(−)-trans-Δ⁹-tetrahydrocannabinol · dronabinol · MW 314.47 g/mol · C₂₁H₃₀O₂ · CAS 1972-08-3 · ChEMBL465
Δ⁹-Tetrahydrocannabinol. The principal psychoactive constituent of Cannabis sativa · a lipophilic classical cannabinoid that mimics the endocannabinoids anandamide and 2-AG · a low-efficacy partial agonist at CB1. Context names: weed, hash, dabs, dronabinol (Marinol/Syndros, the isolated FDA-approved isomer).
THC is a classical cannabinoid that hijacks the body's own retrograde signaling system. Its primary pharmacology is partial agonism at the CB1 receptor — a Gi/o-coupled GPCR concentrated presynaptically on axon terminals throughout the CNS (highest density of any GPCR in the brain: basal ganglia, hippocampus, cerebellum, cortex). Unlike cocaine (DAT blockade) or DMT (5-HT2A agonism), THC does not flood a synapse with a monoamine — it silences neurotransmitter release by mimicking the endocannabinoids anandamide and 2-arachidonoylglycerol (2-AG).
THC's tricyclic ring and pentyl tail insert into the long, lipid-facing orthosteric pocket of CB1, reaching in laterally from the membrane. Engagement rotates the "twin-toggle" Phe2003.36/Trp3566.48 switch, opening the intracellular G-protein cavity.
Active CB1 couples to Gi/o: adenylyl cyclase is inhibited, cytosolic cAMP and PKA tone fall. This is the canonical brake — the opposite of the Gs stimulation many stimulants drive downstream.
Liberated Gβγ subunits inhibit N/P/Q-type voltage-gated Ca²⁺ channels and activate GIRK K⁺ channels. Presynaptic Ca²⁺ influx drops → vesicle fusion is suppressed → neurotransmitter release falls.
Endocannabinoids are made on demand postsynaptically and travel backward to presynaptic CB1 (DSI/DSE). THC bypasses this "on-demand" control, tonically dampening both GABAergic and glutamatergic terminals — hence its diffuse, state-dependent effects.
THC binds well (~40 nM) but activates submaximally: intrinsic efficacy is far below full agonists like CP-55940, WIN-55212 or synthetic "spice" cannabinoids. In high-CB1-reserve tissue it behaves near-full; in low-reserve tissue it can act as a partial antagonist.
THC is a partial agonist at CB2 (~36 nM, immune/microglial), with weaker actions at TRPV/TRPA channels, PPARγ, GPR55 and 5-HT3A. CBD (non-intoxicating) modulates this profile — a negative allosteric CB1 modulator that blunts THC anxiety.
The behavioral signature — the tetrad of hypolocomotion, catalepsy, hypothermia and analgesia in rodents — is entirely CB1-dependent and abolished in CB1 knockouts or by the antagonist rimonabant. Critically, CB1 is essentially absent from the brainstem respiratory nuclei. This single anatomical fact is why THC — unlike opioids at the µ-receptor — has no ceiling-breaking respiratory depression and no documented lethal overdose on its own.
THC is extremely lipophilic (logP ≈ 5.7), which dominates its PK. Inhalation delivers it to arterial blood within seconds — peak plasma during the act of smoking, effects in minutes — with bioavailability of roughly 10–35% (heavily technique-dependent). Oral dosing is the opposite animal: slow, erratic absorption plus heavy hepatic first-pass drop bioavailability to ~4–12%, and onset lags 30–120 minutes. That delay is why edible over-consumption ("I don't feel anything, I'll take more") is the classic route to a bad time.
Metabolism cascade: hepatic CYPs hydroxylate THC at C11 to an active metabolite, which is then oxidized to an inactive acid that anchors every drug test.
11-hydroxy-THC (11-OH-THC, marked ★) is the crux of the oral-vs-smoked difference. It is fully psychoactive — roughly equipotent with THC at CB1 and crosses the blood–brain barrier readily. After oral dosing, hepatic first-pass generates 11-OH-THC in near 1:1 ratio with parent THC, so edibles deliver a stronger, longer, more body-heavy and sometimes more hallucinatory experience than the plasma-THC numbers alone predict. Smoked THC largely skips first-pass, so the 11-OH-THC contribution is smaller.
