This mouse study found that psilocybin changes inhibitory signalling in the medial frontal cortex by reducing the activity of somatostatin interneurons and increasing parvalbumin interneuron activity. The effects appear to act through 5-HT1A receptors on somatostatin cells and may contribute to psilocybin’s longer-lasting behavioural effects.
Psychedelics show therapeutic potential for treating psychiatric disorders. While studies have emphasized the roles of cortical pyramidal cells, GABAergic neurons also express serotonin receptors and are therefore likely targets of psychedelics. In this study, we determine the effect of psilocybin on the activity dynamics of major GABAergic cell types in the mouse medial frontal cortex. Psilocybin reduces the firing of somatostatin-expressing interneurons, but increases the activity of parvalbumin-expressing interneurons. This cell type-specific response is unlikely to involve vasoactive intestinal peptide-expressing interneurons. Instead, pharmacological blockade and conditional knockout experiments demonstrate that psilocybin acts on the 5-HT1A receptor at SST interneurons, which contributes to the drug's long-term behavioral effects. Collectively, the results reveal that the classic psychedelic psilocybin alters cortical inhibition in a cell type-specific manner.
Psilocybin is increasingly studied as a potential treatment for psychiatric disorders, and clinical trials have reported rapid antidepressant effects that may last for weeks to months. However, the neurobiology underlying these effects remains incomplete. Earlier work has largely focused on excitatory pyramidal neurons, even though cortical GABAergic interneurons also express serotonin receptors and are likely to be direct targets of psychedelics. The authors note that previous studies and transcriptomic data had hinted at interneuron involvement, but there was still no clear in vivo picture of how psilocybin affects the firing dynamics of major interneuron classes, especially in a clinically relevant compound being developed for human use. Davoudian and colleagues therefore set out to determine how psilocybin alters the activity of the main inhibitory interneuron populations in the mouse medial frontal cortex. They aimed to test whether psilocybin differentially affects somatostatin-expressing (SST), parvalbumin-expressing (PV), and vasoactive intestinal peptide-expressing (VIP) interneurons, and to identify the receptor mechanism responsible for any SST effect. They also examined whether the relevant interneuron receptor contributes to psilocybin’s behavioural effects, particularly on stress-related paradigms. The study combines cell type-specific extracellular electrophysiology, two-photon calcium imaging, pharmacological blockade, conditional knockout experiments, and behavioural testing to link circuit-level effects with outcomes relevant to stress and antidepressant-like responses.
Papers cited by this study that are also in Blossom
Nutt, D. J., Erritzoe, D., Carhart-Harris, R. L. · Cell (2020)
Vollenweider, F. X., Preller, K. H. · Nature Reviews Neuroscience (2020)
Carhart-Harris, R. L., Giribaldi, B., Watts, R. et al. · New England Journal of Medicine (2021)
Goodwin, G. M., Aaronson, S. T., Alvarez, O. et al. · New England Journal of Medicine (2022)
The researchers studied awake, head-fixed mice and focused on the medial frontal cortex, specifically the ACAd and medial MOs regions. For electrophysiology, they used Neuropixels 1.0 probes to record spontaneous spiking before and after intraperitoneal psilocybin (1 mg/kg) or saline. They used Cre-dependent ChR2 reporter lines to opto-tag SST neurons in Sst Cre;Ai32 mice and PV neurons in Pvalb Cre;Ai32 mice, allowing genetically identified units to be distinguished during recording. Recording sessions included a 30-minute pre-drug baseline followed by a 60-minute post-drug period. Units were isolated by spike sorting and manual curation, and only cells meeting predefined quality criteria were included. For VIP interneurons, where opto-tagging was not successful, the authors switched to two-photon calcium imaging. Vip Cre mice received AAV-Flex-GCaMP6f in the medial frontal cortex, and somatic calcium transients were recorded through a chronic glass window in layer 2/3. Mice received both psilocybin and saline in a counter-balanced within-subject design separated by at least one week. Similar imaging approaches were also used for SST and PV interneurons in layer 2/3 to compare cell type-specific effects. Calcium event rates were extracted from fluorescence time courses after motion correction and neuropil subtraction. To test mechanism, the authors analysed publicly available single-cell transcriptomic data from the Allen Institute to compare serotonin receptor expression across VIP, SST, and PV interneurons. They then used WAY-100635, a 5-HT 1A receptor antagonist, administered systemically 10 minutes before psilocybin, and also generated SST-1A cKO mice by crossing Sst Cre mice with Htr1a fl/fl mice to delete 5-HT 1A receptors selectively from SST interneurons. RNAscope fluorescence in situ hybridisation was used to verify the extent of knockout. To assess behavioural relevance, the researchers tested fear extinction, extinction retention, fear renewal, tail suspension, and head-twitch response. Psilocybin or saline was given systemically, and behavioural assays were separated by at least two weeks in cohorts that underwent multiple tests. Mixed-effects models were used for electrophysiology and imaging, while two-factor ANOVAs and post hoc comparisons were used for behavioural outcomes.
Psilocybin reduced the spiking of SST interneurons in the medial frontal cortex. In Sst Cre;Ai32 mice, 27 opto-tagged SST cells were recorded from 9 psilocybin-treated animals and 12 opto-tagged cells from 3 saline-treated animals. SST firing fell from 5.5±1.1 Hz pre-psilocybin to 3.7±0.8 Hz post-psilocybin, whereas saline produced no clear change (5.0±1.1 Hz to 3.9±0.8 Hz). The mixed-effects model showed a significant treatment effect (P = 0.028). Using a z-score threshold of -2, 18.5% of SST neurons showed a substantial reduction after psilocybin, compared with none after saline. By contrast, PV interneurons showed the opposite pattern. In Pvalb Cre;Ai32 mice, 27 opto-tagged PV cells were recorded after psilocybin and 26 after saline. PV firing increased from 6.1±1.4 Hz pre-drug to 11.6±1.6 Hz post-psilocybin, while saline showed a smaller change (11.3±2.1 Hz to 13.2±2.0 Hz). The treatment effect in the mixed-effects model was described as a trend rather than conventionally significant (P = 0.062). About half of PV interneurons showed a substantial rise in firing after psilocybin. Transcriptomic analysis suggested that VIP interneurons predominantly express Htr2c, SST interneurons preferentially express Htr1a, and PV interneurons primarily express Htr2a. In the calcium imaging experiments, psilocybin did not measurably change VIP interneuron calcium event rates (psilocybin: -3.3±2.2%; saline: -2.8±2.6%; P = 0.8). In contrast, SST calcium event rates decreased after psilocybin (psilocybin: -10.5±6.7%; saline: 26.1±13.3%; P = 0.003), consistent with the electrophysiology. PV calcium event rates in layer 2/3 did not change appreciably (psilocybin: 15.8±8.8%; saline: 15.0±8.1%; P = 0.98). When the 5-HT 1A receptor was blocked with WAY-100635, psilocybin no longer suppressed SST firing. In fact, SST activity slightly increased after antagonist plus psilocybin (7.8±2.1% change; P = 0.028). In SST-1A cKO mice, RNAscope confirmed a marked reduction in Htr1a-positive SST cells compared with controls, although the knockout was incomplete. In these knockout mice, psilocybin had no effect on SST firing (P = 0.4). Behaviourally, psilocybin reduced freezing during fear extinction and extinction retention in control mice, but these effects were absent in SST-1A cKO mice (interaction P = 0.002 for extinction; P = 0.001 for retention). There was no interaction for fear renewal, although both treatment and genotype had main effects. In the tail suspension test, psilocybin reduced immobility in control mice 24 hours later, but this effect was absent in SST-1A cKO mice (interaction P = 0.035). In contrast, psilocybin robustly increased head-twitch responses, and this acute effect was unchanged by SST-specific 5-HT 1A receptor deletion (interaction P = 0.98).
