Metabolomics and integrated network analysis reveal roles of endocannabinoids and large neutral amino acid balance in the ayahuasca experience

Ayahuasca is a traditional Amazonian psychoactive brew that combines Banisteriopsis caapi (providing β-carbolines such as harmine, harmaline and tetrahydroharmine) with Psychotria viridis (providing N,N-dimethyltryptamine, DMT). Previous work shows that β-carbolines inhibit monoamine oxidase (MAO) and thereby permit orally active DMT, and that 5-HT2A receptor activation is central to psychedelic effects. However, the cascade of downstream neurochemical and peripheral metabolic changes triggered by ayahuasca in humans remains poorly characterised; most existing metabolomic work has examined the brew itself or used animal models rather than profiling human plasma after ingestion. Madrid-Gambin and colleagues set out to define the human plasma metabolomic signature produced by ayahuasca consumption and to relate those changes to both subjective experience and plasma alkaloid concentrations. The study combined targeted LC-MS/MS metabolomics with measures of ayahuasca alkaloids and a validated questionnaire of altered states of consciousness, then applied multivariate and integrated network analyses and pathway enrichment to identify metabolic pathways associated with the ayahuasca experience.

This was a within-subject observational study of 23 healthy adult members (14 male, 9 female; mean age 37.9 ± 10.3 years) of the Dutch Santo Daime church. Participants attended a customary ayahuasca ceremony organised by the church; the research team did not handle production, dosing or administration. Each participant was assessed on two consecutive days: a baseline visit the day before the ceremony and a post-consumption visit the following day, with blood sampling and psychological assessment. Exclusion criteria included MRI contraindications, pregnancy and use of (medicinal) substances in the prior 24 hours. Subjective effects were quantified using the 5-Dimension Altered States of Consciousness Rating Scale (ASC), a 94-item visual analogue self-report instrument that yields 11 subscales (for example: experience of unity, blissful state, insightfulness, complex imagery, audio-visual synaesthesia, impaired control and cognition, anxiety). Blood was sampled at baseline and 90 minutes after drinking ayahuasca; plasma alkaloid concentrations (DMT, harmine, harmaline, tetrahydroharmine) were measured by LC-MS/MS using validated calibration ranges and reported lower limits of quantitation. Targeted metabolomics assessed 164 biomarkers across several chemical families (including 12 endocannabinoid-related compounds, 42 amino-acid-related markers, steroids, energy metabolites, acylglycerols, ceramides, lysophosphatidylcholines and choline metabolites) by LC-MS/MS. Data processing removed metabolites with >80% of values below detection; missing values were imputed as half the dataset minimum prior to principal component analysis for outlier detection. Statistical comparisons of pre/post metabolite levels used paired two-tailed Wilcoxon tests with Benjamini-Hochberg false discovery rate (FDR) correction (FDR p < 0.05 considered significant). Predictive modelling employed multilevel partial least squares discriminant analysis (mPLS-DA) on paired samples with repeated double cross-validation and recursive variable selection; model significance was assessed by 200-permutation testing. To integrate metabolomics, alkaloids and subjective measures the investigators used regularised canonical correlation analysis (rCCA) and constructed clustered heat maps and a similarity-based network (edges retained for similarity ≥ 0.3). Finally, enrichment of significant metabolites was performed using the FELLA package to map compounds to KEGG pathways via a diffusion-based propagation algorithm and visualised in Cytoscape.

