirin irreversibly acetylates Ser530 (COX-1) near the active site, sterically blocking substrate access; other NSAIDs competitively occupy the same channel without covalent modification.
- 5-Lipoxygenase, aided by FLAP, converts arachidonate to 5-HPETE and then LTA4, the branch point for the potent bronchoconstrictor cysteinyl leukotrienes (LTC4/D4/E4) and the neutrophil chemoattractant LTB4.
- Eicosanoids act locally through G-protein-coupled receptors (EP, IP, TP, FP, DP, BLT, CysLT) and are inactivated within one pass through the lungs — true paracrine, not endocrine, signals.
- Key idea: one branched oxidative cascade from a single 20-carbon fatty acid generates a family of short-range, receptor-specific messengers whose selective enzymatic and receptor targeting is the pharmacological basis of anti-inflammatory and anti-asthmatic therapy.
Where this lives today
Eicosanoid pharmacology remains one of the largest single categories of prescribed and over-the-counter drugs. NSAIDs (ibuprofen, naproxen, indomethacin) act as reversible competitive COX inhibitors for pain and inflammation; low-dose aspirin remains first-line secondary prevention for myocardial infarction and stroke through the irreversible platelet mechanism worked out in Example 2; selective COX-2 inhibitors (celecoxib) were designed using crystal structures of the COX-2 side pocket to spare gastric COX-1 while retaining anti-inflammatory efficacy, though cardiovascular risk from suppressed endothelial PGI2 limited their use (rofecoxib's withdrawal in 2004 being the clinical demonstration of the TXA2/PGI2 balance problem in Example 3). Latanoprost and related PGF2α analogs lower intraocular pressure in glaucoma by increasing uveoscleral outflow via FP receptor activation. Misoprostol, a PGE1 analog, protects gastric mucosa and induces labor/uterine contraction via EP/FP receptor activity, exploiting the same PGE2-mediated cytoprotection that NSAIDs disable. On the leukotriene side, montelukast and zafirlukast (CysLT1 receptor antagonists) and zileuton (a direct 5-LOX inhibitor) treat asthma by blocking bronchoconstriction and mucus secretion driven by LTC4/D4/E4, while LTB4-receptor antagonism is an active target in neutrophil-driven inflammatory disease. Fish-oil supplementation, by supplying EPA as a competing 5-LOX/COX substrate, produces the less-inflammatory "3-series"/"5-series" eicosanoids discussed in the limiting-cases table, providing a nutritional lever on the same enzymatic branch point. Beyond pharmacology, eicosanoid biosynthesis is central to research in cardiovascular disease, cancer (COX-2 upregulation in tumors and its link to angiogenesis via PGE2), and reproductive biology (PGF2α-triggered luteolysis), making this single branched pathway one of the most therapeutically mined regions of human biochemistry.
Ask the tutor: How does the geometry of the COX and 5-LOX substrate channels dictate which bis-allylic carbon (C13 vs. C7) gets oxidized, and could that channel geometry be engineered to redirect eicosanoid output?