The authors interpret the findings as evidence that psilocybin differentially modulates major cortical inhibitory cell types rather than acting only on pyramidal neurons. They argue that the drug suppresses SST interneurons while increasing PV interneuron firing in the medial frontal cortex, thereby shifting inhibition along the somatodendritic axis of pyramidal cells. In their model, reduced SST activity would relieve dendritic inhibition and increase dendritic excitability, whereas increased PV activity would strengthen perisomatic inhibition and restrain excessive spike output. Davoudian and colleagues position the SST result as mechanistically important because pharmacological blockade and conditional knockout both point to 5-HT 1A receptors on SST interneurons as the key mediator of psilocybin’s suppressive action on this cell class. They further argue that these SST 5-HT 1A receptors are functionally relevant, because deleting them impaired psilocybin’s longer-term beneficial effects in stress-related behavioural assays, including fear extinction and tail suspension. At the same time, the acute head-twitch response was preserved, suggesting that this receptor population is not required for that immediate psychedelic-like behavioural readout. The authors relate their findings to earlier work showing that transient dendritic disinhibition can facilitate plasticity during learning and memory. They suggest psilocybin may engage a similar mechanism. They also note that the results are broadly consistent with reports of increased dendritic calcium transients after psilocybin. However, they emphasise that the effects may vary by brain region and by finer interneuron subtype. For example, they discuss possible differences in hippocampal circuitry and the possibility that some SST subtypes may respond differently. Several limitations and uncertainties are acknowledged. The study focused on the medial frontal cortex, so the findings may not generalise to other regions. The authors also note that psilocybin did not uniformly suppress all SST neurons and that blocking 5-HT 1A receptors sometimes unmasked increased SST activity, possibly because a subset of SST neurons expresses 5-HT 2A receptors. For VIP interneurons, the absence of an effect may reflect regional differences, the complexity of serotonergic signalling, or masking by reduced serotonergic drive from dorsal raphe neurons. More broadly, the authors argue that future work should define cell type- and subtype-specific responses to psychedelics in greater detail. They also contrast the psilocybin pattern with ketamine, noting that although both can promote dendritic disinhibition, ketamine-like NMDA receptor antagonists reduce perisomatic inhibition rather than increasing PV activity.
We recorded spontaneous spiking activity from genetically identified SST interneurons in the medial frontal cortex of awake, head-fixed mice. We used Sst Cre ;Ai32 mice, which were generated by crossing a Sst Cre mousewith an Ai32 reporter mouse, resulting in Credependent expression of ChR2 in SST interneurons. We performed a craniotomy in the mouse frontal cortex above the ACAd and medial MOs region of the medial frontal cortex and inserted a high-density Neuropixels 1.0 electrodeinto the brain (Figure). We recorded for 30 minutes, injected psilocybin (1 mg/kg, i.p.) or saline via a catheter, and then recorded for another 60 minutes (Figure). The catheter allowed for drug delivery without needle insertion during the recording, which could agitate the animal to cause motion artifacts. At the end of each recording session, we performed "opto-tagging", where we used an optical fiber to deliver trains of brief laser pulses (20-50 ms, 473 nm) onto the cortical surface to elicit spiking in the ChR2-expressing cells. Animals would be sacrificed after the recording for histology. The Neuropixels probe was coated with the fluorescent dye CM-DiI, allowing it to be visualized in fixed brain sections post hoc to confirm placement in the correct brain region (Figureand). (E) Spike raster of an opto-tagged SST interneuron from a Sst Cre ;Ai32 mouse. Blue, period of laser stimulation. Inset, the spike waveform, mean ± s.d. (F) The latency to the first spike after the onset of laser stimulation for opto-tagged SST interneurons. n = 39 cells from 12 mice. (G) Firing rate before and after psilocybin administration, averaged across opto-tagged SST interneurons. Line, mean. Shading, ± SEM. Dashed line, mean firing rate during pre-drug period. n = 27 cells from 9 mice. (H) Left, the activity of opto-tagged SST interneurons before and after saline administration. The firing rate of each neuron was converted to a z-score by normalization based on its pre-drug firing rate. Right, mean firing rates during pre-and post-drug periods for the opto-tagged SST interneurons. The axes use a symmetric logarithmic scale, which includes a linear region (0 to 0.03 Hz) that transitions to a logarithmic region (>0.03 Hz). Each dot represents one neuron. n = 12 cells from 3 mice. We recorded from 12 Sst Cre ;Ai32 animals: 9 mice were given psilocybin (4 males, 5 females) and 3 mice were given saline (1 male, 2 females). We isolated single units by automated spike sorting, manual curation, and then included those that passed four quality metrics: presence ratio, inter-spike-interval violations ratio, amplitude cutoff, and isolation distance (Figure). For psilocybin, we had 1021 cells including 27 opto-tagged SST interneurons (range: 82-330 cells and 1-5 opto-tagged units per animal). For saline, we had 212 cells including 12 optotagged SST interneurons (range: 65-77 cells and 1-8 opto-tagged units per animal). Figureshows the spike raster of an opto-tagged SST interneuron, which was driven to spike by the laser photostimulation. Across all opto-tagged SST interneurons, latency from laser onset to first spike was 7.9±0.4 ms (mean ± SEM; Figure) and the baseline firing rate during pre-drug period was 5.4±0.8 Hz (mean ± SEM). After the administration of psilocybin, SST interneurons on average reduced their spiking activity (Figure). Comparing their firing rates between the pre-drug (-30 to 0 minute) and post-drug period (0 to 60 minute), there was no change after saline injection (pre-saline: 5.0±1.1Hz, postsaline: 3.9±0.8 Hz; Figure), but there was a significant decrease in spike rate after psilocybin administration (pre-psilocybin: 5.5±1.1 Hz, post-psilocybin: 3.7±0.8 Hz; Figure). Some cells were recorded from the same mouse, therefore we fit the drug-evoked change in firing rate for SST interneurons using a mixed effects model, which accounted for the hierarchical structure of the data (main effect of treatment: P = 0.028, mixed effects model; see Methods). If we consider a mean z-score of -2 during the post-drug period to be a substantial decrease in spike rate, then 18.5% of the recorded SST interneurons reduced firing activity following psilocybin administration, whereas none of the recorded SST interneurons exhibited such change after saline (psilocybin: 5 cells with Z < -2, 0 cell with Z > 2, out of 27 SST interneurons; saline: 0 cell with Z < -2, 1 cell with Z > 2, out of 12 SST interneurons). These results show that psilocybin administration led to an acute reduction in the spiking activity of SST interneurons in the mouse medial frontal cortex.
