CB1 Receptor

Coupling, position and what activation actually does animal

CB1 was cloned in 1990 and classified with CB2 in the IUPHAR nomenclature in 2002. It couples through Gi/o, so activation inhibits adenylyl cyclase and reduces cyclic AMP, inhibits voltage-gated calcium channels and opens inwardly rectifying potassium channels. Put those together at a presynaptic terminal and the consequence is mechanical: less calcium entry on an action potential means less vesicle fusion, means less transmitter released. That is the entire elementary operation of the system. Whether the net effect on a circuit is inhibitory or excitatory depends on whether the terminal in question was releasing glutamate or GABA, which is why a single receptor produces the diversity of effects it does and why regional distribution is not a detail.

Sources: Matsuda LA 1990 · Howlett AC 2002 · Kano M 2009 · Wilson RI 2001

Distribution, and the brainstem animal

Herkenham and colleagues mapped CB1 by autoradiography in 1990 and the pattern has been confirmed many times since. Binding is extremely dense in the basal ganglia output structures, in the molecular layer of the cerebellum, in the hippocampus and in association cortex, and comparatively dense in the amygdala. It is conspicuously sparse in the brainstem, and specifically in the medullary nuclei that set respiratory rhythm. The motor, memory, appetite and anxiety effects of cannabinoids follow the dense regions, as does the cerebellar and basal-ganglial ataxia at high exposure. The sparse region is the one with the toxicological consequence.

Sources: Herkenham M 1990 · Howlett AC 2002

Why there is no cannabis respiratory-depression death, and why that is not a safety guarantee human data

Opioid receptors are densely expressed in the brainstem respiratory nuclei, which is the anatomical reason an opioid overdose stops breathing. CB1 is not, which is the anatomical reason that no comparable mechanism of lethal respiratory depression exists for cannabinoid receptor agonism. That is a real, structural, well-founded difference and it deserves to be stated without hedging. It is also routinely over-read. The absence of one lethal mechanism is not the absence of lethality: the synthetic cannabinoid receptor agonists sold as herbal-incense products are associated with deaths and severe illness by entirely different routes, and Trecki, Gerona and Schwartz catalogued them — seizures and status epilepticus, cardiac arrhythmia and myocardial injury, hyperthermia, acute kidney injury, agitated delirium with the trauma and hyperthermic collapse that follows it, and psychosis. Not one of those requires the brainstem. The correct sentence is that cannabinoid agonists do not kill by suppressing respiratory drive, and some cannabinoid agonists kill.

• Mechanism absent: brainstem-mediated respiratory depression.
• Mechanisms documented with synthetic full agonists: seizure, arrhythmia, hyperthermia, agitated delirium, acute kidney injury, psychosis.
• A missing lethal pathway is a fact about anatomy, not a general safety claim about a drug class.

Sources: Herkenham M 1990 · Trecki J 2015

Partial versus full agonism, with numbers contested in vitro

This is the pharmacological content behind the previous section. Δ9-THC binds CB1 with a Ki reported by Showalter and colleagues at approximately 40.7 nM and behaves as a PARTIAL agonist: even at receptor saturation it does not produce the maximal response the receptor is capable of. JWH-018, the 1-pentyl-3-(1-naphthoyl)indole from Huffman s aminoalkylindole series, binds with roughly four-fold higher affinity, reported at approximately 9 nM, and behaves as a FULL agonist. The endogenous ligands sit far weaker at the receptor by affinity: 2-AG is reported around 472 nM, and for scale the kavalactone yangonin was measured by Ligresti and colleagues at approximately 720 nM, in the same band as an endocannabinoid rather than anywhere near a synthetic. Those four numbers on one scale are the most useful thing on this page.

Contested — caveat. Binding affinities from different laboratories, radioligands, tissue sources and assay conditions are not strictly comparable and should be read as bands rather than as points. The 472 nM figure for 2-AG is the value carried in the comparison literature and in the operator datasheet; it was not traced to a single primary measurement during compilation, so it is presented as a scale reference rather than an asserted constant.

