Psilocybin use after childbirth may cause lasting negative effects in mothers and their babies

Using an integrative technique of combining z-normalized results across independent maternal care behavior measurements, both stress and psilocybin exposure clearly increase the risk of impairments in maternal care even 24 h later (main stress effect; F(1,39) = 12.57, p = 0.0010; main psilocybin effect; F(1,39) = 5.487, p = 0.0244; Fig. 2I). However, there was no significant interaction between stress and psilocybin (p = 0.6959), suggesting that the effects of stress and psilocybin were not synergistic, but distinct. Notably, the effect size of social stress' contribution to sustained maternal care impairments (η² = 0.20, large) was greater than that of psilocybin (η² = 0.09, medium). While neither social stress nor psilocybin influenced the sucrose preference of dams, the sucrose preference of saline-treated stressed dams negatively correlated with pup avoidance (p = 0.0005; Supplementary Fig. 2). This correlation was not observed for psilocybin-treated stress dams (p = 0.1041).

Combining z-normalized results across all tests, psilocybin exposure clearly increases the risk of behavioral impairments in this battery two weeks later (main psilocybin effect; F(1,45) = 7.925, p = 0.0072; Fig. 3G). Notably, the effect size of psilocybin-induced increases in long-term behavioral risk (η² = 0.144, large) was greater than that of the social stress (η² = 0.023, small). These effects are specific to the postpartum period as virgin female mice treated with a single dose of psilocybin had no, if not decreased, risk of behavioral impairments in the battery using the z-normalized results (p > 0.05; d = 0.86, large; Fig. S3), and in contrast to parous females, produced an anxiolytic effect in the open field test (p = 0.0451; Supplementary Fig. 3).

Comparing z-normalized results across all tests, offspring reared by socially stressed mothers were found to be at greater behavioral risk than offspring reared by controls (main stress effect; F(1,77) = 4.147; p = 0.0451; η² = 0.046, medium; Fig. 4H) and maternal psilocybin exposure increased the risk of behavioral impairments in offspring regardless of their sex or maternal stress-exposure (main psilocybin effect; F(1,77) = 4.677; p = 0.0337; η² = 0.052, medium). There was also a small effect of sex on offspring behavioral risk (η² = 0.026, small).

To determine whether this direct exposure to psilocin could explain the behavioral deficits in offspring of dams exposed to psilocybin (Fig. 4), psilocin was administered directly to pups on PPD 7, before being reared to adulthood and subjected to the same battery of behavioral tests as the previous experiment (Fig. 5D). Direct exposure of pups to psilocin had no effect on anxiety-like behavior in the open field test, spontaneous alternations in the T maze, sociability, or preference for social novelty in 3-chamber social interaction tests (p > 0.05; Fig. 5E–H). Meanwhile, both male and female pups exposed to psilocin displayed more anhedonia in the sucrose preference test than saline-exposed littermates (main psilocin effect; F(1,45) = 7.203; p = 0.0102; Fig. 5I), although there were no differences between groups in the forced swim test (p > 0.05; Fig. 5J).

While these findings in specific behaviors perfectly replicate the phenotype of offspring of dams exposed to psilocybin, there was only a small, non-significant effect of psilocin when comparing overall behavioral risk (p > 0.05; η² = 0.008, small; Fig. 5K). There was also a moderate effect of sex whereby females were found to be at greater behavioral risk in the battery than males (p > 0.05; η² = 0.048, medium; Fig. 4H). While there was no significant interaction effect (p > 0.05), the effect of psilocin exposure on females appeared to be greater in females (d = 0.33, small–medium) than in males (d = 0.04), much like observed previously.

To assess whether this decreased head-twitch response in the postpartum period compared to the virgin state was due to changes in 5-HT receptor levels, we used real-time qPCR to assess relative differences in the expression of 5-HT receptors that are highly responsive to psilocin in the cortex of postpartum females, virgin females, and virgin male mice (Fig. S5). We found that reproductive state/sex significantly altered the expression of mRNA encoding multiple 5-HT receptors (5-HT1A, 2 A and 2 C) as well as Bdnf regardless of transcript (main effect; F(2,59) = 8.545, p = 0.0006). Planned comparisons highlighted that postpartum female mice displayed downregulated mRNA expression compared to both virgin females (p = 0.0004) and virgin males (p = 0.0347).

Psilocybin significantly increased first pup retrieval latency (p = 0.0438; Fig. 6C), while ketanserin had no effect (p > 0.05). Drug treatment also significantly impacted latency to nurse (p < 0.0001; Fig. 6D), with psilocybin increasing nursing latency both alone (p = 0.0006) and in combination with ketanserin (p = 0.0021), although ketanserin alone did not affect latency to nurse (p > 0.05).

