Sex-specific role of the 5-HT2A receptor in psilocybin-induced extinction of opioid reward

Abstract

Emerging evidence suggests that classical psychedelics may offer therapeutic potential for opioid use disorder (OUD) by alleviating key hallmarks such as altered reward processing and dependence. However, the mechanisms behind these effects remain unclear. Our data demonstrate that a single administration of the psychedelic psilocybin (PSI) reduces conditioned behavior and withdrawal induced by the opioid oxycodone (OXY) in male mice but not in females, and this effect is mediated via the 5-HT_2A receptor (5-HT_2AR). We show that the sex-specific attenuation of OXY preference is driven by 5-HT_2AR activation in frontal cortex pyramidal neurons projecting to the nucleus accumbens (NAc). Additionally, PSI modulates epigenomic regulation following repeated OXY exposure and induces sex-specific NAc dendritic structural plasticity independently of 5-HT_2AR. Notably, female frontal cortex and NAc show fewer changes at gene enhancer regions in response to PSI, repeated OXY, or combined PSI-OXY treatment compared to males, with the frontal cortex exhibiting more pronounced sex differences than the NAc at the epigenomic level. Together, these results provide new insights into the neural and epigenetic mechanisms of psychedelic-induced plasticity in OUD, while also highlighting sex differences in PSI's modulation of reward pathways and its therapeutic potential.

Fig. 1. Post-acute effects of PSI on unconditioned behaviors.

a Dose-response effect of PSI on HTR behavior. HTR counts correspond to the first 30 min after injection with PSI (1 mg/kg) or vehicle in male (n = 6 per group) and female (n = 5-6 per group) wild-type mice (dose F [3,36] = 29.67, p < 0.001; sex F [1,36] = 16.27, p < 0.001; interaction F [3,36] = 3.94, p < 0.01). b Effect of pretreatment with the 5-HT_2AR antagonist volinanserin (vol) (0.1 mg/kg) or vehicle 15 min prior to the administration of PSI (1 mg/kg) on HTR behavior. HTR counts correspond to the first 30 min after injection with PSI or vehicle in male (n = 6 per group) and female (n = 5-6 per group) wild-type mice (drug F [3,39] = 60.37, p < 0.001; sex F [1,39] = 8.89, p < 0.01; interaction F [3,39] = 9.00, p < 0.001). c Brain concentration of psilocin after a single administration of PSI (1 mg/kg) in male (n = 4-5 per group) and female (n = 5 per group) wild-type mice (time F [3,31] = 27.44, p < 0.001; sex F [1,33] = 0.16, p > 0.05; interaction F [3,31] = 0.68, p > 0.05). Arrow indicates time of administration of PSI. d Blood concentration of psilocin after a single administration of PSI (1 mg/kg) in male (n = 5 per group) and female (n = 4–5 per group) wild-type mice (time F [3,31] = 117.1, p < 0.001; sex F [1,33] = 0.85, p > 0.05; interaction F [3,31] = 1.07, p > 0.05). Arrow indicates time of administration of PSI. e and f Lack of effect of PSI on exploratory behavior in an open field. Behavior was tested 24 h after a single administration of PSI (1 mg/kg) or vehicle in male (n = 8 per group) and female (n = 5-6 per group) wild-type mice (e: time F [17,391] = 36.40, p < 0.001; drug F [1,23] = 0.09, p > 0.05; sex F [1,23] = 1.45, p > 0.05; interaction F [17,391] = 0.67, p > 0.05; see Supplementary Table 1 for additional statistical comparisons) (f: drug F [1,23] = 0.07, p > 0.05; sex F [1,23] = 0.23, p > 0.05; interaction F [1.23] = 0.0002, p > 0.05). g and h Lack of effect of PSI on dark-light choice test. Behavior was tested 24 h after a single administration of PSI (1 mg/kg) or vehicle in male (n = 6 per group) and female (n = 6 per group) wild-type mice (g: time F [2,60] = 12.99, p < 0.001; drug F [1,60] = 1.29, p > 0.05; sex F [1,60] = 1.06, p > 0.05) (h: drug F [1,20] = 1.98, p > 0.05; sex F [1,20] = 5.93, p < 0.05; interaction F [1,20] = 0.55, p > 0.05). i and j Lack of effect of PSI on exploratory preference during the novel-object recognition test. Behavior was tested 24 h after a single administration of PSI (1 mg/kg) or vehicle in male (n = 6 per group) and female wild-type (n = 5-6 per group) mice (i: preference F [1,20] = 110.0, p < 0.001; drug F [1,20] = 0.28, p > 0.05; interaction F [1,20] = 1.04, p > 0.05) (j: preference F [1,18] = 17.65, p < 0.001; drug F [1,18] = 0.005, p > 0.05; interaction F [1,18] = 0.001, p > 0.05) (i and j: sex F [1,38] = 2.26, p > 0.05; see Supplementary Table 2 for additional statistical comparisons). k and l Sex-specific effect of PSI on exploratory time during the novel-object recognition test. Behavior was tested 24 h after a single administration of PSI (1 mg/kg) or vehicle in male (n = 6 per group) and female (n = 5-6 per group) wild-type mice (k: exploration F [1,20] = 0.09, p > 0.05; drug F [1,20] = 0.55, p > 0.05; interaction F [1,20] = 0.001, p > 0.05) (l: exploration F [1,18] = 34.11, p < 0.001; drug F [1,18] = 5.45, p < 0.05; interaction F [1,18] = 0.21, p > 0.05) (k and l: sex F [1,38] = 6.67, p < 0.05; see Supplementary Table 3 for additional statistical comparisons). Statistical analysis was performed using two-way ANOVA (a, b, c, d, f, h, i, j, k,l) and/or three-way ANOVA (e, g, i, j, k, l) (⁺p < 0.05, ⁺⁺p < 0.01, ⁺⁺⁺p < 0.001, n.s., not significant) followed by Bonferroni’s multiple comparison test (*p < 0.05, **p < 0.01, ***p < 0.001, n.s., not significant). Data show mean ± s.e.m.