In addition to SST interneurons which constitute ~20% of GABAergic cells in the mouse frontal cortex, PV interneurons are the other major population that accounts for ~40% of the inhibitory interneurons. We therefore next asked how psilocybin affects the spiking activity of PV interneurons. We generated Pvalb Cre ;Ai32 mice by crossing a Pvalb Cre mouse (36) with a Ai32 reporter mouse for Cre-dependent expression of ChR2 in PV interneurons. We used the same Neuropixels and opto-tagging approach, recording before and after the administration of psilocybin (1 mg/kg, i.p.) or saline. We recorded from 6 Pvalb Cre ;Ai32 animals: 3 mice were given psilocybin (2 males, 1 female) and 3 mice were given saline (1 male, 2 females). For psilocybin, we had 338 cells including 27 opto-tagged PV interneurons (range: 99-127 cells and 5-11 opto-tagged units per animal). For saline, we had 504 cells including 26 opto-tagged PV interneurons (range: 106-205 cells and 4-17 opto-tagged units per animal; Figure). Across all opto-tagged PV interneurons, the latency from laser onset to first spike was 6.0±0.3 ms (mean ± SEM; Figureand) and the baseline firing rate during pre-drug period was 8.6±1.3 Hz (mean ± SEM). In contrast to what we observed in SST interneurons, the PV interneurons had an increase in firing rate that occurred ~15 minutes after psilocybin injection (Figure). Comparing the preand post-drug firing rates revealed a significant elevation of spike rate after the administration of psilocybin (pre-saline: 11.3±2.1 Hz, post-saline: 13.2±2.0 Hz; pre-psilocybin: 6.1±1.4 Hz, postpsilocybin: 11.6±1.6 Hz; main effect of treatment: P = 0.062, mixed effects model; Figureand 1N, S4, S5). About half of the recorded PV interneurons had a substantial rise in spiking activity after psilocybin administration (psilocybin: 14 cells with Z > 2, 1 cell with Z < -2, out of 27 PV interneurons). Based on probe reconstruction, the majority of recorded SST and PV interneurons were located in layer 5 (Figure). Therefore, psilocybin has opposing effects on the spike rate of the two major populations of GABAergic neurons in the medial frontal cortex.
We were particularly interested in psilocybin's suppressive effect on the SST interneurons. SST interneurons target the dendritic compartments of pyramidal cells, so a reduction in their activity should disinhibit dendrites. This can result in a higher level of dendritic electrogenesis, potentially enabling psilocybin's effect on structural neural plasticity. To gain insight into the potential mechanism, we analyzed the publicly available single-cell sequencing data set from the Allen Institute for Brain Science. We examined transcripts encoding serotonin receptor subtypes, Htr1a, Htr2a, Htr2b, Htr2c, and Htr3a, in VIP, SST, and PV interneurons (Figure). The expression of serotonin receptor transcripts in the mouse frontal cortex has cell typespecific preference. VIP interneurons predominantly express Htr2c (65% of cells, relative to 6% and 8% for Htr1a and Htr2a). SST interneurons express Htr1a more than other subtypes (39% of cells, relative to 18% and 15% for Htr2a and Htr2c). PV interneurons primarily express Htr2a (47% of cells, relative to 4% and 35% for Htr1a and Htr2c). VIP interneurons additionally express Htr3a, but should not contribute to drug action because psilocin has negligible binding to this ionotropic serotonin receptor. Htr2c encodes the 5-HT 2C receptor, thought to be a G q -coupled receptor that increases neuronal excitability. By contrast, Htr1a encodes the 5-HT 1A receptor, which is a G i -coupled receptor known to decrease neuronal excitability. Thus, there are two plausible mechanisms for psilocybin's suppressive effects on SST interneurons. One, psilocybin may act via 5-HT 2C receptors at VIP interneurons, which would increase their firing, leading to inhibition of SST interneurons. Two, psilocybin may act via 5-HT 1A receptors at SST interneurons to inhibit directly their firing. These mechanisms are not mutually exclusive. To evaluate the two mechanisms, we first measured the drug-evoked activity change in VIP interneurons. We initially tried using the Neuropixels and opto-tagging approach, but were unable to reliably find opto-tagged units in Vip Cre ;Ai32 mice (n = 7 animals, 616 cells with no opto-tagged unit; data not shown). We therefore pursued an alternative strategy using twophoton microscopy to image somatic calcium transients. To express a genetically encoded calcium indicator in VIP interneurons, we injected AAV-CAG-Flex-GCaMP6f into the medial frontal cortex of Vip Cre mice. Through a chronically implanted glass window, we imaged GCaMP6f-expressing VIP interneurons in layer 2/3 of the medial frontal cortex of awake, headfixed mice (Figure). We imaged for 30 minutes, administered psilocybin (1 mg/kg, i.p.) or saline, and then imaged for another 90 minutes. Each mouse would receive both psilocybin and saline, delivered one week apart with the order randomized in a counter-balanced design. From the time-lapse images, we extracted fluorescence transients from each cell body and used an automated algorithm to quantify the calcium event rate for each cell (see Methods). Figureshows the somatic calcium transients from three VIP interneurons before and after the administration of saline or psilocybin. We imaged 6 Vip Cre animals (2 males, 4 females) to measure from 53 VIP interneurons (range: 4-11 cells per animal). We did not detect any effect of psilocybin on the calcium event rates in VIP interneurons (psilocybin: -3.3±2.2%; saline: -2.8±2.6%; main effect of treatment: P = 0.8, mixed effects model; Figureand). For completeness, we also imaged somatic calcium transients from SST and PV interneurons in layer 2/3 of the medial frontal cortex using Sst Cre and Pvalb Cre mice, respectively (Figureand). In line with our electrophysiology results, we observed a reduced rate of calcium events in SST interneurons following psilocybin administration (psilocybin: -10.5±6.7%; saline: 26.1±13.3%; n = 117 cells from 7 animals including 5 males, 2 females; main effect of treatment: P = 0.003, mixed effects model; Figure). However, we did not see appreciable changes in calcium event rate in PV interneurons following psilocybin administration (psilocybin: 15.8±8.8%; saline: 15.0±8.1%; n = 135 cells from 5 animals including 3 males, 2 females; main effect of treatment: P = 0.98, mixed effects model;
). The discrepancy between the electrophysiological and imaging results for PV interneurons is likely because two-photon imaging recorded from layer 2/3 cells, missing the deeper neurons that were preferentially captured by the Neuropixels probe (Figure). Nevertheless, the imaging data demonstrate that VIP interneurons did not have appreciable change in activity after psilocybin administration. Therefore, VIP interneurons are unlikely the reason for the psilocybin-induced suppression of SST interneurons.
