Lasting effect of psilocybin on sociability can be blocked by DNA methyltransferase inhibition

17 18 The recent renaissance in research on psychedelics such as psilocybin has highlighted their 19 therapeutic potential including their lasting influences on brain function. Here we report that a 20 single systemic administration of the serotonergic psychedelic psilocybin can durably promote 21 social behaviour in the Cntnap2-knockout mouse model of autism. This effect can be blocked 22 by pharmacological inhibition of DNA methyltransferase I, indicating an epigenetic mechanism 23 underlying the long-lasting effect of psilocybin. 24 25 26

27 28 The serotonergic psychedelic psilocybin has been shown to induce behavioural alterations 29 outlasting the acute psychedelic effect. Clinical studies have revealed a therapeutic potential 30 of psilocybin for the treatment of psychiatric disorders 1 and suggested beneficial effects on 31 behavioural traits associated with and common to several neuropsychiatric conditions 2. 32 Intriguing but mechanistically unexplained are the therapeutic effects that last at least several 33 weeks, for instance the alleviation of depression in treatment-resistant patients following a 34 single dose of psilocybin-assisted treatment3–5. Reduced sociability is a common characteristic 35 of all types of autism under the classification autism spectrum disorder (ASD) and this altered 36 trait is replicated in several mouse models that mimic the genetics of ASD 6–8. Here, we 37 examined the effect of a single systemic injection of psilocybin in altering the social behaviour 38 of the Cntnap2 knockout mouse model8 of autism spectrum disorder and the durability of this 39 effect. 40 41 42

43 44 Psilocybin durably restores sociability in Cntnap2-/- homozygous mutant mice. 45 46 We administered a single systemic injection of 1 mg/kg psilocybin or saline to adult Cntnap2 47 homozygous mutant (Cntnap2-/- KO) and Cntnap2 homozygous wild-type isogenic control 48 (Cntnap2+/+ WT) mice at 8 weeks of age. We assessed the social behaviour of these mice 49 using the three-chamber social behaviour test at 1 day, 1 week, and 2 weeks after the single 50 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted March 11, 2025. ; https://doi.org/10.1101/2025.03.10.642385doi: bioRxiv preprint 2 injection, corresponding to ages of postnatal week 8, 9 and 10 respectively. We quantified 51 sociability as the amount of time a mouse approaches and explores a cup containing another 52 mouse versus an empty cup. Similar to the reduced sociability shown in several previous 53 studies8–10, saline treated homozygous mutant KO mice exhibited a reduced sociability 54 compared to the isogenic wild-type control group (p = 0.042, Cntnap2-/-Vehicle vs Cntnap2+/+Vehicle, 55 Bonferroni’s post-hoc). Importantly, however, psilocybin -treated homozygous mutant KO 56 group showed a significant increase in sociability compared to the vehicle-treated Cntnap2 -/-57 Vehicle KO control group (p < 0.001, Cntnap2-/-Psilocybin vs Cntnap2-/-Vehicle, Bonferroni’s post-hoc, 58 Fig 1a) and this effect was sustained for at least two weeks following the single psilocybin 59 administration. Sociability of Cntnap2-/-Psilocybin KO group was restored to a level similar to the 60 isogenic wild-type control group (p = 0.796, Cntnap2 -/-Psilocybin vs Cntnap2+/+Vehicle, Bonferroni’s 61 post-hoc). The total time a mouse spent exploring any cup (either mouse-containing or empty) 62 did not significantly differ between treatment of the same genotype groups (p > 0.999, 63 Cntnap2-/-Psilocybin vs Cntnap2-/-Vehicle; p = 0.390, Cntnap2+/+Psilocybin vs Cntnap2+/+Vehicle, 64 Bonferroni’s post-hoc; Extended Data Fig 1a), indicating that the difference in sociability is 65 due to a specific deficit in social behaviour rather than a variability in the drive to explore. We 66 did not detect any significant effect of sex on our observations (p = 0.850, repeated measure 67 ANOVA with Bonferroni’s post-hoc). 68 69 70 71 72 Figure 1. Psilocybin durably restores sociability in Cntnap2-/- homozygous KO mice. 73 a. A single injection of psilocybin rescues the social deficits in Cntnap2-/- KO mice for up to two 74 weeks (Extended Tables 1a, 2). b. Psilocybin’s sociability rescuing effect is observed in 75 Cntnap2-/- KO mice tested for the first time at two weeks after a single treatment. c-d. The 76 significantly higher repetitive behaviour displayed by Cntnap2 -/- KO mice (c) is not altered by 77 psilocybin ( d). e-f. The significant hyperactivity displayed by Cntnap2-/- KO mice ( e) is not 78 altered by psilocybin ( f). Mean ± SEM shown. Individual data point shown. For detailed 79 statistics see Extended Data Table 1-3. *p < 0.05, ** p < 0.01, ***p < 0.001. 80 81 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted March 11, 2025. ; https://doi.org/10.1101/2025.03.10.642385doi: bioRxiv preprint 3 Further, we treated a subset of Cntnap2-/- KO mice with a single administration of psilocybin 82 or saline vehicle, and tested for their sociability for the first time at two weeks after treatment. 83 Here we also observed significantly higher sociability in psilocybin treated Cntnap2-/- KO mice 84 (p = 0.025, Cntnap2 -/-Psilocybin vs Cntnap2-/-Vehicle; Fig 1b, Extended Data Table 2a). This 85 observation indicated that our observation of psilocybin effect was indeed a sustained rescue 86 of social behaviour rather than an effect of a repeated testing design. Given this observation, 87 we next sought to examine each timepoints of our repeated measure observation, and 88 detected the greatest psilocybin rescuing effect at 1 week post-administration (at 1 day: p = 89 0.007, at 1 week: p < 0.001, at 2 weeks: p = 0.028, Cntnap2-/-Psilocybin vs Cntnap2-/-Vehicle, Fig 1a, 90 Extended Data Table 2b). 91 92 No detectable effects of psilocybin on two other behavioural measures. 93 94 Then, we explored whether psilocybin also affected other aspects of behaviour that are not 95 directedly related to sociability but are possibly affected in ASD model mice8,11. We examined 96 repetitive behaviour by quantifying self-grooming activity during habituation. We observed that 97 Cntnap2-/- homozygous mutant KO mice displayed significantly higher number of self-98 grooming episodes compared to Cntnap2+/+ homozygous WT mice during the final habituation 99 session prior to drug treatments (p = 0.002, Cntnap2+/+ vs Cntnap2-/-, Mann Whitney, Fig 1c). 100 No significant difference was detected in the number of self -grooming episodes displayed by 101 Cntnap2-/- KO mice at 1 day after psilocybin treatment (p = 0.329, Cntnap2-/-Vehicle vs Cntnap2-102 /-Psilocybin, Mann Whitney, Fig 1d) or at 1 week after drug treatment (p = 0.461, Cntnap2-/-Vehicle 103 vs Cntnap2-/-Psilocybin, Mann Whitney, Fig 1d). We therefore concluded that repetitive behaviour, 104 although increased in Cntnap2-/- KO mice, is not affected by psilocybin. 105 106 Further, we tracked the positions of the test mice during the habituation phase without the 107