The physical picture — arachidonate, released and branched
Arachidonic acid is a 20-carbon fatty acid with four cis double bonds at positions 5, 8, 11, and 14 counting from the carboxyl carbon (20:4 Δ5,8,11,14; equivalently an ω-6 fatty acid). In the resting cell it is almost never free: it is esterified, overwhelmingly at the sn-2 position of membrane glycerophospholipids, particularly phosphatidylcholine and phosphatidylinositol. A rise in intracellular Ca²⁺ — following, for instance, thrombin binding to platelet PAR receptors or antigen cross-linking of mast-cell IgE receptors — recruits cytosolic phospholipase A2 (cPLA2) to the membrane via its C2 domain, where it hydrolyzes the sn-2 ester bond and liberates arachidonate directly into the vicinity of the downstream oxygenases, which are themselves membrane-associated proteins of the endoplasmic reticulum and nuclear envelope. From this single liberated fatty acid, two oxidative enzyme families compete for substrate. Cyclooxygenase (prostaglandin H synthase, PGHS) is a bifunctional heme enzyme: a cyclooxygenase active site performs the committed double dioxygenation and cyclization, and a spatially distinct peroxidase active site reduces the resulting hydroperoxide. Its immediate product, PGH2, is an unstable bicyclic endoperoxide that lasts only minutes before downstream isomerases and synthases convert it, in a tissue-specific manner, to the classical prostaglandins (PGE2, PGF2α, PGD2), prostacyclin (PGI2), or thromboxane A2 (TXA2). The 5-lipoxygenase (5-LOX) pathway instead uses a non-heme iron center to abstract a hydrogen at C7 of arachidonate, insert O2 at C5, and dehydrate the resulting hydroperoxide into the allylic epoxide leukotriene A4 (LTA4) — the branch point for the leukotrienes. The prostanoid numbering system encodes real chemical information worth memorizing because it recurs throughout the pharmacology: the letter (E, F, D, I) denotes the substitution pattern on the cyclopentane ring, and the subscript (1, 2, 3) denotes the number of non-ring double bonds in the side chains, which in turn reflects the number of double bonds in the precursor fatty acid (dihomo-γ-linolenic acid 20:3 gives the "1" series, arachidonic acid 20:4 gives the physiologically dominant "2" series, and eicosapentaenoic acid 20:5, from fish oil, gives the anti-inflammatory "3" series).
The core ideas — stoichiometry, kinetics, and receptor logic
The cyclooxygenase reaction is a double dioxygenation with a well-defined stoichiometry that every mechanism must reproduce:
where: PGG2 is the 15-hydroperoxy endoperoxide, formed by insertion of one O2 at C11 (initiating the bicyclic ring closure) and a second O2 at C15.
where: the peroxidase site of the same bifunctional enzyme reduces the 15-hydroperoxide to a 15-hydroxyl, consuming two electrons typically supplied by a co-substrate (often the enzyme's own heme cycle regenerating via cellular reductants). Reading: cyclooxygenase is often loosely called "COX" as if it did one thing, but the balanced equations show it is genuinely bifunctional — a cyclooxygenase active site (radical chemistry, C–C bond formation, two new C–O bonds) and a peroxidase active site (two-electron reduction) operating in series within one polypeptide. The 5-lipoxygenase stoichiometry is simpler, a single dioxygenation:
where: 5-HPETE is 5-hydroperoxy-6,8,11,14-eicosatetraenoic acid, formed by hydrogen abstraction at C7 and O2 addition at C5 with double-bond migration (Δ6,8,11,14 → conjugated Δ6,8 diene retained, Δ5 hydroperoxide installed). The essential kinetic idea distinguishing aspirin from other NSAIDs is a rate law for covalent versus reversible inhibition. For a simple competitive (reversible) inhibitor:
where: is maximal velocity, the Michaelis constant for arachidonate, inhibitor concentration, its dissociation constant; recovery of activity requires only that fall (drug clearance), since no bond is broken. For aspirin, inhibition is irreversible and time-dependent, following pseudo-first-order covalent inactivation kinetics:
where: is the fraction of active enzyme remaining, the second-order acetylation rate constant; because the modification is covalent (a serine hydroxyl converted to an acetate ester), enzyme activity is restored only by synthesis of new enzyme protein — for anucleate platelets, that means for the ~10-day platelet lifetime.
Goal. Show, step by step with electron/radical flow, how the enzyme converts linear arachidonate into the bicyclic PGG2 endoperoxide using one tyrosyl radical and two molecules of O2, explaining why the pro-S hydrogen at C13 is removed and why the product ring system (a cyclopentane fused bridge with a 9,11-endoperoxide) is essentially the only product formed.