SST interneurons express 5-HT 1A receptors, which are G i -coupled receptors that should reduce neuronal excitability. To determine if the 5-HT 1A receptor is involved in psilocybin's drug action, we used the same Neuropixels and opto-tagging approach to record from SST interneurons in the medial frontal cortex of Sst Cre ;Ai32 mice, with an additional step to inject the 5-HT 1A receptor antagonist WAY-100635 (2 mg/kg, i.p., administered via a catheter) 10 minutes prior to psilocybin administration (1 mg/kg, i.p., administered via a second catheter) or saline (Figureand). For saline, we recorded from 6 animals (4 males, 2 females) to obtain 1582 cells including 22 opto-tagged SST interneurons (range: 107-599 cells and 1-8 optotagged units per animal). The mean firing rates for frontal cortical SST interneurons were 6.9±1.0 Hz and 5.9±1.0 Hz before and after saline, respectively (Figure). For psilocybin, we recorded from 6 animals (5 males, 1 female) to obtain 692 cells including 33 opto-tagged SST interneurons (range: 58-145 cells and 1-8 opto-tagged units per animal). In these animals pretreated with the 5-HT 1A receptor antagonist, psilocybin no longer reduced the firing of cortical SST interneurons (pre-psilocybin: 7.4±1.2 Hz, post-psilocybin: 7.9±1.0 Hz; Figure). In fact, there was now a slight but significant increase in spike rate of SST interneurons after WAY-100635 and psilocybin (7.8±2.1%, mean ± SEM; main effect of treatment: P = 0.028, mixed effects model). The results support a role for the 5-HT 1A receptor. However, the experiment has a couple caveats. Although WAY-100635 is a well-characterized 5-HT 1A receptor antagonist, the compound has off-target effect at the dopamine D4 receptor. Moreover, WAY-100635 was administered systemically, so it can affect numerous cell types including cortical excitatory neurons that also express 5-HT 1A receptors. (H) AAV was used to induce expression of ChR2 in SST interneurons in the medial frontal cortex of SST-1A cKO mice. Spiking activity of ChR2-expressing SST interneurons was recorded using a Neuropixels probe.
The experimental timeline includes pre-drug (-30 -0 min) and post-drug (0 -60 min) periods. (J) Left, the activity of opto-tagged SST interneurons before and after saline administration in SST-1A cKO mice. The firing rate of each neuron was converted to a z-score by normalization based on its pre-drug firing rate. Right, mean firing rates during pre-and post-drug periods for the opto-tagged SST interneurons. The axes use a symmetric logarithmic scale, which includes a linear region (0 to 0.03 Hz) that transitions to a logarithmic region (>0.03 Hz). Each dot represents one neuron. n = 12 cells from 6 animals. To test further the role of the 5-HT 1A receptor, we leveraged a Htr1a fl/fl mouse line that was generated in a prior study. We generated Sst Cre/+ ;Htr1a fl/fl mice (henceforth referred to as SST-1A cKO) and Sst +/+ ;Htr1a fl/fl littermates as controls (Figure). The original study validated the conditional knockout using immunohistochemistry and slice electrophysiology. To characterize the knockout in our experimental context, we used RNAscope fluorescence in situ hybridization to visualize Htr1a and Sst transcripts around DAPI-positive nuclei in the medial frontal cortex (Figure). In control animals, 44±4% of Sst-expressing cells contained Htr1a transcripts (mean ± SEM; n = 1626 SST cells from 6 animals including 4 females, 2 males;
). This percentage is close to the value derived from the single-cell sequencing data, which showed Htr1a transcript in 39% of the frontal cortical SST interneurons (Figure). By contrast, in SST-1A cKO animals, only 14±5% of Sst-expressing cells contained Htr1a transcripts (n = 1396 SST cells from 4 animals including 3 males, 1 female; P = 0.0095, twotailed Wilcoxon rank-sum test). There are a couple reasons why the knockout was substantial but incomplete. Because the Htr1a gene lacks introns, the floxed construct placed loxP sites flanking the entire coding sequence, spanning a large distance of 1.3 kb. In addition, the chromatin state in frontal cortical SST interneurons may limit Cre recombinase access to the loxP sites. Both factors could contribute to reduced Cre-mediated recombination efficiency. The characterization confirmed that the Htr1a fl/fl conditional knockout mice can be used for cell typespecific manipulations in the mouse medial frontal cortex. We injected AAV-EF1a-DIO-hChR2(H134R)-EYFP into the medial frontal cortex to express ChR2 selectively in SST interneurons of SST-1A cKO animals (Figure). We used the Neuropixels and opto-tagging approach to record from SST interneurons, which now lacked Htr1a expression (Figure). For saline, we recorded from 6 animals (2 males, 4 females) to obtain 886 cells including 12 opto-tagged SST interneurons (range: 76-261 cells and 1-4 optotagged units per animal). There was no effect of saline on the mean firing rate of SST interneurons in SST-1A cKO mice (pre-saline: 5.2±1.4 Hz, post-saline: 5.1±1.0 Hz; Figure). For psilocybin, we recorded from 11 animals (7 males, 4 females) to obtain 927 cells including 25 opto-tagged SST interneurons (range: 17-200 cells and 1-5 opto-tagged units per animal). In mice with cell type-specific deletion of 5-HT 1A receptors, psilocybin had no effect on the activity of SST interneurons in the medial frontal cortex (pre-psilocybin: 10.4±3.2 Hz, postpsilocybin: 13.2±2.8 Hz; main effect of treatment: P = 0.4, mixed effects model; Figure). Thus, results from the pharmacological blockade and conditional knockout experiments converge to pinpoint the 5-HT 1A receptor on SST interneurons as an important target for psilocybin's drug action on SST interneurons.