of a social partner. In line with previous reports of hyperactivity in Cntnap2 -/- KO 108 mice8, we observed that Cntnap2 -/- KO mice displayed significantly higher distance travelled 109 during the 10 minutes observation period compared to wildtype (p = 0.001, Cntnap2 +/+ vs 110 Cntnap2-/-, Dunn’s post-hoc, Fig 1e). Despite drug treatment, no significant difference was 111 detected in the distance travelled by Cntnap2-/- KO mice at 1 day after drug treatment (p = 112 0.518, Cntnap2-/-Vehicle vs Cntnap2-/-Psilocybin, Mann Whitney) or at 1 week after drug treatment 113 (p = 0.976, Cntnap2-/-Vehicle vs Cntnap2-/-Psilocybin, Mann Whitney). We therefore concluded that 114 psilocybin has no lasting effect on spontaneous locomotory drive. 115 116 Psilocybin does not enhance social behaviour in the Cntnap2 genotypes without social 117 deficit. 118 119 Interestingly, the single administration of psilocybin reduced sociability in Cntnap2+/+ isogenic 120 wild-type group, at 1 day post-treatment (p = 0.001, Cntnap2 +/+Vehicle vs Cntnap2 +/+Psilocybin). 121 Psilocybin-treated isogenic wild-type mice remained lower in sociability scores at 1 week (p = 122 0.155, Cntnap2+/+Vehicle vs Cntnap2+/+Psilocybin) and at 2 weeks (p = 0.207 , Cntnap2+/+Vehicle vs 123 Cntnap2+/+Psilocybin) after treatment although no longer statistically significantly (Fig 1a). In line 124 with the idea that CNTNAP2 is a recessive gene 12, heterozygous Cntnap2 +/-Vehicle group did 125 not show significantly altered sociability compared to wildtype Cntnap2+/+Vehicle group (p = 1.000, 126 Cntnap2+/-Vehicle vs Cntnap2+/+Vehicle, Bonferroni’s post-hoc , Extended Data Table 1b). 127 Cntnap2+/- heterozygous mice also displayed reduced sociability after psilocybin treatment 128 (Extended Data Fig 1b ), even though not statistically significantly. These observations 129 indicate that while the prosocial effect of psilocybin has potential to rescue abnormal 130 behavioural traits, in the absence of behavioural deficits psilocybin does not enhance and may 131 even reduce sociability. 132 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted March 11, 2025. ; https://doi.org/10.1101/2025.03.10.642385doi: bioRxiv preprint 4 Lack of correlation between sociability and the acute effect of psilocybin. 133 134 Psilocybin or its active metabolite psilocin, like other psychedelic 5HT2AR agonists, trigger an 135 acute head twitch response (HTR)13 that lasts for the duration of the agonist’s bioavailability14. 136 Immediately after psilocybin application, w e assayed the HTR for a 20 minutes observation 137 period to confirm the normal functioning of 5HT2ARs in psilocybin-treated mice (Extended 138 Data Fig 1c , Extended Data Table 4a). No difference in HTR was observed between the 139 different genotype groups without psilocybin (p = 0.825, Cntnap2+/+Vehicle vs Cntnap2-/- Vehicle; p > 140 0.999, Cntnap2 +/+Vehicle vs Cntnap2 +/-Vehicle, Dunn’s post-hoc ), and psilocybin significantly 141 increased HTR in all genotype groups (p < 0.001, Cntnap2+/+Vehicle vs Cntnap2+/+Psilocybin; p = 142 0.001, Cntnap2-/-Vehicle vs Cntnap2-/-Psilocybin; p = 0 .006, Cntnap2+/-Vehicle vs Cntnap2+/-Psilocybin; 143 Dunn’s post-hoc) to a similar level (p > 0.9999, Cntnap2+/+Psilocybin vs Cntnap2-/-Psilocybin; p > 144 0.9999, Cntnap2 +/+Psilocybin vs Cntnap2+/-Psilocybin; Dunn’s post-hoc). Even though sex -related 145 differences in HTR have been reported for other psychedelic compounds and analoqgues15,16, 146 we did not detect any significant effect of sex on the HTR (p = 0.452, Cntnap2-/-Psilocybin::Male vs 147 Cntnap2-/-Psilocybin::Female, Mann Whitney; Extended Data Table 4b) which is similar to a lack of 148 sex-specific effects on psilocybin-induced HTR as observed by others 17. Moreover, w e 149 observed no significant correlation between acute HTR and sociability levels in the Cntnap2-/- 150 KO mice (Extended Data Table 4c ). This indicates that the reduced sociability in Cntnap2 -/- 151 KO mice cannot be explained by an altered acute function of 5HT2ARs. 152 153 DNA methyltransferase inhibition blocks psilocybin-induced sociability changes in 154 Cntnap2-/- homozygous mutant KO mice to a similar extent as 5HT2AR antagonism. 155 156 Given that the observed sociability-rescuing effect outlasts the bioavailability of psilocybin, we 157 reasoned that the lasting alterations in sociability following psilocybin administration may 158 originate from mechanisms other than sustained 5HT2AR activation18. We reasoned that 159 epigenetic alterations may provide a potential mechanism for the observed lasting social deficit 160 rescuing effects of psilocybin. We focused on changes in DNA methylation in view of reports 161 showing differential DNA methylation patterns in autism spectrum disorder19,20. 162 163 We pretreated mice with systemic administration of RG10821–23 - a small-molecule inhibitor of 164 DNA methyltransferase I (DNMT1) - to test whether DNMT inhibition (DNMTi) prevents the 165 effects from psilocybin administration. We then behaviourally tested the mice for sociability at 166 1 day and 1 week after treatment (Fig. 2). When pretreated with vehicle, psilocybin treatment 167 rescues the reduced sociability in homozygous KO mice (p < 0.001, Cntnap2 -/-Vehicle::Vehicle vs 168 Cntnap2-/-Vehicle::Psilocybin, Bonferroni’s post-hoc), similar to our earlier observations . DNMTi 169 pretreatment significantly reduced the psilocybin-induced increased sociability in Cntnap2 -/- 170 KO mice (p = 0.003, Cntnap2 -/-Vehicle::Psilocybin vs Cntnap2-/-DNMTi::Psilocybin, Bonferroni’s post-hoc) 171 and psilocybin treatment is no longer effective (p = 1.000, Cntnap2 -/-DNMTi::Psilocybin vs Cntnap2-172 /-DNMTi::Vehicle, Bonferroni’s post-hoc). DNMTi pretreatment is similarly effective at 1 day and 1 173 week after administration ( Extended Data Table 6). This indicates that de novo DNA 174 methylation is required for the sustained effect of psilocybin on sociability in Cntnap2 -/- KO 175 mice. DNMTi pretreatment itself did not affect sociability in Cntnap2-/- KO mice (p = 1.000, 176 Cntnap2-/-Vehicle::Vehicle vs Cntnap2-/-DNMTi::Vehicle, Bonferroni’s post-hoc), or in Cntnap2+/+ isogenic 177 WT mice (p = 0.526, Cntnap2 +/+Vehicle::Vehicle vs Cntnap2 +/+DNMTi::Vehicle, Repeated measure 178 ANOVA). Additionally, we confirmed that DNMTi pretreatment in isogenic WT mice (Extended 179 Data Table 7) does not have an effect on the number of self -grooming episodes (p = 0.222, 180 Cntnap2+/+Vehicle::Vehicle vs Cntnap2+/+DNMTi::Vehicle, Mann Whitney), or total distance travelled (p = 181 0.128, Cntnap2+/+Vehicle::Vehicle vs Cntnap2+/+DNMTi::Vehicle, Mann Whitney). 