Estradiol — Reference

Glossary, abbreviations, enzyme list, PK parameter definitions, route table, key papers, and a chemical-structure gallery. The lookup companion to the synthesis and evidence pages.

This page is a reference index for the estradiol project. synthesis.html is the narrative; evidence.html is per-claim primary-source quotes with confidence tags; this page defines the terms, abbreviations, enzymes, PK parameters, and paper references the other two assume. If you hit “E1S,” “SULT1E1,” “AUC,” or “Kuhl 2005” in the synthesis and want a one-line definition, look here. C1–C5 confidence tags on claims throughout the project use the framework defined in §2.

1Confidence framework

Adapted from RESEARCH-BEST-PRACTICES.md §6. Every load-bearing factual claim in the synthesis and evidence pages carries one of these badges. The receiver should treat C1 as a quotable fact, C2 as a defensible estimate, C3–C5 as best-guess derivations or unknowns to be flagged when carried forward.

C1 Primary source directly supports the claim. Verbatim quote available, replicated or with independent confirmation. Trust as quotable fact.
C2 Credible secondary source or strong inference from primary literature. Single high-quality review, or two converging compilations, or strong indirect evidence. Treat as defensible working number.
C3 Inferred from related sources, modeling assumption, or compiled estimate. Arithmetic is correct given the assumed inputs, but the inputs themselves are uncertain. Show work; do not present as measured.
C4 Anecdotal, single-case, or preliminary abstract-only evidence. Conference abstract, n=1 case report, or pre-publication preprint. Use cautiously and never as the only support for a downstream claim.
C5 Genuinely unknown / no good data located. Often a question the synthesis would like to answer but cannot. Useful for marking research gaps; never propagated as if known.

Roughly: C1 is “the paper says it,” C2 is “the field consensus is X,” C3 is “if A and B are right then X follows,” C4 is “one group reported X,” C5 is “nobody knows.” The Ropponen et al. 2005 SHBG dose-response is C1; “EE is ~100× more hepatically potent per mole than E2” is C2 (the per-endpoint number ranges 75–1000×); the cytoplasmic E1/E2 ratio of 11 derived from Williamson 1967 plus a Nernst calculation is C3; the Bar 2024 protein-S finding is C4 (conference abstract); the human absolute sublingual E2 bioavailability is C5 (only marmoset measured).

2Estrogen abbreviations

Naming conventions for the molecules referenced throughout. “Estrogen” (or “oestrogen”) is the class; the human endogenous members are E1/E2/E3 (and E4 during pregnancy); circulating sulfates and glucuronides are inactive reservoir species. Anchor IDs are glossary-<abbrev> for cross-page linking.

E217β-Estradiol The principal endogenous human estrogen. C18 steroid; aromatic A-ring with phenolic C3-OH; saturated D-ring with secondary C17β-OH. The active species at ERα and ERβ. Both metabolic vulnerabilities (C3 phenol → SULTs/UGTs; C17 secondary alcohol → HSD17B2 oxidation to E1) sit on this scaffold. Premenopausal serum total E2 ~30–400 pg/mL across the cycle; postmenopausal <20 pg/mL; typical HRT target 50–200 pg/mL.

E1Estrone C17-ketone form of E2; ~5× weaker at ERα. Interconverts with E2 via HSD17B1 (reductive, NADPH) and HSD17B2 (oxidative, NAD⁺). Liver / gut / endometrium oxidize E2 → E1; ovary / breast / placenta reduce E1 → E2. Plasma E1/E2 ratio is a route signature: ~1 on transdermal/IM, ~5–7 on oral.

E3Estriol E2 plus a C16α-OH. The dominant pregnancy estrogen, made by the fetal-placental unit. Weak agonist at ER (~1–10% of E2 potency). Short hepatic half-life; used clinically mainly as vaginal cream for genitourinary syndrome.

E4Estetrol E2 plus C15α-OH and C16α-OH (two extra hydroxyls on the D ring). Produced by fetal liver during pregnancy. Modern oral estrogen (Estelle COC, approved 2021) with hepatic-sparing properties (~10–15× weaker hepatic ERα agonist than E2). Used in combination contraceptives as a lower-VTE alternative to EE.

E1SEstrone-3-sulfate (estrone sulfate) The C3 sulfate of estrone. Inactive at ER; doesn't cross cell membranes passively. The major circulating estrogen species in essentially every hormonal state — plasma E1S sits 3–15× higher than free E1+E2 in men, premenopausal women, postmenopausal women, HRT users, and pregnancy. Reservoir for slow STS-mediated reactivation in target tissues. MCR ~150 L/day (vs ~1400 L/d for E2). Plasma t½ ~10–12 h. Sold as a drug (Premarin component, esterified estrogens).

E2-3-GEstradiol-3-glucuronide C3 glucuronide of E2. Formed primarily by UGT1A10 in gut wall and UGT1A1/1A3 in liver. Hydrophilic; excreted in urine and bile. The major species in the “gut conjugate” bucket of oral E2 mass balance (~60–75% of total gut conjugates).

E2-17-GEstradiol-17-glucuronide C17 glucuronide of E2 — the only conjugate at the C17 position. Formed by UGT2B7 (the only UGT with appreciable activity on the C17 secondary alcohol of E2). Cholestatic in high doses (a classical model compound for drug-induced cholestasis). ~20–30% of gut conjugates.

EE17α-Ethinylestradiol (ethinylestradiol) E2 with a 17α-ethinyl substituent (C≡CH). The C17β-OH is retained, but the carbinol H is gone, so HSD17B2 cannot oxidize EE to an “E1-equivalent.” This is the structural reason EE has ~45% oral bioavailability vs E2's ~5%, and ~100× per-mole hepatic SHBG induction potency (75–1000× depending on endpoint). The classical combined-oral-contraceptive estrogen. Doses 15–50 μg/d. Does not meaningfully bind SHBG (~50× weaker than E2). Not used in modern HRT due to disproportionate hepatic effects and VTE risk.

E2V / EVEstradiol valerate C17 ester of E2 with pentanoic (valeric) acid. Oral E2V hydrolyzes to E2 + valeric acid in gut wall and plasma within ~30 min, so oral E2V is pharmacokinetically equivalent to oral E2 after a molecular-weight correction (E2V is ~76% E2 by mass; 1.5 mg E2V ≈ 1.15 mg E2). Also widely used as IM depot (5 mg q5d delivers Cmax ~200–400 pg/mL).