Fig. 2. Sex-specific effects of PSI on OXY-induced CPP via 5-HT_2AR.

a Timeline of the experimental design. b OXY-induced CPP in male (n = 56-63 per group) and female (n = 41-47 per group) mice (drug F [1,203] = 92.78, p < 0.001; sex F [1,203] = 0.10; p > 0.05; interaction F [1,203] = 0.88, p > 0.05). c PSI reverses OXY-induced CPP post-acutely in male wild-type mice (n = 22-26 per group) (OXY F [1,87] = 0.22, p > 0.05; PSI F [1,87] = 0.02, p > 0.05; interaction F [1,87] = 7.76, p < 0.01). d PSI does not reverse OXY-induced CPP post-acutely in female mice (n = 20-27 per group) (OXY F [1,93] = 12.72, p < 0.001; PSI F [1,93] = 0.28, p > 0.05; interaction F [1,93] = 0.05, p > 0.05). e, f PSI facilitates extinction of OXY preference in male wild-type mice (n = 26 per group). Mice conditioned with OXY were assessed for within-subject changes on day 5 (e: pre-vehicle, red square; f: pre-PSI, red square) and day 6 (e: 24 h post-vehicle, white square; f: 24 h post-PSI, green square) (e: t₂₅ = 1.73, p > 0.05; f: t₂₅ = 3.84, p < 0.001). g and h PSI does not affect extinction of OXY preference in female wild-type mice (n = 24–25 per group). Mice conditioned with OXY were assessed for within-subject changes on day 5 (g: pre-vehicle, red circle; h: pre-PSI, red circle) and day 6 (g: 24 h post-vehicle, white circle; h: 24 h post-PSI, green circle) (g: t₂₃ = 1.56, p > 0.05; h: t₂₄ = 1.48, p > 0.05). i OXY-induced CPP in male (n = 19-20 per group) and female (n = 17-19 per group) 5-HT_2AR-KO mice (drug F [1,71] = 44.43, p < 0.001; sex F [1,71] = 8.88; p < 0.01; interaction F [1,71] = 2.54, p > 0.05). j PSI does not affect OXY-induced CPP post-acutely in male 5-HT_2AR-KO mice (n = 9-10 per group) (OXY F [1,35] = 10.71, p < 0.01; PSI F [1,35] = 0.30, p > 0.05; interaction F [1,35] = 0.11, p > 0.05). k PSI does not affect OXY-induced CPP post-acutely in female 5-HT_2AR-KO mice (n = 8-10 per group) (OXY F [1,32] = 11.44, p < 0.01; PSI F [1,32] = 0.40, p > 0.05; interaction F [1,35] = 0.002, p > 0.05). l, m PSI does not affect extinction of OXY preference in male 5-HT_2AR-KO mice (n = 10 per group). Mice conditioned with OXY were assessed for within-subject changes on day 5 (l: pre-vehicle, red circles; m: pre-PSI, red circles) and day 6 (l: 24 h post-vehicle, white circles; m: 24 h post-PSI, green circles) (l: t₉ = 0.66, p > 0.05; m: t₉ = 0.14, p > 0.05). n and o PSI affects extinction of OXY preference in female 5-HT_2AR-KO mice (n = 9-10 per group) (n: pre-vehicle, red circle; o: pre-PSI, red circle) and day 6 (n: 24 h post-vehicle, white circle; o: 24 h post-PSI, green circle) (n: t₈ = 0.98, p > 0.05; o: t₉ = 4.40, p < 0.01). Statistical analysis was performed using two-way ANOVA (⁺⁺p < 0.01, ⁺⁺⁺p < 0.001, n.s., not significant) followed by Bonferroni’s multiple comparison test (b–d, i–k), or paired Student’s t-test (e–h, l–o) (*p < 0.05, **p < 0.01, ***p < 0.001, n.s., not significant). Data show mean ± s.e.m.