182 183 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted March 11, 2025. ; https://doi.org/10.1101/2025.03.10.642385doi: bioRxiv preprint 5 184 185 Figure 2. Psilocybin’s sociability rescuing effect is blocked by DNMTi. 186 a. Psilocybin’s sociability rescuing effect can be blocked by pretreatment of DNMTi to a similar 187 extent as 5HT2AR antagonism. b. The pretreatment blockade effects are sustained at 1 week 188 after treatment. Mean ± SEM shown. Individual data point shown. For detailed statistics see 189 Extended Data Table 5-6. ** p < 0.01. 190 191 192 Psilocybin is known to act predominantly through the activation of serotonin 2A receptors 193 (5HT2ARs)24–26. To test whether the observed effects of psilocybin were indeed initiated by 194 5HT2AR activation, we also performed similar experiments using pretreatment with a specific 195 5HT2AR antagonist volinanserin (MDL100907). We observed that 5HT2AR antagonist 196 pretreatment prevented the psilocybin-induced increased sociability in Cntnap2 -/- KO mice (p 197 = 0.001, Cntnap2 -/-Vehicle::Psilocybin vs Cntnap2 -/-5HT2AR-ant::Psilocybin groups, Bonferroni’s post-hoc; 198 Fig 2, Extended Data Table 5 ). When comparing mice pretreated with DNMTi or 5HT2AR 199 antagonist, we found no significant difference in psilocybin’s effect on sociability (p = 1.000, 200 Cntnap2-/-DNMTi::Psilocybin vs Cntnap2 -/-5HT2AR-ant::Psilocybin groups, Bonferroni’s post-hoc ; Fig 2, 201 Extended Data Table 5), which indicates that DNMTi is as effective as 5HT2AR antagonist 202 in blocking psilocybin’s effect on rescuing sociability. Further, we observed that the blocking 203 effect of both pretreatment groups were sustained at 1 week after treatment (Fig 2b ; 204 Extended Data Table 6 ). This indicated that both de novo DNA methylation and 5HT2AR 205 activation are required for the sustained effect of psilocybin on sociability in Cntnap2-/- KO mice. 206 207 While pretreatment with 5HT2AR antagonist effectively abolished the psilocybin -induced 208 head-twitch response (p 210 0.999, Cntnap2 -/-Vehicle::Psilocybin vs Cntnap2 -/-DNMTi::Psilocybin groups, Dunn’s post-hoc) or with 211 vehicle (p > 0.999, Cntnap2 -/-Vehicle::Psilocybin vs Cntnap2-/-DNMTi::Psilocybin groups, Dunn’s post-hoc; 212 Extended Data Fig 2, Extended Data Table 8). These observations indicate that the effect 213 of pretreatment with DNMTi did not interfere with the normal acute effect of 5HT2AR activation 214 by psilocybin, or spontaneous head-twitch events in Cntnap2-/- KO mice (p > 0.999, Cntnap2-215 /-Vehicle::Vehicle vs Cntnap2-/-DNMTi::Vehicle groups, Dunn’s post-hoc; Extended Data Fig 2, Extended 216 Data Table 8). 217 218 219 220 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted March 11, 2025. ; https://doi.org/10.1101/2025.03.10.642385doi: bioRxiv preprint 6

221 222 Taken together, our observations demonstrate that psilocybin rescues social deficits in the 223 Cntnap2 knock-out mouse model of altered sociability in ASD, and this effect is sustained for 224 at least up to two weeks. Our results further suggest that the long-lasting effect of psilocybin 225 depends on de novo DNA methylation dynamics, since pretreatment of DNA 226 methyltransferase inhibitor is as effective in blocking psilocybin’s therapeutic effect as 227 antagonism of psilocin’s main binding target 5HT2A receptor. Our work here supports the 228 rapidly expanding body of experimental evidence that psychedelics have the capability to exert 229 sustained effects on brain function and behaviour, and an involvement of epigenetic 230 mechanisms in this durable effect. 231 232 Despite this promising therapeutic effect, we also note that psilocybin does not improve 233 sociability in mice that do not show behavioural deficits, such as that of the Cntnap2 isogenic 234 WT or heterozygote groups here. Our observations even indicate that psilocybin exerts 235 opposing effects, reducing sociability in the absence of abnormal sociability traits . Thus, it 236 appears that psilocybin is not a general sociability enhancer, but rather is only beneficial when 237 in deficit, a clinically relevant behavioural measure for brain disorders or mental health issues. 238 When viewing psilocybin administration as a pharmacological perturbation, its counteracting 239 effects suggest the existence of a setpoint that governs social behaviour. Our observation that 240 DNA methylation plays a role in psilocybin’s sustained social deficit rescuing effect further 241 suggests that the setpoint for social behaviour may be controlled and/or reconfigured through 242 epigenetic plasticity mechanisms. In line with this, others have observed that a single 243 administration of other serotonergic psychedelic such as lysergic acid diethylamide (LSD) into 244 control mice also did not enhance social behaviour27. However, social enhancement could be 245 achieved through repeated LSD treatments27, and repeated LSD treatments into control mice 246 has been separately reported to result in detectable changes in DNA methylation in the brain28. 247 Considering this together with our observations, these experimental evidence collectively 248 suggest that different behavioural setpoints may exist for normal vs deficit states, which may 249 be adjusted with varying levels of ease. 250 251 Separately, the single administration of another serotonergic psychedelic 1-(2,5-dimethoxy-4-252 iodophenyl)-2-aminopropane (DOI) led to detectable changes in chromatin organisation and 253 enhancer activity, associated with increased synaptogenesis29. Psilocybin may similarly lead 254 to epigenomic changes including alteration of DNA methylation dynamics, which may underpin 255 our observations here. In the case of the socially reduced Cntnap2-/- homozygous mutant KO 256 mice, we observed that pretreatment with DNMTi is not the exclusive mechanism to block 257 psilocybin’s therapeutic effects, but it is as effective as pretreatment with 5HT2AR antagonism. 