Psilocybin modulates functional connectivity of the amygdala during emotional face discrimination

This randomized, double-blind, placebo-controlled, crossover study (n=18) analyzed the acute effects of psilocybin (11.2mg/70kg) on brain activity and connectivity during the perceptual discrimination of emotional faces in healthy participants. Psilocybin decreased connectivity between the right amygdala and the right frontal pole while processing happy faces and decreased connectivity between the left striatum and the right amygdala while processing angry faces, thereby acting as a key modulator of the amygdala and emotional processing.

Abstract

“Introduction: Recent studies suggest that the antidepressant effects of the psychedelic 5-HT2A receptor agonist psilocybin are mediated through its modulatory properties on prefrontal and limbic brain regions including the amygdala.

Methods: To further investigate the effects of psilocybin on emotion processing networks, we studied for the first-time psilocybin’s acute effects on amygdala seed-to-voxel connectivity in an event-related face discrimination task in 18 healthy volunteers who received psilocybin and placebo in a double-blind balanced cross-over design. The amygdala has been implicated as a salience detector especially involved in the immediate response to emotional face content. We used beta-series amygdala seed-to-voxel connectivity during an emotional face discrimination task to elucidate the connectivity pattern of the amygdala over the entire brain.

Results: When we compared psilocybin to placebo, an increase in reaction time for all three categories of affective stimuli was found. Psilocybin decreased the connectivity between amygdala and the striatum during angry face discrimination. During happy face discrimination, the connectivity between the amygdala and the frontal pole was decreased. No effect was seen during discrimination of fearful faces.

Discussion: Thus, we show psilocybin’s effect as a modulator of major connectivity hubs of the amygdala. Psilocybin decreases the connectivity between important nodes linked to emotion processing like the frontal pole or the striatum. Future studies are needed to clarify whether connectivity changes predict therapeutic effects in psychiatric patients.”

Authors: O. Grimm, Rainer Kraehenmann, Katrin H. Preller, E. Seifritz & Franz X. Vollenweider

Summary

Abstract

We studied the effects of psilocybin on the amygdala in 18 healthy volunteers in a double-blind balanced cross-over design to elucidate the connectivity pattern of the amygdala over the entire brain.

We found that psilocybin increases reaction time for all three categories of affective stimuli, decreases connectivity between important nodes linked to emotion processing, and has no effect during discrimination of fearful faces.

1. Introduction

Psilocybin modulates serotonergic neurotransmission in the brain, especially in the limbic system and related hubs in the prefrontal cortex. Its effects on face processing may be a testable proxy for the dysfunctions seen in affective disorders.

Psilocybin modulates emotional face processing, and has been demonstrated in temporally fine-grained EEG-studies, but has only been rarely studied in fMRI. We used an event-related paradigm to study the effects of psilocybin on the seed-to-voxel connectivity of the amygdala during emotional face processing. We hypothesized that psilocybin alters the connectivity pattern of the amygdala to the limbic network.

2.1. Subjects

Twenty-five healthy, right-handed subjects were recruited through advertisements placed in local universities. They received either placebo or 0.16 mg/kg oral psilocybin in two separate imaging sessions at least 14 days apart, and their mood state and subjective experience of altered consciousness were assessed.

2.2. Experimental paradigm

Subjects completed an emotional face discrimination task in the MRI scanner. They were presented with five different male and female faces and three different emotional facial expressions, and had to decide whether a face had an emotional or a neutral expression during a 2000 millisecond presentation.

2.4. fMRI data analysis

All three tasks were convolved with the default SPM HRF, and a high-pass filter was used to attenuate low-frequency components.

2.5. Connectivity processing

We modeled every face cue as a separate event and convolved it with a canonical hemodynamic response function. The resulting parameter estimates were sorted into stage-specific beta series. After preprocessing, the images were fed into the CONN-toolbox where the mean timeseries from the left and right amygdala were extracted. The noise was corrected using an aCompCor-strategy and the beta values were sorted by condition and concatenated to generate a beta series for each condition.

2.6. Statistical analysis of fMRI

We calculated a repeated-measure ANOVA with emotional face and its corresponding neutral face as within-subject factors, and then did a seed-to-voxel connectivity analysis for angry, happy and fearful faces separately.

