IUPAC: 2-(2-chlorophenyl)-2-(methylamino)cyclohexan-1-one · C13H16ClNO · MW 237.73 g/mol · CAS 6740-88-1 · ChEMBL CHEMBL742
Ketamine. Racemic arylcyclohexylamine — a use-dependent open-channel blocker of the NMDA-type glutamate receptor. Simultaneously a battlefield anesthetic, a veterinary sedative, a rapid-acting antidepressant (the S-enantiomer, esketamine/Spravato), and a widely used recreational dissociative. Street/context names: K, Special K, Ket, Vitamin K, ket, the "K-hole".
Ketamine is not a receptor agonist or a transporter substrate. It is a physical plug. Its primary action is non-competitive, use-dependent open-channel block of the NMDA-type ionotropic glutamate receptor — a heterotetramer of two obligatory GluN1 subunits plus two GluN2 subunits (GluN2A/2B dominate forebrain). Ketamine binds inside the ion-conducting pore, in the channel vestibule near the asparagine (N-site) ring that also coordinates the Mg2+ block, and physically occludes cation flux. It does not touch the glutamate or glycine orthosteric sites; ChEMBL curates it as an NMDA-receptor negative allosteric modulator.
"Use-dependent" is the load-bearing phrase. The channel must be open for ketamine to reach its site — which requires glutamate + the co-agonist glycine/D-serine bound and the membrane depolarized enough to relieve the resting Mg2+ block. Ketamine therefore preferentially silences NMDA receptors that are most active, and it is partially trapped: the channel can close around the bound blocker, so unblock is slow. This is why the drug's effects outlast its free plasma concentration and why the pharmacology is fundamentally activity-gated.
Protonated ketamine enters the open channel and lodges near the GluN1/GluN2 asparagine ring deep in the vestibule, blocking Na+/Ca2+ influx. The cryo-EM structure (PDB 7EU7) places S-ketamine in this central pore cavity.
Block requires prior channel opening; the gate can shut on the trapped blocker. Slow, voltage-dependent unblock means high-frequency, strongly depolarized synapses are silenced preferentially — a built-in activity filter.
Ketamine shows modest selectivity for GluN2B-containing receptors, which are enriched extrasynaptically and on fast-spiking GABAergic interneurons — the substrate for the disinhibition hypothesis of its antidepressant action.
By preferentially blocking NMDA receptors on tonically active GABAergic interneurons, ketamine disinhibits pyramidal cells, triggering a glutamate surge onto AMPA receptors — the paradoxical "excitation via inhibition of inhibition".
The AMPA-receptor throughput activates BDNF/TrkB and mTORC1 signaling, driving rapid synaptogenesis in the mPFC within hours. This transient molecular cascade — not steady-state NMDA block — is the leading model for the sustained antidepressant effect.
(2R,6R)-hydroxynorketamine (HNK) reproduces antidepressant-like effects in rodents with weak NMDA block, implicating an AMPA-facilitating, largely NMDA-independent arm of the mechanism that is still debated.
Secondary pharmacology is real but higher-affinity only relative to the weak NMDA numbers: ketamine inhibits HCN1 (Ih) channels (contributing to hypnosis), is a weak μ/κ-opioid and muscarinic ligand, and interacts with monoamine transporters and nicotinic receptors at high concentrations. None of these approach the potency needed to explain the core dissociative or anesthetic effect, which is NMDA-driven.
Ketamine's clinical behavior is dictated by first-pass metabolism. Given IV it is ~100% available with onset in seconds; taken orally, hepatic CYP3A4/CYP2B6 demethylation slashes systemic exposure to roughly 16–20% — but that same first pass generates high levels of norketamine, an NMDA-active metabolite, so oral/insufflated ketamine feels qualitatively different from IV. Intranasal esketamine (Spravato) sits in between (~48% bioavailable) and is the FDA-approved antidepressant route.
Metabolism cascade: N-demethylation is the committed first step; the norketamine that results is itself active before being hydroxylated to the hydroxynorketamines (HNKs).
Norketamine (marked ★) retains roughly one-third of ketamine's NMDA-blocking potency and accumulates especially after oral dosing, extending analgesia and dissociation. (2R,6R)-hydroxynorketamine (HNK, ★) is the metabolite that fuelled the "NMDA-independent antidepressant" hypothesis — it is a weak channel blocker yet produces antidepressant-like effects in rodents (Zanos et al., 2016, Nature), though its role in humans remains contested. CYP2B6 and CYP3A polymorphisms and inhibitors (e.g., ritonavir, clarithromycin, grapefruit) meaningfully alter exposure and effect duration.