We next asked if the 5-HT 1A receptors on SST interneurons may play a role in psilocybin's acute and long-term behavioral effects. We tested SST-1A cKO and control mice on three behavioral assays, with the tests spaced at least 2 weeks apart to allow for drug washout. Fear extinction is a preclinical behavioral assay that provides insight into stress-related psychopathology. Psilocybin reduces freezing not only in the extinction session that immediately follows, but also in the extinction retention and fear renewal sessions days later. For the current experiment, on day 1, we paired auditory tones with foot shocks for fear conditioning in context A (Figure; see Methods). On day 3, we administered psilocybin (1 mg/kg, i.p.) or saline, and then 30 minutes later had the first extinction session where only auditory tones were presented in context B. On day 4, we tested extinction retention by performing a second extinction session in context B. Finally, on day 12, we tested fear renewal, which is an extinction session performed in context C. Both SST-1A cKO and control mice readily acquired conditioned freezing (main effect of tones: P < 0.001, mixed-effects model; n = 14-20 mice per group; Figure). In control mice, a single dose of psilocybin (1 mg/kg, i.p.) significantly reduced freezing during fear extinction compared to saline treatment. However, this psilocybin-facilitated extinction was absent in SST-1A cKO mice (interaction effect of treatment and genotype: P = 0.002, two-factor ANOVA; Figure). Similarly, psilocybin's effect on freezing during extinction retention was abolished in SST-1A cKO mice (interaction effect of treatment and genotype: P = 0.001, twofactor ANOVA; Figure). There was no interaction effect detected for fear renewal (main effect of treatment: P = 0.004, main effect of genotype: P = 0.004, two-factor ANOVA; Figure). The tail suspension test measures stress-induced behavioral despair and is used for antidepressant screening (48) (Figure). Psilocybin significantly reduced immobility time 24 hours after administration compared to saline treatment in control mice, but this effect was absent in SST-1A cKO animals (interaction effect of treatment and genotype: P = 0.035, twofactor ANOVA; n = 18-24 per group; Figure). We note that saline-treated SST-1A cKO animals had a low level of immobility, which may occluded detection of any further decrease induced by psilocybin. This resilient phenotype at baseline for the SST-1A cKO animals is consistent with earlier work demonstrating that constitutively disinhibiting SST interneurons leads to an antidepressive-like state. Finally, the head-twitch response is an assessment of the acute drug action of psychedelics, with the head-twitch potency of a compound in mice relating to its hallucinogenic potency in humans (50) (Figure). As expected, psilocybin induced robust increase in the number of head twitches in control mice, and this acute behavioral effect was unaffected by the knockout of 5-HT 1A receptors in SST interneurons (interaction effect of treatment and genotype: P = 0.98, two-factor ANOVA; n = 7-16 per group; Figure). Collectively, these results demonstrate that the 5-HT 1A receptors on SST interneurons modulate psilocybin's long-term effects on stressrelated behaviors, but do not contribute to its acute head-twitch response in mice.
This study reveals that psilocybin differentially affects the firing activity of GABAergic cell populations. In the medial frontal cortex, psilocybin reduced the spiking activity of SST interneurons while increasing the firing of PV interneurons. Pharmacological blockade and conditional knockout experiments demonstrate that the psilocybin-induced suppression of SST interneurons is mediated by the 5-HT 1A receptors expressed on the SST interneurons. This modulation of GABAergic inhibition is functionally important, because deleting the 5-HT 1A receptors from SST interneurons impaired the long-term ameliorating effects of psilocybin on stress-related behaviors. In adult animals, learning is accompanied by a dynamic reorganization of cortical inhibitory tone, marked by cell type-specific changes in interneuron activity. In particular, SST interneurons target the dendrites of pyramidal cells to regulate excitability. Transient reduction of SST interneuron activity and the resulting disinhibition have emerged as an important mechanism gating synaptic plasticity during learning, including for the encoding of fear memory. Our results suggest that psychedelics may leverage a similar dendritic disinhibition mechanism. We showed that psilocybin suppresses the activity of SST interneurons in this study. The expected downstream consequence is increased dendritic excitability, consistent with recent report of an elevated rate of dendritic calcium transients in the pyramidal tract subtype of pyramidal cells after psilocybin administration. Overall, our findings support a model in which psilocybin shifts inhibition along the somatodendritic axis of cortical pyramidal cells: suppressing SST interneurons to reduce dendritic inhibition, while activating PV interneurons to increase perisomatic inhibition and restrict excessive spiking output. We focused the current study on the medial frontal cortex, because this mouse brain region responds robustly to psilocybin administration, as identified by brain-wide c-Fos mapping. However, the effects of psychedelics are likely to vary across brain regions. For example, psychedelics affect the physiology of the hippocampus, which contains microcircuits composed of PV, SST, and VIP interneurons that are analogous to those in the neocortex. PV interneurons in the ventral hippocampus express 5-HT 2A receptors, which are critical for mediating the acute anxiolytic effects of psychedelics. In contrast to the cortex, hippocampal SST interneurons appear to express also 5-HT 2A receptors preferentially, contributing to slow oscillatory activity. Even within cortical regions, psychedelic effects may vary among finer interneuron subtypes. SST interneurons can be subdivided into at least 8 subtypes based on genetic markers and morphological features. Our results indicate that psilocybin does not uniformly suppress the activity of all SST interneurons. Moreover, when 5-HT 1A receptor was blocked pharmacologically, psilocybin increased the activity of SST neurons, possibly because a fraction of the interneurons expresses 5-HT 2A receptorsand their influence was unmasked by the manipulation. highlighting the need for future studies to classify cell typespecific responses to psychedelic drugs in greater detail. We did not detect psilocybin-evoked activity changes in frontal cortical VIP interneurons. VIP interneurons express the G q -coupled 5-HT 2C receptors, but their effect on cellular excitability can vary by brain region and cell type, as shown in studies of the amygdala, claustrum, ventral hippocampus, and piriform cortex. Several possibilities could explain the lack of effect in frontal cortical VIP interneurons. The G q -coupled pathway in cortical VIP interneurons may not engage ion channels to alter excitability. In addition, VIP interneurons express abundant 5-HT 3 receptors, which are normally activated by endogenous serotonin to increase excitability. Classic psychedelics silence the serotonergic neurons in the dorsal raphe, potentially diminishing serotonergic drive onto VIP interneurons. This could mask any excitatory effects mediated through the 5-HT 2C receptors, resulting in an overall lack of activity change. In summary, our findings illuminate how specific receptors and GABAergic cell types within the cortex orchestrate the complex effects of psilocybin. The results are consistent with psilocybin causing a coordinated shift in the inhibitory tone along the somatodendritic axis of pyramidal cells. Interestingly, NMDA receptor antagonists such as the rapid acting antidepressant ketamine also induce dendritic disinhibition. However, NMDAR receptor antagonists additionally reduce perisomatic inhibition, which differs sharply from the increased PV interneuron activity after psilocybin administration. Future studies will continue to uncover these
