Arachidonic acid metabolism and the eicosanoid pathways describe how a 20-carbon polyunsaturated fatty acid released from cell membrane phospholipids is converted by three enzyme systems into prostaglandins, thromboxanes, leukotrienes, and hydroxyeicosatetraenoic acids. Phospholipase A2 liberates arachidonic acid; cyclooxygenases and lipoxygenases then produce short-lived local hormones that run inflammation, fever, pain, blood clotting, and airway tone. Roughly 1 to 2 percent of dietary energy arrives as arachidonic acid itself, with another 5 to 10 percent as linoleic acid that humans convert to arachidonic acid at only 0.2 to 2 percent efficiency. The rest is stored esterified in membrane phospholipids until a stimulus triggers release.
The three-enzyme architecture of arachidonic acid release
Arachidonic acid is not floating free in the cytoplasm. It sits esterified at the sn-2 position of membrane glycerophospholipids, and nothing happens until an enzyme cuts it loose. Cytosolic phospholipase A2 alpha (cPLA2α) does most of that work. It needs calcium and phosphorylation to reach the membrane, which is why calcium ionophores and cell stress reliably trigger eicosanoid output within seconds.
Secretory phospholipase A2 (sPLA2) adds a second channel. It acts on the outer leaflet, is amplified during sepsis, and can feed arachidonic acid to adjacent cells. This transcellular delivery matters: platelets lack a functional 5-lipoxygenase route, yet they generate leukotriene C4 by borrowing neutrophil-derived leukotriene A4. The eicosanoid system is therefore cooperative, not cell-autonomous.
Reacylation competes with release. Acyl-CoA synthetases and lysophospholipid acyltransferases pull arachidonic acid back into phospholipids, so free arachidonic acid sits at low nanomolar concentrations in resting cells. A measurable eicosanoid burst requires release to outpace reuptake by roughly tenfold. In plain terms: the system is built to stay off, and a burst requires two enzyme systems to outrun a recycling one.
On the output side, three enzyme families act on the same substrate and compete for it. Cyclooxygenase (COX) yields prostaglandins and thromboxanes. Lipoxygenase (LOX) yields leukotrienes and hydroxyeicosatetraenoic acids. Cytochrome P450 yields epoxyeicosatrienoic acids and hydroxylated products.
Cyclooxygenase: prostaglandins, thromboxanes, and the 30-fold selectivity gap
COX-1 is expressed constitutively in gastric mucosa, platelets, and kidney collecting ducts. COX-2 is induced by cytokines and growth factors, with baseline expression in kidney macula densa and brain hippocampus. Both enzymes bis-oxygenate arachidonic acid to prostaglandin G2, then reduce it to prostaglandin H2, which downstream synthases convert into prostaglandin E2, prostaglandin D2, prostaglandin F2α, prostacyclin, and thromboxane A2.
The kinetics separate the isoforms. COX-1 converts arachidonic acid with a Km around 2 to 5 micromolar. COX-2 has a Km near 0.3 to 0.5 micromolar, so at low substrate concentrations COX-2 outpaces COX-1 by roughly five to tenfold. Aspirin exploits a different difference: it acetylates serine 530 in COX-1 and serine 516 in COX-2, but COX-2 retains catalytic activity converting some substrates while COX-1 does not. That is the mechanistic basis for low-dose aspirin sparing gastric prostaglandin production more than once thought.
