Role of 5-HT2A receptors in the effects of ayahuasca on ethanol self-administration using a two-bottle choice paradigm in male mice

Alcohol (ethanol) use disorder (AUD) remains a prevalent, relapsing condition with limited pharmacological options and high rates of relapse. Previous human and preclinical work has suggested that ayahuasca, a traditional brew containing the 5-HT2A-preferring agonist N,N-dimethyltryptamine (DMT) together with β-carbolines, may reduce alcohol-related behaviours; religious ayahuasca users have shown lower prevalence of AUD and animal studies from this group have reported blockade of ethanol-induced sensitisation and conditioned place preference. Nevertheless, it is unclear whether activation of 5-HT2A receptors is necessary for ayahuasca’s apparent anti-drinking effects, and whether treatment administered during abstinence can modify subsequent voluntary ethanol consumption.

Serra and colleagues used male Swiss mice (≈3 months, 35–40 g) from an in-house colony. Animals were group-housed except during intermittent individual self-administration sessions. A lyophilised batch of ayahuasca obtained from a Santo Daime member was quantified for DMT and β-carbolines (DMT 0.4 mg/100 mg; tetrahydroharmine 3.07 mg/100 mg; harmine 3.85 mg/100 mg; harmaline 0.17 mg/100 mg). The selective 5-HT2A antagonist R(+)-MDL100,907 (M100) was also used. Ayahuasca (100 mg/kg, intragastric) and M100 (1 mg/kg, intraperitoneal) were each diluted in saline; ethanol for drinking was 10% v/v in water. The behavioural paradigm was an intermittent access two-bottle choice procedure. Ninety mice underwent acquisition with 15 sessions over 30 days: every other day they were individually housed for 15 h with access to one bottle of water and one bottle of 10% ethanol. Animals averaging <1.5 g/kg ethanol intake were excluded after an initial 15 days, leaving 42 mice that continued for a total of 15 acquisition sessions. Those 42 animals were randomly assigned to four treatment groups (n = 9–11 per group): Veh-Veh, M100-Veh, Veh-Aya, and M100-Aya. The experimental timeline comprised three treatment phases, each followed by three ethanol re-exposure tests. In treatment sessions, given every 3 days, mice received i.p. M100 or vehicle and immediately afterwards i.g. ayahuasca or vehicle; 30 minutes later they were placed individually in the self-administration apparatus but with only water available. The first treatment phase included five treatments over 15 days; the two subsequent phases were shorter (three treatments over 9 days each). Re-exposure tests began 48 h after the last treatment in each phase and replicated the acquisition conditions (one water bottle and one 10% ethanol bottle) on alternate days for three sessions. Bottle side was alternated, consumption was weighed before and after sessions to calculate ethanol intake in g/kg and ethanol preference (%), and leak/evaporation was controlled with empty-cage measures (<0.1 mL/day). Animals were weighed daily. Statistical analyses used mixed-measures ANOVAs (time and solution as within-subject factors; M100 and ayahuasca as between-subject factors), with additional ANOVAs restricted to re-exposure tests and Tukey’s post hoc comparisons. Normality and homogeneity were checked (Shapiro–Wilk, Levene). For some restricted analyses the authors applied a more stringent alpha (1%) to reduce type I error. GraphPad Prism and IBM SPSS were used for analyses and figures.

