Updated September 16, 2026: New evidence complicates, rather than resolves, this mechanism question. Du et al. (Chinese Medical Journal, published online September 15, 2026) report that in mice, blocking 5-HT2A with either ketanserin or the selective antagonist MDL100907 did not prevent psilocybin from facilitating fear extinction, a second independent behavioral paradigm, beyond the forced-swim-test result discussed above, in which psilocybin’s effect appears not to require 5-HT2A. In the MDL100907 arm specifically, blockade also failed to prevent psilocybin from increasing neuroplasticity markers, BDNF expression and new neuron growth, in the hippocampus. The authors’ own conclusion states that the therapeutic and hallucinogenic effects “can be dissociated.” But the identical extinction paradigm has produced the opposite result before: a 2024 study using the same antagonist, MDL100907, in a comparable fear-extinction protocol found that it fully blocked psilocybin’s effect on extinction retention, at a lower psilocybin dose, 1 milligram per kilogram against the new study’s 2.5. The dose difference is one plausible explanation for the divergent results, though neither study tested a range wide enough to confirm it. Two studies, similar methods, opposite conclusions, is not resolution. It is the same unsettled mechanism this piece described in June, now with sharper and more directly contradictory evidence on both sides.
The most valuable idea in psychedelic drug development is the psychedelic without the trip: a compound that keeps the rapid, durable antidepressant effect while removing the hours-long hallucinatory experience that makes these drugs expensive to administer. The strategy rests on one assumption: that the antidepressant effect and the hallucination both run through the same receptor, 5-HT2A, a serotonin receptor that psychedelics activate. If that is true, a sufficiently clever 5-HT2A agonist could separate the two effects, keeping the benefit and dropping the trip. A growing body of preclinical work suggests that assumption is shakier than the strategy needs it to be, and that the receptor pharmacology may not cooperate with the commercial plan.
The strategy and its premise
Psychedelics activate the 5-HT2A receptor, which produces their hallucinatory effects, and clinical trials have established psilocybin and related compounds as rapid-acting, durable antidepressants. The commercial response, pursued by a number of companies developing so-called non-hallucinogenic psychedelics, is to engineer 5-HT2A agonists, often through biased signaling, that retain the neuroplastic and antidepressant effects without the subjective experience. The entire approach assumes that 5-HT2A is the seat of both the benefit and the hallucination, and that the two can be pried apart at that single receptor.
The complication
The role of 5-HT2A in the antidepressant effect is contested in the preclinical literature. In one comparative study, researchers tested psychedelics in mice lacking the 5-HT2A receptor. The antidepressant-like effects of two other psychedelics disappeared in these mice, but psilocybin’s effect was preserved. That implies psilocybin works at least partly through a mechanism other than 5-HT2A. That alone muddies the clean picture the non-hallucinogenic strategy depends on. If the benefit does not live entirely at 5-HT2A, then a compound engineered for narrow 5-HT2A selectivity, in order to strip out the trip, might also strip out whatever else carries the effect.
That raises the obvious question of which other receptor, and the candidate the field would least like to hear is 5-HT2B. Psilocybin’s active metabolite has affinity there, and if a non-5-HT2A receptor such as 5-HT2B proves to be a necessary contributor to the antidepressant effect, it points to the one receptor developers most want to keep out of a central-nervous-system drug.
Why 5-HT2B is the wrong receptor to need
5-HT2B agonism is the mechanism behind drug-induced cardiac valve disease. It is why the fen-phen combination was withdrawn, and why ergot-derived dopamine agonists carry valvulopathy warnings. As a rule, drug developers work to eliminate 5-HT2B activity from central-nervous-system compounds precisely to avoid that risk. So if psilocybin’s antidepressant effect partly requires 5-HT2B engagement, the field faces a real bind: the receptor that may be necessary for the benefit is the same one developers most want to avoid for safety. It cannot simply be designed out without potentially taking the efficacy along with it.
Two bets get harder
The possibility presses on two of the field’s commercial strategies at once.
The first is the non-hallucinogenic compound itself. If efficacy depends on receptors beyond 5-HT2A, a molecule engineered for clean 5-HT2A selectivity, in order to remove the hallucination, could also miss the receptors that carry the therapeutic effect. The result could be a drug that is both non-hallucinogenic and non-efficacious. The selectivity that removes the trip could remove the treatment.
The second is repeated dosing, and microdosing in particular. A single supervised psilocybin session is one exposure. The microdosing model, frequent low doses taken over time and itself under active study, is where any 5-HT2B-mediated cardiac risk would accumulate. If the benefit and 5-HT2B engagement travel together, the safety calculus for any repeated-dosing psychedelic gets worse, not better.
What this rodent evidence doesn’t prove
These limits are heavy, and they matter. This is preclinical work, in rodent models, using the forced swim test, a crude and much-criticized proxy for antidepressant activity. Receptor mechanisms in rodents do not always translate to humans. The receptor pharmacology of psychedelics remains unsettled. Even 5-HT2A’s necessity is disputed across studies, and antagonist experiments carry their own off-target confounds. No single result resolves the mechanism. None of this refutes the non-hallucinogenic strategy. Biased agonism and other approaches remain live, and the human efficacy of non-hallucinogenic compounds is still being tested directly. A claim that a non-5-HT2A receptor is a necessary contributor to the antidepressant effect, based only on a rodent assay, is a strong one. It needs replication and evidence of human relevance before it should move anyone’s strategy.
A bet on an unsettled mechanism
The field’s most important commercial idea, keep the benefit and drop the trip, rests on a clean mechanistic story that the preclinical evidence keeps muddying. The complication is the live possibility that 5-HT2B, a receptor developers normally avoid for cardiac safety, is part of how psilocybin works. If that is right, the non-hallucinogenic compounds risk losing efficacy along with the hallucination, and any repeated-dosing model inherits a cardiac-safety question. None of this is established in humans, and none of it should be over-read from a rodent assay. But it is a reminder that the psychedelic-without-the-trip thesis is a bet on a mechanism the field has not pinned down, and that the receptor pharmacology may not cooperate with the commercial plan. The appeal of separating a molecule’s benefit from its hallucination is obvious, and the capital chasing it is real. The persistent message from the biology is that the separation may be harder than the thesis assumes, because the benefit may not live where the strategy needs it to.