Chemistry and Structure-Activity Relationships of Psychedelics

This work is a narrative, literature-based synthesis rather than an original experimental study. Nichols integrates data from diverse sources: historical human psychopharmacology reports, animal behavioural experiments (including drug discrimination, head-twitch response, hyperthermia and others), smooth muscle assays (rat fundus, uterus, sheep umbilical artery), receptor binding studies (for example displacement of radioligands such as [125I]DOI or [3H]ketanserin), cell-based functional assays (phosphoinositide turnover, Ca2+ mobilisation, IP3 accumulation), in silico docking into homology models of the 5-HT2A receptor, and site-directed mutagenesis where available. The chapter compares affinities (Ki values), functional potencies (EC50), efficacies (Emax) and in vivo potencies (ED50, behavioural substitution) when reported, but Nichols notes that many early or recreational compounds lack comprehensive molecular pharmacology. He draws on examples where radioligands and PET tracers have been developed (e.g. [125I]DOI, [11C]Cimbi-36) and summarises studies that used metabolic or pharmacokinetic data to explain toxicity (for example for NBOMe analogues). Where structural hypotheses were tested, the author reports results from synthesised analogues, chiral separations and conformationally constrained molecules to infer active binding poses.

Tryptamines: Nichols reports that tryptamines most closely resemble serotonin and that oxygen substitution at the 4- or 5-position generally enhances 5-HT2A agonist and psychedelic activity. Example values include N,N-dimethyltryptamine (DMT) with a reported Ki of about 75 nM at rat cortical 5-HT2A sites, 5-MeO-DMT increasing affinity to ~14 nM and psilocin (4‑OH) to ~6 nM. Ring fluorination studies produced mixed effects: 6- or 7-fluoropsilocin gave similar affinities to psilocin but reduced potency, whereas fluorination of 5‑MeO‑DMT increased EC50 values substantially (e.g. to 7.9 and 18.1 µM for some congeners) without large changes at 5-HT2C. Replacing the indole NH with heteroatoms (thiophene, benzofuran) usually reduced 5-HT2A affinity and functional substitution in discrimination assays, though some indazole derivatives (for example AL-38022A) were highly potent full agonists with EC50s in the sub‑ to low‑nanomolar range and ED50s in the 0.04–0.05 mg/kg range in animal discrimination tasks. Side‑chain and N‑alkylation effects: N‑alkylation patterns influence affinity and oral activity. Nichols summarises that certain N,N‑disubstitutions (dimethyl, diethyl, N‑methyl‑N‑isopropyl, N,N‑diisopropyl) showed relatively high affinities (low tens of nM) in displacement assays, and that larger N‑alkyl groups or alpha‑methylation (e.g. a‑methyltryptamines like AMT) can confer oral activity by blocking MAO deamination. Chiral effects are notable: S‑(+)-enantiomers of some a‑methyltryptamines are more active at 5-HT2A and as human hallucinogens than their R counterparts. Certain constrained side‑chain analogues (trans‑cyclopropylindoles) showed stereoselective affinities at 5-HT2 subtypes. Ergolines: LSD is unusually potent in humans yet shows pharmacology at 5-HT2A that is not dramatically different from simpler ligands. Nichols emphasises that small modifications to the diethylamide moiety or stereochemistry greatly reduce potency; many amide changes cause ~10‑fold losses. Epimerisation at C8 or reduction of key double bonds abolishes psychedelic activity. Some 2‑halogenated lysergic derivatives (for example 2‑bromo‑LSD/BOL‑148) are 5-HT2A antagonists and can block LSD effects in humans; BOL compounds have also shown clinical activity for cluster headache. Substituting N(6) (the N(6)‑alkyl series) produced several congeners (AL‑LAD, ETH‑LAD) that were psychoactive in humans and had potencies paralleling animal discrimination data. Virtual docking and molecular dynamics point to the diethylamide nestling into an extracellular pocket bounded by extracellular loop 2, notably Leu‑229, suggesting a sterically constrained receptor region that explains sensitivity to amide substitutions. Phenethylamines: Mescaline is a low‑potency prototype; the phenethylamine SAR is the most extensively explored due to synthetic tractability. The 2,4,5‑substitution pattern (or 2,5‑dimethoxy with a hydrophobic 4‑substituent) generally confers highest potency. Hydrophobic 4‑substituents (methyl, halogens, alkylthio, CF3) markedly increase potency and seem to enhance CNS penetration and interaction with a complementary hydrophobic pocket within the 5-HT2A orthosteric site. Structural constraints show the side chain favours an extended trans conformation; conformationally constrained analogues (for example benzocyclobutene R‑71, also known as TCB‑2) achieved very high affinity (Ki ~0.26 nM) and high functional potency (EC50 18 nM, Emax 97%). Alpha‑methylation (amphetamines) typically increases duration and potency in the R‑(-) enantiomer, which is also the more active form at 5-HT2A. Heterocyclic and tethered analogues: Tethering methoxy oxygens into dihydrofuran or dihydropyran rings could dramatically enhance affinity and potency for 2,4,5‑derived phenethylamines; for example, a difurano compound had Ki ~0.48 nM at human 5-HT2A. By contrast, similar tethering strategies applied to 3,4,5‑mescaline analogues lowered efficacy and in vivo potency, indicating different binding poses for different substitution patterns. N‑benzyl (NBOMe/NBxx) series and imaging ligands: Introduction of an N‑benzyl moiety to phenethylamines produced exceptionally high 5-HT2A affinity and potency (25I‑NBOMe and analogues), with some derivatives deployed as radioligands and PET tracers (e.g. 25B‑NBOMe/[11C]Cimbi‑36). These NBOMe compounds are not orally active but produce strong effects via buccal absorption; several fatalities have been associated with their recreational use. Metabolic studies cited show rapid O‑demethylation followed by fast glucuronidation, and the authors discuss the hypothesis that glucuronide metabolites — or interindividual differences in demethylating P450 activity — may contribute to observed toxicity. Nichols also reports that addition of N‑benzyl groups enhances 5-HT2A activity in tryptamines and that constrained N‑benzyl phenethylamines can be highly selective 5-HT2A agonists in vitro and active in behavioural models. Mechanistic and modelling findings: Homology models, docking and mutagenesis have identified specific receptor residues (for example Ser‑239/Ser‑242 region, Phe‑340) implicated in ligand interactions, but the chapter stresses that without a crystallographic structure of 5-HT2A complexed with a phenethylamine the full binding poses remain uncertain. Functional assays reveal that different ligands can produce differing intracellular signalling patterns (functional selectivity), and that many psychedelics have off‑target receptor affinities that could influence physiological effects.