Step 1 — Radical initiation. The heme prosthetic group of COX, in its resting ferric state, reacts with a peroxide (often a trace of the enzyme's own product, PGG2 itself, in an autocatalytic "priming" step) to form an oxoferryl (Fe(IV)=O) porphyrin radical cation intermediate (Compound I chemistry, structurally analogous to horseradish peroxidase). This high-valent iron–oxo species oxidizes a nearby tyrosine, Tyr385, to a tyrosyl radical by one-electron transfer, generating a phenoxyl radical with substantial spin density on the ortho and para ring carbons.
Step 2 — Hydrogen abstraction at C13. Tyr385 sits directly above C13 of bound arachidonate in the hydrophobic cyclooxygenase channel. The tyrosyl radical abstracts the pro-S hydrogen at C13 — a bis-allylic position flanked by the Δ11 and Δ14 double bonds — because bis-allylic C–H bonds are especially weak (BDE ≈ 75 kcal/mol, versus ~100 kcal/mol for an ordinary sp³ C–H) owing to delocalization of the resulting radical over both adjacent π systems. This generates a pentadienyl radical delocalized over C11–C15.
Step 3 — First O2 addition (C11) and ring closure. Molecular O2, itself a diradical, adds preferentially at C11 of the delocalized radical (the site favored by the protein's steric constraints, which orient the substrate in a characteristic "L-shaped" conformation folding C8–C12 back toward C13–C20). The resulting C11-peroxyl radical is positioned to attack C9 intramolecularly, forming the new C9–C11 bond of the cyclopentane-fused ring and generating a carbon radical at C8. Concertedly, the peroxyl oxygen also bridges to C9, closing the characteristic bicyclic (dioxabicyclo[2.2.1]) endoperoxide — this is the step that converts a linear fatty acid into a fused-ring prostanoid skeleton, and it happens because the enzyme's channel geometry holds C8–C12 in a folded, near-eclipsed conformation that makes the 5-membered ring closure geometrically favorable.
Step 4 — Second O2 addition (C15) and radical termination. The carbon radical generated at C15 (after rearrangement of the allylic system from the C8 radical through the remaining Δ14 alkene) reacts with a second O2 molecule to install the C15-hydroperoxide, and the resulting peroxyl radical abstracts a hydrogen (regenerating Tyr385 as a phenol, completing the catalytic cycle) to give the neutral hydroperoxide product PGG2. Two molecules of O2 and two radical propagation steps have converted arachidonate into a bicyclic endoperoxide with a pendant hydroperoxide — matching exactly the stoichiometry written above.
Step 5 — Peroxidase reduction to PGH2. PGG2 diffuses (without leaving the enzyme, in physiological turnover) to the spatially separate heme peroxidase site, where the 15-hydroperoxide is reduced by two electrons to the 15-hydroxyl, releasing PGH2 and regenerating the ferric resting state (or, if reduction is incomplete, regenerating the oxoferryl species that re-initiates Step 1 in the next catalytic cycle — this autocatalytic coupling is why COX activity shows a lag phase that peroxide "primes" away).
Go deeper · Derivation B · 2 minThe 5-LOX to LTA4 pathway: non-heme iron, hydrogen abstraction, and epoxide formation
Goal. Trace, with explicit electron/proton flow, how 5-lipoxygenase converts arachidonate into 5-HPETE and then how a dehydration step converts the hydroperoxide into the reactive allylic epoxide LTA4 — the true branch point of leukotriene biosynthesis.
Step 1 — Iron activation. Resting 5-LOX carries a mononuclear non-heme Fe(II) center coordinated by three histidines, one carboxylate (Asp/Ile C-terminus), and typically a hydroxide/water ligand. An external oxidant (a lipid hydroperoxide, or the initial traces of the enzyme's own 5-HPETE product) oxidizes this to catalytically active Fe(III)–OH, which is the hydrogen-abstracting species — directly analogous in logic, though not in cofactor identity, to the tyrosyl-radical abstraction used by COX.
