Monoamine oxidase inhibitors in South American hallucinogenic plants: tryptamine and beta-carboline constituents of ayahuasca

Ayahuasca is a traditional Amazonian hallucinogenic brew made by boiling the vine Banisteriopsis caapi with leaves of various admixture plants, commonly Psychotria viridis or Diplopterys cabrerana. Earlier chemical surveys identified three beta-carboline alkaloids in B. caapi—harmine, harmaline and tetrahydroharmine (THH)—and N,N-dimethyltryptamine (DMT) in the admixture plants. Because DMT is inactive orally due to degradation by visceral monoamine oxidase (MAO), it has been hypothesised that the beta-carbolines act as reversible MAO inhibitors to protect DMT from deamination and thereby enable the oral activity of the brew; however, this mechanism had not been experimentally demonstrated in ayahuasca preparations themselves. McKenna and colleagues set out to characterise the alkaloid constituents of multiple ayahuasca brews and their source plants and to test directly whether ayahuasca and its beta-carboline constituents inhibit MAO in vitro. The study combined ethnobotanical sampling of brews and plants from Peru with chromatographic identification and quantitation (two‑dimensional thin‑layer chromatography (TLC), high‑performance liquid chromatography (HPLC), and gas chromatography/mass spectrometry (GC/MS)), and assessed MAO inhibition using a rat‑liver cytosol assay to explore structure–activity relationships among selected beta‑carbolines and mixtures approximating ayahuasca composition.

Field sampling comprised eight ayahuasca brews obtained from ayahuasqueros around Iquitos, Pucallpa and Tarapoto in Peru, preserved either undiluted or diluted 1:1 with methanol to arrest fermentation. Stem samples of several Banisteriopsis caapi cultivars and leaves of admixture plants (primarily Psychotria viridis and a Diplopterys cabrerana clone) were also collected and vouchered when possible. Phytochemical analysis used two‑dimensional TLC to profile crude brews and purified alkaloid fractions, with visualisation under short‑ and long‑wave UV and Ehrlich's reagent for tryptamines. Quantitative determinations were performed on a Varian HPLC system with reverse‑phase column and UV detection at 260 nm; calibration used external standards of harmine, harmaline, THH and DMT at known concentrations. Sample preparation procedures differed by material type (crude brews, methanol‑preserved brews, dried B. caapi stems, and methanol extracts of admixture leaves) but generally involved methanolic extraction, filtration, solvent removal, acid/base partitioning for plant extracts, and injection of aliquots in replicate runs. GC/MS (Finnigan model 1020) was used to confirm the identity of DMT in admixture leaf extracts by matching retention time and mass spectra to authentic DMT standard. Uncommon admixture plants were screened for alkaloids using precipitation reagents (Meyer's, Valser's, Dragendorff's) and TLC when only limited material was available. MAO inhibition assays employed whole rat‑liver cytosol prepared from perfused Wistar rat livers. The enzyme assay used 14C‑labelled 5‑hydroxytryptamine (5‑HT) as substrate, incubated at 37.5°C; reaction products were extracted into ethyl acetate and quantified by liquid scintillation counting. Stock solutions of selected beta‑carbolines were prepared in 0.1 N HCl and assayed over serial 1:10 dilutions to generate concentration–response data; boiled blanks and active‑enzyme, no‑inhibitor controls were run in parallel. Two ayahuasca brews (Don Fidel no. 1 and Don Juan no. 2) were reconstituted from methanol‑preserved samples, quantified by HPLC (total alkaloid 3.5 mg/ml and 4.8 mg/ml respectively), diluted serially, and tested for MAO inhibition. Assays also compared activity of (a) equimolar mixtures of harmine, harmaline and THH and (b) an "ayahuasca analogue" mixture with the same total alkaloid concentration but with relative proportions matching measured brews (approximately 69% harmine, 26% THH, 4.6% harmaline); DMT was not included in the analogue mixtures.

