Repeated low doses of psilocybin increase resilience to stress, lower compulsive actions, and strengthen cortical connections to the paraventricular thalamic nucleus in rats

Psilocybin, a serotonergic psychedelic prodrug that is converted to the psychoactive psilocin, has shown efficacy in treating depression, end-of-life anxiety and substance-use disorders when given as single high doses combined with psychotherapy. An alternative practice, commonly called microdosing, consists of repeated administrations of very low doses that do not produce overt psychedelic effects; human observational studies report mood, well‑being and cognitive benefits but are vulnerable to expectation and placebo effects and lack rigorous biological validation. The authors therefore sought to establish an objective, translational psilocybin ‘‘microdosing’’ regimen in rats based on in vivo 5-HT2A receptor occupancy, and to test whether repeated low doses produce measurable behavioural and neurobiological effects. Kiilerich and colleagues aimed to determine a dose that occupies less than 20% of brain 5-HT2A receptors (a threshold previously linked to absence of overt psychedelic effects), then to apply that dose intermittently over multiple weeks and assess behaviours relevant to anxiety, anhedonia and compulsivity, together with receptor and synaptic markers postmortem. The study combined in vivo PET receptor-occupancy experiments, in vitro receptor pharmacology and functional assays, a repeated low-dose treatment regimen, multiple behavioural tests (open-field, pre-pulse inhibition of the acoustic startle reflex, sucrose preference, elevated plus maze) and autoradiographic quantification of 5-HT receptors and synaptic vesicle protein 2A (SV2A). The authors hypothesised that repeated low doses would alter behaviours linked to mood, anxiety and compulsivity and produce detectable receptor or synaptic changes.

Male Long Evans rats were used across two main experiments (the extracted text reports n = 78 for experiment 1 and n = 18 for experiment 2). Housing, light/dark cycle and enrichment were standardised; rats in experiment 1 were food-restricted for motivation in startle tests, whereas animals for receptor binding studies had ad libitum food. Ethical approval and adherence to guidelines are reported. To define a microdose, a PET occupancy study used the 5-HT2A-selective tracer [18F]MHMZ. Four dose conditions (vehicle, 0.05, 0.20 and 1.00 mg/kg psilocybin, subcutaneous) were administered and animals scanned 40 minutes later with dynamic PET. Non-displaceable binding potential (BPND) in frontal cortex was calculated with a Logan reference tissue model using cerebellum as reference, and percent occupancy was derived from dose-induced changes in BPND relative to vehicle. The extracted text states group numbers as "n = 18" in this section but does not clearly report how that total was distributed across the four dose groups. Psilocin pharmacology was characterised in vitro. Autoradiographic competition binding assays quantified psilocin Ki against 5-HT1A, 5-HT2A, 5-HT2C and 5-HT7 receptors in region-appropriate brain sections. Functional activity (EC50 and efficacy) was assessed in HEK293 cell lines stably expressing the receptors using Ca2+ (Fluo-4) assays for 5-HT2A/2C and cAMP assays for 5-HT1A/5-HT7. A behavioural microdosing regimen was established by administering aliquots of psilocybin (prepared at 1 mg/mL and diluted) subcutaneously at 0.05 mg/kg every second day. The text contains inconsistent statements about total regimen length (it refers both to 21 days and 24 days), so the precise number of treatment days is not unambiguously reported in the extracted material. Two cohorts were used in experiment 1 with some procedural differences and incomplete blinding: cohort 1 (handled by an experimenter with male pheromones) experienced a problematic sucrose preference test and received a single high dose at the end; cohort 2 (experimenter with female pheromones) did not. Experiment 2 was performed by a blinded experimenter. Behavioural testing comprised (a) an open-field test (familiar and novel variants) with automated tracking of locomotion, rearing and grooming, (b) acoustic startle reflex (SR) and pre-pulse inhibition (PPI) testing in a sound-attenuated chamber (PPI is a measure of sensorimotor gating, with reduced PPI used as an endophenotype of schizophrenia), (c) repeated sucrose preference tests to probe anhedonia, and (d) elevated plus maze for anxiety-like behaviour. Wet-back shakes, a stereotyped behaviour indicative of 5-HT2A activation, were scored in a separate dose–response arena test. After euthanasia by decapitation, brains were frozen and sectioned for autoradiography of 5-HT1A, 5-HT2A/2C, 5-HT7 (using radioligands specified in the text) and SV2A ([3H]UCBJ) to estimate synaptic density. Specific binding ratios were calculated relative to cerebellum or corpus callosum reference regions depending on section orientation. Statistical approaches included one-way and two-way ANOVAs with Dunnett, Šídák or other corrections for multiple comparisons within experiments; reported significance thresholds were P < 0.05, P < 0.01 and P < 0.001. The authors note they did not apply correction across separate experiments or different measurements defined a priori.