Ninety mice began acquisition; 42 mice that averaged ≥1.5 g/kg ethanol intake after the initial acquisition period were retained and allocated to four groups (n = 9–11 per group). A mixed-measures ANOVA across all time-points found significant main effects of time [F(7,266) = 3.537; p < 0.01] and solution (ethanol vs water) [F(1,38) = 149.962; p < 0.0001], and significant interactions involving time, solution, ayahuasca and M100 treatments (several interactions with p < 0.05 to p < 0.01). During the last week of acquisition (D25–D30) all groups consumed significantly more ethanol than water (p < 0.0001). During the first re-exposure (R1), the Veh-Veh, M100-Veh and M100-Aya groups maintained significantly higher ethanol versus water consumption (p < 0.0001 for each). By contrast, the Veh-Aya group (ayahuasca only) increased water consumption relative to control (p = 0.0052) and showed no significant ethanol–water difference in R1 (p > 0.05). In R2 and R3 the Veh-Veh, M100-Veh and M100-Aya groups continued to prefer ethanol (p < 0.0001), whereas the Veh-Aya group consumed significantly less ethanol than Veh-Veh (p < 0.05 for both tests) and showed increased water intake versus Veh-Veh (p < 0.01); the Veh-Aya group did not differ between water and ethanol in R2–R3 (p > 0.7). When ethanol intake (g/kg) was examined across all phases, only a significant effect of time was detected [F(7,266) = 3.376; p < 0.01], and Veh-Veh increased intake in later re-exposures versus early acquisition (D1–D6, p < 0.05). Restricting the analysis to re-exposure tests with a stricter alpha revealed a significant time × ayahuasca interaction [F(2,76) = 6.023; p < 0.01], and post hoc tests showed Veh-Aya consumed significantly less ethanol than Veh-Veh in R2 and R3 (p < 0.01). Preference (%) analyses showed a main effect of time [F(7,266) = 3.678; p < 0.001] and interactions including time × ayahuasca and ayahuasca × M100. All groups preferred ethanol (>50%) during acquisition with no between-group differences. During re-exposures the Veh-Aya group showed significantly lower ethanol preference compared to its own last acquisition period (D25–D30, p < 0.01) and compared to Veh-Veh across all re-exposure tests (restricted analysis: ayahuasca main effect F(1,38) = 12.28; p < 0.01; ayahuasca × M100 interaction F(1,38) = 13.55; p < 0.001). The Veh-Aya group also had lower preference than the M100-Veh and M100-Aya groups in R1 and R2 (p < 0.05). Pretreatment with M100 prevented the reduction in ethanol intake and preference produced by ayahuasca: the M100-Aya group did not differ significantly from Veh-Veh in ethanol consumption/preference during re-exposures. The M100-Veh group tended to show lower intake and preference than Veh-Veh but these differences were not statistically significant. Total water consumption during treatment days and body weight across the protocol showed only main effects of time, with no significant post hoc group differences reported. The authors report that bottle leak/evaporation was minimal (<0.1 mL/day).

Serra and colleagues interpret their results as showing that administration of ayahuasca during ethanol abstinence reduced subsequent ethanol self-administration and ethanol preference in male mice tested in an intermittent two-bottle choice paradigm. They note the effects emerged at a dose (100 mg/kg) previously found to block ethanol-related behavioural sensitisation and conditioned place preference, and that locomotor impairment is unlikely to explain the findings given prior strain-matched data. The pattern of increased water intake in ayahuasca-treated animals, observed already in the first re-exposure, is taken as compatible with a shift in consumptive choice away from ethanol; the authors propose this could reflect pairing of ayahuasca’s rewarding properties with the water bottle during treatment sessions, an interpretation supported by prior conditioned place preference data. The investigators emphasise that pretreatment with the selective 5-HT2A antagonist M100907 blocked the anti-drinking effects of ayahuasca, indicating a necessary role for 5-HT2A receptor activation in these effects. They also acknowledge that M100 alone did not significantly reduce ethanol intake in this paradigm, and that in M100 + ayahuasca groups a non-significant trend toward reduced intake remained, implying additional mechanisms may contribute. Candidates discussed include 5-HT2C receptor actions, β-carbolines from Banisteriopsis caapi, and other DMT-mediated pathways; the authors previously reported that the β-carboline-containing extract of B. caapi alone can block ethanol-conditioned place preference. Serra and colleagues contrast their findings with a recent negative report in rats, highlighting differences in species and, importantly, in treatment timing: the present protocol administered ayahuasca during abstinence with later drug-free re-exposures, whereas the other study gave ayahuasca immediately before ethanol availability. The authors suggest the therapeutic context (experiencing ayahuasca in the absence of ethanol, and the treatment environment) may modulate its effects. Key limitations acknowledged include the use of a single dose of ayahuasca and a single dose of M100907, and the fact that the chosen ayahuasca dose can produce reward in mice; they call for dose–response studies, including non-rewarding doses, to clarify mechanisms and translational relevance. Overall, the authors conclude their data support further investigation of ayahuasca and 5-HT2A agonists as adjunctive pharmacotherapies for AUD, while noting mechanistic complexity and the need for additional research.