Twenty-three participants provided baseline and post-ayahuasca plasma for both metabolomics and alkaloid assays. The analysed brew contained 0.14 mg/ml DMT, 4.50 mg/ml harmine, 0.51 mg/ml harmaline and 2.10 mg/ml tetrahydroharmine. The mPLS-DA model discriminating pre- and post-consumption samples achieved a classification rate of 100.0% (95% CI 85.2–100.0) and was significant by permutation testing (p < 0.01). Multivariate modelling identified 31 differential metabolites; paired Wilcoxon tests (FDR-corrected p < 0.05) detected differences for 20 metabolites. Most altered metabolites belonged to endocannabinoid and amino-acid metabolism. After ayahuasca intake, circulating N-acyl-ethanolamine endocannabinoids (for example anandamide and related N‑acyl‑ethanolamines) increased, whereas several monoacylglycerols/2‑acyl‑glycerols (including 2‑AG species) decreased. Markers of serotonin metabolism fell, notably 5‑hydroxyindoleacetic acid (5HIAA) and the ratios 5HIAA/tryptophan and 5HIAA/serotonin. In amino-acid metabolism, levels of certain large neutral amino acids (LNAAs) showed mixed changes: tyrosine and derived ratios (tyrosine/phenylalanine, tyrosine/LNAA) decreased, while relative indices such as leucine/LNAA and isoleucine/LNAA rose; glutamate, glutamine and α‑hydroxybutyrate also increased. Certain steroid metabolites (cortisone, 20α‑DHE and 20β‑DHE) were elevated after consumption. The rCCA-based integrated analysis yielded five clusters separating alkaloid/dose variables and three clusters of subjective ratings (termed Effects A, B and C). Nine of the eleven ASC subscales (all except impaired control and cognition, and anxiety) showed strong correlations (similarity up to |0.7|) with metabolic marker clusters. A cluster containing tryptophan/LNAA, 5HIAA/tryptophan, 5HIAA/serotonin and several monoacylglycerols was positively correlated with nine ASC subscales, most strongly with Effects A (mystical-dissolution themes: experience of unity, blissful state, insightfulness). By contrast, a group of amino acids including tyrosine, phenylalanine, isoleucine and the sum of LNAAs was inversely correlated with the ASC clusters; tyrosine showed the strongest negative association with the ‘‘experience of unity’’. Many of these relationships were not linked to plasma alkaloid concentrations: the alkaloid cluster (notably tetrahydroharmine, which had the highest number of connections) correlated positively with several endocannabinoids, ceramides, hexosylceramides and lysophosphatidylcholines, but only one endocannabinoid species, DGLEA (dihomo-γ-linolenoylethanolamide), connected both to alkaloids and to subjective effects. Behavioural variables such as number of ceremonies, years in Santo Daime and days since last ceremony showed no association with metabolic markers. Enrichment analysis mapped 11 of the 31 significant metabolites onto KEGG-derived network elements (243 items), including valine, leucine, isoleucine, tyrosine, tryptophan, phenylalanine, glutamate and serotonin. The propagated network highlighted crosstalk between branched-chain amino acid (BCAA) and aromatic amino acid (AAA) pathways converging on glutamate and serotonin, and placed these within modules such as melatonin biosynthesis from tryptophan and dopamine biosynthesis from tyrosine (via tyrosine 3‑monooxygenase). Additional connected pathways suggested by KEGG entries included synaptic vesicle cycle, gap junction, mTOR signalling and pathways related to addiction (amphetamine, cocaine, alcoholism).

Madrid-Gambin and colleagues interpret their findings as evidence of a distinct peripheral metabolic signature associated with the ayahuasca experience, characterised chiefly by (i) alterations in serotonin metabolism consistent with MAO inhibition, (ii) modulation of peripheral endocannabinoid species (increased N‑acyl‑ethanolamines, decreased 2‑acyl‑glycerols), and (iii) dysregulation of large neutral amino acid balance. The decrease in 5HIAA and lowered 5HIAA/tryptophan and 5HIAA/serotonin ratios accords with β‑carboline‑mediated MAO inhibition, which would be expected to increase synaptic serotonin; the investigators note that this biochemical pattern correlated positively with many subjective effects on the ASC. For endocannabinoids, the pattern of increased N‑acyl‑ethanolamines could be mechanistically linked to inhibition of fatty acid amide hydrolase (FAAH) by β‑carbolines, while DMT’s action at 5‑HT2A receptors may influence 2‑AG levels; the authors caution that peripheral endocannabinoid changes may reflect multiple influences (circadian, stress, inflammation or steroid changes) and that peripheral measures do not necessarily mirror central concentrations. The single peripheral timepoint and known dissociations between plasma and central measures (for example cerebrospinal fluid or brain tissue) limit inferences about brain endocannabinoid signalling. Regarding amino acids, the investigators highlight that both branched-chain and aromatic amino acids share transport mechanisms across the blood–brain barrier, so shifts in peripheral LNAA balance (notably higher relative tryptophan/LNAA and lower tyrosine and BCAAs) could bias precursor availability for serotonin versus catecholamine synthesis. This pattern—tryptophan/LNAA positively associated with subjective ‘‘mystical’’ and unity-type effects, and tyrosine/BCAAs inversely associated—suggests a role for competition at transporters in shaping neurotransmitter-dependent aspects of the ayahuasca experience. Pathway enrichment linked these amino-acid changes to neurotransmitter biosynthesis (serotonin, dopamine), mTOR signalling and addiction-related pathways, which the authors propose might help explain putative therapeutic effects in mood and addiction contexts. Key limitations acknowledged include the observational ceremony context (set and setting not controlled), a single post‑consumption pharmacokinetic timepoint for alkaloid measures, and the restriction to peripheral plasma metabolites. The authors therefore advise caution in extrapolating peripheral metabolite changes to central neurochemistry and state that further work is needed to characterise pharmacokinetics, time courses and causal relationships before these metabolic signatures can be linked definitively to therapeutic mechanisms.

This study provides the first targeted human plasma metabolomics profile of the ayahuasca experience, identifying changes in serotonin metabolism, endocannabinoid species (N‑acyl‑ethanolamines and 2‑acyl‑glycerols) and the balance of large neutral amino acids. Integrated network analysis linked circulating tryptophan/LNAA and tyrosine/BCAA balances to distinct subjective effects (notably ‘‘unity’’, ‘‘blissful state’’ and ‘‘insightfulness’’), and many of these associations appeared independent of peripheral plasma alkaloid concentrations. The authors conclude that a crosstalk between circulating LNAAs, neurotransmitter precursor availability and subjective phenomenology may contribute to the neurophysiological and potentially therapeutic effects reported after ayahuasca use, while emphasising the need for further controlled and time-resolved studies.