Psilocybin and ketanserin independently produced acute reductions in contact with the full litter (main psilocybin effect; F(1,42) = 7.143, p = 0.0107; main ketanserin effect; F(1,42) = 7.403, p = 0.0094; Fig. 6E), although only ketanserin significantly reduced nursing (main ketanserin effect; F(1,42) = 7.270, p = 0.0100; Fig. 6F) while only psilocybin significantly reduced sniffing/licking of pups (main psilocybin effect; F(1,42) = 5.885, p = 0.0196; Fig. 6G). Both psilocybin and ketanserin independently showed acute increases in stress-related behaviors (main psilocybin effect; F(1,42) = 8.104, p = 0.0067; main ketanserin effect; F(1,42) = 12.14, p = 0.0012; Fig. 6H), while neither impacted fragmented care compared to saline (p > 0.05; Fig. 6I).

Using an integrative technique of combining z-normalized results across independent maternal care behavior measurements, both acute psilocybin and ketanserin exposure increase the risk of impairments in maternal care (main psilocybin effect; F(1,42) = 10.36, p = 0.0025; main ketanserin effect; F(1,42) = 5.739, p = 0.0211; Fig. 2I). However, there was no significant interaction between psilocybin and ketanserin (p = 0.9886), suggesting that the effects of psilocybin and ketanserin were not synergistic, but distinct. Notably, the effect size of psilocybin's contribution to acute maternal care impairments (η² = 0.18, large) was greater than that of ketanserin (η² = 0.10, medium).

C57BL/6 J (B6) mice were used for all experiments. Mice were maintained on a 12/12 light/dark cycle (lights on at 0300 h) and given access ad libitum to food and water. All behavioral testing was conducted during the active (dark) phase, 1 h after lights off, under dim red light. All behavioral coders were blind to drug treatment conditions throughout all experiments. For maternal behavior coding, experimenters administered the stress or control paradigm within minutes of data collection in distinct vivarium rooms and therefore were not blind to stress/control treatment. All procedures were conducted in compliance with the University of California, Davis Institutional Animal Care and Use Committee. Adult virgin female and male mice were obtained from Jackson Laboratories (Stock # 000664) at six weeks of age. Male intruders were virgin B6 mice, at least 60 days of age, bred in our colony and maintained in a separate vivarium room from experimental females. Four independent cohorts of nulliparous female mice (N = 49) were harem bred with stud males for 17 days, after which subjects were randomly assigned to one of two model groups: control (N = 22) or social stress (N = 27) and transferred to two separate vivarium rooms. All females were then transferred to a modified home cage containing Beta Chip bedding (Newco, CA), a nestlet, food and water ad libitum, and monitored daily for the presence of pups (designated postpartum/postnatal day 0). On postpartum day 3, an additional modified cage was attached to the home cage, referred to as the escape cage. The escape cage was designated as the cage without the dam's nest and litter but was otherwise identical to the home cage. The modified caging was identical to standard shoebox caging except that modified cages included a PVC portal (4 cm diameter) that connected the two shoebox cages. Access to each compartment was controlled using a PVC pipe cap. Dams and litters were weighed daily from PPDs 4–8 at the start of behavioral testing (Supplementary Table 1).

To examine the effects of psilocybin on maternal care behaviors, control and social stress groups were further assigned to psilocybin or saline treatment groups on PPD 7 by an experimenter who was never present during maternal behavior testing. As pregnancies were not timed and mice gave birth on different dates, group assignment was pseudo-randomized to balance the treatment-assignments on a given day. The final 2 × 2 study design consisted of the following groups: control + saline (N = 11), stress + saline (N = 16), control + psilocybin (N = 11), and stress + psilocybin (N = 11). On PPD 9, dams and their litters were transferred to a clean, standard shoebox cage. Offspring were weaned at PPD 21 and housed in same-sex groups until adulthood (87–150 days old) before behavioral testing. Offspring were tested in 7 independent cohorts by 8 experimenters. No more than one male and one female offspring from each litter were used for behavioral testing. After weaning, all dams were exposed to a behavioral test battery to assess cognitive and emotional behavior from PPDs 23–28.

To determine whether offspring were directly exposed to psilocybin or psilocin via breast milk after maternal psilocybin treatment, a separate group (N = 6) of nulliparous mice were harem bred as described above and allowed to deliver pups in standard shoebox cages. On PPD 6, litters were separated from dams and kept warm on a heating pad set to low for 2 h. All dams were injected with saline immediately before pup reunion and acute maternal behavior was recorded for 1 h. On PPD 7, the procedure was repeated, however all dams were injected with psilocybin (2 mg/kg i.p.). Brain tissue was collected from at least 5 independent litters at 0 min (just before reunion), 30 min, 60 min, or 120 min after the initiation of nursing at reunion ( ~ 5 min after reunion). Pups were euthanized by decapitation and brain tissue from two pups was pooled for each sample. The dams (N = 5), as well as virgin female mice (N = 6) treated with psilocybin 2 h prior, were euthanized by cervical dislocation, and their brains removed, flash-frozen, and stored at −80 °C until analysis.