258 This suggests that psilocybin-induced changes in DNA methylation dynamics is downstream 259 of 5HT2AR activation. Future work will be necessary to identify the signalling pathways 260 involved. 261 262 Further, we noted a general decline in sociability across postnatal weeks 8-10 for all 263 experimental groups in line with previously reported age-dependent reduction in social-related 264 behaviour in mice 30. This decline is much more significant in Cntnap2 KO mice (p = 0.003 , 265 Cntnap2-/-vehicle:1 day vs Cntnap2-/- vehicle:1 week; p = 0.532, Cntnap2-/- vehicle: 1 week vs Cntnap2-/- vehicle: 266 2 weeks; p = 0.008, Cntnap2-/- vehicle:1 day vs Cntnap2-/- vehicle: 2 weeks; T test) and is delayed following 267 psilocybin treatment (p = 0.120, Cntnap2 -/-psilocybin:1-day vs Cntnap2 -/-psilocybin: 1 week ; p = 0.040 , 268 Cntnap2-/-psilocybin: 1 week vs Cntnap2-/-psilocybin: 2 weeks; p = 0.002, Cntnap2-/-psilocybin:1 day vs Cntnap2-/-269 psilocybin: 2 weeks; T test). It will be interesting to investigate whether psilocybin extends or even re-270 opens the critical developmental window where experience-dependent components of 271 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted March 11, 2025. ; https://doi.org/10.1101/2025.03.10.642385doi: bioRxiv preprint 7 sociability traits are the most plastic, and indeed, latest experimental observations by others 272 indicate that psychedelics has the potential to re-open the critical period for social reward 273 learning31. This appears to be in line with our observations here that while psilocybin was 274 effective in rescuing social deficits, it did not alter the increased self- grooming repetitive 275 behaviour or hyperactivity associated with this ASD model 8, which may not involve critical 276 periods. 277 278 279

280 281 Animals. 282 283 Experimental procedures at Nanyang Technological University Singapore were performed in 284 accordance with approved protocols following review by the Institutional Animals Care and 285 Use Committee. Experimental procedures performed at Imperial College London UK in 286 accordance with the United Kingdom Animal Scientific Procedures Act (1986), under Home 287 Office Personal and Project Licences following appropriate ethical review. Cntnap2 -/- 288 homozygous mutant KO8 transgenic mice (JAX017482, Jackson Labs USA) were maintained 289 on a C57bl6/J background. Offsprings were weaned at P21 and housed in groups of up to five 290 animals per cage after weaning. All animals were maintained in individually ventilated cages, 291 on a 12/12 h light/dark cycle (light on between 07:00-19:00) at 21 ± 2ºC and 55 ± 10% humidity. 292 Water and food were provided ad libitum. 8-10 weeks old mice of both genders were used for 293 behavioural experiments. Mice were group-housed where possible, with the same littermate 294 mice both before and after treatments. Mice in the same cage typically included treatments of 295 different conditions and/or included untreated companion mice. All experiments were 296 conducted during the light phase. 297 298 Drug administration. 299 300 Psilocybin (Usona Institute US) was dissolved in sterile saline solution (0.9% NaCl) to 301 appropriate concentrations. 8-week old offspring were intraperitoneally injected either with 302 saline or psilocybin (1 mg/kg) 24 hours prior to behavioural testing. We chose this dose of 303 psilocybin based on our previous observations 32 as well as observations from several other 304 groups17,33. For the duration of the acute drug effect, each mouse was placed in a separate 305 chamber to assess the head-twitch response. 306 307 Head-twitch response. 308 309 Head-twitch events were evaluated immediately after intraperitoneal administration of either 310 saline or psilocybin for a period of 20 minutes. The mouse was placed into a customized 311 behavioural box immediately after injection and was free to explore. The number of head-312 twitch events were counted by direct observation in 5-minute bins. The box is of size 40 cm X 313 20 cm X 20 cm [L x W x H] and the floor was covered with clean bedding. Experimenters were 314 not blind to the treatment for head-twitch counts. 315 316 DNA methyltransferase I inhibition. 317 318 We used the small -molecule DNMT1 inhibitor RG108 (Tocris, UK) to inhibit DNA 319 methyltransferase I in vivo21–23. RG108 was dissolved in 10% DMSO then diluted using saline 320 (final DMSO received by the mice was 0.75%). Two doses of 10 mg/kg RG108 or vehicle was 321 systemically administered via intraperitoneal injections per test animal, first dose at 60 min 322 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted March 11, 2025. ; https://doi.org/10.1101/2025.03.10.642385doi: bioRxiv preprint 8 prior to psilocybin or saline treatment, second dose together with psilocybin or saline treatment. 323 We selected RG108 due to its low cytotoxicity and its selective binding to DNA 324 methyltransferase I. 325 326 5HT2A receptor antagonism. 327 328 We used the selective 5HT2A receptor antagonist MDL100907 (Tocris, UK) to inhibit 5HT2A 329 receptor activity in vivo34. MDL100907 was dissolved in 10% DMSO then diluted using saline. 330 One dose of 1 mg/kg MDL100907 or vehicle was systemically administered via intraperitoneal 331 injections per test animal at 15 min prior to psilocybin or saline treatment. 332 333 Behavioural testing. 334 335 Social behaviour was tested in a custom -built Plexiglass three-chamber box with matt white 336 acrylic floor. The total arena (60 x 40 x 22 cm [L x W x H]) was divided in three smaller evenly 337 sized chambers (20 x 40 x 22 cm [L x W x H]) interconnected by 4 x 4 cm doors. Transparent 338 cylinders (10.5 cm internal diameter, 11 cm external diameter, 16 cm length) with 1 -cm slits 339 were used as cups. The box was placed in a darkened and quite room, illuminated from above 340 with infrared LEDs located 1 meter over the arena. 341 342 Age-matched, sex -matched non-littermate mice with no social behavioural deficits (i.e. no 343 homozygous mutant KO mice) from the same genetic background were used as social 344 partners. 