For correction of multiple testing during second-level statistics, we used cluster-wise whole-brain analysis with topological FDR correction (clustero0.001).

2.7. Statistical analysis of behavioural effects

We computed a linear mixed model for brain-behaviour analysis using psilocybin versus placebo and corrected for the three tests applied twice.

The work flow of the analysis is as follows: emotional faces recognition paradigm, beta weights concatenation, mean time-series extraction, and correlation with whole-brain event-specific brain response.

3.1. Behavioural effects

Psilocybin significantly decreased reaction time while performing the face discrimination task, but correct performance was not altered. There was no significant interaction between drug and face type.

The hit rate and reaction time for three categories of emotional faces are given. The neutral faces belong to one of three sessions.

3.2. Main effect of amygdala connectivity during face discrimination

We calculated an F-contrast ANOVA including all emotional faces and found several regions with strong connectivity to the amygdalae. The amygdalae showed stronger connectivity to the striatum, the lateral occipital cortex and the frontal pole.

3.3. Effect of psilocybin intake during discrimination of emotional faces

Psilocybin decreased coupling between left striatum and right amygdala during the angry versus neutral face contrast, but did not affect self-report questionnaires.

4. Discussion

The whole-brain right amygdala seed-to-voxel significant clusters in the different face conditions for the comparison psilocybin4-placebo are shown in Table 3.

This study looked at connectivity patterns during three facial emotion discrimination tasks while participants were probed with either psilocybin or placebo. The study hypothesized that psilocybin modulates emotional face discrimination both on a behavioural level as well as on a neural level.

In our behavioural analyses, psilocybin increased reaction times for recognition of emotional faces compared to placebo, and this effect was independent of the face emotion. There was no effect of psilocybin on the correct emotional face recognition rate (hit rate). In an event-related EEG-study, psilocybin reduced activity in the limbic and fusiform areas of the brain in response to fearful and neutral faces, whereas happy faces showed a reduction in activity later (211 – 242 ms), pointing to a temporally specific effect of psilocybin on emotional face processing.

We cannot explain the different connectivity results in the angry and happy (versus neutral) conditions by temporal selective emotional processing, but it might recruit different neuronal networks.

We detected a decrease in connectivity between the right amygdala and left striatum during psilocybin intake in the angry versus neutral faces condition. This result underlines the role of serotonin in mediating between the evaluation systems of amygdala and striatum at the moment of emotional valence evaluation. The striatum is often discussed within the neurotransmitter dopamine, but psilocybin increases striatal dopamine concentrations and a recent microdialysis study found an increase of dopamine in the nucleus accumbens after infusion of psilocin.

Psilocybin reduces the neurophysiological response across all emotional face valences, and this effect could be seen as modulation of selective attention. However, the current study could not differentiate between general and selective attention, and the correct hit rate was relatively low.

Although the results show no correlation between seed-to-voxel connectivity clusters and reaction time, it is plausible that the connectivity represents a rapid response.

During psilocybin intake, we found decreased connectivity to the striatum during angry faces and increased connectivity to the frontal pole during happy faces. This can be interpreted from the background of the amygdala’s well known role in emotional processing and emotional attention. Psilocybin’s effect on the amygdala can be explained by its 5HT2A-agonistic properties. However, a previous study found that psilocybin’s impact on facial expression decreased when 5HT2A-blocking by ketanserin.

In rats, the amygdala, striatum and medial prefrontal cortex are implicated in decision-making processes. Psilocybin may affect the connectivity of these regions during face discrimination, which explains why we did not find any effects between emotional faces.

Although our paradigm did not detect face emotion-specific effects, it demonstrated a highly specific spatial connectivity to valid hubs of an affective salience system, namely amygdala, striatum and frontal pole. This points to the potential of psilocybin as a tool, probe and therapeutic agent for targeting the limbic system.

Psilocybin induces changes in attention and salience attribution, which are key to depressive symptoms formation, recovery and relapse. However, relatively little is known about the underlying large-scale connectivity changes induced by psilocybin.

Contributors

RK, FW and ES designed the study, RK executed the experiments and OG and RK did the neuroimaging analysis.