Fig. 3. Sex-specific effects of PSI on somatic signs of OXY withdrawal via 5-HT_2AR.

a Timeline of the experimental design. b Post-acute effects of PSI on somatic signs of OXY withdrawal in male wild-type mice (n = 8-9 per group) (OXY F [1,29] = 81.32, p < 0.001; PSI F [1,29] = 28.74, p < 0.001; interaction F [1,29] = 25.55, p < 0.001). c Post-acute effects of PSI on somatic signs of OXY withdrawal in female wild-type mice (5–10 per group) (OXY F [1,28] = 20.29, p < 0.001; PSI F [1,28] = 0.09, p > 0.05; interaction F [1,28] = 0.66, p > 0.05). d Post-acute effects of PSI on somatic signs of OXY withdrawal in male 5-HT_2AR-KO mice (n = 3-6 per group) (OXY F [1,17] = 26.11, p < 0.001; PSI F [1,17] = 0.08, p > 0.05; interaction F [1,17] = 0.71, p > 0.05). e Post-acute effects of PSI on somatic signs of OXY withdrawal female 5-HT_2AR-KO mice (3-4 per group) (OXY F [1,11] = 31.90, p < 0.001; PSI F [1,11] = 0.89, p > 0.05; interaction F [1,11] = 0.64, p > 0.05). f and g Post-acute effects of PSI on OXY-induced antinociception in male wild-type mice (n = 6 per group) (f: drug F [3,20] = 66.90, p < 0.001; time F [6,120] = 17.53, p < 0.001; interaction F [18,120] = 4.40, p < 0.001) (g: OXY F [1,20] = 494.0, p < 0.001; PSI F [1,20] = 0.82, p > 0.05; interaction F [1,20] = 0.84, p > 0.05). h and i Post-acute effects of PSI on OXY-induced antinociception in female wild-type mice (5-6 per group) (h: drug F [3,19] = 27.30, p < 0.001; time F [6,114] = 26.29, p < 0.001; interaction F [18,114] = 8.30, p < 0.001) (i: OXY F [1,19] = 123.8, p < 0.001; PSI F [1,19] = 0.007, p > 0.05; interaction F [1,19] = 0.10, p > 0.05). Maximum possible effect (MPE). Statistical analysis was performed using two-way ANOVA (b–e, g, i) or two-way repeated measured ANOVA (f, h) (⁺⁺⁺p < 0.001) followed by Bonferroni’s multiple comparison test (*p < 0.05, **p < 0.01, ***p < 0.001, n.s., not significant). Data show mean ± s.e.m.