The quantified comparison that matters clinically: celecoxib's COX-2 selectivity is about 30-fold at therapeutic concentrations, etodolac about 10-fold, and ibuprofen under 1-fold. A 30-fold gap is enough to suppress joint prostaglandin E2 while leaving gastric COX-1 largely intact. The price appears in the cardiovascular column, where prostacyclin-to-thromboxane balance shifts.
| Isoform | Resting expression | Km for arachidonic acid | Main products | Clinical consequence of inhibition |
|---|---|---|---|---|
| COX-1 | Platelets, gastric mucosa, kidney | 2 to 5 micromolar | Thromboxane A2, gastric prostaglandin E2 | Bleeding risk, gastropathy |
| COX-2 | Induced; kidney and brain baseline | 0.3 to 0.5 micromolar | Prostacyclin, prostaglandin E2 at inflammation | Fluid retention, hypertension, thrombotic risk |
| 5-LOX | Neutrophils, mast cells | Substrate-limited by FLAP | Leukotriene B4, cysteinyl leukotrienes | Airway protection in asthma |
Lipoxygenase branches and the FLAP handoff
5-Lipoxygenase requires two co-factors: calcium and 5-lipoxygenase activating protein (FLAP). FLAP is not an enzyme. It is a membrane anchor that presents arachidonic acid to 5-LOX and allows the two-step conversion to leukotriene A4. Zileuton blocks the enzyme directly, while FLAP inhibitors block the handoff. The distinction explains zileuton's short half-life of about 2.5 hours and thrice-daily dosing.
Leukotriene A4 is unstable and branches. Leukotriene A4 hydrolase makes leukotriene B4, a potent neutrophil chemoattractant active at picomolar to nanomolar concentrations. Leukotriene C4 synthase conjugates glutathione to make leukotriene C4, the parent of the cysteinyl leukotrienes. Cysteinyl leukotrienes constrict human bronchi at concentrations near 10 nanomolar, roughly 1,000-fold more potent than histamine on a molar basis.
The 12-lipoxygenase and 15-lipoxygenase branches are less clinically exploited but mechanistically important. 15-LOX products, including lipoxin A4, are anti-inflammatory and promote resolution. The conversion of arachidonic acid toward lipoxins versus leukotrienes is one place where the pathway can be steered rather than merely blocked.
Cytochrome P450 epoxygenases and the underappreciated third branch
The CYP4A and CYP4F families omega-hydroxylate arachidonic acid to 20-hydroxyeicosatetraenoic acid, a potent vasoconstrictor. CYP2C and CYP2J epoxygenases produce epoxyeicosatrienoic acids, which dilate arterioles and inhibit inflammation. Soluble epoxide hydrolase degrades these epoxides within minutes, so their vasodilatory effect is tightly localized.
This branch is where the counterintuitive finding lives. Blocking soluble epoxide hydrolase raises epoxy-eicosatrienoic acid levels and lowers blood pressure in animal models, but human trials have not reproduced a consistent antihypertensive effect. The pathway hypothesis is sound; the clinical translation remains unproven and should not be presented as established.
Arachidonic acid metabolism timeline: from stimulus to resolution
| Time after stimulus | Event | Measurable marker |
|---|---|---|
| 0 to 30 seconds | Calcium rise; cPLA2α translocates to membrane | Free arachidonic acid rises 5 to 10-fold |
| 1 to 5 minutes | COX-1 and 5-LOX act; prostaglandin H2 and leukotriene A4 form | Prostaglandin E2 and leukotriene B4 detectable in tissue fluid |
| 5 to 30 minutes | Thromboxane A2 and cysteinyl leukotrienes peak; platelet aggregation and bronchoconstriction | Urinary 11-dehydro-thromboxane B2 rises |
| 1 to 6 hours | COX-2 induction; sustained prostaglandin E2 output | COX-2 protein detectable by Western blot |
| 6 to 24 hours | Neutrophil influx driven by leukotriene B4; 15-LOX lipoxin synthesis begins | Lipoxin A4 rises; neutrophil count falls |
| 24 to 72 hours | Resolution phase; macrophage clearance of apoptotic neutrophils | Specialized pro-resolving mediators peak |
| 3 to 7 days | Return to baseline esterified arachidonic acid pool | Free arachidonic acid back to nanomolar baseline |
Dietary and tissue variation in arachidonic acid pools
Human plasma arachidonic acid sits around 100 to 300 micromolar when measured as total fatty acid, but free arachidonic acid is under 1 micromolar at rest. The ratio matters more than the absolute amount, because phospholipid pools act as buffer.