345 346 Animals were acclimatised to the testing room for at least one hour before the start of 347 experiment on each day. Behavioural testing took place over 4 consecutive days in the first 348 week. Each test mouse was placed into the centre chamber and allowed to freely explore and 349 habituate to the three-chamber setup for 10 minutes over 3 consecutive days, doors to the 350 side chambers were kept open. Social partner mice were individually pre-habituated to the 351 cups also for 10 minutes over 3 consecutive days. On day 4, each test animal was allowed to 352 freely explore all three empty chambers for 10 minutes, then driven to the central chamber 353 with the doors closed while an age-, size -, and gender -matched unfamiliar non-littermate 354 mouse (S) was placed into a cup in one side chamber, and an empty cup (C) was placed in 355 the opposite side chamber. The side of the social partner chamber was randomised. The test 356 mouse was then allowed to freely explore chambers for 10 minutes. Re-testing sessions at 357 day 7 and day 14 after administration were performed using the day 4 protocol, with new 358 stranger mice that the test mouse never encountered previously. Sex of the test mice were 359 randomised where possible. 360 361 All behavioural sessions were video-recorded using a CMOS camera (Basler acA2000-362 165umNIR) and Basler software (Basler AG, Germany). Video recordings were used to track 363 the position of the body of the animal in each chamber, as well as to manually time the duration 364 of the test mouse’s social exploratory behaviour by a blinded well -trained observer. We 365 observed that the test mice frequently spent time in a side chamber while avoiding the 366 interaction object, therefore we manually quantified the actual interaction duration of the test 367 mouse with the cup quantified as either sniffing or crawling on a cup that is either empty or 368 containing another mouse. We then derived a normalised sociability index using the equation: 369 370 𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆 𝑆𝑆 𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 = 𝐼𝐼𝑖𝑖𝑆𝑆 𝑖𝑖𝐼𝐼𝑆𝑆𝑆𝑆 𝑆𝑆𝑆𝑆𝑆𝑆𝑖𝑖 𝑖𝑖𝑑𝑑 𝐼𝐼𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑖𝑖 𝑤𝑤 𝑆𝑆𝑆𝑆ℎ 𝑆𝑆 − 𝐼𝐼𝑖𝑖 𝑆𝑆𝑖𝑖𝐼𝐼 𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑖𝑖 𝑖𝑖𝑑𝑑 𝐼𝐼𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑖𝑖 𝑤𝑤 𝑆𝑆𝑆𝑆ℎ 𝐶𝐶 𝐼𝐼𝑖𝑖𝑆𝑆 𝑖𝑖𝐼𝐼𝑆𝑆𝑆𝑆 𝑆𝑆𝑆𝑆𝑆𝑆𝑖𝑖 𝑖𝑖𝑑𝑑 𝐼𝐼𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑖𝑖 𝑤𝑤 𝑆𝑆𝑆𝑆ℎ 𝑆𝑆 + 𝐼𝐼𝑖𝑖𝑆𝑆 𝑖𝑖𝐼𝐼𝑆𝑆𝑆𝑆 𝑆𝑆𝑆𝑆𝑆𝑆𝑖𝑖 𝑖𝑖𝑑𝑑 𝐼𝐼𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑖𝑖 𝑤𝑤 𝑆𝑆𝑆𝑆ℎ 𝐶𝐶 371 372 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted March 11, 2025. ; https://doi.org/10.1101/2025.03.10.642385doi: bioRxiv preprint 9 Position tracking and analysis. 373 374 We used DeepLabCut35,36 to track the positions of the randomly selected test mouse during 375 the habituation phase. Each mouse was labelled by a trained network to identify at least 4 376 body parts including the snout, two ears, tail base. Body positions were then calculated from 377 the label coordinates and analysed using custom codes in MATLAB. Movement speed is 378 calculated as the difference between the body positions between two consecutive image 379 frames. 380 381 Chamber occupancy during the habituation phase of the test were calculated. No consistent 382 side preference were detected either for an experimental group or for the same mouse on 383 different test sessions. 384 385 Grooming analysis. 386 387 Episodes of grooming behaviour were manually scored by an experimenter who was blind to 388 the genotyping and treatment conditions. 389 390 Statistical analysis. 391 392 Head twitch response data were analysed using non-parametric Kruskal -Wallis test with 393 alpha-adjusted Dunn’s multiple post-hoc comparisons, or Mann Whitney test as a priori 394 specified. Data related to sociability index were analysed via three-factorial ANOVA with 395 repeated measures on one factor and followed up on by Bonferroni -corrected multiple 396 comparisons. Calculations were carried out using GraphPad Prism and SPSS, respectively. 397 Correlations (Spearman’s rho) were calculated using MATLAB. 398 399 400

401 402 We thank R Wood, B Glenn, C Day, D Macdonald, N Chua, K Lee, J Lim, Z Yong for 403 assistance with animal husbandry. We thank MJ Hossen for technical assistance. We thank 404 G Pearson, I Mollinedo-Gajate, T Lyons and all other members of the Knopfel lab for 405 discussions and feedback on the manuscript. We thank CC Ng, EPY Wong, R Dienzo, MF 406 Basir, BW Ng, P Teo, KL Lim, S Ponniah, A Tashiro, K Wong, G Augustine, W Tay, QJ Liew, 407 YS Yip, CY Chin, J Ng, C Rantle, R Tyacke, D Nutt for administrative and infrastructure support. 408 This work was supported by: US National Institute of Health BRAIN Initiative Grant 409 (5U01NS099573) to TK, Hong Kong STEM Initiative Funds (Hong Kong Jockey Club Charities 410 Trust) to TK, Research Matching Grant (Government of Hong Kong, TFD2023-P05-02) to TK, 411 Lee Kuan Yew Postdoctoral Fellowship administered by Nanyang Technological University 412 Singapore (022506-00001) to CS, and Open Fund Young Individual Research Grant ( MOH-413 001720) administered by the Singapore Ministry of Health’s National Medical Research 414 Council to CS . This work was supported by the Investigational Supply Program of Usona 415 Institute US. 416 417 418 Contributions 419 420 CS and TK designed the study. CS and TC performed the experiments. CS, TK and TB 421 analysed the data. CS and TK supervised the project and wrote the manuscript with inputs 422 from all authors. 423 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted March 11, 2025. ; https://doi.org/10.1101/2025.03.10.642385doi: bioRxiv preprint 10 424 Ethics declarations and competing interests 425 426 The authors report no competing interests. 427 428 Data availability 429 430 Data is available from the corresponding author upon reasonable request. 431 432 433