Supplementary materials include Figure#1DEX022, Bimeda) and carprofen (5 mg/kg, s.c.; #059149, Covetrus) for anti-inflammatory and analgesic purposes. Anesthesia was induced with 2-3% isoflurane and the mouse was positioned in a stereotaxic apparatus (Model 900, David Kopf Instruments). Anesthesia was maintained with 1-1.5% isoflurane throughout the procedure. Body temperature was kept at 38°C using a far-infrared warming pad (#RT-0515, Kent Scientific). Petrolatum ophthalmic ointment (Optixcare Eye Lube, #062143, Covetrus North America) was applied to protect the eyes. The hair on the head was shaved. An incision was made to remove the skin and the periosteum was cleared. The scalp was disinfected by wiping with ethanol pads and povidoneiodine. We targeted the recording location to be ACAd and medial MOs subregion of medial frontal cortex (AP: +1.7 mm, ML: -0.25 mm, relative to bregma). To reach the target, we would center the craniotomy at ML = -0.5 mm, with intent to insert the probe in the medial portion of the craniotomy. We would mark the spot for the center of craniotomy with ink (e.g., skin-safe pen). A small burr hole was made above the cerebellum using a 0.7 mm burr (#19007-07, Fine Science Tools) and a handheld dental drill (#HP4-917, Foredom). A 0.86 mm self-tapping bone screw (#19010-10, Fine Science Tools) was placed through the skull bone into the cerebellum to serve both as a ground and to provide mechanical support for the head plate implant. A custom stainless steel head plate (fabricated by eMachineShop; design available at) was affixed onto the skull using a quick adhesive cement system (Metabond, #S396, Parkell) while leaving a 5-mm diameter area clear around the planned recording site and also leaving sufficient metal contact in the screw to make contact with reference wire in the future. Any area not covered be Metabond was then covered with silicone elastomer (#10006546, Smooth-On, Inc.). At the end of surgery, animal was given carprofen (5 mg/kg, s.c.) immediately and then again once on each of the following 3 days. The mouse would recover for at least 1 week after surgery before the recording day. Specifically for electrophysiology with SST Cre/+ ;Htr1a fl/fl mice, we required virally mediated expression of ChR2 in SST interneurons. An additional surgery was performed. During this surgery, instead of marking or denting the spot for the medial frontal cortex, a burr hole was made above the ACAd and medial MOs subregion of medial frontal cortex using the burr and handheld dental drill. AAV1-EF1a-DIO-hChR2(H134R)-EYFP-WPRE-HGHpA virus was delivered intracranially into the brain by inserting a pipette pulled from a borosilicate glass capillary and using an injector (Nanoject II Auto-Nanoliter Injector, Drummond Scientific). Injections were done using 4.6 nL pulses with 20 s interval between each pulse. To reduce backflow of the virus, we waited 5-10 min after completing an injection at one site before retracting the pipette to move on to the next site. A total of 4 injections (totaling 200 nL) sites were targeted, corresponding to the vertices of a 0.2 mm-wide square centered at the coordinates for the targeted brain region. Throughout the procedure, the brain surface was kept moist with artificial cerebrospinal fluid (ACSF; in mM: 135 NaCl, 5 HEPES, 5 KCl, 1.8 CaCl2, 1 MgCl2; pH: 7.3). After injections, the craniotomies were covered with silicone elastomer. The skin was sutured (#1265B, Surgical Specialties Corporation). Mice would recover for at least 7 day prior to the surgery to implant the head plate, and we would wait a minimum of 3 weeks to allow for viral-mediated expression. For two-photon imaging, surgeries were performed involving the same pre-and post-operative procedures. A burr hole was made to inject 92 nL of AAV1-CAG-Flex-GCaMP6f-WPRE-SV40 (1:20 diluted in PBS) into the ACAd and medial MOs subregion of medial frontal cortex (AP: 1.5 mm, ML: -0.4 mm, DV: -1.0 mm). After 2-3 weeks, the mouse underwent a second procedure, with the same pre-and post-operative procedures, to implant a glass window for imaging. An incision was made to expose the skull, and the surface was cleaned to remove connective tissues. The dental drill was used to make a ~3-mm-diameter circular craniotomy above the previously targeted location at the medial frontal cortex. ACSF was used to immerse the exposed dura in the craniotomy Coherent) was connected to a patch cable with a 200 μm core, 0.5 NA optical fiber (M104L01, ThorLabs), with an unjacketed end that was mounted on another micromanipulator on an inverted arm in the ring platform system. The optical fiber was moved to position above the medial frontal cortex aimed at the craniotomy. The OpenEphys software was used to trigger a PulsePal (#1102, Sanworks) to drive the analog input in the back panel of the OBIS laser control unit to produce pulses at 1 Hz and ~25 mW/mm 2 per trial. Typical pulse duration was 20 ms, but in a small subset of trials the duration was up to 50 ms. Each trial lasted for 1 s, with inter-trial interval of 950 -980 ms, and at least 500 trials were conducted per session. This trigger from the OpenEphys was split and also routed to the trig input of the IMEC Neuropixels PXIe acquisition module card, so the timing of the laser stimulation can be aligned to the spike times. After completion of an electrophysiological recording, mice were immediately perfused with PBS, followed by paraformaldehyde solution (PFA, 4% (v/v) in PBS). The brains were extracted and further fixed in 4% PFA at 4°C for 12 -24 hr. Subsequently, 40-µm-thick coronal sections were obtained using a vibratome (#VT1000S, Leica) and mounted on slides including Vectashield containing DAPI (#H-1200-10, Vector Laboratories) with glass coverslips. Sections were imaged using a wide-field fluorescence microscope (BZ-X810, Keyence). The Neuropixels probe was cleaned between recordings by soaking in 1% Tergazyme solution (Alconox) overnight, then in DI water, and then isopropyl alcohol (AB07015, AmericanBio) before being allowed to air-dry prior to the next session. Analysis of electrophysiology data. We used the wrapper software SpikeInterfaceto execute an analysis pipeline for preprocessing, spike sorting, and quality metric calculation. Preprocessing included (1) passing the data through a high pass filter with cutoff at 400 Hz, ( detecting and removing bad sites based on noise level, which discard mostly the dorsally positioned recording sites that were not in the brain, (3) phase shift correction, to account for small delays between sites, and (4) common median reference subtraction, to reduce noise particularly the laser-induced artifacts, which were applied to the data from the entire session. Spikes were