Medicinal Chemistry & Drug Development

This report summarises what Blossom’s database shows about the medicinal chemistry and drug development of psychedelics. It is worth being clear at the outset: this is not a condition page and not a treatment. It is about the molecules themselves, how they are designed, how they are changed to make better drug candidates, and how far that engineering has actually got. The honest theme throughout is that the chemistry is advancing quickly while several of its biggest promises remain unproven.

A note before the evidence

This page is a research summary, not medical advice, and nothing here is a recommendation to take any psychedelic, experimental or otherwise. Most of the novel compounds described below exist only in cells and animals, and even the most advanced are early in human testing. "A promising molecule" is not "a safe or effective medicine", and the gap between the two is measured in years of trials.

What this field is actually trying to do

Medicinal chemistry treats a psychedelic as a set of adjustable properties rather than a fixed substance. Change the structure and you change the potency, the duration, the receptor preferences and the safety. The clearest example is the design goal that runs through the whole field: hit the serotonin 2A receptor, which carries the therapeutic and psychedelic effects, while avoiding the 2B receptor and the hERG ion channel, whose stimulation and blockade respectively are linked to heart-valve damage and dangerous arrhythmias[1]Front Pharmacol (2024), review: serotonergic hallucinogens act on the heart (contraction, rate, arrhythmia risk), the 5-HT2B/cardiac design constraint. A good deal of real progress is exactly this: molecules engineered to keep the useful target and drop the dangerous ones, such as halogenated tryptamines with reduced 5-HT2B and hERG affinity[2]preprint (2026), 2-halogenated tryptamines: 2-halogenation cuts 5-HT2A/5-HT2B and hERG affinity (safer cardiac profile) while preserving 5-HT6 (preclinical).

The toolkit for doing this has improved dramatically. Where older psychedelic chemistry was largely trial and error, the field now has cryo-EM structures of the relevant receptors[3]Nature (2024), cryo-EM 5-HT1A structures yield a 5-HT1A-selective 5-MeO-DMT analogue without hallucinogenic-like effects in mice and computational models accurate enough to screen large libraries before anything is synthesised[4]Science (2024), AlphaFold2 models of 5-HT2A support prospective large-library docking with hit rates similar to experimental structures. This is the same structure-based, rational-design approach that transformed the rest of pharmacology, arriving, belatedly, in psychedelics.

The big bet: a therapy without the trip

The most ambitious idea in the field is the "non-hallucinogenic neuroplastogen": a molecule that keeps the brain-plasticity or antidepressant effect of a psychedelic while removing the subjective experience. The appeal is obvious, a pill you could take at home without a day of supervision. And it is not pure fantasy: Ariadne, a phenylalkylamine that differs from a psychedelic by a single methylene group, is a non-hallucinogenic 5-HT2A agonist[5]ACS Chem Neurosci (2022), Ariadne: a one-methylene change yields a non-hallucinogenic 5-HT2A agonist with lower 5-HT2A signalling efficacy; a 5-HT1A-selective analogue of 5-MeO-DMT kept antidepressant-like effects in mice without the hallucinogenic-like behaviour[3]Nature (2024), cryo-EM 5-HT1A structures yield a 5-HT1A-selective 5-MeO-DMT analogue without hallucinogenic-like effects in mice; and zalsupindole, a non-hallucinogenic neuroplastogen, produced plasticity in rats comparable to ketamine, psilocybin and DMT[6]ACS Chem Neurosci (2025), zalsupindole, a non-hallucinogenic neuroplastogen, drives prefrontal plasticity comparable to ketamine/psilocybin/DMT in rats (preclinical).

The honest problem is that the whole programme rests on a premise that has not been tested in humans and is actively disputed within the field: that the therapeutic benefit and the subjective experience are separable. The animal evidence relies on surrogates, a head-twitch behaviour standing in for "hallucination", a swim test for "antidepressant". Those proxies may or may not track what matters in people, and a large strand of clinical research argues the opposite, that the quality of the experience predicts the benefit. So the non-hallucinogen compounds are best read as a fascinating, well-motivated experiment whose central claim is still open, not as a settled improvement.