Step 2 — Hydrogen abstraction at C7. The Fe(III)–OH center abstracts the pro-S hydrogen from C7 of arachidonate, a bis-allylic position flanked by the Δ5 and Δ8 double bonds (note this is a different bis-allylic carbon from the C13 used by COX, reflecting a different substrate-binding orientation in the two enzyme families). Abstraction is again facile because of the low bond dissociation energy of bis-allylic C–H, and it reduces the iron center back to Fe(II)–OH2 while generating a pentadienyl radical over C5–C9.
Step 3 — Regiospecific O2 insertion at C5. Molecular O2, held in a hydrophobic pocket positioned over C5, adds to the delocalized radical specifically at C5 (rather than at C9, the other terminus of the pentadienyl system), dictated by the geometry of the substrate-binding channel, which is narrower and shorter than the COX channel and admits arachidonate "tail-first." The resulting C5-peroxyl radical abstracts a hydrogen (from a second substrate molecule or from a reduced iron/tyrosine relay) to give the neutral product 5-HPETE, with the double-bond pattern rearranged to the conjugated cis,trans-diene Δ6,8 plus the retained Δ11 and Δ14 alkenes.
Step 4 — Dehydration to LTA4. 5-LOX itself (or, in some schemes, a second turnover at the same active site) catalyzes elimination of water from 5-HPETE: the C5 hydroperoxide oxygen is retained as an epoxide oxygen bridging C5–C6, while a proton is lost from C10 and the double-bond system reorganizes into the conjugated triene Δ6,8,10 characteristic of leukotrienes. The leaving group is the distal peroxide oxygen (as hydroxide, or trapped by active-site residues), and the mechanism is best described as an E1cb-like elimination: base removal of the C10 proton generates a carbanion/allylic anion stabilized by the adjacent triene system, which pushes electrons through to expel the peroxide oxygen as the epoxide ring closes at C5–C6. The product, leukotriene A4 (LTA4), is an unstable allylic epoxide — reactive enough that in the absence of a downstream enzyme it hydrolyzes non-enzymatically in water to a mixture of 6-trans isomers of LTB4 (5,6-diHETE regioisomers), which is why cells channel LTA4 directly to LTA4 hydrolase or LTC4 synthase rather than releasing it into free solution.
Step 5 — Branch point. LTA4 hydrolase (a zinc metalloenzyme) opens the epoxide with regiospecific water attack at C12 to give LTB4, a potent neutrophil chemoattractant. Alternatively, LTC4 synthase (a membrane glutathione transferase family member) opens the same epoxide with the thiol of glutathione attacking at C6, installing a thioether linkage and generating LTC4, the parent cysteinyl leukotriene, which is then trimmed extracellularly (loss of glutamate, then glycine) to LTD4 and LTE4 — the pharmacologically dominant components of what was historically called "slow-reacting substance of anaphylaxis."
Worked examples
Example 1 (foundational) — Counting the oxygens and mapping COX stoichiometry onto molecular formulas. Arachidonic acid has the molecular formula C20H32O2 (a carboxylic acid, so it already carries two oxygens in the –COOH group). Confirm the molecular formula of PGH2 from the stated stoichiometry and check it against the known structure. From the two dioxygenation steps in the derivation:
Each O2 adds two oxygen atoms without loss of any atoms (the cyclization forms new C–C and C–O bonds but breaks no C–H bonds net — one H is abstracted at C13 but the radical chain ultimately reincorporates an H at the terminal peroxyl step, so H count is conserved). Thus PGG2 has 2 + 4 = 6 oxygens... but by convention PGG2 is drawn as C20H32O5, because one of the four added oxygens ends up as part of the internal ring-oxygen bridge shared between two carbons (C9 and C11 share a single bridging O–O unit contributing only one new "extra" oxygen beyond the ring carbons it connects, while the second O2 contributes a full hydroperoxide –OOH at C15 contributing two oxygens). Explicitly:
Reduction of the C15 hydroperoxide to a hydroxyl in the peroxidase step converts PGG2 (C20H32O5) to PGH2 (C20H32O4) by loss of one oxygen atom (as water, using 2 electrons and 2 protons):
This matches the accepted structure of PGH2: a bicyclic endoperoxide (2 oxygens in the 9,11-bridge) plus a single C15 hydroxyl (1 oxygen) plus the carboxylic acid (2 oxygens) = 5 oxygens total, consistent with molecular formula C20H32O5 for PGH2 as conventionally tabulated in enzyme databases (some sources write PGH2 with 5 oxygens counting the carboxylate, and PGG2 with 6 — the internal bookkeeping detail matters less than the two-dioxygenation, one-reduction logic, which is the mechanistically load-bearing fact).