Receptor occupancy and behavioural activation: PET with [18F]MHMZ showed dose-dependent 5-HT2A receptor occupancy. Frontal cortex BPND was 2.01 ± 0.13 in controls, 1.64 ± 0.13 after 0.05 mg/kg (reported as 19 ± 9% occupancy, not significant), 1.55 ± 0.17 after 0.2 mg/kg (21 ± 16% occupancy, P = 0.028) and 1.26 ± 0.35 after 1.0 mg/kg (41 ± 19% occupancy, P < 0.001). In a separate behavioural dose–response, wet-back shakes followed an inverted U-shaped curve; low doses did not increase shakes, whereas 1 mg/kg produced a clear increase (8.5 ± 5.8 per 10 min, P = 0.01). Based on these data, the authors defined a microdose as 0.05 mg/kg subcutaneously, because it evoked <20% 5-HT2A occupancy and did not elicit overt psychedelic-like behaviour. Psilocin pharmacology: In vitro autoradiography reported psilocin Ki values of 80 ± 28 nM at 5-HT1A, 15 ± 2 nM at 5-HT2A, 12 ± 8 nM at 5-HT2C and 524 ± 272 nM at 5-HT7 (measured against specific radioligands in region-appropriate sections). Functional assays in HEK293 cells produced EC50 estimates of 727 ± 45 nM (5-HT1A), 36 ± 8 nM (5-HT2A), 11 ± 0.3 nM (5-HT2C) and 45 ± 59 nM (5-HT7). Psilocin behaved as a partial agonist at all four receptors with efficacies ranging across targets. The authors infer that a dose producing ~20% 5-HT2A occupancy would also engage 5-HT2C substantially (potentially more than 5-HT2A) and show modest 5-HT7 activation, with negligible 5-HT1A activation at that dose. Behaviour in familiar open-field and startle/PPI tests: Repeated low-dose treatment (0.05 mg/kg every second day) did not alter locomotion, time in centre, or rearing counts in a familiar 40‑minute open-field. Grooming frequency, defined as uninterrupted grooming episodes, decreased by 14% in microdosed animals versus controls (unpaired two‑tailed t-test, P = 0.016), while total grooming time was not changed. Basal acoustic startle (SR) was intensity‑dependent and plateaued at 110 dB; SR at 110 dB did not differ between groups (45 ± 12% control vs 45 ± 9% treated). There were no group differences in SR with or without pre-pulses and no difference in percentage PPI. Latency to peak startle and body-weight gain were similar between groups. Short-term habituation to startle showed the expected reduction in controls (P = 0.007) but not in psilocybin-treated animals; however, the authors attribute this to high variance in the treated group rather than a clear effect. Sucrose preference (anhedonia measure): In experiment 2, baseline sucrose preference was 81 ± 0.3% on the second baseline test. During the treatment period, control animals that received repeated subcutaneous injections showed a decline in sucrose preference to a mean of 50 ± 27% (mixed-effects model F(1,64,11.08) = 4.94, P = 0.03), indicating stress-induced reduction in preference. By contrast, psilocybin-treated animals did not show this decline; instead they exhibited an immediate within-treatment increase to 92.7 ± 5.2% on their first within-treatment test, and preference returned to baseline after treatment cessation. The authors interpret this as increased resilience to stress-induced anhedonia during the dosing period, while noting the effect did not persist after treatment ended. Novel open-field and elevated plus maze: Repeated low doses produced no changes in centre entries or time in the novel open-field, and no change in open- versus closed-arm preference in the elevated plus maze, supporting the lack of anxiogenic or anxiolytic effects. Grooming frequency was further reduced by 48% in the elevated plus maze without changes in locomotion or rearing, replicating the grooming finding. Tolerance and receptor expression: An acute challenge with 1 mg/kg psilocybin after the repeated low‑dose regimen produced comparable wet-back shake responses in previously treated and control animals, indicating no behavioural desensitisation. Postmortem autoradiography showed no changes in 5-HT2A or 5-HT2C receptor levels in prefrontal cortex, striatum or choroid plexus. 