To determine whether the acute effects of psilocybin on maternal care behaviors were dependent on engagement with the serotonin system, a separate group (N = 45) of nulliparous mice were harem bred as described above and allowed to deliver pups in standard shoebox cages. Dams were further assigned to one of the following treatment groups to enable a 2 × 2 study design on PPD 7 by an experimenter who was never present during maternal behavior testing: saline + saline (N = 12), ketanserin + saline (N = 11), saline + psilocybin (N = 12), and ketanserin + psilocybin (N = 11). Adequate blocking of the acute effects of psilocybin by ketanserin were validated by a head-twitch response assay and compared to yoked virgin females.

To assess the relative expression of genes encoding serotonin receptors that have a high affinity for psilocin, a separate group (N = 6) of nulliparous mice were harem bred as described above and allowed to deliver pups in standard shoebox cages. Postpartum mice alongside age-matched virgin male (N = 6) and virgin female mice (N = 6) were euthanized by cervical dislocation and brains rapidly excised. Cortical tissue was carefully removed from the rest of the brain tissue, placed into RNAse free microcentrifuge tubes, flash-frozen in liquid nitrogen and stored at −80 °C until analysis.

To assess the effects of direct psilocin exposure on PPD 7 on adult behaviors, a separate group (N = 14) of nulliparous mice were harem bred as described above and allowed to deliver pups in standard shoebox cages. On PPD 7, litters were separated from dams and kept warm on a heating pad set to low. No more than two males and two females from each litter were used for experiments, enabling a 2 × 2 study design of the following groups: male + saline (N = 14), male + psilocin (N = 14), female + saline (N = 12), and female + psilocin (N = 12). Mice were injected with psilocin slowly for approximately 1 min, then the needle held in the intraperitoneal space for another 1 min to ensure adequate absorption, before being returned to the heating pad. Mice were observed for an additional 15 min before reunion.

Psilocybin and psilocin were synthesized in-house and dissolved in saline to yield a final dose of 2 mg/kg⁵⁸. As we were restricted to assessing just a single dose of psilocybin in this paradigm due to feasibility, dose selection was based on literature precedent^19,59,60. Ketanserin tartrate (Sigma-Aldrich) was dissolved in saline to yield a final salt-corrected dose of 1 mg/kg ketanserin. Dose was selected based on experience and pilot data ensuring minimal motoric impairments. All drugs were administered to adults at a 5 mL/kg injection volume using a 28-gauge needle. Psilocin or saline was administered to pups at a 5 mL/kg injection volume using a 5/16", 31-gauge needle. All drugs were administered by intraperitoneal injection (i.p).

Once offspring were weaned and aged to adulthood (minimum 12 weeks old), they were exposed to a larger battery of standardized tests to assess multiple functional domains that could be relevant to several neuropsychiatric disorders; social, emotional and cognitive behavior (open field test, novel object recognition, T-maze spontaneous alternation, social interaction, sucrose preference, forced swim test). This test battery began at least 11 weeks following maternal psilocybin exposure on PPD 7. The battery order was determined by increasing degrees of stress to the animal, in compliance with the University of California, Davis Institutional Animal Care and Use Committee guidelines.

All data were analyzed using GraphPad Prism (v10.5). Maternal behavior composite scores were calculated by summing the total number of observations in which the subject was engaged in a behavior or set of behaviors in physical contact with pups and dividing by the total number of possible observations (20) multiplied by 100. Fragmented care was calculated by summing the number of observations where the dam transitioned between behavior groups (nursing, nest-building, nest-related and non-litter related) out of 19 possible transitions^15,62. The maternal stress-related behavior score represents a sum of behaviors that diverge from the stereotyped pattern of caregiving described in rodents. These behaviors include partial litter contact (contact with some but not all pups), abnormal in-nest behaviors (out of pup contact or contact that prevents pups from accessing nipples), or stress-related behaviors outside of the nest such as digging, and self-grooming or inactivity (Supplementary Data 1). Latency data were analyzed using a Kaplan-Meier method whereby median latencies to retrieve the first pup across the three intruder days were plotted using a cumulative incidence plot (simple survival analysis in GraphPad Prism), and a Log-rank Mantel–Cox test used to make statistical comparisons between groups⁶³ followed by a Bonferroni multiple comparisons test when significant. Pup circling and nest relocation data were analyzed using contingency tables, and statistical comparisons between groups made using a Chi-square (and Fisher's exact) test.