11-nor-9-carboxy-THC (THC-COOH) is pharmacologically inactive but lingers for weeks in lipid stores and is the analyte immunoassays and GC-MS detect (as the glucuronide). CYP2C9 poor metabolizers (e.g. CYP2C9*3/*3) clear THC more slowly and show markedly higher exposure — relevant when THC is co-taken with CYP2C9 substrates such as warfarin, where competition can raise INR and bleeding risk.
Because CB1 sits presynaptically on nearly every neuron class, THC's effect is not a single-transmitter surge but a region-specific modulation of release probability. The subjective experience is the sum of disinhibition and suppression across distinct circuits.
THC does not bind dopamine machinery directly. Instead, CB1 on GABAergic interneurons in the VTA suppresses their firing, disinhibiting dopaminergic projections to the nucleus accumbens. The resulting DA rise is modest and slow compared with stimulants — which is why cannabis reward is real but its addiction liability is lower (cannabis use disorder ~9% lifetime, higher with adolescent/daily use).
Dense CB1 expression in hippocampus and neocortex dampens glutamatergic and GABAergic transmission, degrading working memory encoding and the sense of temporal flow ("time dilation"). This is the mechanistic basis of acute short-term memory impairment and the divergent, associative "stoned" cognition.
THC's effect on anxiety is dose-dependent and inverted-U: low doses are anxiolytic (CB1 dampening of amygdala threat circuits), high doses are anxiogenic and can trigger panic, paranoia and transient psychotomimetic states, especially in naïve users and with high-THC / low-CBD product. Set, setting, dose and CBD content all move the inflection point.
CB1 in the hypothalamus and mesolimbic feeding circuits drives the hyperphagic "munchies"; CB1 in the dorsal vagal complex underlies THC's acute antiemetic action (exploited by dronabinol/nabilone in chemotherapy). Autonomic effects — tachycardia, conjunctival vasodilation ("red eyes"), orthostatic hypotension, dry mouth — are CB1-mediated peripheral and central sympathetic/baroreflex shifts.
The through-line is efficacy, not just affinity. THC's low intrinsic activity gives it a soft ceiling: past a point, more receptor occupancy yields diminishing additional effect. This is exactly what a full agonist lacks — and the reason synthetic full-agonist cannabinoids (next section) are a categorically different, far more dangerous drug class despite sharing THC's target.
Evidence-based, non-moralistic. THC has no known lethal dose on its own — the real dangers are impairment, acute distress, specific chronic syndromes, and above all mistaking THC for the synthetic full agonists that do kill.
| Target | Affinity | Rel. | Mechanism |
|---|---|---|---|
|
CB1
Cannabinoid receptor 1 (CNR1)
|
Ki ≈ 40 nM
range 2.9–41 nM
|
Partial agonist | |
|
CB2
Cannabinoid receptor 2 (CNR2)
|
Ki ≈ 36 nM
range 3.3–72 nM
|
Partial agonist | |
|
GPR55
Putative cannabinoid receptor
|
agonist (µM)
low potency
|
Agonist | |
|
TRPV / TRPA
Thermo-TRP channels
|
EC₅₀ µM
weak modulator
|
Modulator | |
|
5-HT3A
Serotonin 3A ionotropic receptor
|
µM
allosteric
|
Neg. modulator |
THC's binding is a textbook hydrophobic-effect event: a near-desolvated, greasy ligand slides laterally out of the membrane into a lipid-facing groove. Burying that surface releases ordered water, a favorable solvent-entropy gain (positive ΔSsolv) that offsets the enthalpic modesty of a pocket lined with aromatics and few polar contacts — hence high affinity from a small, flexible molecule.
The pentyl tail and the receptor's Phe200/Trp356 "twin-toggle" pay the price on the other side of the ledger: agonist engagement collapses part of the receptor's conformational microstate distribution toward the active, G-protein-coupled ensemble — a negative ΔSconf the FlexAID∆S tENCoM term captures as vibrational-mode rigidification on binding.
Partial vs full agonism is an entropy story. A full agonist (AM11542, CP-55940, "spice") drives a deeper Shannon-entropy collapse — a tighter, more fully ordered active state and stronger transducer coupling. THC only partially narrows that ensemble: residual conformational entropy is retained, the receptor samples active and inactive-leaning states, and the readout is submaximal efficacy with a soft ceiling — the molecular reason THC is safe where its full-agonist cousins are not.