5-HT7 and synaptic marker changes in PVT: Autoradiographic quantification revealed increased 5-HT7 receptor expression in the paraventricular nucleus of the thalamus (PVT): cohort 2 showed a 16% increase in 5-HT7 in the PVT (P = 0.01). Adjacent-section analysis of SV2A binding ([3H]UCBJ) demonstrated a 60% increase in specific binding ratio in the PVT of psilocybin-treated animals compared with controls (P < 0.001), with no change in hippocampus or cingulate cortex. These findings are presented as consistent with increased presynaptic inputs, potentially of cortical origin, to the PVT after repeated low-dose psilocybin.

Kiilerich and colleagues interpret their results as validation of a reproducible low‑dose psilocybin regimen in male rats that is pharmacologically defined by <20% 5-HT2A occupancy and absence of overt psychedelic behaviour. The regimen did not produce anxiogenic effects or deficits in sensorimotor gating (no PPI impairment), nor did it induce behavioural or receptor desensitisation to 5-HT2A agonists. Importantly, repeated low doses reduced compulsive-like self-grooming and increased resilience to stress-induced reductions in sucrose preference during the treatment period, suggesting acute protective effects against stress-induced anhedonia. The authors link the behavioural findings to neurobiological changes in the paraventricular nucleus of the thalamus (PVT), where they observed increased 5-HT7 receptor expression and elevated SV2A, a proposed synaptic density marker. They propose that increased presynaptic 5-HT7 and SV2A reflect strengthened corticothalamic glutamatergic inputs to the PVT, and that PVT modulation may underlie improved approach‑avoidance and motivational behaviours reported anecdotally in human microdosing studies. Psilocin pharmacology measured in vitro supports engagement of 5-HT2C and, to a lesser extent, 5-HT7 receptors at the chosen low dose, which the authors suggest may contribute to the observed effects. Several limitations are acknowledged in the extracted text. Only male rats were studied, so sex differences are not addressed. The sucrose preference result reflects resilience during dosing and normalisation after cessation, leaving uncertain whether effects persist long term; the authors note sucrose preference is a single index of anhedonia and encourage testing in validated depression models (for example chronic mild stress). Methodological heterogeneities are also apparent: some behavioural cohorts were not blinded, cohort differences and a problematic sucrose test in cohort 1 may have influenced certain measures, and the extracted text contains inconsistent reporting of total regimen duration (21 vs 24 days). The authors further acknowledge that human microdosing studies are confounded by expectation and blinding challenges and argue that animal models can provide objective insight into underlying biology. Finally, the investigators suggest that the combination of behavioural resilience to stress, reduced compulsive actions and increased synaptic markers in the PVT provides a plausible mechanism by which repeated low doses of psilocybin could affect motivation and compulsivity, warranting further research including sex comparisons, depression-model testing and experiments to map the precise circuit-level origin of increased PVT inputs.