Sensitization to the prosocial effects of 3,4-methylenedioxymethamphetamine (MDMA)

Male Swiss Webster mice (age 7–10 weeks) were used across nine independent experiments (total n = 172). Mice were group-housed except where noted, maintained on a 12-h light/dark cycle, and all procedures were approved by the institutional animal care committee. Drugs were administered intraperitoneally at established doses: MDMA 7.8 mg/kg, D-amphetamine 2 mg/kg, 5-HT2A R antagonist M100 (R(+)-MDL100,907) 1 mg/kg, and the 5-HT2A R agonist DOI 1 mg/kg; vehicles were saline. The primary behavioural assay was a 10-min dyadic social interaction test conducted in a clear plexiglass arena after a 25-min isolation period following drug or vehicle injection. Three behaviours were timed by observers blind to conditions and summed into a total social interaction score: anogenital investigation, general investigation (non-anogenital sniffing, following, grooming), and adjacent lying (huddling). For social-sensitization experiments, pairs of mice (the unit of analysis) received MDMA or saline every 48 h for four sessions (typically days 1, 3, 5, 7) with several cohorts used to test familiarity controls, persistence (off-drug pairing day 11 and drug challenge on day 21), and the role of social context (mice isolated versus paired for 2 h after MDMA treatment). To probe 5-HT2A R involvement, mice received M100 or saline 15 min before MDMA on days 1, 3 and 5, then MDMA without pretreatment on day 7 to test development. Separate cohorts received M100 only on day 7 to test expression. A parallel set of mice received DOI, MDMA or saline across the treatment sessions to assess sufficiency. Locomotor activity was measured in an open-field by photobeam breaks for 1 h after injection across four treatments; a D-amphetamine challenge occurred two weeks after the last MDMA/saline treatment to assess cross-sensitization. Microdialysis probes were implanted unilaterally targeting the nucleus accumbens to measure 5-HT overflow; eight mice underwent surgery with samples collected every 20 min before and after MDMA on the first and fourth treatment days (three mice were removed from analysis due to technical failures). 5-HT was quantified by HPLC with electrochemical detection. Statistical analyses treated pairs as the unit for social tests and used two-factor ANOVAs (treatment × test day) with Sidak's multiple comparisons for longitudinal sensitization; mixed-design or repeated-measures ANOVAs were used where appropriate. Between-day social comparisons were treated as between-subjects because novel pairings were used each day. Alpha was set at 5% and error bars represent SEM.