Nichols interprets the assembled evidence as showing that SAR for classic 5-HT2A‑acting psychedelics are well developed across tryptamines, ergolines and phenethylamines, yet important mechanistic mysteries remain. He highlights several unresolved issues: why LSD is such a potent human psychedelic despite only modestly exceptional in vitro pharmacology, and why small changes in lysergamide amide substituents dramatically alter psychopharmacology. The discussion credits advances in molecular pharmacology, docking and mutagenesis for clarifying receptor–ligand interactions (for example the complementary pocket for LSD’s diethylamide), but stresses that homology models have limitations and that a crystal structure of the 5-HT2A receptor with a phenethylamine bound would be transformative. The author positions these SAR findings relative to earlier research by noting that much of the early human pharmacology lacked molecular data, so contemporary receptor and signalling assays have been essential to reframe older observations. Nichols emphasises the growing appreciation of functional selectivity — that different agonists can preferentially engage distinct intracellular effectors — as a likely explanation for some differences in human phenomenology and for the imperfect correlations between receptor affinity and reported psychopharmacology. Key limitations acknowledged include the paucity of comprehensive receptor and signalling data for many historically used or recently emerged compounds, reliance on animal behavioural assays and in vitro proxies for human effects, and the absence (at the time of writing) of a solved 5-HT2A crystal structure. The author also notes practical and funding limitations: many important experiments (for example exhaustive receptor‑by‑pathway characterisations for single molecules) would be large, complex undertakings and there is little government funding for psychedelic research despite increasing clinical interest. Implications discussed by Nichols include the need for more complete pharmacological profiling (affinity, functional potency across signalling pathways, and off‑target screens) to understand the molecular basis of psychedelic effects, and the potential to exploit SAR knowledge to design ligands with desired profiles (for example peripheral 5-HT2A agonists for glaucoma that lack hallucinogenic CNS penetration). He emphasises that resolving ligand–receptor interactions and signalling biases could inform both therapeutic development and safety assessments.

Nichols concludes that while SAR for serotonin 5-HT2A‑acting psychedelics are now well elaborated for the three principal chemotypes, significant questions persist. Functional selectivity and off‑target receptor interactions complicate simple structure–potency correlations, and there has been no systematic correlation of specific intracellular signalling pathways with aspects of human psychedelic phenomenology. The author argues that comprehensive studies examining both affinity and pathway‑specific functional potency across likely brain receptors would be necessary to fully understand psychedelic molecular pharmacology, but notes that such endeavours are daunting and underfunded despite rising clinical interest.