ECEstradiol cypionate C17 ester of E2 with 3-cyclopentylpropionic acid. IM depot; longer-acting than EV (t½ ~6–7 days). Standard transfeminine HRT regimen: 5 mg q14d → Cmax ~100–150 pg/mL averaged over cycle.

EUnEstradiol undecylate C17 ester of E2 with undecanoic acid (C11). Very long-acting SC/IM depot; t½ on the order of weeks (~22 days extrapolated from testosterone undecanoate data). Used in some transfeminine HRT regimens at 25–100 mg/month. Not currently FDA-approved; sourced from compounding pharmacies or international markets.

EEnEstradiol enanthate C17 ester of E2 with heptanoic acid. Intermediate-acting IM depot (t½ ~4–5 days). Component of Mesigyna combined injectable contraceptive (E2 enanthate + algestone acetophenide).

E2MATEEstradiol-3-sulfamate E2 with a sulfamate (−OSO₂NH₂) group at C3. Originally designed as an oral E2 prodrug intended to escape hepatic first-pass via carbonic-anhydrase-II binding in red blood cells. However, the C3 sulfamate turned out to act as an irreversible transition-state inhibitor of STS (steroid sulfatase), so the compound was repurposed as a clinical STS inhibitor for endometriosis (PGL2001).

EC508Estradiol-17β-(1-(4-(aminosulfonyl)benzoyl)-L-proline) Experimental oral E2 prodrug. Structure: E2 esterified at C17 with an L-proline whose nitrogen is acylated by a 4-sulfamoylbenzoyl group. The para-sulfamoylbenzoyl “head” binds carbonic anhydrase II inside red blood cells (Kd ~nM), sequestering the molecule past the liver; plasma esterases slowly hydrolyze the ester to release free E2. Animal data report ~100% oral bioavailability with minimal hepatic effects; no human trial data as of 2026.

CEEConjugated equine estrogens (Premarin) Mixture of estrogen sulfates extracted from pregnant mare urine. Principal components: estrone sulfate (~50%), equilin sulfate (~25%), 17α-dihydroequilin sulfate, and ~7 other minor steroids. Oral; relies on STS for reactivation in target tissues. Standard HRT doses 0.3–1.25 mg/d. SHBG-induction potency ~1.4–2.5× oral micronized E2 on a weight basis. Largely supplanted by oral micronized E2 in modern HRT.

2-OH-E2 / 4-OH-E2Catechol estrogens CYP1A1/1A2/1B1/3A4 hydroxylation products at C2 or C4 of the A ring. Reactive; 4-OH-E2 has DNA-damaging potential (the proposed mechanism linking estrogen exposure to breast carcinogenesis). Rapidly methylated by COMT to 2-MeO-E2 / 4-MeO-E2 or conjugated. Minor metabolic fate (~5% of oral E2 dose).

2-MeO-E22-Methoxyestradiol COMT methylation product of 2-OH-E2. Pharmacologically interesting (anti-angiogenic, anti-tubulin); investigated as an oncology agent (panzem). Not a classical estrogen receptor agonist.

DHT5α-Dihydrotestosterone Not an estrogen, but appears throughout the synthesis as the comparator that drove a load-bearing error: DHT has SHBG Kd ~1 nM, whereas E2's SHBG Kd is ~10–30 nM. Older AI synthesis drafts used DHT's value for E2, which inflated SHBG buffering effects ~20×. Corrected in this draft using Avvakumov 2010 (PMID 19748550).

TTestosterone Primary endogenous androgen. Aromatized by CYP19 to E2 (in adipose, gonads, brain, etc.); 5α-reduced by SRD5A1/2 to DHT. Δ⁴-3-keto A-ring. Relevant here because peripheral aromatization is a major source of E2 in postmenopausal women (the bulk of postmenopausal E2 originates from peripheral T → E2 conversion in adipose tissue).

3Steroid nomenclature primer

Just enough vocabulary to make the rest of the project intelligible. All steroids share the same four-ring skeleton; substituent identity, ring saturation, and stereochemistry distinguish the families.

A, B, C, D rings The four fused rings of the steroid skeleton, numbered (carbon numbering 1–17) and lettered (ring labels A–D) from the bottom-left clockwise. Rings A, B, C are six-membered cyclohexanes; ring D is a five-membered cyclopentane. Estrogens are distinguished from androgens by an aromatic A ring (with no C19 methyl); androgens have a saturated A ring and a C19 angular methyl. Both retain the C18 angular methyl on C13 between rings C and D.

C3 position Bottom-left carbon of ring A. Bears the C3 hydroxyl in estrogens (phenolic, attached to the aromatic ring) or the C3 ketone in androgens / Δ⁴-3-ones. Phenolic-OH is the site for SULT and UGT conjugation in estrogens; it is also the position where the prodrug strategies (E2MATE, EC508 indirectly via blocked sulfation) attempt their chemistry.

C17 position Top-right carbon of ring D. Bears the C17β-OH in E2 / testosterone / DHT or the C17-ketone in E1. The aliphatic secondary alcohol is the substrate for HSD17B1/2 oxidoreduction and the site for UGT2B7 glucuronidation and for ester prodrugs (valerate, cypionate, undecylate — all C17 esters). EE's distinguishing feature is a 17α-ethinyl substituent here.

α / β face convention Stereochemistry at any chiral steroid carbon. β (beta) means the substituent points up from the mean plane of the steroid ring system (toward the viewer in standard depictions); α (alpha) means it points down. So “17β-estradiol” has the C17 hydroxyl up, “17α-estradiol” (a weaker isomer) has it down; “17α-ethinylestradiol” has the C17 ethinyl group down and the C17-OH still up. The C18 angular methyl is β (up) by definition; the C19 angular methyl (in androgens) is also β.

Phenolic vs aliphatic OH A phenolic hydroxyl is attached to an aromatic ring (like the C3-OH of estrogens); pKa ~10. Phenolic-OH conjugates with sulfate (SULTs) and glucuronic acid (UGTs); it is not oxidized by classical alcohol dehydrogenases. An aliphatic hydroxyl is attached to a saturated carbon (like the C17-OH of E2); pKa ~16. Aliphatic-OH is oxidized to a ketone by short-chain dehydrogenases like HSD17B2 and can also be glucuronidated (by UGT2B7 specifically at the C17 of E2). The chemistry difference is the entire reason the metabolic fate of C3 and C17 diverges.