434 435 1. Nutt, D., Erritzoe, D. & Carhart-Harris, R. Psychedelic Psychiatry’s Brave New World. 436 Cell 181, 24–28 (2020). 437 2. Danforth, A. Psychedelic -Assisted Therapy for Social Adaptability in Autistic Adults. 438 Curr. Top. Behav. Neurosci. 56, 71–92 (2022). 439 3. Davis, A. K. et al. Effects of Psilocybin-Assisted Therapy on Major Depressive Disorder: 440 A Randomized Clinical Trial. JAMA Psychiatry 78, 481–489 (2021). 441 4. Carhart-Harris, R. L. et al. Psilocybin with psychological support for treatment-resistant 442 depression: six-month follow-up. Psychopharmacology (Berl.) 235, 399–408 (2018). 443 5. Goodwin, G. M. et al. Single-Dose Psilocybin for a Treatment-Resistant Episode of 444 Major Depression. N. Engl. J. Med. 387, 1637–1648 (2022). 445 6. Spencer, C. M., Alekseyenko, O., Serysheva, E., Yuva-Paylor, L. A. & Paylor, R. 446 Altered anxiety-related and social behaviors in the Fmr1 knockout mouse model of fragile X 447 syndrome. Genes Brain Behav. 4, 420–430 (2005). 448 7. Qin, L. et al. Social deficits in Shank3-deficient mouse models of autism are rescued 449 by histone deacetylase (HDAC) inhibition. Nat. Neurosci. 21, 564–575 (2018). 450 8. Peñagarikano, O. et al. Absence of CNTNAP2 leads to epilepsy, neuronal migration 451 abnormalities, and core autism-related deficits. Cell 147, 235–246 (2011). 452 9. Peñagarikano, O. et al. Exogenous and evoked oxytocin restores social behavior in 453 the Cntnap2 mouse model of autism. Sci. Transl. Med. 7, 271ra8 (2015). 454 10. Levy, D. R. et al. Dynamics of social representation in the mouse prefrontal cortex. Nat. 455 Neurosci. 22, 2013–2022 (2019). 456 11. Cording, K. R., Tu, E. M., Wang, H., Agopyan-Miu, A. H. C. W. & Bateup, H. S. Cntnap2 457 loss drives striatal neuron hyperexcitability and behavioral inflexibility. BioRxiv Prepr. Serv. 458 Biol. 2024.05.09.593387 (2024) doi:10.1101/2024.05.09.593387. 459 12. Strauss, K. A. et al. Recessive symptomatic focal epilepsy and mutant contactin-460 associated protein-like 2. N. Engl. J. Med. 354, 1370–1377 (2006). 461 13. Halberstadt, A. L., Chatha, M., Klein, A. K., Wallach, J. & Brandt, S. D. Correlation 462 between the potency of hallucinogens in the mouse head-twitch response assay and their 463 behavioral and subjective effects in other species. Neuropharmacology 167, 107933 (2020). 464 14. Jones, N. T. et al. Transient Elevation of Plasma Glucocorticoids Supports Psilocybin-465 Induced Anxiolysis in Mice. ACS Pharmacol. Transl. Sci. 6, 1221–1231 (2023). 466 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted March 11, 2025. ; https://doi.org/10.1101/2025.03.10.642385doi: bioRxiv preprint 11 15. Dunlap, L. E. et al. Identification of Psychoplastogenic N,N -467 Dimethylaminoisotryptamine (isoDMT) Analogues through Structure- Activity Relationship 468 Studies. J. Med. Chem. 63, 1142–1155 (2020). 469 16. Jaster, A. M. et al. Differences across sexes on head-twitch behavior and 5-HT2A 470 receptor signaling in C57BL/6J mice. Neurosci. Lett. 788, 136836 (2022). 471 17. Shao, L.-X. et al. Psilocybin induces rapid and persistent growth of dendritic spines in 472 frontal cortex in vivo. Neuron 109, 2535-2544.e4 (2021). 473 18. Vargas, M. V. et al. Psychedelics promote neuroplasticity through the activation of 474 intracellular 5-HT2A receptors. Science 379, 700–706 (2023). 475 19. Tremblay, M. W. & Jiang, Y.-H. DNA Methylation and Susceptibility to Autism 476 Spectrum Disorder. Annu. Rev. Med. 70, 151–166 (2019). 477 20. Nardone, S. et al. DNA methylation analysis of the autistic brain reveals multiple 478 dysregulated biological pathways. Transl. Psychiatry 4, e433 (2014). 479 21. Day, J. J. et al. DNA methylation regulates associative reward learning. Nat. Neurosci. 480 16, 1445–1452 (2013). 481 22. LaPlant, Q. et al. Dnmt3a regulates emotional behavior and spine plasticity in the 482 nucleus accumbens. Nat. Neurosci. 13, 1137–1143 (2010). 483 23. Argyrousi, E. K. et al. Effects of DNA methyltransferase inhibition on pattern separation 484 performance in mice. Neurobiol. Learn. Mem. 159, 6–15 (2019). 485 24. Madsen, M. K. et al. Psychedelic effects of psilocybin correlate with serotonin 2A 486 receptor occupancy and plasma psilocin levels. Neuropsychopharmacol. Off. Publ. Am. Coll. 487 Neuropsychopharmacol. 44, 1328–1334 (2019). 488 25. Chiu, Y.-T. et al. A suite of engineered mice for interrogating psychedelic drug actions. 489 BioRxiv Prepr. Serv. Biol. 2023.09.25.559347 (2023) doi:10.1101/2023.09.25.559347. 490 26. González-Maeso, J. et al. Hallucinogens recruit specific cortical 5-HT(2A) receptor -491 mediated signaling pathways to affect behavior. Neuron 53, 439–452 (2007). 492 27. De Gregorio, D. et al. Lysergic acid diethylamide (LSD) promotes social behavior 493 through mTORC1 in the excitatory neurotransmission. Proc. Natl. Acad. Sci. U. S. A. 118, 494 e2020705118 (2021). 495 28. Inserra, A. et al. Modulation of DNA methylation and protein expression in the 496 prefrontal cortex by repeated administration of D-lysergic acid diethylamide (LSD): Impact on 497 neurotropic, neurotrophic, and neuroplasticity signaling. Prog. Neuropsychopharmacol. Biol. 498 Psychiatry 119, 110594 (2022). 499 29. de la Fuente Revenga, M. et al. Prolonged epigenomic and synaptic plasticity 500 alterations following single exposure to a psychedelic in mice. Cell Rep. 37, 109836 (2021). 501 30. Nardou, R. et al. Oxytocin-dependent reopening of a social reward learning critical 502 period with MDMA. Nature 569, 116–120 (2019). 503 31. Nardou, R. et al. Psychedelics reopen the social reward learning critical period. Nature 504 618, 790–798 (2023). 505 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted March 11, 2025. ; https://doi.org/10.1101/2025.03.10.642385doi: bioRxiv preprint 12 32. Mollinedo-Gajate, I. et al. Psilocybin Rescues Sociability Deficits in an Animal Model 506 of Autism . http://biorxiv.org/lookup/doi/10.1101/2020.09.09.289348 (2020) 507 doi:10.1101/2020.09.09.289348. 508 33. Hesselgrave, N., Troppoli, T. A., Wulff, A. B., Cole, A. B. & Thompson, S. M. 509 Harnessing psilocybin: antidepressant-like behavioral and synaptic actions of psilocybin are 510 independent of 5-HT2R activation in mice. Proc. Natl. Acad. Sci. U. S. A. 118, e2022489118 511 (2021). 512 34. de la Fuente Revenga, M., Vohra, H. Z. & González -Maeso, J. Automated 513 quantification of head-twitch response in mice via ear tag reporter coupled with biphasic 514 detection. J. Neurosci. Methods 334, 108595 (2020). 515 35. Mathis, A. et al. DeepLabCut: markerless pose estimation of user -defined body parts 516 with deep learning. Nat. Neurosci. 21, 1281–1289 (2018). 517 36. Nath, T. et al. Using DeepLabCut for 3D markerless pose estimation across species 518 and behaviors. Nat. Protoc. 14, 2152–2176 (2019). 