Psychedelics and the human receptorome

The empirical core of the paper combines new high-throughput binding data from the NIMH‑PDSP with receptor data extracted from the literature. NIMH‑PDSP assayed a panel that the paper describes variously as 51 sites for the primary set of compounds and, when combined with literature data, a total of 67 receptors, transporters and ion channels across the thirty-five drugs. The PDSP component comprised assays of a set including phenylalkylamines, tryptamines and at least one ergoline; the extracted text reports sixteen phenylalkylamines, eight tryptamines and one ergoline among the compounds assayed by PDSP. In addition, ten compounds (including ergolines, salvinorin A, ibogaine, THC and morphine) were incorporated from the literature; the latter three are noted as incompletely assayed and should be treated cautiously. Binding assay workflow at NIMH‑PDSP followed a two-stage strategy: a primary screen at 10 μM (10,000 nM) to identify >50% inhibition “hits”, followed by secondary concentration–response assays (serial dilutions across nM concentrations) to estimate Ki (equilibrium dissociation constant) values for hits. Each reported Ki was calculated from at least three replicated assays. Activity (functional) assays were also performed for twenty-five compounds at 5-HT2A and 5-HT2C using Ca2+ mobilisation (FLIPR TET); these activity assays used cell lines with high receptor expression, so partial agonists may appear to have considerable agonist activity. Activity estimates (Emax, EC50) were derived from single experiments done in quadruplicate, with reported means and variance. Because raw Ki values span many orders of magnitude, the study log-transformed affinities into pKi-like values (higher pKi = higher affinity) and then applied a bespoke normalisation intended to factor out absolute potency differences between drugs. In this normalisation, each drug's highest affinity is scaled to a common maximum (reported as 4.0), and Ki values reported as >10,000 nM (the practical assay ceiling) are treated as zero for the normalised metric (npKi). The normalised scale is intended so that each unit corresponds approximately to one order of magnitude of Ki; Ray uses an npKi threshold of about 2.0 (a 100-fold drop from the maximum) as a heuristic limit of likely perceptible receptor interaction. To quantify overall ‘‘breadth’’ (inverse selectivity) of a drug's profile, three indices were used: B (simple sum of npKi across receptors), B_sq (square-root of sum-of-squares, which gives greater weight to higher affinities and is favoured by the author), and B_exp (an exponential weighting). Proportional participation of receptor groups was calculated as the ratio of the sum-of-squares for that group to the total sum-of-squares across all receptors (Bp). The author also examined the effect of assay truncation (when a drug's best measurable Ki is relatively weak) on interpretable receptor profiles and flagged several compounds whose profiles are truncated by the PDSP assay limits.

Across the combined data set of thirty-five drugs and the assayed receptor panel, Ray reports that psychedelic and related compounds interact broadly: 42 of 49 commonly assayed sites produced at least one measurable hit among the compounds. PDSP directly assayed twenty-five compounds across a large panel (reported as 51 sites for the PDSP set) and literature sources added ten further drugs, yielding raw Ki data for thirty-five drugs at sixty-seven sites in total. Normalised affinity patterns (npKi) show diverse, drug-specific profiles. Some phenylalkylamines are relatively selective but few drugs are exclusively 5-HT2-selective; DOB and MEM are cited as the closest to 5-HT2-selective prototypes, whereas DOI—often used experimentally as a prototypical 5-HT2 agent—was found to be among the least selective compounds. Using the perceptibility heuristic (npKi ≈2.0 as the 100-fold cutoff), many drugs have multiple receptors in the putatively perceptible range. For example, potent phenylalkylamines such as DOB and DOI show measurable affinities across many receptors, but DOB's interactions cluster around the 5-HT2 family while DOI has a broader set of roughly nineteen receptors in the perceptible range. Ranking by breadth indices places the thirty-five drugs along a continuum from promiscuous to selective; Ray concludes that B_sq is the most meaningful summary statistic. At the receptor level, summing across drugs identifies the receptors most frequently and strongly engaged by the set: in descending order the top receptors were 5-HT2B, 5-HT1A, 5-HT7, 5-HT1D, 5-HT2A, 5-HT2C and alpha2C, with some variation among the three breadth metrics. Specific noteworthy observations include: LSD has the strongest collective interaction with the five dopamine receptors and with the ten assayed 5-HT receptors; DMT is the only drug whose best hit was at 5-HT7 and also shows the strongest interaction with an individual dopamine receptor (Dmax); mescaline's best hit was an adrenergic receptor (alpha2C); ibogaine's best hit was a sigma receptor; MDMA's best hit was an imidazoline receptor. Salvinorin A is highly selective for kappa opioid receptor (KOR), and THC shows highest affinity at CB1 and CB2 in the available data. The analysis also highlights methodological limitations: seven drugs (TMA, mescaline, TMA‑2, DIPT, MDMA, 5‑MeO‑DIPT, ibogaine) had best-hit Ki values >100 nM such that the assayable dynamic range was insufficient to characterise a full 100-fold perceptible range, leading to truncated receptor profiles. Morphine and THC were not broadly assayed by PDSP and their profiles are incomplete in the combined data set. Activity assays at 5-HT2A and 5-HT2C indicate most compounds behave as full agonists under the high-expression assay conditions, but psilocin, MDMA, DOM and three control drugs showed lower Emax values.

Ray interprets the principal finding as a reframing of how psychedelics should be characterised pharmacologically: rather than being highly selective 5-HT2 agents, many psychedelics interact with a wide variety of receptors. The NIMH‑PDSP data show that thirty-five drugs together emphasise many receptor systems beyond 5-HT2A/C, with substantial engagement of 5-HT1 subtypes, 5-HT7, adrenergic, dopaminergic and other receptors. Phenylalkylamines tend to be more selective than tryptamines and ergolines, but selectivity is relative rather than absolute; DOI, frequently treated in the literature as a paradigmatic 5-HT2-selective ligand, is highlighted as notably non-selective and the author suggests that conclusions from work assuming DOI's selectivity may need re-evaluation. The diversity of multi-receptor interaction patterns is presented as a plausible explanation for qualitative differences in subjective effects across drugs and as an opportunity: the panel of drugs with differing receptor emphasis could serve as pharmacological probes to investigate the contribution of specific receptor systems to human subjective states. Ray emphasises that studies seeking a truly 5-HT2-selective pharmacological tool should consider DOB or MEM rather than DOI. Methodological caveats and limitations are acknowledged. Assay ceilings (Ki reported as >10,000 nM) and differences between PDSP and literature data produce incomplete or truncated receptor profiles for several compounds; ibogaine, morphine and THC are flagged as incompletely characterised. Functional assays used cell lines with high receptor expression, which can inflate apparent efficacy for partial agonists. Finally, the normalisation and breadth metrics are heuristic tools—the choice of B_sq was justified by regression against human potency data, but Ray notes that different indices weight affinities differently and that some equivalences implicit in simple summation may not be realistic. Ray concludes by proposing that the present receptorome data be integrated with human pharmacology in follow-up work to further probe how particular receptor interactions contribute to the mind-altering effects of these compounds.