Diet shifts the precursor supply. A Western diet delivers roughly 100 to 300 mg of arachidonic acid per day from meat and eggs, plus 10 to 20 grams of linoleic acid. The delta-6-desaturase step converting linoleic acid to gamma-linolenic acid is rate-limiting and slow, which is why linoleic acid loading raises tissue arachidonic acid only modestly in adults. In infants, delta-6-desaturase activity is higher, so the same linoleic acid load produces more arachidonic acid.
Tissue distribution is uneven. Brain phospholipids contain 8 to 12 percent arachidonic acid by weight and turn over slowly. Platelets and neutrophils hold high concentrations for rapid release. Adipose tissue stores very little, so the body has no large inert reservoir; supply depends on continuous synthesis and dietary intake.
Common misconceptions about eicosanoids
The first misconception is that all prostaglandins drive inflammation. Prostaglandin E2 does promote fever and pain sensitization through EP2 and EP4 receptors, but prostaglandin D2 and lipoxin A4 actively resolve inflammation. Blocking COX therefore removes brakes as well as accelerators, which is part of why chronic nonsteroidal use carries risks beyond the stomach.
The second misconception is that arachidonic acid is inherently inflammatory. In isolation, free arachidonic acid is a signaling molecule that can activate protein kinase C and modulate ion channels. The inflammatory output depends on which enzymes are expressed in that cell at that moment.
The third misconception is that COX-2 is purely inducible. It is constitutive in kidney macula densa and brain, which explains why COX-2 selective inhibitors raise blood pressure and why they carry renal warnings even in patients without inflammation.
The fourth is the omega-3 substitution fallacy. Eicosapentaenoic acid competes with arachidonic acid at the COX and LOX active sites, but it produces weaker or different products. Replacing 20 percent of membrane arachidonic acid with eicosapentaenoic acid reduces thromboxane A2 output without abolishing prostaglandin E2 synthesis, so the effect is partial, not absolute.
What to actually do with this pathway map
- When choosing a nonsteroidal anti-inflammatory, match isoform selectivity to risk. Patients with gastroduodenal ulcer history do better with a 30-fold COX-2 selective agent plus gastroprotection; patients with established cardiovascular disease should avoid high-dose selective agents.
- Measure aspirin response rather than assume it. Urinary 11-dehydro-thromboxane B2 is the standard functional readout, and roughly 5 to 10 percent of patients show high residual levels on standard dosing.
- For aspirin-exacerbated respiratory disease, target the 5-LOX branch with a leukotriene receptor antagonist rather than a COX inhibitor, because COX blockade shunts arachidonic acid toward leukotrienes.
- Consider dietary fatty acid balance in chronic inflammatory conditions. Lowering linoleic acid intake below 4 percent of energy and raising eicosapentaenoic acid intake to 1 to 2 grams per day measurably shifts the product profile within 8 to 12 weeks.
- Avoid combining COX-2 selective agents with high-dose omega-3 supplements without monitoring, because both reduce thromboxane and the combined bleeding effect is not well quantified.
Limits of current evidence
Species differences limit translation. Mouse macrophage 5-lipoxygenase expression and human differ substantially, so mouse knockout data on leukotrienes do not always predict human pharmacology. Zileuton's modest clinical effect in asthma trials despite potent enzyme inhibition suggests leukotriene output is not the sole determinant of airway inflammation.
Resolution biology is promising but immature. Lipoxin A4 and resolvin E1 have convincing rodent data at nanomolar doses, but human dosing studies remain small and short. No specialized pro-resolving mediator has achieved regulatory approval as of 2026.
Threshold reconciliation: the 30-fold COX-2 selectivity figure, the 10 nanomolar cysteinyl leukotriene bronchoconstriction threshold, and the 100 to 300 micromolar plasma total arachidonic acid value are measured under different assay conditions and should not be combined into a single dose-response curve. They describe separate steps in a pathway, not one unified potency scale.