Aromatic A-ring The defining feature of an estrogen. Aromatization is catalyzed by CYP19A1 (aromatase), which converts the C19 methyl of an androgen to nothing (loss of CH₃ as formaldehyde) and dehydrogenates the A ring. The aromatic ring makes the C3 substituent a phenol (above) and locks the A-ring in a planar conformation that is essential for ER binding.

Δ⁴-3-keto A double bond between C4 and C5 of the A ring combined with a C3 ketone — the canonical androgen / progestogen / glucocorticoid A-ring substitution pattern (testosterone, progesterone, cortisol). Distinguishes androgens from DHT (which is fully saturated, no Δ⁴) and from estrogens (aromatic A ring, neither Δ⁴ nor 3-keto).

4Enzymes & receptors

Each enzyme or receptor is listed by its standard gene symbol with a one-line summary of its position, reaction, tissue distribution, and clinical relevance for estrogen pharmacology. Anchor IDs enzyme-<symbol>.

17β-Hydroxysteroid dehydrogenases (HSD17Bs)

HSD17B1Type 1 17β-HSD — reductive, NADPH-driven Catalyzes E1 → E2 (and androstenedione → testosterone). Highly expressed in placental syncytiotrophoblast, ovarian granulosa cells, and breast tissue. Drives the “target tissues reawaken estrogen” arm of the SULT/STS + HSD pump. Cytoplasmic NADPH/NADP⁺ ratio (~100) favors reduction.

HSD17B2Type 2 17β-HSD — oxidative, NAD⁺-driven Catalyzes E2 → E1 (and testosterone → androstenedione). Highly expressed in adult liver, gut, endometrium, placenta. Drives the “liver inactivates estrogen” arm. Cytoplasmic NAD⁺/NADH ~700 in fed hepatocyte (Williamson 1967) strongly favors oxidation. The blocked target for EE: 17α-ethinyl removes the C17 carbinol H, so HSD17B2 cannot oxidize EE.

HSD17B4Type 4 17β-HSD — peroxisomal, fatty-acid β-oxidation Multifunctional peroxisomal enzyme; oxidative for estrogens but the major in-vivo role is fatty-acid oxidation. Mutations cause D-bifunctional protein deficiency.

HSD17B12Type 12 17β-HSD — broadly expressed, weakly reductive Reductive on E1 → E2 in many tissues; also a fatty-acid elongation enzyme. May contribute to background tissue E2 levels.

Sulfotransferases (SULTs) and steroid sulfatase

SULT1E1Estrogen sulfotransferase 1E1 Lowest Km among human SULTs for estrogen substrates (~5–20 nM for E2, ~4 nM for EE). Conjugates the C3 phenol to give E1S / E2S. Expressed in liver, intestine, endometrium, placenta. Wins over SULT1A1 at physiological [E2] despite lower abundance because of the affinity difference (low Km). Shows substrate inhibition above ~100 nM (the “bell-shaped” SULT1E1 curve). Primary route of EE inactivation in gut wall.

SULT1A1Broad-spectrum aryl sulfotransferase Highly expressed in liver and gut. Km for E2 ~2.4 μM (~100× higher than SULT1E1's). Dominates at supra-physiological substrate concentrations; contributes minor fraction of estrogen sulfation at HRT/physiological [E2]. Potently inhibited by EE — one source of EE's drug-drug interaction footprint.

STSSteroid sulfatase Single isoform (X-linked gene). Hydrolyzes E1S → E1, DHEAS → DHEA, and other steroid sulfates. Expressed in breast, endometrium, brain, placenta, liver. The basis of the E1S “reservoir” concept: STS in target tissues reactivates the inert circulating sulfate to active free estrogen. STS inhibitors (E2MATE / PGL2001, irosustat / STX-64) are clinical investigational compounds for endometriosis and estrogen-dependent breast cancer. STS deficiency causes X-linked ichthyosis.

UDP-glucuronosyltransferases (UGTs)

UGT1A1Hepatic bilirubin / steroid glucuronosyltransferase Major hepatic UGT for the C3 phenol of E2; also responsible for bilirubin conjugation (Gilbert syndrome / Crigler-Najjar). Activity on E2 is ~10× lower than gut UGT1A10 on a per-protein basis. UGT1A1 polymorphism (UGT1A1*28) affects bilirubin handling and may modulate estrogen and EE metabolism.

UGT1A8Intestinal phenolic UGT Expressed predominantly in intestine (small bowel and colon). Conjugates C3 phenol of E2 with appreciable activity; one of the gut-specific isoforms responsible for the high gut-wall first-pass extraction.

UGT1A10Intestinal phenolic UGT — high-activity on E2 Predominantly intestinal expression (Strassburg 1998, 2000). ~10× higher E2-glucuronidation activity than hepatic UGT1A1 in recombinant enzyme assays (Basu 2004, J Biol Chem). The primary molecular basis of the ~40% gut-wall conjugation observed for oral estrogens. Acts on C3.

UGT2B7C17 glucuronosyltransferase The only UGT that conjugates the C17 secondary alcohol of E2 with appreciable activity (Lépine 2004, JCEM). Hepatic and extrahepatic. Forms E2-17-glucuronide (~20–30% of E2 gut-conjugate composition). Also conjugates morphine and many NSAIDs.

Cytochromes P450

CYP1A1Cytochrome P450 1A1 — extrahepatic catechol estrogen formation Catalyzes 2-hydroxylation of E2 in lung, breast, and other extrahepatic tissues. Inducible by aryl hydrocarbon receptor agonists (cigarette smoke, dietary indoles).

CYP1A2Cytochrome P450 1A2 — hepatic 2-hydroxylation Dominant hepatic 2-hydroxylator of E2. Induced by smoking (~50% higher CYP1A2 activity in smokers — one reason smoking reduces estrogen exposure).

CYP1B1Cytochrome P450 1B1 — 4-hydroxylation The dominant 4-hydroxylator of E2 (forms 4-OH-E2, the DNA-reactive catechol). Extrahepatic, including breast tissue. Implicated in estrogen-related carcinogenesis.

CYP3A4Cytochrome P450 3A4 — major hepatic/intestinal hydroxylase Performs C2, C4, and C16α hydroxylations of E2; primary CYP for EE oxidative metabolism (~22% of EE clearance, Lorbek 2018 PBPK). Inducers (carbamazepine, rifampicin, St John's wort) accelerate EE/E2 clearance, the basis of well-known contraceptive interactions.