519 520 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted March 11, 2025. ; https://doi.org/10.1101/2025.03.10.642385doi: bioRxiv preprint Lasting effect of psilocybin on sociability can be blocked by DNA methyltransferase inhibition Overview of Extended Data Extended Figures Extended Figure 1: Parameters related to psilocybin treatment of Cntnap2 mice. Extended Figure 2: Head-twitch response related to the pretreatment experiment. Extended Tables Extended Table 1: Statistics for durability of psilocybin-induced effects on sociability, relates to Fig 1. Extended Table 2 : Statistics for durability of psilocybin- induced effects on sociability by test session, relates to Fig 1. Extended Table 3: Statistics for repetitive behaviour and locomotion, relates to Fig 1. Extended Table 4: Statistics for head-twitch response, and HTR-SI correlations. Extended Table 5: Statistics for either DNMTi or 5HT2AR antagonist pretreatment on psilocybin’s sociability effect, relates to Fig 2. Extended Table 6: Statistics for either DNMTi or 5HT2AR antagonist pretreatment on psilocybin’s sociability effect by test session, relates to Fig 2. Extended Table 7: Statistics for the effects of DNMTi on the behaviour of isogenic wild-type mice. Extended Table 8: Statistics for either DNMTi or 5HT2AR antagonist pretreatment on psilocybin’s effect on HTR. .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted March 11, 2025. ; https://doi.org/10.1101/2025.03.10.642385doi: bioRxiv preprint Extended Data Figure 1. a. Total approach time for at 1 day (left) and 1 week (right) after treatment. b. Sociability at 1 day, 1 week and 2 weeks post-treatment (1 mg/kg psilocybin or vehicle control) for Cntnap2+/- heterozygote mice. c. Head-twitch response for all three Cntnap2 genotype groups. Mean ± SEM. Extended Data Figure 2. Head-twitch response for pretreatment experiments. Mean ± SEM. .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted March 11, 2025. ; https://doi.org/10.1101/2025.03.10.642385doi: bioRxiv preprint Extended Data Table 1. Statistics for durability of psilocybin-induced effects on sociability, relates to Fig 1. Comparison group Test used Main effect P value Degrees of freedom Group <0.001 3, 44 Time <0.001 2, 88 Group 0.049 3, 30 Time 0.023 2, 60 1.000 Extended Data Table 1b. Effect of psilocybin treatment on sociability in Cntnap2(+/-) compared to Cntnap2(+/+) mice Extended Data Fig. 1c Sociability index Cntnap2(+/+) & Psilocybin vs Cntnap2(+/-) & Psilocybin Bonferroni's post-hoc Cntnap2(+/+) & Vehicle vs Cntnap2(+/+) & Psilocybin Bonferroni's post-hoc <0.001 0.106 Bonferroni's post-hoc Cntnap2(+/-) & Vehicle vs Cntnap2(+/-) & Psilocybin Bonferroni's post-hoc 0.368 Repeated measure ANOVA Cntnap2(+/+) & Vehicle vs Cntnap2(+/-) & Vehicle Bonferroni's post-hoc 1.000 Extended Data Table 1a. Effect of psilocybin treatment on sociability in Cntnap2(-/-) compared to Cntnap2(+/+) mice Cntnap2(+/+) & Vehicle vs Cntnap2(-/-) & Vehicle Bonferroni's post-hoc 0.042 Sociability indexFig. 1a Cntnap2(+/+) & Vehicle vs Cntnap2(-/-) & Psilocybin Bonferroni's post-hoc 0.796 Repeated measure ANOVA Cntnap2(-/-) & Vehicle vs Cntnap2(-/-) & Psilocybin .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted March 11, 2025. ; https://doi.org/10.1101/2025.03.10.642385doi: bioRxiv preprint Extended Data Table 2. Statistics for durability of psilocybin-induced effects on sociability by test session, relates to Fig 1. Group Mean Median SEM 15 Cntnap2(-/-) & Vehicle 0.129 0.101 0.050 15 Cntnap2(-/-) & Psilocybin 0.328 0.260 0.079 19 Cntnap2(-/-) & Vehicle 0.374 0.380 0.058 15 Cntnap2(-/-) & Psilocybin 0.598 0.650 0.033 19 Cntnap2(-/-) & Vehicle 0.105 0.200 0.062 15 Cntnap2(-/-) & Psilocybin 0.498 0.500 0.053 15 Cntnap2(-/-) & Vehicle 0.028 0.100 0.104 15 Cntnap2(-/-) & Psilocybin 0.291 0.380 0.080 15 Cntnap2(+/+) & Vehicle 0.497 0.510 0.041 7 Cntnap2(+/+) & Psilocybin 0.151 0.130 0.076 15 Cntnap2(+/+) & Vehicle 0.331 0.260 0.080 7 Cntnap2(+/+) & Psilocybin 0.129 0.020 0.131 10 Cntnap2(+/+) & Vehicle 0.256 0.340 0.093 7 Cntnap2(+/+) & Psilocybin 0.149 0.140 0.088 Extended Data Table 2a. Sociability first tested at 2 weeks after treatment Sociability index Vehicle vs Psilocybin Independent T test 0.025 Comparison group Test used P valueDescriptive statsN Extended Data Table 2b. Sociability by test session Sociability index 1 day: Cntnap2(-/-) & Vehicle vs Cntnap2(-/-) & Psilocybin Independent T test 0.001 7 days: Cntnap2(-/-) & Vehicle vs Cntnap2(-/-) & Psilocybin Independent T test <0.001 14 day: Cntnap2(-/-) & Vehicle vs Cntnap2(-/-) & Psilocybin Independent T test 0.028 1 day: Cntnap2(+/+) & Vehicle vs Cntnap2(+/+) & Psilocybin Independent T test 14 days: Cntnap2(+/+) & Vehicle vs Cntnap2(+/+) & Psilocybin Independent T test 0.207 0.001 7 days: Cntnap2(+/+) & Vehicle vs Cntnap2(+/+) & Psilocybin Independent T test 0.107 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted March 11, 2025. ; https://doi.org/10.1101/2025.03.10.642385doi: bioRxiv preprint Extended Data Table 3. Statistics for repetitive behaviour and locomotion, relates to Fig 1. Genotype Treatment Mean Median SEM Cntnap2(+/+) 14 11.214 10.500 1.668 Cntnap2(-/-) 11 25.909 28.000 3.843 Cntnap2(-/-) Vehicle 5 30.200 26.000 2.577 Cntnap2(-/-) Psilocybin 7 33.857 31.000 3.284 Cntnap2(-/-) Vehicle 6 30.833 26.000 4.246 Cntnap2(-/-) Psilocybin 7 29.714 27.000 5.272 Cntnap2(+/+) 15 29200 28853 1737 Cntnap2(-/-) 22 38099 38158 1294 Cntnap2(+/+) 15 29200 28853 1737 Cntnap2(+/-) 10 32375 33378 2265 Cntnap2(-/-) Vehicle 11 38583 39653 1095 Cntnap2(-/-) Psilocybin 12 36990 37014 1620 Cntnap2(-/-) Vehicle 11 39930 41439 1526 Cntnap2(-/-) Psilocybin 12 40714 38568 2069 0.518 Mann Whitney 0.976 0.329 7 days after treatment 0.461 Test used N Fig. 1c Self- grooming episodes Fig. 1d Self- grooming episodes Test session Comparison group 1 day after treatment Mann Whitney Mann Whitney >0.999 Extended Data Table 3b. Locomotion 0.001 - Before treatment Descriptive stats Extended Data Table 3a. Repetitive behaviour 0.002 P value Before treatment Mann Whitney - Distance travelled Kruskal-Wallis with Dunn's post-hoc Before treatment Fig. 1d Distance travelled Fig. 1e 7 days after treatment 1 day after treatment Mann Whitney .