Example 2 (applied) — Aspirin pharmacokinetics: how a low, once-daily dose achieves selective antiplatelet effects. Low-dose aspirin (75–100 mg) is used specifically for cardioprotection because of a pharmacokinetic asymmetry between platelets and endothelial cells. Aspirin acetylates Ser530 of COX-1 irreversibly, with a plasma half-life of aspirin itself of only about 20 minutes (rapid hydrolysis to inactive salicylate by esterases). Platelets are anucleate and cannot synthesize new COX-1; once acetylated, a platelet's thromboxane A2 output is permanently zero for that platelet's lifetime (8–10 days). Endothelial cells, by contrast, are nucleated and re-synthesize COX-2 (and COX-1) within hours, regenerating their prostacyclin (PGI2) output well before the next dose. Quantitatively, model platelet inhibition as cumulative and irreversible. If a fraction of the circulating platelet pool is newly produced each day (normal platelet turnover replaces about 10% of the pool daily, consistent with a 10-day lifespan), then at steady state under continuous daily low-dose aspirin, the fraction of platelets with functional COX-1 approaches:
meaning roughly 90% of the circulating platelet pool is permanently thromboxane-deficient at any moment during chronic low-dose therapy, even though the drug itself is cleared from plasma within a few hours. Endothelial COX-2 recovers to near 100% activity between doses because of ongoing protein synthesis, so the net effect of the once-daily 75 mg dose is a durable shift of the TXA2:PGI2 balance toward PGI2, favoring vasodilation and reduced platelet aggregation — this asymmetry (irreversible, cumulative platelet inhibition vs. reversible, resynthesized endothelial inhibition) is the mechanistic reason low-dose aspirin is antithrombotic while doses ten times higher (600 mg–1 g, used for analgesia) also inhibit endothelial COX-2/COX-1 enough to blunt the benefit by suppressing PGI2 as well.
Example 3 (exam/competition level) — Deducing an unknown eicosanoid receptor pathway from EC50 and second-messenger data. A pharmacology problem states: a novel eicosanoid X, applied to vascular smooth muscle cells, produces relaxation with EC50 = 3 nM; the effect is blocked by pertussis toxin pretreatment is absent (i.e., Gi is not involved), but is abolished by an adenylate cyclase inhibitor (SQ22536) and mimicked by dibutyryl-cAMP. Furthermore, X is chemically an unstable bicyclic compound with an enol-ether oxygen bridging C6–C9, derived from PGH2, with a half-life in aqueous buffer of about 3 minutes at pH 7.4. Identify X, its receptor, and predict the effect of a thromboxane synthase inhibitor on its production. Reasoning. The short aqueous half-life (~3 min) immediately signals an unstable prostanoid, not a stable prostaglandin like PGE2 (half-life in blood ~30 s to a few minutes but chemically the molecule itself is stable in buffer; it is metabolically inactivated, not chemically unstable). A structural half-life of minutes in buffer points instead to PGI2 (prostacyclin) or TXA2, both of which are chemically labile due to strained ring oxygens (PGI2's enol ether hydrolyzes to 6-keto-PGF1α; TXA2's oxane-oxetane bridge hydrolyzes to inactive TXB2). The explicit mention of "enol-ether oxygen bridging C6–C9" is diagnostic: this is the bicyclic vinyl ether structure unique to prostacyclin, distinguishing it from thromboxane's oxane ring. So X = PGI2. The signaling data confirm this identification independently: PGI2 acts through the IP receptor, a Gs-coupled GPCR, which activates adenylate cyclase and raises cAMP — exactly consistent with the abolition of the effect by an adenylate cyclase inhibitor, the mimicry by a cell-permeant cAMP analog, and the lack of pertussis-toxin sensitivity (ruling out Gi-coupled receptors such as the EP3 prostaglandin receptor). Vascular smooth muscle relaxation via cAMP-dependent protein kinase (PKA)-mediated