CYP19A1Aromatase Catalyzes androgen → estrogen (testosterone → E2, androstenedione → E1) by removing the C19 methyl and aromatizing the A ring. Expressed in ovary, placenta, adipose tissue, bone, brain, breast. Source of most circulating E2 in postmenopausal women (peripheral conversion in adipose). Target of aromatase inhibitors (anastrozole, letrozole, exemestane).

COMTCatechol-O-methyltransferase Methylates the C2 (or C4) hydroxyl of catechol estrogens to give 2-MeO-E2 (or 4-MeO-E2). Detoxifies the reactive catechols. COMT polymorphism (Val158Met) affects catechol estrogen handling and has been investigated in breast-cancer-risk studies.

Receptors and binding proteins

ESR1Estrogen receptor α (ERα) The classical estrogen receptor; nuclear; ligand-activated transcription factor. Mediates most reproductive-tissue estrogen effects (uterus, breast, bone, liver). Kd for E2 ~0.1–0.5 nM in most assays; hepatic ERα Kd ~0.25 nM in rat hepatocyte (Vickers 1989). EE binds with affinity similar to E2; estrone binds ~5× weaker. The therapeutic target for SERMs (tamoxifen, raloxifene) and the ER-positive breast cancer pathway.

ESR2Estrogen receptor β (ERβ) Second estrogen receptor; expressed predominantly in ovary, prostate, colon, vasculature, CNS. Lower abundance in classical reproductive targets. Distinct transcriptional program from ERα; sometimes opposes ERα effects. Less load-bearing for HRT pharmacology.

SHBGSex hormone-binding globulin Plasma glycoprotein. Binds E2 (Kd ~10–30 nM), testosterone (Kd ~3–5 nM), DHT (Kd ~1 nM, highest affinity) with one high-affinity site per monomer. ~38% of plasma E2 is SHBG-bound; ~60% albumin-bound; ~2% free. Synthesized in liver under transcriptional control by ERα; SHBG induction by oral estrogens is the standard marker of hepatic estrogenic exposure. Baseline ~30–90 nmol/L; rises 2–5× on oral E2 and 5–10× on EE COCs. Does not bind EE meaningfully.

AlbuminPlasma carrier ~60% of plasma E2 reversibly bound (Kd ~17 μM, low affinity but high capacity). The “bioavailable” fraction is often defined as free + albumin-bound (since albumin binding is loose enough that dissociation occurs in capillary transit). EE binding is ~97–98% to albumin, near-zero to SHBG.

CAIICarbonic anhydrase II Cytosolic carbonic anhydrase abundant in red blood cells. Not an estrogen-pathway enzyme directly — but the molecular target of the sulfamate prodrug strategy. Aryl sulfamates bind CAII with sub-nM affinity in RBC cytoplasm, sequestering the drug past the liver's first-pass extraction. EC508 and E2MATE exploit this. The therapeutic concept: red blood cells as a portable drug reservoir.

SRD5A1 / SRD5A25α-Reductase (types 1 and 2) Not estrogen-pathway, but referenced as the comparator for DHT and as the analogue for how a single enzyme can amplify a hormone's tissue specificity. Reduces testosterone to DHT (the most potent endogenous androgen at AR). Type 2 is the prostate/scalp isoform targeted by finasteride/dutasteride.

5PK / PD parameters glossary

Standard pharmacokinetic and pharmacodynamic parameters as used throughout the project. Anchor IDs param-<abbrev>.

FBioavailability Fraction of an administered dose that reaches systemic circulation as parent drug. By convention F=1 for IV. For oral E2, the “~5% bioavailability” figure refers specifically to free E2 AUC — total estrogen (E2 + E1 + E1S + glucuronides) is much higher (~20–40%), and the hepatic estrogenic effect is driven by the high portal-vein concentration during absorption regardless. F is route-dependent: oral E2 ~5%, sublingual E2 ~10% (marmoset, Kuhnz 1993), transdermal ~10% (of nominal), IM ester depot ~100% (of released E2), oral EE ~45%.

CmaxMaximum plasma concentration The peak plasma concentration reached after a dose. Distinct from steady-state Cmax (which can be 1.5–2× higher after accumulation). For 1 mg oral E2 single dose: ~35 pg/mL (LC-MS/MS, Doll 2022). For 1 mg sublingual E2: ~144 pg/mL. For 30 μg oral EE single dose: ~70 pg/mL. The Goldzieher 1990 EE single-dose curves show inter-individual Cmax spread of ~5× in n=24 women on the same dose.

TmaxTime to Cmax Time post-dose at which Cmax occurs. Oral E2: ~8 h (slow because of E1S reservoir loading). Sublingual E2: ~1 h. Oral EE: ~1.5 h. Transdermal: steady-state, no defined Tmax. IM esters: days (E2V ~2 d, EC ~4–5 d, EUn ~6–14 d).

AUCArea under the plasma concentration-time curve Integrated systemic exposure. Specified by interval: AUC₀–ₐ (single dose, total exposure), AUC₀–ₘ (one dosing interval at steady state), AUC₀–ₐ₈ₐ (24 hour). Doll 2022 reports AUC₀–₈ for SL E2 = 1.8× oral E2 in the same subjects (1 mg single dose each route).

t½Half-life Time for plasma concentration to decline by 50%. Critical distinctions: Distribution t½ — early rapid phase, drug equilibrating between compartments; Elimination (terminal, β-phase) t½ — late phase governed by metabolic clearance; Apparent t½ — observed log-linear terminal slope, which for oral E2 reflects E1S reservoir release (13–20 h), not free-E2 metabolism (IV bolus free-E2 t½ is 1–2 h). Oral EE single-dose t½ 12–18 h; steady-state t½ extends to 17–24 h from deep-compartment loading.

VdApparent volume of distribution The hypothetical plasma volume needed to contain the body's drug burden at the observed plasma concentration. Vd = dose/C0. For E2: large (extensive tissue distribution into adipose and other lipophilic compartments). For EE: PBPK estimate ~4 L/kg (Vss); central compartment V₂/F ~24 L; deep peripheral compartment V₃/F ~1330 L (Klipping 2012 pop-PK).