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted March 11, 2025. ; https://doi.org/10.1101/2025.03.10.642385doi: bioRxiv preprint Extended Data Table 4. Statistics for head-twitch response, and HTR-SI correlations. Genotype Treatment Mean Median SEM Kruskal-Wallis <0.001 Cntnap2(+/+) Vehicle 9 2.444 2.000 0.341 Cntnap2(-/-) Vehicle 15 6.067 6.000 0.740 Cntnap2(+/+) Vehicle 9 2.444 2.000 0.341 Cntnap2(+/-) Vehicle 6 3.000 2.000 1.072 Cntnap2(+/+) Vehicle 9 2.444 2.000 0.341 Cntnap2(+/+) Psilocybin 7 22.429 23.000 1.462 Cntnap2(-/-) Vehicle 15 6.067 6.000 0.740 Cntnap2(-/-) Psilocybin 15 21.267 22.000 1.300 Cntnap2(+/-) Vehicle 6 3.000 2.000 1.072 Cntnap2(+/-) Psilocybin 6 19.500 18.500 2.332 Cntnap2(+/+) Psilocybin 7 22.429 23.000 1.462 Cntnap2(-/-) Psilocybin 15 21.267 22.000 1.300 Cntnap2(+/+) Psilocybin 7 22.429 23.000 1.462 Cntnap2(+/-) Psilocybin 6 19.500 18.500 2.332 Male Cntnap2(-/-) Psilocybin 6 19.167 21.500 2.151 Female Cntnap2(-/-) Psilocybin 9 22.667 22.000 1.546 Cntnap2(+/+) Psilocybin 0.110 Cntnap2(-/-) Psilocybin 0.143 Cntnap2(+/-) Psilocybin 0.111 Cntnap2(+/+) Psilocybin 0.081 Cntnap2(-/-) Psilocybin 0.763 Cntnap2(+/-) Psilocybin 0.880 Cntnap2(+/+) Psilocybin 0.213 Cntnap2(-/-) Psilocybin 0.766 Cntnap2(+/-) Psilocybin 0.406 P value / Adjusted P value Test used N Mann Whitney Dunn's multiple comparison >0.999 Extended Data Table 4b. HTR by sex 0.006 0.001 >0.999 Dunn's multiple comparison >0.999 Comparison group Dunn's multiple comparison Dunn's multiple comparison Descriptive stats Dunn's multiple comparison Extended Data Table 4a. Head twitch response Dunn's multiple comparison <0.001 Extended Data Fig. 1d HTR 0.825 Dunn's multiple comparison rho = 0.083 2 weeks after treatment Spearman's rho rho = -0.538 rho = 0.084 rho = 0.421 0.452 Extended Data Table 4c. HTR-SI correlations 1 day after treatment Spearman's rho rho = -0.655 rho = -0.397 rho = 0.714 1 week after treatment Spearman's rho rho = -0.699 rho = 0.085 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted March 11, 2025. ; https://doi.org/10.1101/2025.03.10.642385doi: bioRxiv preprint Extended Data Table 5. Statistics for pretreatment on psilocybin’s effect on sociability, relates to Fig 2. Genotype Pretreatment Group <0.001 6, 57 Time 0.001 1, 57 Cntnap2(-/-) Vehicle Vehicle Cntnap2(-/-) Vehicle Psilocybin Cntnap2(-/-) Vehicle Psilocybin Cntnap2(-/-) DNMTi Psilocybin Cntnap2(-/-) Vehicle Psilocybin Cntnap2(-/-) 5HT2AR-ant Psilocybin Cntnap2(-/-) DNMTi Psilocybin Cntnap2(-/-) DNMTi Vehicle Cntnap2(-/-) Vehicle Vehicle Cntnap2(-/-) DNMTi Vehicle Cntnap2(-/-) Vehicle Vehicle Cntnap2(-/-) 5HT2AR-ant Vehicle Cntnap2(-/-) DNMTi Psilocybin Cntnap2(-/-) 5HT2AR-ant Psilocybin Degrees of freedom Bonferroni's post-hoc 1.000 Bonferroni's post-hoc 0.001 Extended Data Table 5. Effect of pretreatment on psilocybin’s effect on sociability in Cntnap2(-/-) mice Fig. 2 Sociability index Repeated measure ANOVA Bonferroni's post-hoc Comparison group Test used Main effect P value Comparison group Bonferroni's post-hoc 1.000 Bonferroni's post-hoc 1.000 Bonferroni's post-hoc 1.000 <0.001 Bonferroni's post-hoc 0.003 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted March 11, 2025. ; https://doi.org/10.1101/2025.03.10.642385doi: bioRxiv preprint Extended Data Table 6. Statistics for pretreatment on psilocybin’s sociability effect by test session, relates to Fig 2. Genotype Pretreatment Treatment Mean Median SEM Vehicle Vehicle 10 0.207 0.220 0.038 Vehicle Psilocybin 10 0.514 0.530 0.060 Vehicle Psilocybin 10 0.514 0.530 0.060 DNMTi Psilocybin 11 0.258 0.290 0.065 Vehicle Psilocybin 10 0.514 0.530 0.060 5HT2AR-ant Psilocybin 9 0.177 0.210 0.059 DNMTi Psilocybin 11 0.258 0.290 0.065 DNMTi Vehicle 8 0.318 0.330 0.041 Vehicle Vehicle 10 0.207 0.220 0.038 DNMTi Vehicle 8 0.318 0.330 0.041 Vehicle Vehicle 10 0.207 0.220 0.038 5HT2AR-ant Vehicle 9 0.251 0.230 0.039 DNMTi Psilocybin 11 0.258 0.290 0.065 5HT2AR-ant Psilocybin 9 0.177 0.210 0.059 Vehicle Vehicle 10 0.077 0.110 0.066 Vehicle Psilocybin 10 0.475 0.525 0.090 Vehicle Psilocybin 10 0.475 0.525 0.090 DNMTi Psilocybin 11 0.155 0.140 0.080 Vehicle Psilocybin 10 0.475 0.525 0.090 5HT2AR-ant Psilocybin 9 0.166 0.190 0.069 DNMTi Psilocybin 11 0.155 0.140 0.080 DNMTi Vehicle 8 0.144 0.155 0.082 Vehicle Vehicle 10 0.077 0.110 0.066 DNMTi Vehicle 8 0.144 0.155 0.082 Vehicle Vehicle 10 0.077 0.110 0.066 5HT2AR-ant Vehicle 9 0.143 0.100 0.050 DNMTi Psilocybin 11 0.155 0.140 0.080 5HT2AR-ant Psilocybin 9 0.166 0.190 0.069 Independent T test 0.435 Independent T test 0.918 Descriptive stats N Independent T test Independent T test Independent T test Test used Sociability index 0.001 <0.001Independent T test Independent T test 0.007 0.432 7 days post- treatment Independent T test 0.537 Independent T test Independent T test 0.926 0.067 0.450 Independent T test 0.008 Independent T test P value 1 day post- treatment Comparison group 0.005 Independent T test <0.001 Independent T test 0.362 Extended Data Table 6. Sociability by test session Fig. 2 Cntnap2(-/-) Cntnap2(-/-) .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted March 11, 2025. ; https://doi.org/10.1101/2025.03.10.642385doi: bioRxiv preprint Extended Data Table 7. Statistics for the effects of DNMTi on behaviour of Cntnap2(+/+) mice. Genotype Pretreatment Treatment Vehicle Vehicle 0.526 DNMTi Vehicle 0.230 Genotype Pretreatment Treatment Mean Median SEM Vehicle Vehicle 7 9.571 7.000 2.608 DNMTi Vehicle 7 11.857 11.000 1.844 Vehicle Vehicle 7 29791 27924 1521 DNMTi Vehicle 7 25134 23588 2602 Degrees of freedom P value Locomotion 1 day after treatment Cntnap2(+/+) Self-grooming episodes Mann Whitney Mann Whitney 0.222 0.128 Descriptive stats Extended Data Table 7b. Effect of DNMTi pretreatment on other behaviour of Cntnap2(+/+) mice Comparison group Test used NTest session 1, 12 1, 12 Sociability index Cntnap2(+/+) Comparison group Extended Data Table 7a. Effect of DNMTi pretreatment on sociability of Cntnap2(+/+) mice Test used Main effect Group Time Repeated measure ANOVA P value .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted March 11, 2025. ; https://doi.org/10.1101/2025.03.10.642385doi: bioRxiv preprint Extended Data Table 8. Statistics for either DNMTi or 5HT2AR antagonist pretreatment on psilocybin’s effect on HTR. Genotype Pretreatment Treatment Mean Median SEM Kruskal-Wallis <0.001 Cntnap2(-/-) Vehicle Vehicle 10 5.500 4.000 1.529 Cntnap2(-/-) Vehicle Psilocybin 10 27.200 27.000 2.542 Cntnap2(-/-) Vehicle Psilocybin 10 27.200 27.000 2.542 Cntnap2(-/-) 5HT2AR-ant Psilocybin 9 0.333 0.000 0.236 Cntnap2(-/-) Vehicle Psilocybin 10 27.200 27.000 2.542 Cntnap2(-/-) DNMTi Psilocybin 11 28.636 26.000 3.231 Cntnap2(-/-) Vehicle Vehicle 10 5.500 4.000 1.529 Cntnap2(-/-) 5HT2AR-ant Vehicle 9 0.778 1.000 0.278 Cntnap2(-/-) Vehicle Vehicle 10 5.500 4.000 1.529 Cntnap2(-/-) DNMTi Vehicle 8 6.750 6.000 1.461 Test used Adjusted P valueN Dunn's multiple comparison Dunn's multiple comparison Dunn's multiple comparison Dunn's multiple comparison Comparison group Descriptive stats Extended Data Table 8. Head twitch response Extended Data Fig. 2 HTR Dunn's multiple comparison 0.335 >0.999 0.0208 0.999 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted March 11, 2025. ; https://doi.org/10.1101/2025.03.10.642385doi: bioRxiv preprint