reduction of intracellular Ca²⁺ and myosin light chain kinase activity is the well-established mechanism of PGI2-induced vasodilation, complementary to its cAMP-mediated inhibition of platelet activation. Because PGI2 and TXA2 are both derived from the same PGH2 pool but are produced by different tissue-specific synthases (prostacyclin synthase in endothelium, thromboxane synthase in platelets), a thromboxane synthase inhibitor blocks the platelet branch and, since the two synthases compete for a shared, diffusion-limited PGH2 pool, tends to shunt additional PGH2 substrate toward prostacyclin synthase in adjacent endothelial cells — predicting an increase in PGI2 production, not a decrease, upon thromboxane synthase inhibition. This substrate-shunting phenomenon is exactly what is observed pharmacologically with thromboxane synthase inhibitors and is the mechanistic rationale for their (limited) clinical interest as antithrombotic agents that spare or even enhance the protective PGI2 signal.
Limiting cases, scope & common traps
| Scenario | What happens | Why |
|---|---|---|
| COX-1 vs. COX-2 selectivity | COX-1: constitutive, GI/platelet; COX-2: inducible, inflammation/pain | COX-2 active site has a larger side pocket (Val523 vs. Ile523 in COX-1) permitting selective inhibitors (coxibs) to bind without blocking COX-1 |
| Aspirin at very low dose (≈75 mg) | Selective, cumulative platelet inhibition; endothelial COX-2 recovers | Irreversible acetylation + platelet anucleation vs. nucleated endothelial resynthesis (Example 2) |
| Fish-oil-derived (ω-3, EPA) eicosanoids | "3-series" prostaglandins/thromboxanes and "5-series" leukotrienes, generally less potent | EPA (20:5) competes with arachidonate for COX/LOX but its extra double bond changes downstream product potency, e.g., TXA3 is a weak platelet activator vs. TXA2 |
| Leukotriene pathway with 5-LOX knocked out | No LTB4, no cysteinyl leukotrienes; COX branch unaffected | Independent enzyme families sharing only the substrate pool, not catalytic machinery |
| Zileuton / 5-LOX inhibitors in asthma | Blocks all leukotriene synthesis at the first committed step | Direct competitive inhibition of 5-LOX iron-dependent hydrogen abstraction |
| Montelukast (a CysLT1 antagonist) | Blocks LTC4/D4/E4 action, not their synthesis | Receptor antagonist downstream of 5-LOX; leukotrienes are still made, just not sensed by bronchial smooth muscle |
| High-dose NSAID overdose | GI ulceration, renal impairment | Loss of protective COX-1-derived PGE2/PGI2 in gastric mucosa and renal microvasculature |
The chemist behind it
Sune Bergström, together with his student Bengt Samuelsson, isolated and structurally characterized the first prostaglandins from seminal vesicle and sheep prostate extracts in the 1950s–60s, establishing the cyclopentane-ring nomenclature system (E, F, A, B series) still used today, work that required painstaking chromatographic purification of biologically active lipids present at nanogram quantities in tissue. John Vane, working independently, showed in 1971 that aspirin and related NSAIDs act by inhibiting prostaglandin biosynthesis — explaining, for the first time, a century-old drug's mechanism of action — and subsequently co-discovered prostacyclin (PGI2) and its opposing relationship to thromboxane A2 in vascular biology, establishing the conceptual framework (opposing local mediators controlling clot formation) that underlies modern antiplatelet pharmacology. All three shared the 1982 Nobel Prize in Physiology or Medicine for these discoveries. Bengt Samuelsson went on to elucidate the leukotriene pathway in the late 1970s, isolating LTA4, LTB4, and LTC4/D4/E4 from stimulated leukocytes and establishing the lipoxygenase branch as mechanistically and pharmacologically distinct from the cyclooxygenase branch — foundational work behind the subsequent development of leukotriene-pathway asthma drugs.