Fig. 4. Post-acute effects of PSI on natural reward behavior.

a, b Post-acute effects of PSI on saccharine preference (a) and total fluid consumption (b) in male (n = 6 per group) and female (n = 6 per group) wild-type mice (a: drug F [1,28] = 3.23, p > 0.05; sex F [1,28] = 4.08, p > 0.05; interaction F [1,28] = 1.15, p > 0.05) (b: drug F [1,28] = 2.91, p > 0.05; sex F [1,28] = 0.0002, p > 0.05; interaction F [1,28] = 0.01, p > 0.05). c Effect of repeated OXY administration on PSI-induced HTR behavior in male wild-type mice (n = 4 per group) (OXY F [1,12] = 0.04, p > 0.05; PSI F [1,12] = 20.34, p < 0.001; interaction F [1,12] = 0.03, p > 0.05). d Effect of repeated OXY administration on PSI-induced HTR behavior in female wild-type mice (n = 3-4 per group) (OXY F [1,11] = 1,64, p > 0.05; PSI F [1,11] = 191.0, p < 0.001; interaction F [1,11] = 9.36, p < 0.05). Statistical analysis was performed using two-way ANOVA (a–d) (⁺⁺⁺p < 0.001, n.s., not significant) followed by Bonferroni’s multiple comparison test (*p < 0.05, n.s., not significant). Data show mean ± s.e.m.

Fig. 5. Neural circuit-specific restoration of 5-HT_2AR-NAc expression.

a Schematic representation of the dual AAVretro and AAV strategy to selectively restore 5-HT_2AR expression in frontal cortex pyramidal neurons of 5-HT_2AR^LSL/LSL mice projecting to the NAc. b Representative micrographs of eYFP expression in mouse frontal cortex neurons following injection of AAVretro-EF1α-mCherry-IRES-Flp in the NAc, and AAV-EF1α-fDIO-eYFP in the frontal cortex. Mice received stereotaxic injections of mCherry- and eYFP-tagged (top panels), mCherry-tagged only (middle panels), and eYFP-tagged only (bottom panels) constructs. White arrowheads indicate mCherry-positive cells, whereas yellow arrowheads mark cells that are positive for both mCherry and eYFP. c, d Virally-mediated restoration of 5-HT_2AR expression in frontal cortex pyramidal neurons of 5-HT_2AR^LSL/LSL mice. Animals (5-HT_2AR^LSL/LSL mice and wild-type controls) received mock viral vector surgeries or stereotaxic administration of AAVr-Flp-mCherry into the NAc, and AAV-fDIO-Cre-eYFP or the AAV-fDIO-eYFP control viral vector into the frontal cortex. Representative immunoblots with anti-5-HT_2AR and anti-GAPDH antibodies in mouse frontal cortex samples (c). Expression of 5-HT_2AR, 5-HT_1AR, D₁R and the housekeeping gene Gapdh mRNAs in frontal cortex samples was assessed by RT-qPCR (4-11 per group) (5-HT_2AR F [3,29] = 10.51, p < 0.001; 5-HT_1AR F [3,18] = 0.59, p > 0.05; D₁R F [3,18] = 1.06, p > 0.05;; Gapdh F [3,32] = 1.01, p > 0.05) (d). Scale bar represents 20 µm (b). Statistical analysis was performed using one-way ANOVA followed by Bonferroni’s multiple comparison test (d) (*p < 0.05, ***p < 0.001, n.s., not significant). Data show mean ± s.e.m.

Fig. 6. Frontal cortex 5-HT_2AR-NAc circuit partially mediates PSI’s modulation of opioid seeking.

a OXY-induced CPP in male 5-HT_2AR^LSL/LSL mice co-injected with AAVr-Flp-mCherry and either AAV-fDIO-Cre-eYFP or AAV-fDIO-eYFP (n = 4-7 per group) (F [2,16] = 7.17, p < 0.01). b Timeline of the experimental design. c PSI does not reverse OXY-induced CPP post-acutely in male 5-HT_2AR^LSL/LSL mice co-injected with AAVr-Flp-mCherry and AAV-fDIO-Cre-eYFP (n = 3-17 per group) (OXY F [1,36] = 8.36, p < 0.01; PSI F [1,36] = 0.73, p > 0.05; interaction F [1,36] = 0.43, p > 0.05). d, e PSI facilitates extinction of OXY preference in male 5-HT_2AR^LSL/LSL mice co-injected with AAVr-Flp-mCherry and AAV-fDIO-Cre-eYFP (n = 16 per group). Mice conditioned with OXY were assessed for within-subject changes on day 5 (d: pre-vehicle, red hexagon; e: pre-PSI, red hexagon) and day 6 (d: 24 h post-vehicle, white hexagon; e: 24 h post-PSI, green hexagon) (d: t₁₅ = 0.99, p > 0.05; e: t₁₅ = 2.45, p < 0.05). Statistical analysis was performed using one-way (a) or two-way (c) ANOVA (⁺⁺p < 0.01) followed by Bonferroni’s multiple comparison test (a, c), or paired Student’s t-test (d, e) (*p < 0.05, n.s., not significant). Data show mean ± s.e.m.