Chemical profiling showed that harmine, harmaline and THH were the major beta‑carbolines present in all analysed ayahuasca brews; harmalol was detected in only one sample. DMT was detected in the admixture plants and in most brews, but was absent from the single Tarapoto brew examined, which employed Psychotria carthagenensis rather than the more common Psychotria viridis. Two‑dimensional TLC, HPLC and GC/MS confirmed the identities: Psychotria viridis leaf samples consistently contained DMT (reported between 1 and 1.6 mg/g dry weight), and the Diplopterys cabrerana sample (Plowman 6040) contained DMT at about 1.74 mg/g dry weight with a trace of 5‑hydroxy‑DMT. Quantitative HPLC revealed substantial variation in alkaloid content across brews and plant material. The Pucallpa lyophilised samples showed notably higher total alkaloid content than samples from Iquitos and Tarapoto; one Pucallpa sample was reported as 75.7 mg alkaloid/g dry weight, comprised approximately of 76% harmine, 10.6% THH and 7.6% DMT. Using the average of five Pucallpa samples, the authors calculated that a 100‑ml dose could contain about 728 mg total alkaloid (467 mg harmine, 160 mg THH, 41 mg harmaline and 60 mg DMT), substantially exceeding levels reported in an earlier study by at least an order of magnitude. Analysis of dried B. caapi stems from different cultivars showed wide variation in total alkaloid content, from 1.8 to 13.6 mg/g, with no simple correlation to vernacular cultivar designations. Screening of less common admixture plants found that Abuta grandifolia yielded a positive alkaloid test and is reported to contain palmatine (a benzylisoquinoline), whereas Justicia pectoralis showed no alkaloids in the tested collection. In vitro MAO assays produced three principal findings. First, ayahuasca preparations were potent inhibitors of MAO in the rat‑liver cytosol system: Don Fidel no. 1 and Don Juan no. 2, quantified at 3.5 mg/ml and 4.8 mg/ml total alkaloid respectively, inhibited MAO strongly even when serially diluted across many orders of magnitude. The extracted text reports that Don Fidel still produced greater than 40% inhibition at a highly diluted concentration reported as 10^-5 of full strength; Don Juan's sample exceeded 50% inhibition even at a high dilution, although the exact dilution for that result is ambiguously reported in the extraction. Second, assays of single beta‑carboline standards showed structure–activity relationships: fully aromatic and dihydro‑beta‑carbolines were substantially more potent MAO inhibitors than fully saturated analogues; methoxy substitution tended to increase potency relative to hydroxyl substitution at comparable positions; certain methylations (for example N‑methylation) enhanced activity in some analogues. Third, mixtures approximating ayahuasca composition were assayed: an equimolar mixture of harmine, harmaline and THH and an "ayahuasca analogue" (same total alkaloid but proportions matching measured brews) had similar inhibitory potencies (reported IS0 values are presented in the paper but the extracted text does not clearly show the full numeric exponents). The authors conclude that the combined activity of harmine and THH can account for most of the MAO inhibition observed with the brews, and that the overall activity of mixtures is not greater than that of the most active component(s) present.

McKenna and colleagues interpret these findings as the first direct empirical evidence that ayahuasca preparations inhibit monoamine oxidase in vitro and that the degree of inhibition correlates with the concentration of MAO‑inhibiting beta‑carbolines. They note that the high levels of beta‑carbolines and DMT measured in their Pucallpa samples—considerably greater than levels reported in an earlier study—support the ethnopharmacological hypothesis that beta‑carbolines render DMT orally active by preventing its deamination, thereby providing a plausible pharmacological mechanism for the brew's psychoactivity. The authors caution that the present results derive from an in vitro rat‑liver cytosol assay and represent only an initial step toward understanding ayahuasca pharmacology. They state explicitly that in vivo studies in animals and humans are required before firm conclusions about the brew's effects in living organisms can be drawn. Variation in alkaloid composition between regions and between B. caapi cultivars is emphasised, with preparation methods (length and repetition of boiling, concentration) and environmental factors proposed as contributors to observed differences in alkaloid load; the extracted text indicates that Pucallpa preparations are simmered and concentrated for many hours, which may account for higher alkaloid concentrations relative to other regional methods. Additional implications discussed by the authors include the possibility that ayahuasca made from B. caapi without DMT‑containing admixtures would have a different pharmacology and that at substantially higher concentrations the beta‑carbolines alone could produce psychotropic effects. They also highlight that non‑tryptamine admixtures such as Abuta grandifolia (containing palmatine) could modulate the brew's effects in ways that are not yet characterised and call for further pharmacological work on these admixture plants. Finally, the report acknowledges limitations in botanical sampling (the need for broader systematic sampling of B. caapi cultivars and consideration of environmental variables) and the ambiguity of some numerical assay details in the extracted text; they reiterate the necessity of follow‑up in vivo and clinical investigations to clarify the pharmacodynamics and safety of ayahuasca.