Origins · 1 minSir John Robert Vane and how this was discovered
Vane's key experimental innovation was the "cascade superfusion bioassay" — perfusing the effluent of one stimulated tissue directly over a series of indicator tissues, each sensitive to a different eicosanoid, allowing real-time detection of unstable, short-lived lipid mediators (like PGI2, half-life minutes) that conventional isolation chemistry would have destroyed before analysis. This bioassay technique made it possible to detect prostacyclin's existence before its chemical structure was even known, purely from its distinctive, transient, tissue-specific biological activity.
Why are eicosanoids not considered classical hormones?
Classical endocrine hormones are synthesized in specialized glands, stored (often in secretory granules), released into the bloodstream, and act on distant target tissues after transport, sometimes over minutes to hours. Eicosanoids fail every criterion of that definition except release: they are synthesized on demand from membrane phospholipid precursors rather than pre-made and stored, they act locally on the cell that made them (autocrine) or immediately adjacent cells (paracrine), and most have half-lives of seconds to a few minutes in blood because they are rapidly inactivated (largely by a single pass through the pulmonary circulation, where 15-hydroxyprostaglandin dehydrogenase oxidizes the crucial C15 hydroxyl). This local, transient signaling logic — production tightly coupled to an immediate stimulus, action restricted to nearby cells, and rapid self-inactivation — is why eicosanoids are classified as "local hormones" or paracrine/autocrine mediators rather than true circulating endocrine hormones.
What determines whether a cell makes prostacyclin or thromboxane from the same PGH2 precursor?
Both PGI2 and TXA2 come from the identical substrate, PGH2, so the deciding factor is not substrate availability but which terminal synthase enzyme a given cell expresses. Vascular endothelial cells express high levels of prostacyclin synthase, a cytochrome P450 enzyme that isomerizes PGH2 into the enol-ether bicyclic structure of PGI2. Platelets instead express thromboxane synthase, another cytochrome P450 family member, which rearranges PGH2 into the oxane-oxetane bicyclic structure of TXA2. Because these two cell types sit adjacent to each other at the vessel wall–blood interface, and because PGH2 itself can diffuse short distances between cells before reacting further, the local ratio of prostacyclin synthase to thromboxane synthase activity across the endothelium–platelet interface functionally sets the vasodilation/anti-aggregation versus vasoconstriction/pro-aggregation balance — the physiological push–pull exploited in Example 3.
How does aspirin differ mechanistically from ibuprofen if both are called COX inhibitors?
Aspirin (acetylsalicylic acid) transfers its acetyl group covalently onto Ser530 of the COX active-site channel, forming a permanent ester bond; the modified enzyme is catalytically dead for the lifetime of that protein molecule, and activity returns only through new enzyme synthesis. Ibuprofen and most other NSAIDs are reversible competitive inhibitors: they occupy the same hydrophobic channel through noncovalent interactions (hydrogen bonding to Arg120, van der Waals contacts along the channel), competing with arachidonate for binding, but they do not form a covalent bond and dissociate once drug concentration falls. This distinction has real pharmacokinetic consequences: aspirin's antiplatelet effect persists for the ~10-day platelet lifespan after a single dose (Example 2), whereas ibuprofen's COX inhibition tracks its plasma concentration and disappears within hours of the drug's clearance.