MCRMetabolic clearance rate Volume of plasma cleared of drug per unit time. Independent of dose for linear PK. For E2 in women: MCR ~1400 L/day (or 1360 ± 40 L/d/m² BSA-normalized; Longcope 1968). For E1: ~1900 L/d/m². For E1S: only ~150 L/d (~80 L/d/m²; Ruder 1972) — the low MCR is why E1S accumulates as the dominant circulating estrogen species. For EE: ~400 L/d, ~3–4× slower than E2.

KmMichaelis constant Substrate concentration at which enzyme velocity is half-maximal (V₃⁄₂). Low Km = high affinity for substrate. SULT1E1 Km for E2 ~5–20 nM (lowest among human SULTs for estrogens). SULT1A1 Km for E2 ~2.4 μM (~100× higher, so SULT1E1 wins at physiological [E2]). HSD17B1, HSD17B2, UGT2B7 Km values typically in the μM range for E2.

KdDissociation constant Equilibrium dissociation constant of a binding interaction (receptor-ligand, protein-drug). Low Kd = high affinity. ERα Kd for E2: ~0.1–0.5 nM (Vickers 1989 reports 0.25 nM in rat hepatocyte). SHBG Kd for E2: ~10–30 nM (Avvakumov 2010). SHBG Kd for DHT: ~1 nM (the value erroneously used for E2 in earlier drafts). CAII Kd for aryl sulfamates: ~nM.

EC50Half-maximal effective concentration Concentration producing 50% of maximal observed response in a dose-response curve. Distinct from Kd: EC50 captures the full transduction chain (binding + signalling + transcription + protein turnover) while Kd is just binding. SHBG induction EC50 for E2 in HepG2: 0.5–2.5 μM (Kalme 1999) — orders of magnitude higher than the receptor Kd. The v4 model uses ~1500 pg/mL (5.5 nM) as a calibration target for in vivo human SHBG dose-response, not a measured EC50.

n / Hill coefficientSteepness of dose-response Slope parameter in the Hill equation E/Emax = Cⁿ / (EC50ⁿ + Cⁿ). n=1: standard Michaelis/Langmuir saturation. n>1: cooperative, switch-like response. n<1: shallow, graded. For hepatic ERα → SHBG in vivo: not measured; default modeling assumption n≈1; could plausibly be 0.7–1.5.

f_uFree (unbound) fraction Fraction of plasma drug not bound to plasma proteins; the species that crosses cell membranes and binds intracellular receptors. E2: ~2–3% free, ~38% SHBG-bound, ~60% albumin-bound (premenopausal women, follicular phase). EE: ~2% free, ~97–98% albumin-bound, essentially 0% SHBG-bound. Free fraction shifts with SHBG concentration: rising SHBG (from oral estrogen) crashes free E2 fraction but leaves EE free fraction unchanged — the explanation for EE's preserved activity during COC-induced SHBG rise.

IIV / CVInter-individual variability / coefficient of variation For most oral estrogen PK parameters, CV is 30–60% across populations. Goldzieher 1990 n=24 EE single-dose study showed ~5× spread in Cmax. Bioavailability variation drives most of this (CYP3A4 / SULT1E1 / UGT polymorphisms). The model curves shown in the synthesis are central estimates; real cohort data have ~2–4× the dispersion of the drawn lines.

OR / RR / HROdds ratio / risk ratio / hazard ratio Standard epidemiologic effect-size measures used in the VTE section. ESTHER (Canonico 2007): oral E2 adjusted OR 4.2 (95% CI 1.5–11.6) vs no HRT; transdermal E2 adjusted OR 0.9 (0.4–2.1). Pooled meta-analyses (Scarabin, Vinogradova) dilute the ESTHER signal to RR ~1.5–1.8.

6Routes & formulations

Standard estrogen formulations encountered in HRT and contraception. Numbers are central estimates synthesized from Kuhl 2005, Stanczyk 2013, and route-specific primary literature; full per-claim sourcing is on evidence.html.

Route / formulation	Typical dose	Cmax (E2)	F	Hepatic first-pass	E1/E2 ratio
Oral E2 (micronized)	1–4 mg/d	35–80 pg/mL	~5%	Heavy	5–7:1
Oral E2V (estradiol valerate)	1–4 mg/d (E2V)	~35–95 pg/mL	~5% (as E2)	Heavy	~5:1
Sublingual E2	0.5–2 mg q12h	100–450 pg/mL	~10% direct (+swallowed)	Partial (mixed SL+oral)	2–3:1
Buccal E2	0.5–2 mg	~100–200 pg/mL	~10–15%	Partial	2–3:1
Transdermal patch	25–100 μg/d	50–100 pg/mL	~5–10%	Minimal	~1:1
Transdermal gel	0.5–1.5 mg/d	50–100 pg/mL	~10%	Minimal	~1:1
SC E2 injection (aqueous)	2–6 mg q3–7d	100–200 pg/mL	~100%	Minimal	~1:1
IM E2 valerate (E2V)	5 mg q5–7d	200–700 pg/mL (peak)	~100% (depot)	Minimal	~1–2:1
IM E2 cypionate (EC)	5 mg q14d	100–340 pg/mL (peak)	~100% (depot)	Minimal	~1:1
IM E2 enanthate (EEn)	10 mg q4w (combined inj.)	100–200 pg/mL	~100% (depot)	Minimal	~1:1
SC E2 undecylate (EUn)	25–100 mg/mo	150–250 pg/mL	~100% (depot, slow)	Minimal	~1:1
Implantable E2 pellet	25–100 mg pellet, 3–6 mo	80–200 pg/mL	~100%	Minimal	~1:1
Vaginal E2 (cream / ring / tablet)	10–25 μg/d local	5–20 pg/mL systemic	low / variable	Minimal at low dose	~1:1
Oral CEE (Premarin)	0.3–1.25 mg/d	100–300 pg/mL E1 (eq.)	varied	Heavy	high (mostly E1)
Oral EE (in COCs only)	15–50 μg/d	40–200 pg/mL EE	~45%	Heavy + recirculates	n/a (no E1 made)
Transdermal EE patch (COC)	20–30 μg/d	30–60 pg/mL EE	N/A (systemic)	still heavy (intrinsic)	n/a

7Key papers & sources

One-line summary of each primary or anchor reference cited across the project, grouped by topic. Anchor IDs paper-<firstauthor>-<year>. URLs are PubMed or PMC where available; absent URLs reflect papers cited via secondary sources rather than directly accessed. Confidence tags refer to the load-bearing claim each paper supports.