A single molecular action — activity-gated NMDA block — fans out into four clinically distinct effect domains depending on dose, route, and the circuits engaged. Sub-anesthetic doses dissociate and antidepress; anesthetic doses produce "dissociative anesthesia" with preserved airway reflexes and cardiovascular tone, which is exactly why ketamine is a field and pediatric anesthetic.
Disruption of NMDA-dependent corticothalamic and cortico-cortical integration uncouples sensory input from self-referential processing. Subjectively: analgesia, depersonalization/derealization, distorted body schema, and — at higher doses — the fully immersive, immobile "K-hole". This functional disconnection, not sedation, is the signature of a dissociative anesthetic.
A single sub-anesthetic infusion (0.5 mg/kg IV over 40 min) can lift treatment-resistant depression and acute suicidality within hours — a timescale no monoaminergic antidepressant matches. The working model is interneuron-NMDA block → cortical disinhibition → AMPA/BDNF/mTORC1-driven synaptogenesis in the mPFC and hippocampus. Intranasal esketamine (Spravato) received FDA approval in 2019 for treatment-resistant depression under a restricted REMS program.
Ketamine indirectly raises catecholamine tone (blocking reuptake and increasing central sympathetic outflow), producing tachycardia, hypertension, and bronchodilation while largely sparing respiratory drive and pharyngeal reflexes when used alone. This cardiovascular-stable, airway-protective profile is why it is favored in trauma, prehospital, and low-resource anesthesia — and why it becomes dangerous only when paired with genuine respiratory depressants.
The S-(+)-enantiomer (esketamine) is roughly 3–4× more potent at NMDA and carries the anesthetic/antidepressant development, while R-(−)-ketamine is being investigated for possibly longer-lasting, lower-dissociation antidepressant effects. Racemic "ketamine" is a 50:50 mix; the metabolite (2R,6R)-HNK derives from the R-side and further complicates any clean enantiomer story.
Emergence phenomena — vivid dreams, agitation, hallucinations on waking — track the same dissociative pharmacology and are attenuated by benzodiazepine co-administration in clinical settings (a controlled trade-off, not a recreational recommendation; see harm reduction below).
Evidence-based, non-moralistic. Ketamine has a wide anesthetic margin and is relatively safe alone — the serious harms come from chronic uropathy, injury during dissociation, escalating redosing, and combination with other depressants.
| Target | Affinity | Rel. | Mechanism |
|---|---|---|---|
|
NMDA-R
Glutamate [NMDA] receptor · human
|
Ki = 420 nM
PCP-site displ. · pChEMBL 6.38
|
Open-channel block | |
|
GluN1/GluN2B
NMDA-R GRIN1/GRIN2B · human
|
IC50 = 3.2 µM
|
Channel blocker | |
|
GluN1/GluN2A
NMDA-R GRIN1/GRIN2A · human
|
IC50 = 1.3 µM
|
Channel blocker | |
|
S / R enantiomers
NMDA PCP-site · rat cortex
|
S ≈ 0.3 / R ≈ 1.4 µM
|
S > R block | |
|
HCN1
Hyperpolarization-activated Ih channel
|
IC50 ~8–16 µM
|
Inhibitor | |
|
μ-opioid
Opioid receptor (secondary)
|
Ki ~28–42 µM
|
Weak agonist | |
|
mAChR
Muscarinic ACh receptor (secondary)
|
Ki ~45 µM
|
Antagonist |
The channel-block pose is an unusually clean entropy problem. Ketamine binds a pre-formed, water-filled pore that only opens under agonist + depolarization, so association is gated by the receptor's own conformational cycle, not by folding a flexible site around the ligand. With just two rotatable bonds and a rigid chlorophenyl–cyclohexanone scaffold, ketamine surrenders little internal conformational entropy on binding — the dominant −TΔS term is the desolvation of the pore vestibule and the displacement/ordering of channel waters near the N-site ring.
In FlexAID∆S terms, the Shannon-entropy collapse on binding is concentrated in the solvent and side-chain microstates lining the vestibule rather than in the ligand. "Trapping" then reads as a second, slower entropic gate: the closing gate lowers the configurational entropy of the bound state, raising the barrier to escape. A rigorous ΔG decomposition here should weight ΔSsolvent and pore-water release far more heavily than ligand ΔSconf — the opposite of a floppy-ligand orthosteric binder.