Phytochemical analyses of Banisteriopsis caapi and Psychotria viridis

Callaway and colleagues frame this study against longstanding ethnobotanical use of Banisteriopsis caapi and Psychotria viridis as the principal plant ingredients in hoasca/ayahuasca brews. Earlier work has identified harmala alkaloids in B. caapi and N,N-dimethyltryptamine (DMT) in P. viridis, but the authors note incomplete characterisation of how alkaloid concentrations vary across plants, regions and time of day. The study therefore set out to perform a broad phytochemical survey of these two species across many sites in Brazil, and to examine whether DMT content in P. viridis leaves varies over a 24-hour cycle. The investigators emphasise that producing a geographically wide, same-day collection was intended to provide a representative picture of ingredients commonly used by members of the União do Vegetal (UDV).

Plant material was collected on a single day (7 October 1995) from 22 sites across Brazil between 06:00 and 09:00 by experienced UDV members. Samples were taken largely from garden specimens maintained near UDV temples, some of which had been cultivated for up to 20 years. All material was dried at ambient temperature (25–35°C), stored in paper envelopes in a dark, dry place, and analysed later. In addition to the broad regional collection, leaf samples from a single P. viridis plant were taken from opposing stems at three-hour intervals over roughly 24 hours to assess circadian variation in DMT. Chemical analysis used high-performance liquid chromatography (HPLC) with fluorescence detection, with a C-8 column and a mobile phase composed of 20% methanol, 20% acetonitrile and 60% 0.1 M ammonium acetate buffer adjusted to pH 6.9. Individual chromatographic signals were confirmed by liquid chromatographic mass spectrometry (LC–MS) using a thermospray probe coupled to a quadrupole mass spectrometer. Sample preparation differed by species. For B. caapi, approximately 30 g of each sample was milled and dried; 100 mg of milled material was sonicated in 2 ml methanol for 10 minutes, soaked for 24 hours in the dark, centrifuged and the supernatant diluted 100-fold in mobile phase prior to injection for harmine, harmaline and tetrahydroharmine (THH) analysis. For P. viridis, whole leaves (0.60–2.60 g) were homogenised in a solvent mixture of 67% methanol, 11% acetonitrile and 22% 0.1 M ammonium acetate at pH 8.0 for DMT analysis. The extracted text contains inconsistent sample counts: the abstract reports 32 B. caapi and 36 P. viridis samples, whereas results passages refer to B. caapi N = 33 and P. viridis N = 37. The text does not resolve this discrepancy.

Banisteriopsis caapi samples uniformly contained the three harmala alkaloids analysed: harmine, harmaline and tetrahydroharmine (THH). Across the B. caapi set (reported in the body as N = 33), harmine was the dominant component in nearly every sample, typically present at approximately one order of magnitude higher concentration than harmaline (about 10:1). THH concentrations were more variable and did not show a clear relationship with harmine or harmaline. A few minor chromatographic signals were seen in most samples, but harmine, THH and harmaline were consistently the major components. Two samples from older plants (seven and nine years old) were notably low in all harmala alkaloids. The investigators also describe two locally recognised varietal forms, mariri caupuri and mariri tucunaca, and report a non-significant trend toward higher harmala levels in caupuri samples compared with tucunaca. Psychotria viridis exhibited a broad range of DMT concentrations (the body text reports N = 37). One sample contained no detectable DMT, and eight samples contained less than 0.60 mg/g. The authors affirm confidence in species identification despite low or absent DMT in some specimens. Time-course sampling from a single P. viridis plant revealed a circadian pattern in leaf DMT: levels rose until about 06:00, declined between 06:00 and 09:00, then increased again to a peak around 18:00, followed by a sharp decline to basal levels by 21:00. Analytical confirmation of chromatographic peaks was made using LC–MS. Numerical effect sizes, confidence intervals or formal statistical tests are not detailed in the extracted results beyond the descriptive comparisons noted above.

The authors interpret their findings as demonstrating substantial natural variability in the key alkaloids of B. caapi and P. viridis across plants and sites. They highlight the consistent dominance of harmine in B. caapi and the relatively fixed proportional relationship between harmine and harmaline, contrasted with more variable THH levels. For P. viridis, the observed wide range of leaf DMT — including specimens with very low or undetectable amounts — indicates that DMT content can vary greatly even within material used by a single religious community. Callaway and colleagues discuss possible explanations for the circadian fluctuation in P. viridis DMT, suggesting hypotheses such as a protective role against solar (UV-B) radiation or a role in absorbing solar radiation, but they present these as speculative. They also note that subjective differences reported between local B. caapi varieties (caupuri versus tucunaca) are not explained by the measured alkaloid profiles and remain unresolved. Limitations acknowledged in the text include an explicit statement that these analyses are not complete and that not everything about these plants and their derivatives can be known from the present work. The authors also emphasise that most samples were garden specimens maintained near UDV temples, which may affect generalisability to wild populations. They present the survey as the largest of its kind to date and suggest the results provide a more detailed baseline for understanding the chemical variability of the plants used to prepare hoasca/ayahuasca.

The study is presented as the largest phytochemical survey to date of Banisteriopsis caapi and Psychotria viridis. The investigators conclude that both species show broad natural variability in their principal alkaloids, that harmine is the consistently dominant harmala in B. caapi, and that P. viridis DMT content can vary markedly across specimens and over the day. While acknowledging the work is not exhaustive, the authors hope the findings will contribute to further insights into the plants’ roles in traditional preparations and stimulate additional research.