Master reviews

Kuhl H. 2005. Pharmacology of estrogens and progestogens: influence of different routes of administration. Climacteric 8 Suppl 1:3–63. PMID 16112947. The single most-cited reference for estrogen pharmacology. Aggregates ~60 PK studies across oral / sublingual / vaginal / transdermal / IM ester / implant / intranasal routes; provides comparative hepatic-potency table (SHBG, CBG, TBG, angiotensinogen) with oral E2 = 1 and oral EE ≈ 100–500 weight-basis. The anchor for “oral E2 F≈5%, oral EE F≈45%, oral CEE SHBG ≈ 1.4–2.5× oral E2” and most route-comparison numbers in the synthesis. C1

Stanczyk FZ, Archer DF, Bhavnani BR. 2013. Ethinyl estradiol and 17β-estradiol in combined oral contraceptives: pharmacokinetics, pharmacodynamics and risk assessment. Contraception 87(6):706–727. PMID 23375353. Comprehensive comparative review of EE vs E2 in COCs. The anchor for EE bioavailability range (38–48%) and the canonical comparative SHBG and VTE numbers across COC formulations. C1

Stanczyk FZ et al. 2024. Update on PK and clinical considerations of oral estradiol. Contraception (update to 2013 review). Modern follow-up incorporating LC-MS/MS-era data. Used as cross-check anchor for oral E2 steady-state numbers. C2

Oral / sublingual E2 PK

Doll TG et al. 2022. Pharmacokinetics of sublingual versus oral estradiol in transgender women. Endocr Pract 28(3):237–243. PMID 34781041. Modern LC-MS/MS crossover (n=10, single 2 mg dose each route, 1-week washout): SL Cmax 144 pg/mL at Tmax ~1 h vs PO 35 pg/mL at ~8 h; AUC₀–₈ ratio SL/PO = 1.8×. The load-bearing modern anchor for sublingual E2 PK. C1

Cirrincione LR et al. 2021. Sublingual estradiol is associated with higher estrone concentrations than transdermal or injectable preparations in transgender women and gender nonbinary adults. LGBT Health 8(2):125–132. PMID 33651641. Cross-sectional steady-state observation: SL E2 produces oral-like elevated E1, transdermal preserves physiologic E1/E2 ratio (~1). Direct evidence that sublingual is NOT liver-sparing. C1

Price TM, Blauer KL, Hansen M, Stanczyk F, Lobo R, Bates GW. 1997. Single-dose pharmacokinetics of sublingual versus oral administration of micronized 17β-estradiol. Obstet Gynecol 89(3):340–345. PMID 9052581. Original sublingual vs oral E2 PK comparison; n=6 RIA. Established biphasic SL shape and ~2.5× AUC ratio. C1

Kuhnz W, Gansau C, Mahler M. 1993. Pharmacokinetics of estradiol, free and total estrone, in young women following single intravenous and oral administration of 17β-estradiol. Arzneimittelforschung 43(9):966–973. Clean IV vs oral comparison giving absolute oral E2 bioavailability ~5%; companion marmoset study gives the only direct sublingual F estimate (~10%). The primary anchor for “oral E2 F ~5%, SL ~10%.” C1 for oral; C2 for SL (marmoset).

Bar et al. 2024. Low-dose sublingual estradiol effect on free protein S in treatment-naïve transgender women. ECE 2024 abstract EP592. endocrine-abstracts.org. Conference abstract reporting clinically significant free protein S drop on low-dose SL E2, consistent with meaningful hepatic exposure. Used to motivate the peak-vs-AUC hepatic-response distinction. C4 (abstract only).

EE PK

Goldzieher JW. 1990. Selected aspects of the pharmacokinetics and metabolism of ethinyl estrogens and their clinical implications. Am J Obstet Gynecol 163(1 Pt 2):318–322. PMID 2196804. Comprehensive EE PK review; Figure 1 reproduces Brody / Turkes / Goldzieher 1989 single-dose curves (n=24, 70 μg EE) used as the synthesis page's EE PK anchor. Documents ~5× Cmax spread across individuals. C1

Back DJ et al. 1982. The gut wall metabolism of ethinyloestradiol. Br J Clin Pharmacol. PMID 7059434; PMC1402099. Portal-vein cannulation in humans dosed with EE: gut-wall conjugation fraction 0.44, hepatic 0.25. The single load-bearing primary measurement for “gut-wall first-pass is large” throughout the project. C1

Back DJ et al. 1981. The pharmacokinetics of ethinyloestradiol in women: effect of intestinal flora suppression. Contraception 24(4):447–469. Earlier paper in the Back / Liverpool series; establishes that antibiotic gut-flora suppression does not significantly affect EE exposure (so enterohepatic recirculation is real but quantitatively modest in humans).

Klipping C et al. 2012. Characterisation of the pharmacokinetics of ethinylestradiol and drospirenone in extended-cycle regimens: population pharmacokinetic analysis. PMC3632974. Pop-PK fit on EE three-compartment model. The source for EE central V₂/F = 24 L, deep V₃/F = 1330 L, CL/F = 25.3 L/h, ka 0.295 h⁻¹. C1

Devineni D et al. 2007. PK overview of EE dose and bioavailability using two transdermal contraceptive systems and a standard combined oral contraceptive. PMC4285808. Comparator study for the 20 μg oral EE COC (Loestrin). Anchor for Cmax 60 pg/mL, Cavg 25 pg/mL, AUC₀–ₘ ~600 pg·h/mL. C1

Zhang H, Wu X, Wang H, Mikus G, Wang W. 2018. Risk-benefit assessment of ethinylestradiol using a physiologically-based pharmacokinetic modeling approach. CPT Pharmacometrics Syst Pharmacol 7(11):756–767. PMC6282492. Simcyp PBPK for EE. Source for the fractional-metabolism split (CYP3A4 22% + CYP2C9/8/1A2 18% + UGT 5% + SULT 37% + other 18%) and Vss = 4 L/kg. Cited throughout as “Lorbek 2018 PBPK.” C2

Rodrigues AD et al. 2022. Drug interactions involving 17α-ethinylestradiol: considerations beyond CYP3A induction and inhibition. Clin Pharmacol Ther 112(1):69–90. Reviews non-CYP3A DDIs of EE, including SULT1A1 inhibition. Anchor for the “EE inhibits SULT1A1” pharmacology.

