Psilocybin induces schizophrenia-like psychosis in humans via a serotonin-2 agonist action

Earlier research has emphasised dopaminergic dysfunction in schizophrenia, but accumulating evidence implicates serotonin (5-HT) systems as well. Clinical observations include altered cortical serotonin receptor binding in schizophrenia, potent 5-HT2 antagonism among atypical antipsychotics, and the fact that classic indoleamine hallucinogens such as psilocybin and LSD can elicit syndromes in healthy people that resemble early stages of schizophrenia, including ego disturbances, affective changes, loosened associations and perceptual alterations. Animal studies suggest that hallucinogens act via 5-HT2 receptors, yet extrapolation to humans is uncertain and some hallucinogens also affect dopaminergic systems, leaving open whether human effects are mediated directly by 5-HT2 receptors, indirectly via dopamine, or both. To address this gap, Vollenweider and colleagues tested whether psilocybin’s psychotomimetic and cognitive effects in healthy volunteers are mediated by 5-HT2 and/or dopamine D2 receptors. Using a placebo-controlled, within-subject design, they examined how pretreatment with a selective 5-HT2 antagonist (ketanserin), a D2 antagonist typical antipsychotic (haloperidol), or a mixed 5-HT2/D2 antagonist atypical antipsychotic (risperidone) altered psilocybin-induced psychological and working memory effects, assessed chiefly with the Altered State of Consciousness (APZ-OAV) rating scale and a delayed response task (DRT).

The study enrolled 25 healthy university-affiliated volunteers screened by psychiatric interview and medical assessment; individuals with personal or first‑degree family histories of major psychiatric disorder or with a history of illicit drug abuse were excluded. Ethical approval and regulatory authorisation for psilocybin were obtained. Psilocybin (0.25 mg/kg, oral) was chosen to produce robust psychological effects lasting roughly 120–180 minutes after a 20–30 minute onset. Antagonists were ketanserin (20 and 40 mg, oral), haloperidol (0.021 mg/kg, intravenous), and risperidone (0.5 and 1.0 mg, oral); the haloperidol and risperidone doses were expected to occupy roughly 55–65% of D2 receptors. Experiment 1 used a within-subject design in 15 subjects (eight male, seven female; mean age 29.7 ± 5.3) split into three groups of five, each group tested at monthly intervals for one antagonist series (ketanserin, haloperidol, or risperidone). For ketanserin and haloperidol pretreatments, antagonist or placebo was given and 75 minutes later subjects received psilocybin (or placebo); for risperidone, pretreatment preceded psilocybin by 90 minutes for kinetic reasons. Psychological state was measured with the APZ-OAV scale and working memory with a visual-manual delayed response task (DRT) immediately prior to treatment and again approximately 80 minutes after, coinciding with peak psilocybin effects. Experiment 2 comprised a separate cohort of 10 subjects (five male, five female; mean age 28.4 ± 4.0) to replicate the ketanserin 40 mg pretreatment finding; the protocol matched Experiment 1 but included APZ-OAV ratings only. The APZ-OAV yields three dimensions: OSE (oceanic boundlessness), VUS (visionary restructuralization), and AIA (dread of ego-dissolution). The DRT involved 32 trials in which participants recalled target positions after a 10 s delay while performing a concurrent category-shift task; response times on correct trials were recorded. Statistical analysis used two-way repeated-measures ANOVAs with pretreatment and drug as within-subject factors, Tukey post hoc tests where indicated, non-parametric Friedman ANOVAs for certain comparisons, and Spearman correlations for associations between reaction-time changes and psychopathology. Significance was set at p < 0.05.

