Dose-dependent changes in global brain activity and functional connectivity following exposure to psilocybin: a BOLD MRI study in awake rats

Abstract

Psilocybin is a hallucinogen with complex neurobiological and behavioral effects. This is the first study to use MRI to follow functional changes in brain activity in response to different doses of psilocybin in fully awake, drug naive rats. We hypothesized that psilocybin would show a dose-dependent increase in activity in the prefrontal cortex and thalamus, while decreasing hippocampal activity. Female and male rats were given IP injections of vehicle or psilocybin in doses of 0.03 mg/kg, 0.3 mg/kg, and 3.0 mg/kg while fully awake during the imaging session. These levels were validated by measuring psilocybin and its metabolite, psilocin. Changes in BOLD signal were recorded over a 20 min window. Data for resting state functional connectivity were collected approximately 35 min post injection. All data were registered to rat 3D MRI atlas with 169 brain areas providing site-specific changes in global brain activity and changes in functional connectivity. Treatment with psilocybin resulted in a significant dose-dependent increase in positive BOLD signal. The areas most affected by the acute presentation of psilocybin were the somatosensory cortex, basal ganglia and thalamus. Males and females showed different sensitivity to psilocybin dose, with females exhibiting greater activation than males at 0.3 mg/kg, especially in thalamic and basal ganglia regions. There was a significant dose-dependent global increase in functional connectivity, highlighted by hyperconnectivity to the cerebellum. Brain areas hypothesized to be involved in loss of sensory filtering and organization of sensory motor stimuli, such as the cortico-striato-thalamo-cortical circuit and the claustrum, showed increased activation at higher doses of psilocybin. Indeed, the general neuroanatomical circuitry associated with the psychedelic experience was affected but the direction of the BOLD signal and pattern of activity between neural networks was inconsistent with the human literature.

Keywords: 5-HT2A receptor; BOLD resting state functional connectivity; cerebellar nuclei; hyperconnectivity; psilocin.

Figure 1

Statistical heat maps of positive BOLD activation. Depicted are 2D coronal sections (a-k) showing the location of brain areas that were significantly different in positive BOLD VoA (red highlight) between vehicle and high dose (3.0 mg/kg) PSI. White matter tracts are highlighted in yellow. The insert in the lower right shows statistical heat maps in transverse and sagittal views. The 3D color-coded reconstructions summarize the major brain areas that were significantly different.

Figure 2

Regional differences in BOLD volume of activation. Shown are bar graphs (mean ± SD) for the dose-dependent change in positive VoA for different brain regions. Data was compared using a matched one-way ANOVA followed by Tukey’s multiple comparison post hoc test. * p < 0.05; ** p < 0.01; *** p < 0.001; **** p < 0.0001.

Figure 3

Regional differences in functional connectivity. Rows (a,b) shown are dot plots and bar graphs (mean ± SD) for the number of degrees or connections to other brain areas for veh 0.03 mg/kg, 0.3 mg/kg and 3.0 m/kg PSI. The violin plot show as an insert at the beginning of row (a) is the total of all degrees between the veh, 0.03 mg/kg 169 different brain areas. Shown in row (c) re functional connections to the cerebellum. Shown are radial representations of the connections or degrees (lines) to the three cerebellar nuclei (center red), e.g., fastigial, dentate and interposed, following vehicle or 3.0 mg/kg PSI. The total network is represented by 52 brain areas as the union of both treatments. All brain areas highlighted in red have direct connections to the cerebellar nuclei under each experimental condition. The significant difference in connectivity (student t-test, *p < 0.05) is shown in the inserted bar graph at the center of the figure. The 52 brain areas comprising the network can be organized in three major brain regions—olfactory bulbs (red), cerebellum (green), and brainstem/pons (yellow), with connections to the thalamus, striatum and hippocampus (all in blue). The data are presented for different brain regions. *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.

Figure 4

Plasma levels of psilocybin, psilocin, and endogenous lipids measured 30 min after PSI injection. (A,B) Average (Mean ± SEM) levels of psilocybin and its metabolite, psilocin. (C–H) Average levels of N-acyl ethanolamine endogenous lipids, including the endocannabinoid (eCB), Anandamide (G). (I,J) Average levels of 2-acyl glycerol endolipids, including the eCB, 2-AG. * p ≤ 0.05.

Figure 5

Somatosensory cortex time course. Shown is the change in BOLD signal over time for the somatosensory cortices in response to vehicle (black line) and 3.0 mg/kg PSI (red line). 250 images were acquired over the 25 min imaging session. Each acquisition is the mean ± SE of vehicle and high dose PSI rats combining the data from the different somatosensory cortices, e.g., barrel field, hindlimb, forelimb, trunk, and upper lip. The 1% threshold is highlighted by the blue lines to account for the normal fluctuations in BOLD signal in the awake rat brain. A two-way ANOVA showed a significant Time x Treatment interaction F_{(2.49, 202.42)} = 4.424, p < 0.0001.

Figure 6

Sex differences in brain activation. Depicted are 2D coronal sections (a-i) showing the location of brain areas that were significantly different in positive BOLD VoA (red highlight) between males and females for the 0.3 mg/kg dose of PSI. Females were greater than males. White matter tracts are highlighted in yellow. The insert in the lower right shows statistical heat maps in transverse and sagittal views the 3D color-coded reconstructions summarize the major brain areas that were significantly different.