Hepatic ERα / SHBG

Vickers AE, Nelson K, McCoy Z, Lucier GW. 1989. Changes in estrogen receptor, DNA ploidy, and estrogen metabolism in rat hepatocytes during a two-stage model for hepatocarcinogenesis using 17α-ethinylestradiol as the promoting agent. Cancer Res 49(23):6512–20. Reports [³H]-E2 binding to rat hepatocyte ER (nuclear/cytosolic exchange assay): Kd = 0.25 nM. The number is real and load-bearing for “hepatic ERα is sub-nM affinity for E2,” but should not be transferred uncritically to in-vivo human SHBG induction (the EC50 for that endpoint is much higher; see Kalme et al.). C1 for receptor Kd in rat hepatocyte.

Kalme T et al. 1999. HepG2 SHBG dose-response to E2. In-vitro HepG2 dose-response for SHBG mRNA induction by E2: responses at 0.5–2.5 μM, far higher than ERα receptor Kd. The basis for distinguishing “receptor occupancy” (sub-nM) from “SHBG induction” (μM) in the round-2 corrections. Replaces the erroneously-cited Selva & Hammond 2009 (which is on HNF-4α / thyroid, not E2). C1

Ropponen A, Aittomäki K, Vihma V, Tikkanen MJ, Ylikorkala O. 2005. Effects of oral and transdermal estradiol administration on levels of sex hormone-binding globulin in postmenopausal women with and without a history of intrahepatic cholestasis of pregnancy. JCEM 90(6):3431–3434. doi:10.1210/jc.2005-0352. Oral E2 valerate 2–4 mg/d raised SHBG +67–171% (controls) / +42–121% (ICP history); transdermal E2 produced no significant change; +MPA lowered SHBG 14–18%. The abstract reports only the aggregate range (no per-dose breakdown). The clinical anchor used to calibrate the v4 model's hepatic-ER → SHBG component; the model's predicted 1 mg→+60% / 2 mg→+120% / 4 mg→~+180% curve lands inside this measured band. (Earlier drafts misattributed first author as “Lindberg” and year as 2003; both corrected — verified via Crossref.) C1

Vihma V, Ropponen A, Aittomäki K, Ylikorkala O, Tikkanen MJ. 2004. Postmenopausal estrogen therapy and serum estradiol fatty acid esters in women with and without previous intrahepatic cholestasis of pregnancy. Ann Med 36(5):393–399. doi:10.1080/07853890410033847, PMID 15478314. Companion paper on the same 40-woman cohort as Ropponen et al. 2005. Documents the dosing protocol used in both reports: oral E2 valerate 2–4 mg/d, or transdermal E2 50–100 μg/d, for 6 weeks in a double-blind crossover (4-week wash-out). This is the source that fixes the dose ladder — there was no 1 mg oral arm, so the “1 mg→+60%” point cited elsewhere is the v4 model's prediction, not measured data. C1

Avvakumov GV, Cherkasov A, Muller YA, Hammond GL. 2010. Structural analyses of sex hormone-binding globulin reveal novel ligands and function. Mol Cell Endocrinol 316(1):13–23. PMID 19748550. Definitive structural / affinity reference for SHBG. SHBG Kd for E2: ~10–30 nM. The reference that corrected the synthesis's earlier erroneous use of DHT's ~1 nM as E2's Kd. (Earlier drafts of this page mislabeled the open-access PMC8144348 — which is in fact Jasuja et al. 2021, iScience, an allosteric-coupling study — as this paper; corrected to PMID 19748550, the paywalled original.) C1

Namkung PC, Stanczyk FZ, Cook MJ, Novy MJ, Petra PH. 1989. Half-life of plasma sex steroid-binding protein (SBP) in the primate. J Steroid Biochem 32(5):675–680. PMID 2500563. Tracer clearance of labeled SBP/SHBG in the primate. Biphasic: fast component t½ ~7.5 h (>90% of label cleared within 24 h), slow terminal component t½ ~3.95 days. Underpins the synthesis §3 “two timescales” argument — plasma SHBG turnover is the multi-day, rate-limiting step between a hepatic E2 pulse and the measured protein change. C2

Hammond GL. 2016. Plasma steroid-binding proteins: primary gatekeepers of steroid hormone action. J Endocrinol 230(1):R13–R25. PMC5064763. Review establishing the free-hormone and bioavailable-hormone frameworks. Albumin's affinity for steroids is 3–4 orders of magnitude lower than SHBG's, but albumin's plasma concentration is ~1000× higher, so the two carriers hold comparable shares of circulating E2. Source for the synthesis §3 protein-binding subsection (free-fraction equation, free- vs bioavailable-hormone distinction). C1

Stegeman BH et al. 2013. Effect of ethinylestradiol dose and progestagen in combined oral contraceptives on plasma SHBG levels. J Thromb Haemost 11(1):203–205. Cross-formulation SHBG rise table by EE dose and progestin (LNG, dienogest, drospirenone, cyproterone, desogestrel). Source for the “SHBG rise correlates with VTE risk across COC formulations” ordering. C1

VTE epidemiology

Canonico M et al. 2007 (ESTHER). Hormone therapy and venous thromboembolism among postmenopausal women: impact of the route of estrogen administration and progestogens. Circulation 115:840–845. PMID 17309934. The ESTHER case-control study. Adjusted OR for VTE: oral estrogen 4.2 (95% CI 1.5–11.6), transdermal 0.9 (0.4–2.1). The load-bearing single-study basis for the “oral vs transdermal route difference is large” argument. C1

Scarabin PY et al. 2003. Differential association of oral and transdermal oestrogen-replacement therapy with venous thromboembolism risk. Lancet 362:428–432. Companion earlier case-control to ESTHER; replicates the route signal in a different French cohort.

Vinogradova Y, Coupland C, Hippisley-Cox J. 2019. Use of hormone replacement therapy and risk of venous thromboembolism: nested case-control studies using QResearch and CPRD databases. BMJ 364:k4810. Large UK primary-care database study. Pooled HR ~1.7 for oral HRT vs none; transdermal HR ~1.0. The modern large-N replication of the ESTHER direction (smaller magnitude due to pooled effects).

