Harnessing psilocybin: antidepressant-like behavioral and synaptic actions of psilocybin are independent of 5-HT2R activation in mice

Abstract

Depression is a widespread and devastating mental illness and the search for rapid-acting antidepressants remains critical. There is now exciting evidence that the psychedelic compound psilocybin produces not only powerful alterations of consciousness, but also rapid and persistent antidepressant effects. How psilocybin exerts its therapeutic actions is not known, but it is widely presumed that these actions require altered consciousness, which is known to be dependent on serotonin 2A receptor (5-HT2AR) activation. This hypothesis has never been tested, however. We therefore asked whether psilocybin would exert antidepressant-like responses in mice and, if so, whether these responses required 5-HT2AR activation. Using chronically stressed male mice, we observed that a single injection of psilocybin reversed anhedonic responses assessed with the sucrose preference and female urine preference tests. The antianhedonic response to psilocybin was accompanied by a strengthening of excitatory synapses in the hippocampus-a characteristic of traditional and fast-acting antidepressants. Neither behavioral nor electrophysiological responses to psilocybin were prevented by pretreatment with the 5-HT2A/2C antagonist ketanserin, despite positive evidence of ketanserin's efficacy. We conclude that psilocybin's mechanism of antidepressant action can be studied in animal models and suggest that altered perception may not be required for its antidepressant effects. We further suggest that a 5-HT2AR-independent restoration of synaptic strength in cortico-mesolimbic reward circuits may contribute to its antidepressant action. The possibility of combining psychedelic compounds and a 5-HT2AR antagonist offers a potential means to increase their acceptance and clinical utility and should be studied in human depression.

Keywords: depression; hallucination; hallucinogen; psychedelic; serotonin.

Fig. 1.

Restoration of hedonic behavior after chronic stress by psilocybin is unaffected by ketanserin. (A) Experimental timeline illustrating when hedonic behaviors were measured in relation to CMMS and drug treatment. (B) CMMS significantly decreased sucrose preference (SP) compared to baseline across all treatment groups: vehicle–vehicle (gray; P = 0.0012; n = 12), ketanserin–vehicle (blue; P = 0.0012; n = 6), vehicle–psilocybin (yellow; P = 0.0012; n = 13), and ketanserin–psilocybin (green; P = 0.0012; n = 7). Treatment with psilocybin (1 mg/kg, i.p.) significantly increased SP compared to values after CMMS, whether animals were pretreated with ketanserin (2 mg/kg; P = 0.042) or a vehicle control (P = 0.0012). Neither injection with vehicle (P = 0.078) nor ketanserin alone (P = 0.87; n = 6) had a significant effect on SP following CMMS. Three-way repeated-measures ANOVA revealed a significant effect of stress (F_2,68 = 60.26, P < 0.0001) and interaction of Stress × Psilocybin (F_2,68 = 4.50, P = 0.015) but not for Stress × Psilocybin × Ketanserin (F_2,68 = 0.05917, P = 0.9426). (C) CMMS significantly decreased preference for female urine compared to baseline: vehicle–vehicle (P = 0.013; n = 6), ketanserin–vehicle (P = 0.0012; n = 4), vehicle–psilocybin (P = 0.0012; n = 5), and ketanserin–psilocybin (P = 0.0012; n = 4). Treatment with psilocybin significantly increased preference for the scent of female urine compared to values after CMMS whether animals were pretreated with ketanserin (P = 0.0024) or vehicle (P = 0.0012). Following CMMS, neither injection with vehicle (P = 0.43) nor ketanserin alone (P = 0.072) had a significant effect on female urine preference. Stress significantly reduced female urine preference in all groups (F_2,30 = 43.41, P < 0.0001), and a three-way repeated-measures ANOVA showed a significant interaction between Stress × Psilocybin (F_2,30 = 4.26, P < 0.024) but not between Stress × Psilocybin × Ketanserin (F_2,30 = 0.8677, P = 0.4302). The figure bars represent the group means ± SEM. The reported post hoc comparisons were corrected with the Holm-Sidak method. *P < 0.05; **P < 0.005; ns, not significant.

Fig. 2.

Pretreatment with ketanserin effectively blocks 5-HT2A activation. (A) Pretreatment with ketanserin (2 mg/kg) 60 min prior to psilocybin (1 mg/kg) administration prevents head twitching during the 15 min immediately following drug injection in mice (ket-psil versus ket-veh, P = 0.26). The difference between groups (F_3,25 = 6.96, P = 0.0015) was driven by the increase in head twitching seen in vehicle–psilocybin animals (veh-veh versus veh-psil, P = 0.0045). (B) In animals treated with psilocybin (n = 14), both with and without ketanserin, there was no effect of head twitching (F_1,12 = 0.53, P = 0.48) but a significant effect of time (F₂,₂₄ = 38.90, P < 0.0001), revealing significant differences between stress and posttreatment SPT in both high head twitching (P = 0.0005) and low head twitching (P = 0.038) mice. Of these also used in the FUST (n = 9), there was no effect of head twitching (F_1,7 = 0.1040, P = 0 0.7565) but a significant effect of time (F_2,14 = 23.05, P < 0.0001), showing a significant difference in female urine preferences between high (P = 0.0007) and low (P = 0.0086) head twitching mice between poststress and posttreatment timepoints. A two-tailed, unpaired t test comparing AMPA/NMDA ratios between high (n = 8) and low (n = 6) head twitching mice found no difference between groups (P = 0.7717). (C) In a separate cohort of animals, psilocybin (10 mg/kg, n = 4) acutely reduced hippocampal local field potential activity in the low-frequency range, with a peak effect around 15 to 25 min following psilocybin injection. This effect was inhibited by pretreatment with ketanserin (2 mg/kg, n = 4). (D) Pretreatment with ketanserin significantly attenuated psilocybin-induced reductions in the delta frequency band 15 to 25 min following psilocybin injection. Two-way ANOVA revealed significant interaction of Frequency × Ketanserin (F_4,30 = 3.66, P = 0.0152), and a post hoc analysis revealed a significant difference between Vehicle–Psilocybin and Ketanserin–Psilocybin in the delta band (t = 4.99, df = 30, P = 0.00012). The figure bars represent the group means ± SEM; *P < 0.05; **P < 0.01; ***P < 0.001.

Fig. 3.

Psilocybin strengthens hippocampal TA-CA1 synapses following CMMS. (A) Example field EPSPs (fEPSPs) from a single stimulation intensity from one hippocampal slice per group, recorded in Mg²⁺-free ACSF (black), after wash-in of DNQX (50 µM; dark gray) and then APV (80 µM; light gray) to isolate AMPA- and NMDAR-mediated components. (B) Mice subjected to CMMS and treated with psilocybin had higher AMPA/NMDA ratios compared to stressed mice given only vehicle (gray, n = 12) or ketanserin (blue, n = 7), regardless of whether they were pretreated with ketanserin (green, n = 7; P = 0.0002) or vehicle (yellow, n = 13; P = 0.0003). Two-way ANOVA showed a significant effect of psilocybin (F_1,34 = 34.79, P < 0.0001) but not ketanserin × psilocybin (F_1,34 = 1.422, P = 0.2414). (C) Psilocybin increased the AMPA/FV ratio of the fEPSP (two-way ANOVA: F_1,34 = 4.378, P = 0.044). (D) Treatment with psilocybin did not change the NMDA/FV ratio of the fEPSP (two-way ANOVA: F_1,34 = 2.077, P = 0.16). AMPA/NMDA, AMPA/FV and NMDA/FV for each animal is shown along with group means ± SEM; *P < 0.05; ***P < 0.0005; ns, not significant.