Repeated intermittent MDMA produced progressive increases in social interaction indicative of sensitization. Initial MDMA treatment did not differ from saline, but subsequent MDMA treatments produced steadily greater total social interaction. Two-factor ANOVA showed significant effects of treatment (F(1,28) = 115.2, p < 0.0001), day (F(3,28) = 50.1, p < 0.0001) and a treatment × day interaction (F(3,28) = 29.32, p < 0.0001). Social interaction after MDMA was significantly greater than controls on days 5 and 7 (p < 0.0001) and increased across MDMA sessions (day 7 > day 5, p = 0.0003). Both active (general investigation) and passive (adjacent lying) components increased across MDMA sessions while anogenital investigation did not; ANOVA effects for day, behaviour and their interaction were all highly significant. Control experiments indicated the effect was not due to familiarity with testing: mice given saline on days 1, 3 and 5 and MDMA only on day 7 showed no increase relative to saline controls (p = 0.7914) and were significantly lower than mice given MDMA across all four sessions (p < 0.0001). Off-drug pairing on day 11 (four days after last treatment) revealed no difference between MDMA- and saline-history mice (t(10) = 0.1591, p = 0.8767), but when challenged with MDMA two weeks later (day 21) previously MDMA-treated mice showed greater social interaction than controls (t(10) = 2.646, p = 0.0245), indicating a lasting drug-elicited enhancement. Social context during treatment was critical: mice isolated during MDMA treatment sessions did not develop sensitized social behaviour and were similar to first-time MDMA-treated mice on day 7 (isolation vs paired: p = 0.0214; isolated vs first-time MDMA: p = 0.9417). Evidence for locomotor or neurochemical sensitization was limited. Locomotor activity showed a significant effect of test day (F(3,30) = 4.688, p = 0.0084) and MDMA-treated mice were more active on day 7 than day 1 (p = 0.0013), but the interaction of treatment and day did not reach significance (trend for treatment: F(1,10) = 4.627, p = 0.0570; interaction F(3,30) = 2.451, p = 0.0828), so locomotor sensitization could not be confirmed. There was no cross-sensitization with D-amphetamine (t(10) = 0.1104, p = 0.9142). Microdialysis in the nucleus accumbens showed that MDMA affected 5-HT overflow across time (F(11,44) = 2.474, p = 0.0165) but there was no effect of treatment day (first vs fourth MDMA; F(1,4) = 0.04376, p = 0.8445), and subject variability limited interpretation. Pharmacological manipulations of 5-HT2A Rs indicated these receptors are necessary but not sufficient for social sensitization. Pretreatment with the 5-HT2A R antagonist M100 (1 mg/kg) on days 1, 3 and 5 inhibited the development of social sensitization: a significant pretreatment effect (F(1,24) = 9.257, p = 0.0056) and pretreatment × day interaction (F(3,24) = 3.129, p = 0.0444) were observed. Mice pretreated with saline showed increased social interaction on day 7 relative to day 1 (p = 0.0242) and higher day-7 interaction than mice pretreated with M100 (p = 0.0358). However, administering M100 only on day 7 did not attenuate expression of MDMA-induced social behaviour in previously sensitized mice (t(6) = 1.31, p = 0.2380). M100 did not produce general sedation in a locomotor test (t(8) = 0.8869, p = 0.401). Conversely, DOI (1 mg/kg), a 5-HT2A R agonist, given on days 1, 3 and 5 did not suffice to produce later sensitization to MDMA on day 7; ANOVA for day 7 treatments was significant (F(3,11) = 12.43, p = 0.0007), but DOI-treated mice did not differ from saline-treated mice (p = 0.9775) and were lower than mice given MDMA across sessions (p = 0.0018).

Curry and colleagues interpret their findings as the first demonstration that animals can sensitise to MDMA's prosocial effects: repeated intermittent MDMA produced a rapid, progressive augmentation of drug-elicited social interaction that could be re-elicited after abstinence. The pattern resembled classical behavioural sensitization in speed of development and re-elicitation by the drug, but key divergences were noted. With the low-frequency dosing used here (every other day), locomotor and neurochemical sensitization were weak or absent, suggesting social sensitization may occur independently from the adaptations usually linked to locomotor sensitization. The investigators propose that 5-HT2A R activation is required for the development of social sensitization but not sufficient on its own: M100 blocked development when given during the treatment phase, yet DOI alone did not substitute for MDMA. They therefore suggest that 5-HT2A R-mediated facilitation of plasticity might interact with other MDMA effects, such as serotonin and oxytocin release and the presence of a social context, to produce lasting changes. Supporting this, social pairing during MDMA treatment was necessary for sensitization to emerge, consistent with the idea that MDMA may augment the salience or reward value of social stimuli via mesolimbic circuits and neuropeptides such as oxytocin. Several limitations were acknowledged. Sample sizes were relatively small, which may have limited detection of locomotor or neurochemical sensitization. Only male mice were studied, leaving open sex differences. The authors also note that sensitization and tolerance can coexist and depend on dose and regimen; higher or more frequent MDMA dosing could produce tolerance rather than sensitization, as seen in other studies and human heavy-use cases, potentially via 5-HT depletion. Finally, the social-context dependence and possible links between social pairing and other sensitization measures require further dissection in future work. Overall, the authors propose that social sensitization to MDMA provides a novel animal model to investigate neural adaptations underlying the drug's social effects and possibly mechanisms that contribute to the persistence of MDMA-assisted therapy outcomes.

This study reports the first evidence that mice sensitize to the prosocial effects of MDMA: sensitization developed rapidly across intermittent treatments, depended on both 5-HT2A R activation and a concurrent social context, and could be re-elicited after a period of abstinence. The investigators suggest that social sensitization may serve as a new model to identify neurobiological changes induced by MDMA that could underlie durable therapeutic effects, and they recommend further work to characterise long-term behavioural change and the specific neural mechanisms responsible.