Why do fish-oil-derived (ω-3) fatty acids produce "less inflammatory" eicosanoids?
Eicosapentaenoic acid (EPA, 20:5 ω-3) competes directly with arachidonic acid (20:4 ω-6) for the same COX and LOX enzymes, since both are 20-carbon polyunsaturated fatty acids that fit the same active-site channels. When EPA is the substrate instead of arachidonate, COX produces the "3-series" prostanoids (PGE3, PGI3, TXA3) and 5-LOX produces "5-series" leukotrienes (LTB5) rather than the "2-series"/"4-series" products from arachidonate. The extra double bond present in EPA-derived products generally reduces their receptor potency — TXA3, for instance, is a substantially weaker platelet activator than TXA2 at the TP receptor — so a diet enriched in EPA shifts the overall eicosanoid output toward less potent pro-inflammatory and pro-thrombotic mediators, without necessarily eliminating prostanoid signaling altogether.
What is the biochemical difference between the E, F, and D letters in prostaglandin names?
The letter designates the substitution pattern on the cyclopentane ring formed during the endoperoxide-to-prostaglandin isomerization step. PGE compounds carry a 9-keto and 11-hydroxyl group; PGF compounds carry a 9,11-diol (both hydroxyls, no ketone); PGD compounds carry a 9-hydroxyl and 11-keto group — essentially a positional swap of the keto/hydroxyl assignment relative to PGE. These differences arise because distinct isomerase enzymes (PGE synthase, PGF synthase, PGD synthase) act on the same PGH2 endoperoxide substrate, opening the 9,11-peroxide bridge and reducing or retaining oxidation states at C9 and C11 differently. Because GPCRs are exquisitely sensitive to substituent geometry, this small structural difference translates into entirely distinct receptor selectivity: PGE2 for EP receptors, PGF2α for the FP receptor, PGD2 for the DP receptor, each triggering different downstream physiology.
Why does 5-lipoxygenase require FLAP (5-lipoxygenase-activating protein) to function efficiently in cells?
5-LOX is a soluble cytosolic enzyme in the resting state, but arachidonic acid, once liberated by phospholipase A2, is a hydrophobic molecule that partitions into the nuclear membrane rather than remaining freely soluble in the cytosol. FLAP is an integral nuclear membrane protein that binds free arachidonic acid and presents it to 5-LOX after the enzyme translocates to the nuclear envelope upon cell activation, functioning essentially as a substrate-presentation scaffold rather than a catalytic partner. Without FLAP, 5-LOX retains full catalytic competence in vitro with solubilized substrate, but cannot efficiently access membrane-embedded arachidonate in intact cells — which is why FLAP inhibitors (an alternative drug strategy to direct 5-LOX inhibitors like zileuton) also effectively suppress leukotriene biosynthesis in asthma and inflammatory disease.
Does inhibiting COX-2 always reduce inflammation without side effects, and why did rofecoxib get withdrawn?
Selective COX-2 inhibitors were designed to spare COX-1-dependent gastric cytoprotection while blocking COX-2-dependent inflammatory prostaglandin synthesis, and they succeeded at that specific goal, causing significantly less GI ulceration than nonselective NSAIDs. However, COX-2 is also constitutively expressed in vascular endothelium, where it contributes meaningfully to prostacyclin (PGI2) synthesis; selectively blocking COX-2 while leaving platelet COX-1-derived thromboxane A2 synthesis untouched shifts the local TXA2:PGI2 balance toward TXA2, favoring platelet aggregation and vasoconstriction. Long-term rofecoxib use was associated with increased cardiovascular events (myocardial infarction, stroke), leading to its market withdrawal in 2004 — a direct clinical consequence of the same PGI2/TXA2 physiological push-pull mechanism explored quantitatively in Example 3.