Dissociating the Hallucinogenic and Neuroplastic Effects of Psilocybin

The researchers used adult C57BL/6J mice and Thy1-GFP-M mice for imaging-based spine analyses, alongside genetically modified mouse lines to manipulate 5-HT 2A R expression in cortical layer 5 pyramidal neurons. Mice were group-housed on a 12 h light/dark cycle and assigned randomly to groups. All procedures were approved by the relevant animal ethics committee. Psilocybin was given by intraperitoneal injection at 0.3, 1, or 3 mg/kg, with saline as the vehicle. To assess hallucination-like behaviour, mice were placed in a cage after injection, video recorded for 20 min, and HTRs were manually annotated by an observer blinded to group allocation. The authors first compared dose effects in wild-type mice, then focused on 1 mg/kg, which appeared to maximise both behavioural and structural effects. To measure neuroplasticity, the researchers implanted cranial windows around postnatal day 60 and performed longitudinal in vivo two-photon imaging of dendritic spines in sparse cortical layer 5 pyramidal neurons. Mice were imaged at baseline and again after treatment at 24 h; additional follow-up imaging examined longer intervals of 7 and 21 days. Spine formation, elimination, density, morphology, and the fate of newly formed versus pre-existing spines were analysed from 3D image stacks in ImageJ. The authors also used a conditional rescue line, a full knock-out line, and a conditional knock-out line to test whether restoring or deleting 5-HT 2A R specifically in layer 5 pyramidal neurons altered HTRs and spine dynamics. Immunohistochemistry confirmed receptor deletion or restoration in the relevant lines. Statistical analyses were done in GraphPad Prism 10, with each mouse treated as one data point. The extracted text says the sample sizes, tests, and p values are reported in the figures and associated tables, but those details are not fully reproduced in the text provided.

In wild-type mice, all tested psilocybin doses induced robust HTRs. The 1 mg/kg and 3 mg/kg doses produced similar HTR counts, and both were significantly greater than the 0.3 mg/kg dose. HTRs rose during the first 5 min after injection, peaked at 5-10 min, and then declined. In vivo spine imaging showed a dose-dependent but saturating neuroplastic effect. Compared with saline, 1 mg/kg and 3 mg/kg psilocybin significantly increased spine formation over 24 h to a similar extent, whereas 0.3 mg/kg did not. Spine elimination over the same short interval was unchanged. As a result, 1 mg/kg and 3 mg/kg produced a net increase in spine density over 2 days. Because both behavioural and spine-formation effects plateaued at 1 mg/kg, this dose was used in later experiments. Longer-term imaging suggested that psilocybin’s effect on spine formation was transient at the level of the immediate 2-day window, but detectable over longer intervals. At 7 and 21 days, spine formation remained elevated in psilocybin-treated mice when measured over those longer intervals, and spine elimination increased significantly by 21 days. This meant that the net gain in spine density became smaller as the imaging interval lengthened. When the authors tracked the fate of individual spines, newly formed spines after psilocybin were more likely to survive than newly formed spines in controls. In control mice, 40.6 ± 3.0% of new spines survived to day 7 and 29.3 ± 4.1% survived to day 21. After psilocybin, survival rose to 65.7 ± 5.6% at day 7 and 51.0 ± 4.0% at day 21. By contrast, pre-existing spines in psilocybin-treated mice were less likely to survive than comparable spines in control mice. Morphology analysis suggested faster maturation of new spines. In controls, most new spines were stubby (65.6 ± 2.9%), few were mushroom-shaped (2.2 ± 2.2%), and the rest were thin. In psilocybin-treated mice, the proportion of new spines that were stubby fell to 46.5 ± 4.8%, while mushroom spines rose to 28.4 ± 1.8%. The authors report that mushroom spines were relatively stable, consistent with more mature synaptic structures. Genetic manipulation of 5-HT 2A R produced a dissociation between HTRs and neuroplasticity. In the full knock-out mice, psilocybin did not induce HTRs and did not change spine formation or elimination. In the conditional rescue mice, receptor expression was restored only in layer 5 pyramidal neurons, yet psilocybin still failed to induce HTRs. However, these rescue mice did show psilocybin-induced increases in spine formation and net spine density gain, although the effect was smaller than in wild-type mice. Newly formed spines were more likely to persist in rescue mice than in full knock-out mice, but pre-existing spine survival stayed at baseline levels. In the conditional knock-out mice, where 5-HT 2A R was deleted specifically from layer 5 pyramidal neurons, psilocybin still induced HTRs, showing that this receptor population was not required for the behavioural response. But psilocybin no longer altered spine formation, elimination, or spine density. Overall, the results indicate that 5-HT 2A R signalling in cortical layer 5 pyramidal neurons is necessary and sufficient for the neuroplastic effects measured here, but neither necessary nor sufficient for psilocybin-induced HTRs.

The authors interpret their findings as showing that psilocybin has separable hallucinogenic and neuroplastic effects at the cellular level. They argue that receptor signalling in cortical layer 5 pyramidal neurons drives the structural plasticity response, including new spine formation, preferential stabilisation of new spines, and faster maturation towards mushroom-shaped spines. At the same time, the HTR cannot be explained solely by 5-HT 2A R activity in this neuron class, because restoring receptors only in these cells did not bring back HTRs, while deleting them did not abolish the response. They place these results in the context of earlier work that had reported mixed findings on whether 5-HT 2A R signalling is required for psychedelic-induced plasticity. Baker and colleagues present their cell type-specific rescue and knock-out strategy as clarifying that layer 5 pyramidal neurons are central for psilocybin-triggered synaptic remodelling. They also note that psilocybin increased elimination of pre-existing spines in wild-type mice, but not in the rescue or knock-out lines, suggesting that reorganisation of older circuits may depend on interactions between layer 5 pyramidal neurons and other cell types or circuits. The authors further interpret the longer-term spine data as suggesting a more complex, possibly homeostatic, pattern than a simple sustained increase in spine density. They note that spine density tended to drift back towards baseline over time, which they say may depend on the context in which psilocybin is administered, such as normal versus pathological brain states. Key limitations and uncertainties acknowledged in the discussion include that the HTR is only an indirect proxy for hallucinations, and that the precise circuits and cell types responsible for the behavioural response remain unresolved. They also caution that their findings may not generalise across all cortical excitatory neurons, given that their rescue and knock-out manipulations were focused on layer 5 pyramidal cells. The authors conclude that their work supports the idea that it may be possible to develop neuroplasticity-promoting therapies without unwanted psychotropic effects, although they present this as a longer-term implication rather than a demonstrated outcome.