Acute psilocybin enhances cognitive flexibility in rats

Abstract

Psilocybin has been shown to improve symptoms of depression and anxiety when combined with psychotherapy or other clinician-guided interventions. To understand the neural basis for this pattern of clinical efficacy, experimental and conceptual approaches that are different than traditional laboratory models of anxiety and depression are needed. A potential novel mechanism is that acute psilocybin improves cognitive flexibility, which then enhances the impact of clinician-assisted interventions. Consistent with this idea, we find that acute psilocybin robustly improves cognitive flexibility in male and female rats using a task where animals switched between previously learned strategies in response to uncued changes in the environment. Psilocybin did not influence Pavlovian reversal learning, suggesting that its cognitive effects are selective to enhanced switching between previously learned behavioral strategies. The serotonin (5HT) 2 A receptor antagonist ketanserin blocked psilocybin's effect on set-shifting, while a 5HT2C-selective antagonist did not. Ketanserin alone also improved set-shifting performance, suggesting a complex relationship between psilocybin's pharmacology and its impact on flexibility. Further, the psychedelic drug 2,5-Dimethoxy-4-iodoamphetamine (DOI) impaired cognitive flexibility in the same task, suggesting that this effect of psilocybin does not generalize to all other serotonergic psychedelics. We conclude that the acute impact of psilocybin on cognitive flexibility provides a useful behavioral model to investigate its neuronal effects relevant to its positive clinical outcome.

Fig. 1. Set-shifting task.

a Schematic representation of the operant set-shifting task. b Example sessions for one animal. Top, two sessions in the early stages of training. Bottom, two sessions during stable behavior. Vertical ticks denote the animal’s choice on each trial (Left or Right; black ticks, correct responses; red ticks, incorrect responses). Note that which port is illuminated on each trial is not shown. The current rule for each trial is shown in the colored bar at the top (yellow, Light rule; pink, Side rule). SL = side (left); SR = side (right); L1 = first light set; L2 = second light set. c Number of sessions until performance stabilized for every animal used in the set-shifting task (n = 35, 18 males and 17 females). Black open circles indicate the point at which each animal reached stability (see Methods and Supplementary Materials for details). Gray solid circles indicate the first session in which that animal received a treatment injection (saline or drug). d Schematic of drug administration and task timeline. Three sessions of stable behavior were used as a baseline for each animal. For main treatment sessions (top) injections were given 20 min prior to the start of the session. For antagonist sessions (bottom), the antagonist drug was injected 10 min prior to the main treatment.

Fig. 2. Psilocybin improves performance in the set-shifting task.

a Average trials-to-criterion across all sets in the baseline period (PRE, open gray circles) and in the acute psilocybin condition (PSI, blue circles, n = 15). ***p = 0.0005, paired t-test. b Number of completed sets in baseline and psilocybin conditions, p = 0.546, Wilcoxon sign-rank test. c Response time in baseline and psilocybin conditions. **p = 0.008, Wilcoxon sign-rank test. d Fraction of correct responses across all sets in baseline and psilocybin conditions. e Number of streaks of consecutive correct responses averaged across all sets, in baseline and psilocybin conditions. *p = 0.025, Wilcoxon sign-rank test. f Average length of streaks averaged across all sets, in baseline and psilocybin conditions. *p = 0.034, paired t-test. g Average trials-to-criterion for baseline (PRE, gray open circles) and saline conditions (SAL, brown circles, n = 14). h Number of completed sets in baseline and saline conditions. i Fraction of correct responses across all sets in baseline and saline conditions.

Fig. 3. No effect of psilocybin in the reversal learning task.

a Schematic diagram of the Pavlovian reversal learning task. b Structure of the reversal learning task, including time of injection. c R-value comparison of responding to the initially appetitive cue (App > Av) between saline-treated (open circles, n = 6) and psilocybin-treated (solid circles, n = 7) animals. Red arrow indicates the session in which contingencies change and treatment is administered. Mixed-effects analysis: main effect of Treatment, p = 0.472; interaction effect, Session x Treatment, p = 0.179. d As in (c), for the initially aversive cue. Mixed-effects analysis: main effect of Treatment, p = 0.834; interaction effect, Session x Treatment, p = 0.993. e R-value in early (first 10) and late (last 10) trials in session 6 of the reversal task, when the reversal happened, for the initially appetitive cue. All individual animals are shown. Two-way ANOVA: main effect of Timepoint, p = 0.009; interaction effect, Treatment x Timepoint, p = 0.487. f As in (e), for the initially aversive cue. Two-way ANOVA: main effect of Timepoint, p = 0.371; interaction effect, Treatment x Timepoint, p = 0.754.

Fig. 4. Treatment with DOI impairs performance in the set-shifting task.

a Average trials-to-criterion for baseline (PRE, gray open circles) and DOI conditions (red circles, n = 12). b Number of completed sets in baseline and DOI conditions. *p = 0.012, Wilcoxon sign-rank test. c Fraction of correct responses averaged across all sets in baseline and DOI conditions. **p = 0.005, Wilcoxon sign-rank test.

Fig. 5. Pre-treatment with the 5HT2A antagonist ketanserin blocks psilocybin’s effect on cognitive flexibility.

a Average trials-to-criterion for baseline (PRE, gray open circles) and ketanserin-psilocybin condition (KET + PSI, orange circles, n = 14). b Number of completed sets in baseline and ketanserin-psilocybin conditions. c Fraction of correct responses averaged across all sets in baseline and ketanserin-psilocybin conditions. d Average trials-to-criterion for baseline (PRE, gray open circles) and ketanserin-saline conditions (KET + SAL, pink circles, n = 11). *p = 0.026, paired t-test. e Number of completed sets in baseline and ketanserin-saline conditions. f Fraction of correct responses averaged across all sets in baseline and ketanserin-saline conditions. *p = 0.05, paired t-test. g Average trials-to-criterion for baseline (PRE, gray open circles) and SB242084-psilocybin conditions (SB + PSI, green circles, n = 11). *p = 0.032, Wilcoxon sign-rank test. h Number of completed sets in baseline and SB242084-psilocybin conditions. i Fraction of correct responses averaged across all sets in baseline and SB242084-psilocybin conditions. j Average trials-to-criterion for baseline (PRE, gray open circles) and SB242084-saline conditions (SB + SAL, teal circles, n = 11). k Number of completed sets in baseline and SB242084-saline conditions. l Fraction of correct responses averaged across all sets in baseline and SB242084- saline conditions. *p = 0.031, paired t-test.