Fig. 7. PSI post-acutely modulates MOR-G protein coupling in a sex-dependent manner.

a Effect of the MOR agonist DAMGO (10 µM) on [³⁵S]GTPγS binding in membrane preparations of frontal cortex samples from male (n = 10) and female (n = 8) wild-type mice (males: DAMGO F [1,36] = 25.90, p < 0.001; PSI F [1,36] = 0.41, p > 0.05, interaction F [1,36] = 0.68, p > 0.05) (females: DAMGO F [1,28] = 17.07, p < 0.01; PSI F [1,28] = 5.29, p < 0.05; interaction F [1,28] = 5.29, p < 0.05) (sex F [1,64] = 7.14, p < 0.01; see Supplementary Table 7 for additional statistical comparisons). b Effect of the MOR agonist DAMGO (10 µM) on [³⁵S]GTPγS binding in membrane preparations of NAc samples from male (n = 8) and female (n = 4) wild-type mice (males: DAMGO F [1,28] = 19.92, p < 0.001; PSI F [1,28] = 0.23, p > 0.05, interaction F [1,28] = 0.23, p > 0.05) (females: DAMGO F [1,12] = 891.0, p < 0.001; PSI F [1,12] = 1.32, p > 0.05; interaction F [1,12] = 1.32, p > 0.05) (sex F [1,40] = 27.53, p < 0.001; see Supplementary Table 8 for additional statistical comparisons). c PSI does not alter MOR (µ-OR) or housekeeping gene rps3 mRNA expression in the frontal cortex of male wild-type mice (n = 5) (µ-OR: t₈ = 0.05, p > 0.05; rps3: t₈ = 1.24, p > 0.05). d PSI reduces MOR (µ-OR) but not housekeeping gene rps3 mRNA expression in the frontal cortex of female wild-type mice (n = 5) (µ-OR: t₈ = 3.14, p < 0.05; rps3: t₁₀ = 1.20, p > 0.05). e PSI does not alter MOR density in the frontal cortex of male (n = 6) or female (n = 6) wild-type mice, as assessed by [³H]naloxone binding assays (PSI F [1,20] = 1.48, p > 0.05; sex F [1,20] = 6.53, p < 0.05; interaction F [1,20] = 0.04, p > 0.05). Agonist-induced [³⁵S]GTPγS binding is presented as net stimulation (DAMGO effect minus basal [³⁵S]GTPγS binding) (a, b). Statistical analysis was performed using two-way and/or three-way ANOVA (⁺p < 0.05. ⁺⁺p < 0.01, ⁺⁺⁺p < 0.001) followed by Bonferroni’s multiple comparison test (a, b, e), or multiple unpaired t-test (c, d) (*p < 0.05, **p < 0.01, ***p < 0.001, n.s., not significant). Data show mean ± s.e.m.

Fig. 8. Post-acute effects of PSI on OXY-induced differential enhancers and their associated genes.

a Number of differential enhancers in the frontal cortex of male and female wild-type mice across treatments: oxycodone-vehicle, oxycodone-psilocybin and vehicle-psilocybin, with the vehicle-vehicle group as the reference. b Number of differential enhancers in the NAc of male and female wild-type mice across treatments: oxycodone-vehicle, oxycodone-psilocybin and vehicle-psilocybin, with the vehicle-vehicle group as the reference. c–f Overlap between the differential enhancer-like genes under oxycodone-vehicle (OXY-VEH) and oxycodone-psilocybin (OXY-PSI) treatment in frontal cortex (c, d) and NAc (e, f) samples of male (c, e) and female (d, f) wild-type mice. g–o Top KEGG terms associated with the groups of differential enhancer-linked genes in c–f. Top 10 terms are listed if there are more than 10 terms. The complete lists of terms are included in Supplementary Data 1. Only groups that yield enriched terms are included. Male (n = 4-6) and female (n = 3–6) mice were treated with OXY (3 mg/kg) or vehicle once a day for three days, received a single administration of PSI (1 mg/kg) or vehicle 24 h after the last treatment, and tissue samples (frontal cortex and NAc) were collected 24 h after (n = 3-6 mice per group). See Supplementary Data 1 for additional statistical information.