Sources: Showalter VM 1996 · Huffman JW 2005 · Ligresti A 2012 · Mechoulam R 1995 · Van Kush Family Research Institute 2026*

A second axis: the side chain can flip agonist to antagonist, not just move the number contested in vitro

The table above already puts one axis on display — how tightly a ligand binds, from the endocannabinoids at the weak end to JWH-018 at the tight end — but affinity by itself says nothing about what a ligand does once bound, and the classical cannabinoid scaffold supplies a clean demonstration that the second question does not track the first. Δ9-tetrahydrocannabivarin (THCV) is the propyl homolog of Δ9-THC: the same resorcinol-terpenoid scaffold, the same pyran closure, the side chain shortened from THC's five-carbon pentyl to three-carbon propyl. Thomas and colleagues (2005) reported that THCV displaces [3H]CP55940 from CB1 with a Ki of 75.4 nM in mouse brain membranes (and 62.8 nM at CB2 in CHO cells expressing the human receptor) — respectable affinity, in the same band as Δ9-THC itself — but behaves as a competitive antagonist rather than an agonist at either receptor: it blocked CP55940-, WIN55212-, anandamide- and methanandamide-induced [35S]GTPγS binding, and it antagonised Δ9-THC's own inhibition of electrically evoked contractions in the mouse vas deferens. Shorten the side chain by two carbons and the molecule keeps binding CB1 but stops activating it. Martin and colleagues (1999) established the same principle from a wider systematic series of THC side-chain variants, delineating agonists, partial agonists and antagonists across the set rather than one uniform cannabimimetic profile that merely varies in strength. Read against the partial-versus-full agonist table above, the honest description of this receptor's pharmacology has two axes, not one: how well a ligand binds, and what it does once bound — and the second is a property of the specific molecule, not something the scaffold or the chain length predicts by itself.

Contested — caveat. The 75.4 nM and 62.8 nM figures are the values reported by Thomas et al. 2005 for mouse brain CB1 and human CB2 respectively; functional antagonism (blockade of agonist-stimulated [35S]GTPγS binding and of THC-induced vas deferens inhibition) was demonstrated in the same paper and is not a separate inference from the binding number.

Sources: Thomas A 2005 · Martin BR 1999 · Showalter VM 1996

The absence of a ceiling is a toxicological property human data

A partial agonist has a maximum achievable effect that is set by the molecule, not by the dose. Once every CB1 receptor is occupied by Δ9-THC, the response stops rising, and any further exposure adds duration and distribution rather than intensity of receptor signalling. A full agonist has no such internal limit: the response scales with occupancy all the way up, so the dose determines the effect without an intrinsic brake. This is why the synthetic cannabinoid products were, and are, a categorically different hazard from cannabis rather than a stronger version of it — and why the difference cannot be managed by "using less" of an unmarked, inhomogeneously sprayed product whose active content varies by orders of magnitude between and within packages. Efficacy class is not a potency footnote. It is the variable that decides whether overshooting the intended dose produces an unpleasant few hours or a seizure.

Sources: Huffman JW 2005 · Trecki J 2015 · Showalter VM 1996

Desensitisation and downregulation with chronic agonism human data

CB1 is subject to agonist-driven desensitisation and downregulation, and this has been measured in people rather than only inferred from rodents. Hirvonen and colleagues used positron-emission tomography with a CB1 radioligand in chronic daily cannabis smokers and found reduced CB1 receptor availability in cortical regions, correlated with years of use, and reversible: availability returned toward control values over roughly four weeks of monitored abstinence. Two things follow. Chronic exogenous full or partial agonism moves the receptor substrate itself, so the pharmacology a heavy user is operating in is not the pharmacology a naive user is operating in. And the same adaptation is what appears in the chronic MAGL-blockade animal work, which is the bridge between this page and the argument against maximal enzyme inhibition on the magl page.

Sources: Hirvonen J 2012 · Schlosburg JE 2010