All other maternal behaviors and non-maternal dam behaviors were analyzed using 2way ANOVAs either comparing the effects of stress across the three intruder days, comparing the interaction between stress and psilocybin exposure or comparing the interaction between ketanserin and psilocybin exposure on behavioral outcomes. All manual behavioral coding was conducted by independent raters trained to >90% agreement during practice sessions that were blind to group assignment.

Virgin female behavioral battery test data were analyzed using t-tests comparing saline- to psilocybin-treated animals. Yoked virgin female and postpartum female head-twitch response data were analyzed using a 3way ANOVA comparing the effects of reproductive state, ketanserin, and psilocybin exposure, followed by Šídák's multiple comparisons test comparing the means of groups that differed by only one factor.

All offspring behavioral data were analyzed either using 3way ANOVAs comparing the effects of maternal stress, maternal psilocybin exposure, and offspring sex on behavioral outcomes, or 2way ANOVAs comparing the effects of psilocybin exposure and offspring sex on behavioral outcomes. Significant interaction effects involving sex from 3way ANOVAs were followed up with 2way ANOVAs split by sex to determine the nature of the sex differences. For the psilocybin dose-escalation headtwitch response experiment, groups were analyzed by 2way ANOVA, followed by planned comparisons between postpartum females and all other groups. Given that no mice exhibited head-twitches in response to saline, these data were removed from analysis to prevent a forced interaction effect between dose and group. The relationship between pup avoidance and sucrose preference was analyzed using a two-tailed Pearson correlation analysis.

Behavioral battery data for postpartum females, virgin females, or offspring were also analyzed using an integrative technique to calculate a "behavioral risk score", whereby a higher score indicates more behavioral impairments across the complete battery of tests. Results from each test were first converted to z-scores following removal of any statistical outliers. Z-scores are standardized scores indicating how many standard deviations the result is above or below the mean of all observations made in the test, using the following equation: z = (X – μ)/σ, where z is the standardized score, X is the measured result in the behavioral task, and μ and σ are respectively the mean and standard deviation of all results measured in the behavioral task. To calculate the integrated behavioral risk score, directionality was established for each behavior such that higher z-scores indicated behavioral impairments specific to each task. For example, z-normalized data for results in the T-maze were multiplied by −1 such that mice that performed fewer alternations, and therefore showed worse working memory, were attributed a higher z-score indicative of this impairment. Immobility in the forced swim test, however, is associated with an increased depressive-like phenotype and therefore z-normalized immobility measures were multiplied by 1. Finally, sociability and social novelty scores were first averaged before integration with other tests to ensure equal weighting across behavioral domains. An overall behavioral risk score was then calculated for each mouse by taking the average of z-normalized results across all tests: Behavioral risk score = \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\frac{{Z}_{{Test}1}+{Z}_{{Test}2}+{Z}_{{Test}3}+\,\cdot \cdot \cdot \,}{{Number\; of\; tests}}$$\end{document}ZTest1+ZTest2+ZTest3+⋅⋅⋅Numberoftests, where Z_{Test 1} etc. is the average z-score of all animals in that test⁶⁴.

Similarly, a maternal care behavioral risk score was calculated by averaging z-normalized results for independent maternal care behaviors; arched back nursing, sniffing/licking, avoidance (when available), stress-related behaviors and fragmented care. Note that the maternal behavior risk score calculated for the experiment examining psilocybin/psilocin transfer to offspring does not include an avoidance measure as these mice were housed in standard cages.

Effect sizes were calculated for all behavioral risk scores; η² values were calculated for ANOVAs where values around 0.01 were considered small, 0.06 considered medium, and 0.14 considered large⁶⁵, and Cohen's d values were calculated for t tests where values around 0.2 were considered small, 0.5 considered medium, and 0.8 considered large.

Relative quantification values were calculated and normalized by log₂-transformation for each animal and each transcript according to the following formula:\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\log }_{2}{{{\rm{fold}}}}\; {{{\rm{change}}}}={\log }_{2}({2}^{-\Delta \Delta {Ct}})$$\end{document}log2foldchange=log2(2−ΔΔCt)

where ∆∆Ct is the Ct of a given transcript, subtracted from the Ct of the housekeeping gene (beta-2-microglobulin) of the same animal, and then subtracting from that the average Ct values of the given transcript from the reference animals (virgin females).

In all instances, statistical outliers were determined by a ROUT test with Q = 1%. Normal distribution and sphericity were assumed for all ANOVA data. Data were considered statistically significant if p < 0.05.

Further information on research design is available in the linked to this article. Nature Portfolio Reporting Summary