Why the mechanism question is not academic

Whether you can chemically separate trip from therapy depends on how these drugs actually work, and that is unsettled. One influential line of work concludes that a threshold of one specific signalling pathway, 5-HT2A-Gq, is what makes a molecule psychedelic[7]preprint (2023), 5-HT2A-Gq (not beta-arrestin2) efficacy predicts psychedelic potential, and a Gq threshold explains why lisuride is non-psychedelic, which would suggest you could tune a molecule below that threshold and keep other effects. A different line argues that plasticity is driven by activating the 5-HT2A receptor inside the cell, somewhere lipophilic drugs can reach but serotonin cannot[8]Science (2023), psychedelics promote plasticity via INTRACELLULAR 5-HT2A receptors (location bias), which lipophilic drugs reach but serotonin cannot, supported by simulations of how these molecules cross cell membranes[9]molecular-dynamics study, membrane permeation of psychedelic tryptamines (relevant to the intracellular-receptor hypothesis). These accounts are in real tension, and which is right shapes whether the field’s central bet can be won.

The unglamorous half: formulation and optimisation

Much of drug development is not new molecules at all but making known ones usable. This is where the field is most mature. There are extended-release oral ketamine tablets in mid-stage trials[10]Phase 2 (2024), extended-release oral ketamine tablets (R-107) in treatment-resistant depression (BEDROC), a non-dissociative ketamine metabolite in early human testing[11]Clin Pharmacol Ther (2024), Phase 1 of (2R,6R)-hydroxynorketamine, a non-dissociative ketamine metabolite, in healthy volunteers, subcutaneous prodrugs engineered to give a controlled, shorter session[12]Phase 1 (2024), subcutaneous RE104, a 4-OH-DiPT prodrug, safety/PK/PD in humans, transdermal DMT delivery[13]Eur J Pharm Sci (2024), transdermal DMT patch: 77% bioavailability vs IV and a 20-fold half-life extension at sub-hallucinogenic levels (preclinical, mouse), and careful human work on basics like the bioavailability of different LSD salt forms[14]clinical PK (2024), absolute oral bioavailability and bioequivalence of LSD base vs tartrate salt in humans. Even a shorter-acting novel 5-HT2A agonist such as GM-2505[15]J Psychopharmacol (2025), GM-2505 Phase 1 (n=48): a 5-HT2A agonist with a 40-50 min half-life, shorter-acting than psilocybin, optimal 10-15 mg IV is partly a formulation-and-duration play: a more practical session length is a development advantage in its own right.

This work rarely makes headlines, but it is where chemistry most directly enables (or blocks) a real medicine. It also quietly drives the commercial pipeline: a new salt, a prodrug, a deuteration or a novel route is often as much about owning intellectual property as about improving outcomes. The same is true of analogues such as methylone and other entactogen variants of the MDMA scaffold[16]Front Psychiatry (2023), methylone, an MDMA-scaffold entactogen analogue with rapid anxiolytic/antidepressant-like activity (preclinical), which extend a known chemistry into patentable new territory.

Reading this honestly

So how should you read the chemistry story? As a field whose engineering is maturing impressively while its central therapeutic premises remain unproven. The genuine advances are real and worth taking seriously: rational, structure-based design has arrived; chemists can now build molecules that hit the therapeutic receptor while sparing the cardiac ones; and the formulation work needed to turn awkward natural compounds into dosable medicines is well underway. But the headline ambition, a therapy without the trip, depends on a premise that animals cannot settle and humans have not yet tested, and the mechanism that would tell us whether it is achievable is itself contested. Layered on top is a commercial pipeline that compresses the long distance between "a clean molecule in a dish" and "a proven medicine", and a great deal of patent-driven tweaking dressed up as innovation. The most useful thing this literature offers an honest reader is a clear line between what the chemistry has achieved (selectivity, designed-in safety, better formulations, rational design) and what it has merely promised (non-hallucinogenic cures, next-generation breakthroughs). Credit the first; keep waiting, sceptically, on the second.