Fig. 9. K-means clustering of differential enhancers over three conditions (VEH-VEH, OXY-VEH, and OXY-PSI) with VEH-VEH as the reference, and two sexes.

a K-means clustering of differential enhancers in the frontal cortex. b Average pairwise Pearson correlation coefficient (PPCC) for each K-means cluster in (a). c K-means clustering of differential enhancers in the NAc. d Average pairwise Pearson correlation coefficient (PPCC) for each K-means cluster in (b). Male (n = 4-6) and female (n = 3–6) wild-type mice were treated with OXY (3 mg/kg) or vehicle once a day for three days, received a single administration of PSI (1 mg/kg) or vehicle 24 h after the last treatment, and tissue samples (frontal cortex and NAc) were collected 24 h after (n = 3–6 mice per group). See Supplementary Data 2 for additional statistical information.

Fig. 10. Sex-divergent alterations in NAc structural plasticity following PSI are independent of 5-HT_2AR.

a–f Post-acute effects of PSI on dendritic spine structural elements in the NAc of 5-HT_2AR-KO mice and wild-type controls. Samples from male and female animals were collected 24 h after a single injection of PSI (1 mg/kg) or vehicle. Representative image of HSV-mediated transgene expression in the NAc. HSV-GFP was injected in the NAc, and GFP expression was revealed by fluorescence microscopy imaging (a). Representative three-dimensional reconstructions of HSV-injected NAc dendritic segments (b). Total NAc dendritic spine density in male wild-type (n = 44-58 neurons from 3-4 mice) and 5-HT_2AR-KO (n = 15-17 neurons from 3-4 mice) littermates (PSI F [1,130] = 18.18, p < 0.001; genotype F [1,130] = 0.53, p > 0.05; interaction F [1,130] = 0.02, p > 0.05; see Supplementary Table 9 for additional statistical comparisons) (c). Total NAc dendritic spine density in female wild-type (n = 31-35 neurons from 3-4 mice) and 5-HT_2AR-KO (n = 28-33 neurons from 3-4 mice) and littermates (PSI F [1,123] = 34.86, p < 0.001; genotype F [1,123] = 0.19, p > 0.05; interaction F [1,123] = 6.81, p < 0.01; see Supplementary Table 9 for additional statistical comparisons) (d). Stubby, thin and mushroom NAc dendritic spine density in male and female wild-type mice (PSI F [1,468] = 0.56, p > 0.05; spine type F [2,468] = 186.1, p < 0.001; sex F [1,468] = 0.34, p < 0.05; interaction F [2468] = 12.95, p < 0.001; see Supplementary Table 10 for additional statistical comparisons) (e). Stubby, thin and mushroom NAc dendritic spine density in male and female 5-HT_2AR-KO mice (PSI F [1,270] = 2.70, p > 0.05; spine type F [2270] = 35.46, p < 0.001; sex F [1,270] = 3.52, p = 0.06; interaction F [2270] = 2.67, p = 0.07; see Supplementary Table 11 for additional statistical comparisons) (f). g-i Repeated PSI administration affects reward processing in a sex-dependent manner. Timeline of the experimental design (g). PSI-induced CPP in male wild-type mice (n = 12 per group) (t₂₂ = 1.09, p > 0.05) (h). PSI-induced CPP in female wild-type mice (n = 8-9 per group) (t₁₅ = 3.60, p < 0.01) (i). Scale bars represent 20 µm (a) or 10 µm (b). Statistical analysis was performed using two-way (c, d) or three-way (e, f) ANOVA (⁺⁺⁺p < 0.001, n.s., not significant) followed by Bonferroni’s multiple comparison test, or unpaired Student’s t-test (h, i) (*p < 0.05, **p < 0.01, ***p < 0.001. n.s., not significant). Data show mean ± s.e.m.