sorted and putative single units were identified along with drift correction via automated procedures using Kilosort 2.5. The sorted spikes were then manually curated using Phy (). Quality metrics and waveform features were generated via SpikeInterface. The curated units were screened for four quality metrics: a presence ratio of ≥0.9, inter-spike interval violations ratio of <0.5, amplitude cutoff of <0.1, and isolation distance >20. Presence ratio measures the fraction of time during the recording in which a unit has at least one action potential, with time bin width of 60 s. Inter-spike-interval violations ratio measures the fraction of spikes of the unit that occurred in rapid succession within 1.5 ms, relative to the true spikes of the unit. Amplitude cutoff estimates the fraction of missed spikes, i.e. spikes of a unit that was below the spike detection threshold. Isolation distance calculates the distance between a unit and other units based on waveforms. Only manually curated units that additionally met all quality metrics were included for further analysis. To identify opto-tagged neurons, we created peri-stimulus time histograms by aligning spiking activity to the onset of laser stimulation. We classified opto-tagged neurons by considering the latency to spike and reliability of spiking in response to onset of laser stimulation. To compare firing rates during pre-versus post-drug periods, we generated scatter plots with the axes plotted in symlog scale, consisting of a linear region and a logarithmic region with the transition set at 0.03 Hz, using the 'symlog' function in Matplotlib with the linthresh parameter set to 0.03. Histology of electrophysiology data. After electrophysiological experiments, we performed histology and successfully reconstructed probe trajectory in most animals. Mice were perfused with PBS, followed by paraformaldehyde solution (PFA, 4% (v/v) in PBS). The brains were extracted and further fixed in 4% PFA at 4 °C for 12-24 h. Subsequently, 40-µm-thick coronal sections were obtained using a vibratome (VT1000S, Leica) and mounted onto slides with glass coverslips using Vectashield containing DAPI (H-1200-10, Vector Laboratories). The sections were imaged using a wide-field fluorescence microscope (BZ-X810, Keyence). To locate the Neuropixels probe, we used SHARP-TRACKto align the fluorescence images of the coronal sections, including the DiI signals from the probe, with the mouse brain reference atlas in the Allen Common Coordinate Framework. To estimate the location of each unit, we used the channel with the largest voltage deflection in spike waveform. We located 12/12 and 27/27 opto-tagged cells for saline and psilocybin conditions for Sst Cre ;Ai32 animals, and 22/26 and 27/27 opto-tagged cells for saline and psilocybin conditions for Pvalb Cre ;Ai32 animals. Two-photon imaging. Two-photon imaging experiments were performed using a Movable Objective Microscope (MOM, Sutter Instrument) equipped with a resonant-galvo scanner (Rapid Multi Region Scanner, Vidrio Technologies) and a water-immersion 20X objective (XLUMPLFLN, 20x/0.95 N.A., Olympus). ScanImage 2020 software was used to control the microscope for image acquisition. To visualize GCaMP6f, we used an excitation wavelength of 920 nm and emission was collected behind a 475 -550 nm bandpass filter. The laser power measured at the objective was typically ≤120 mW and varied depending on the imaging depth. When imaging the same field of view across days, the laser power was kept the same in each imaging session. Prior to imaging, the mouse was habituated to head fixation in an acrylic tube under the microscope for 3-4 days, increasing the duration each day. We targeted fields of view in the medial portion of the cranial window in order to record from neurons in ACAd and medial MOs. To examine the acute treatment effects, we imaged a single field of view for 30 min to obtain pre-treatment baseline data. Imaging was then paused to inject psilocybin (1 mg/kg, i.p.) or saline (10 mL/kg, i.p.). Almost immediately following injection (<1 min), we imaged the same field of view again for 90 min to acquire post-treatment data. Each animal received both psilocybin and saline, with at least 1 week between imaging sessions and the order of treatment balanced across subjects. The same field of view was imaged for each treatment, providing a precise within-cell estimate of the drug's effect relative to saline. Only cells with robust calcium transients in both imaging sessions were kept for imaging analysis.
The multi-page .tiff image files from each experiment were concatenated and processed with NoRMCorre in MATLAB to correct for non-rigid translational motion. Regions of interest (ROI) corresponding to cell bodies were manually traced using an in-house graphical user interface in MATLAB. For each ROI, the pixel-wise average was calculated at each data frame to generate a fluorescence time course F ROI (t). Since calcium imaging was performed on the same field of view before and after drug injection, a single ROI mask was used to extract calcium signals for all images in an experiment. Next, each ROI was processed to reduce the contribution from background neuropil. Taking each ROI's area and considering a circle with equivalent area that has radius r, a neuropil mask specific to that ROI was created as an annulus with inner radius 2r and outer radius 3r centered on the centroid of the ROI. To exclude neuropil mask pixels that may belong to unselected cell bodies, we calculated the time-averaged signal for each pixel, taking the median amongst pixels in the mask. Pixels were excluded from the neuropil mask if their time-averaged signal was higher than the median. Finally, the remaining pixels in the neuropil mask were averaged per data frame to generate F neuropil (t). Each ROI had the fluorescence from its neuropil mask subtracted as follows: where the neuropil correction factor, c, was set to 0.4. Next, the fractional change in fluorescence ΔF/F(t) was calculated for each ROI by normalizing F(t) against its baseline, F 0 (t), estimated as the 10th percentile within a two-minute sliding window: ∆F/F(t) = (F(t)-F 0 (t))/(F 0 (t)) Calcium events were detected using automated procedure for each cell using a deconvolution "peeling" algorithm. The peeling algorithm uses an iterative template-matching procedure to decompose a ΔF/F(t) trace into a series of elementary calcium events. The template for elementary calcium events was defined on a cell-by-cell basis by calculating the median amplitude and single-exponential decay time constant from calcium transients exceeding 2 standard deviations of the cell's mean signal. Briefly, the algorithm searches a given ΔF/F(t) trace for a match to the template calcium event, subtracts it from the trace (i.e., "peeling"), and successively repeats the matching process until no more event is found. This event detection process outputs the recorded event times with a temporal resolution by the original imaging frame rate. In this way, it is possible to detect multiple calcium events during the same imaging frame (e.g., for large-amplitude transients). For each imaging session, an ROI's calcium event rate was computed by dividing the number of calcium events by the duration of the imaging session. The change in calcium event rate due to drug administration was computed for each ROI using the post-drug minus pre-drug values divided by the pre-drug values.