Oral psilocybin (0.25 mg/kg) produced a transient psychosis-like syndrome in healthy volunteers, with effects emerging 20–30 minutes after ingestion, peaking after an additional 30–50 minutes and lasting around 1–2 hours. On the APZ-OAV, psilocybin significantly increased scores on all three dimensions (OSE, VUS, AIA) across the antagonist experiments, reflecting derealisation and exalted mood (OSE), visual/perceptual alterations including hallucinations and synaesthesia (VUS), and anxious ego-dissolution with thought disorder and suspiciousness (AIA). Pretreatment with the 5-HT2 antagonist ketanserin dose-dependently attenuated these effects: two-way ANOVAs showed significant drug × pretreatment interactions for OSE (F(2,8) = 12.2, p < 0.004), VUS (F(2,8) = 26.3, p < 0.0003) and AIA (F(2,8) = 21.7, p < 0.0006). Specifically, 20 mg ketanserin reduced psilocybin-induced APZ-OAV scores by about 50–70%, while 40 mg largely prevented the effects in four of five subjects (98–100% reduction) and substantially reduced scores in the remaining subject (75–87%). A separate replication cohort confirmed that 40 mg ketanserin blocked psilocybin effects by 87–96% with highly significant interactions (OSE F ≈ 51.7, p < 0.0001; VUS F ≈ 111.7, p < 0.00001; AIA F ≈ 16.2, p < 0.003). Risperidone, a mixed 5-HT2/D2 antagonist, produced dose-dependent attenuation as well: 0.5 mg reduced APZ-OAV scores by about 69–78% and 1.0 mg nearly abolished the psilocybin-induced syndrome (98–99%), with significant drug × pretreatment interactions (OSE F(2,8) = 7.4, p < 0.02; VUS F(2,8) = 8.2, p < 0.01; AIA F(2,8) = 8.5, p < 0.01). By themselves, neither ketanserin nor risperidone altered APZ-OAV scores. In contrast, haloperidol had a limited and differential effect: it reduced the OSE scale by about 54% (drug × pretreatment interaction for OSE F(1,4) = 20.4, p < 0.01) but did not reduce VUS scores (main drug effect F(1,4) = 52.3, p < 0.002; no significant pretreatment or interaction). Unexpectedly, haloperidol amplified AIA scores by 148% when combined with psilocybin; statistical tests showed significant main effects (drug F(1,4) = 14.4, p < 0.02; pretreatment F(1,4) = 8.7, p < 0.04) though no significant interaction in the two-way ANOVA, and confirmatory Friedman tests indicated the increase was greater than with psilocybin alone (p < 0.025). Item analysis suggested this increase reflected increased thought disorder and anxiety. On the DRT, psilocybin lengthened reaction times without reducing accuracy. Ketanserin and risperidone, but not haloperidol, dose-dependently blocked the psilocybin-induced reaction-time increases (ketanserin F(2,8) = 6.0, p < 0.03; risperidone F(2,8) = 4.8, p < 0.04). Antagonists alone did not affect reaction times, and changes in reaction time did not correlate significantly with APZ-OAV psychopathology measures.

Vollenweider and colleagues interpret their results as strong evidence that psilocybin’s psychotomimetic effects in humans are mediated specifically by activation of 5-HT2 receptors, and in particular the 5-HT2A subtype. The finding that a selective 5-HT2 antagonist (ketanserin) and the mixed 5-HT2/D2 antagonist risperidone nearly abolished both the subjective psychosis-like syndrome and the working-memory impairment implicates 5-HT2A agonism as the principal mechanism, validating prior animal data linking hallucinogenic effects to 5-HT2 receptors. The pattern of symptom changes—prominent derealisation and euphoria (OSE), visual hallucinations (VUS), and anxious ego-dissolution and thought disorder (AIA)—is taken to mirror features of early or acute schizophrenia, supporting the notion that serotonergic hyperactivity may contribute to psychotic symptom formation. The selective blockade by ketanserin, which is substantially more potent at 5-HT2A than 5-HT2C receptors, is presented as evidence favouring 5-HT2A over 5-HT2C involvement. By contrast, the typical D2 antagonist haloperidol was substantially less effective at blocking psilocybin-induced phenomena and, notably, increased anxious ego-dissolution (AIA) when combined with psilocybin; this pattern suggests that D2 antagonism alone does not prevent—and may even exacerbate—certain psilocybin-elicited symptoms. The authors note that psilocybin and its active metabolite psilocin do not act directly at D2 receptors, so haloperidol’s partial effects might reflect an indirect modulation of dopaminergic systems; however, the data do not permit inference about whether haloperidol-enhanced AIA arises from serotonergic modulation of dopamine, GABA, or other systems. The demonstration that psilocybin-induced working memory deficits were prevented by 5-HT2 but not D2 antagonism leads the investigators to suggest a possible role for 5-HT2A stimulation in cognitive impairments seen in schizophrenia, consistent with clinical observations that some cognitive deficits respond better to atypical than typical neuroleptics. Finally, the authors propose that psilocybin-induced psychosis could serve as an experimental human model to study aspects of atypical antipsychotic action that are not captured by amphetamine-based (dopamine-centric) models, while acknowledging mechanistic uncertainties that remain.

The study concludes that psilocybin produces schizophrenia-like symptoms and working-memory impairments primarily via stimulation of 5-HT2 receptors, most likely the 5-HT2A subtype. From this, the authors infer that excessive 5-HT2A activity may contribute to psychotic symptom formation and cognitive deficits in at least a subset of patients with schizophrenia, and that selective 5-HT2A antagonists could have therapeutic utility. They further suggest that psilocybin-induced psychosis may provide a useful human model to investigate facets of atypical antipsychotic pharmacology that differ from those revealed by dopaminergic models.