Endogenous PK gold standards

Longcope C, Layne DS, Tait JF. 1968. Metabolic clearance rates and interconversions of estrone and 17β-estradiol in normal males and females. J Clin Invest 47(1):93–106. PMC297151. Tracer kinetic gold standard for E2/E1 MCR and interconversion. MCR(E2) 1360 ± 40 L/d/m² in women; transfer constants E2 → E1 = 0.15, E1 → E2 = 0.05. Every subsequent E2 PK model is calibrated against these. C1

Ruder HJ, Loriaux DL, Lipsett MB. 1972. Estrone sulfate: production rate and metabolism in man. J Clin Invest 51(4):1020–1033. PMC302214. The reference paper for E1S kinetics. MCR(E1S) 157 L/d; transfer constants ρ(E1→E1S) 0.54, ρ(E2→E1S) 0.65, ρ(E1S→E1) 0.21, ρ(E1S→E2) 0.014. Establishes E1S as a low-MCR slow-release reservoir. C1

Longcope C. 1986. [Postmenopausal estrogen production review]. Anchor for postmenopausal endogenous E1 production rate ~40 μg/day (the figure used to correct an earlier 80 μg/d error that conflated premenopausal early-luteal with postmenopausal data).

Düsterberg B, Nieuweboer B. 1985. Pharmacokinetics and biotransformation of estradiol valerate in ovariectomized women. Horm Res 21:145–154. PMID 2987096. IV / IM / oral E2V PK in OVX women. Source for IV E2 bi-exponential half-life (α ~6 min, β ~1–2 h) and IM 5 mg E2V Cmax 667 pg/mL at day 2.

Stricker R et al. 2006. Establishment of detailed reference values for LH, FSH, E2, progesterone during menstrual cycle on the Abbott ARCHITECT analyzer. Clin Chem Lab Med 44(7):883–887. Reference-range anchor for normal menstrual cycle E2: follicular 30–80 pg/mL, mid-cycle peak 200–400, luteal 80–250, postmenopausal <20.

Enzymology

Williamson DH, Lund P, Krebs HA. 1967. The redox state of free nicotinamide–adenine dinucleotide in the cytoplasm and mitochondria of rat liver. Biochem J 103:514. PMID 4291787. Canonical NAD⁺/NADH cytosolic ratio in rat hepatocyte: 725 (fed), 528 (starved), 208 (alloxan-diabetic). Used to argue HSD17B2 equilibrium drives E2 → E1 in liver. Rat, inferred from lactate/pyruvate ratio. C1 for the fed-state value.

Basu NK et al. 2004. Human UDP-glucuronosyltransferases show isozyme-specific differences in their kinetics of glucuronidation toward 17β-estradiol. J Biol Chem 279:28320. PMID 15117964. The load-bearing kinetic reference for “UGT1A10 ~10× more active on E2 than hepatic UGT1A1.” (Earlier drafts misattributed to Strassburg 1998, which only characterized expression patterns.) C1

Strassburg CP et al. 1998, 2000. Expression of human UGT1A locus in intestine. Characterized UGT1A8 / UGT1A10 expression patterns in human intestinal microsomes; established that UGT1A10 is highly expressed in gut wall. Foundational expression-pattern work; not the kinetic Vmax/Km source (that's Basu 2004).

Lépine J et al. 2004. Specificity and regioselectivity of the conjugation of estradiol, estrone, and their catecholestrogen and methoxyestrogen metabolites by human UDP-glucuronosyltransferases. JCEM. Comprehensive UGT activity table across all 19 human UGTs against E2, E1, and catecholestrogens. Establishes the standard regiospecificity (UGT1A1/3/8/10 at C3; UGT2B7 at C17). C1

Schrag ML et al. 2004. Sulfotransferase 1E1 is a low-Km isoform mediating the 3-O-sulfation of ethinyl estradiol. Drug Metab Dispos. Establishes SULT1E1 Km ~4 nM for EE; the basis for the “SULT1E1 dominates EE gut-wall conjugation” argument.

Reed MJ, Purohit A, Woo LWL, Newman SP, Potter BVL. 2005. Steroid sulfatase: molecular biology, regulation, and inhibition. Endocr Rev 26:171–202. Definitive STS review. Source for STS distribution, the SULT1E1 / STS pump concept, and the medicinal-chemistry rationale for sulfamate STS inhibitors (irosustat, PGL2001).

Falany CN. 1997. Enzymology of human cytosolic sulfotransferases. FASEB J 11:206–216. Classical reference for SULT1E1 kinetic parameters in human tissues; source for the Km 5–20 nM range.

Modeling and tools

Plowchalk DR, Teeguarden J. 2002. Development of a physiologically based pharmacokinetic model for estradiol in rats and humans. Toxicol Sci 69(1):60–78. PMID 12215661. Foundational 7-compartment estrogen PBPK; the structural template for all subsequent estrogen PBPK papers. Parameters: SHBG-E2 Kd 1.5 nM (probably too tight by ~10×), albumin Kd 17 μM.

Hartman JH, Knott K, Miller GP (Karelina et al.) 2017. Linking physiologically-based pharmacokinetic and genome-scale metabolic networks to understand estradiol biology. BMC Syst Biol 11 Suppl 7:141. PMC5732473. 18-compartment PBPK with genome-scale liver metabolic network. Closest published precedent for explicit enzyme-level estrogen PBPK.

Aly. 2022. Injectable E2 meta-analysis. transfemscience.org. Three-compartment empirical fits for IM EB / EV / EC / EEn injectable estradiol curves. The de-facto reference for transfeminine IM dosing. C2

Aly. Sublingual estradiol as an alternative to oral estradiol in transfeminine people. transfemscience.org. Comprehensive narrative review of sublingual E2 PK in transfeminine context, aggregating Price 1997, Pines 1999, Devissaguet 1999, Kuhnz 1993, Doll 2022, Cirrincione 2021. The most-cited secondary source for SL E2.

Aly / Violet / Luna. estrannaise.js (MIT, 2025). github.com/WHSAH/estrannaise.js, estrannai.se. Open-source Bayesian three-compartment PK calculator for transfeminine HRT. MCMC-derived parameter posteriors for EV / EEn / EC / EB / EUn IM, EUn SC, and transdermal patch.

Wikipedia / aggregated indices

Wikipedia. Pharmacokinetics of estradiol. en.wikipedia.org/wiki/Pharmacokinetics_of_estradiol. Heavily-referenced parameter table with 100+ primary citations. Used throughout as an index into primary literature (not the primary source itself).

8Chemical structure gallery

Compact line-formula structures for the molecules referenced throughout the project. SVGs are the same renderings used in the synthesis tooltip system. SMILES strings below each are the canonical ones from PubChem where available; complex prodrugs (EC508) are written in a partly-descriptive form for readability.