To quantify the efficiency of the knockout in the SST-1A cKO mice, RNAscope (Advanced Cell Diagnostics, ACD) was conducted using the Multiplex Fluorescent Reagent Kit v2 (#323270) and RNAscope probes for Htr1a (#312301-C2) and Sst (#404631). Mice were anesthetized with isoflurane and transcardially perfused with chilled phosphate buffered saline (PBS), followed by 4% paraformaldehyde (PFA) in PBS. Brains were extracted and fixed in 4% PFA for 24 h at 4°C then moved through a gradient of 10%, 20%, and 30% sterile-filtered sucrose at 4°C, allowing enough time to sink in each solution (16-20 h). Tissue was frozen in optimal cutting temperature (OCT) embedding medium (#4583, Tissue-Tek) and 14 μm coronal sections of the PFC were cut with a cryostat (Thermo Scientific Microm HM 550) and mounted onto ColorFrost Plus slides (#22-230-892, Fisher Scientific). Slides were stored at -80°C until the RNAscope pretreatment protocol and assay. To pretreat the tissue, slides were washed in 1X PBS, baked on a slide warmer at 60°C for 45 min, and post-fixed in 4% PFA at 4°C for 15 min. Sections were dehydrated in 50%, 70%, and 100% ethanol for 5 min each at room temperature, then allowed to dry completely. Hydrogen peroxide (#322335, ACD) was applied to each section and allowed to sit for 10 min at room temperature. Slides were then rinsed in distilled water before incubating in 96°C Target Retrieval Buffer (#322000, ACD) for 8 min. Following a wash in distilled water and an additional dehydration in 100% ethanol for 3 min, slides were dried on the 60°C slide warmer and, once completely dry, barriers were drawn around each tissue section with an ImmEdge Hydrophobic Barrier Pen (#319918, ACD). After the barrier was completely dry, RNAscope Protease III (#322337, ACD) was added to each section and incubated at 40°C for 30 min in a HybEZ™ II Oven (#321711, ACD). Slides were rinsed in distilled water and immediately processed according to the instructions for the RNAscope assay. The C2 (Sst) probe was diluted in the C1 (Htr1a) probe at a 1:50 ratio and 75-125 μl (depending on size of drawn hydrophobic barrier) of the mix was added to completely cover each section. Sections were incubated with the probe mix at 40°C in the HybEZ™ II Oven for 2 h. After Behavioral assays. All behavioral assays were conducted between 10:00 AM and 4:00 PM. To ensure the reproducibility of behavioral findings, assays were performed across multiple cohorts. In one cohort, which included 17 Sst Cre/+ ;Htr1a fl/fl (SST-1A cKO) and 26 sibling Sst +/+ ;Htr1a fl/fl control mice, the same animals underwent all behavioral assays. Both male and female mice were included, and animals were randomly assigned to experimental groups while maintaining a similar sex ratio across groups. The fear extinction test was conducted first, followed by the tail-suspension test two weeks later, and the head-twitch response test another two weeks after that. To minimize potential crosstalk between assays, at least two weeks was maintained between tests for drug washout. For other cohorts, each animal was used in only one behavioral assay. In total, 68 mice underwent the fear extinction test, including 19 control mice treated with saline, Tail suspension test. The tail-suspension test was conducted using a metal hanging apparatus constructed from optical beam parts (Thorlabs). At 24 hr prior to testing, mice received either psilocybin (1 mg/kg, i.p.) or saline (10 mL/kg, i.p.). The following day, in a tall sound-attenuating cubicle (Med Associates), each mouse was suspended by its tail using removable tape, leaving approximately 5 mm of the tail exposed. A small plastic tube was placed around the base of the tail to prevent tail climbing. The suspension lasted for 6 min, during which a video camera (acA1920, Basler) was positioned directly in front of the subject to record movement. The hanging apparatus and cubicle were thoroughly cleaned with 70% ethanol between animals. Immobility time was assessed by a blinded scorer and was defined as periods when the majority of both front and hind limbs remained motionless.
Each mouse received an intraperitoneal injection of either psilocybin (1 mg/kg) or saline (10 mL/kg). Head-twitch responses were measured in groups of 2-3 mice, with psilocybin-and saline-treated mice typically tested simultaneously; however, each mouse was placed in a separate chamber to prevent interactions. Immediately after injection, each mouse was placed into a plexiglass chamber (4'' × 4'' × 4'') with a transparent lid, positioned within a soundattenuating cubicle (Med Associates). A high-speed video camera (acA1920, Basler) was mounted overhead to capture recordings from all chambers simultaneously. Video recordings were collected for 10 min, and chambers were thoroughly cleaned with 70% ethanol between animals. Head-twitch responses were manually scored by an experimenter blinded to the treatment conditions. While we have previously demonstrated that head twitches can be quantified using magnetic ear tags (79), we opted for video recording in this study to avoid potential interference of the ear tags with performance in other behavioral assays. Statistics. Statistical tests were conducted using Python and R studio. For electrophysiological data, mixed effects models were fit for the normalized change in firing rates (post-drug value minus pre-drug value, divided by pre-drug value) using the mixedlm function in statsmodels module in Python. We constructed a model for each cell type with fixed effects using fixed effect of treatment, along with a random effects term (intercept) for nested cell per mouse. For imaging data, mixed effects models were fit for the normalized change in calcium event rates (post-drug value minus pre-drug value, divided by pre-drug value) using the lme4 package in R. We constructed a model for each cell type using fixed effects of treatment, sex, and treatment order, along with a random effects term (intercept) for nested cell per mouse. For behavioral studies, statistical analyses were performed with GraphPad Prism 10.3. In the 4 sessions of fear extinction tests, two-factor ANOVAs were used to assess the interaction between treatment (psilocybin vs. saline) and genotype (control vs. SST-1A cKO) on the percentage of freezing across all CS presentations. Mixed-effects model with Bonferroni's multiple comparisons test was applied to compare psilocybin:control vs. saline:control, as well as psilocybin:SST-1A cKO vs. saline:SST-1A cKO at each tone bin. For the tail-suspension test and head-twitch response, two-factor ANOVAs were used to examine the interaction between treatment (psilocybin vs. saline) and genotype (control vs. SST-1A cKO) on immobility percentage and the number of head-twitch responses. Post hoc comparisons with Bonferroni correction were used to assess differences between psilocybin:control and saline:control, as well as psilocybin:SST-1A cKO and saline:SST-1A cKO. Data and code availability. The data that support the findings and the code used to analyze the data in this study will be made publicly available at.
Create a free account to open full-text PDFs.
Raison, C. L., Sanacora, G., Woolley, J. D. et al. · JAMA (2023)
Schindowski, E. M., Jungwirth, J., Schuldt, A. et al. · EClinicalMedicine (2023)
Shao, L-X,, Liao, C., Gregg, I. et al. · Neuron (2021)
Savalia, N., Shao, L-X,, Kwan, A. C. · Trends in Neuroscience (2021)
Halberstadt, A. L., Chatha, M., Klein, A. K. et al. · Neuropharmacology (2020)
Shao, L-X,, Tan, D., Liao, C., Davoudian, P. A. et al. · Nature Communications (2025)