---

Originally printed in the Yearbook for Ethnomedicine 1997/98

-------------------------------------------------------------

JONATHAN OTT

Pharmahuasca, Anahuasca and Vinho da Jurema:

     Human Pharmacology of Oral DMT Plus Harmine

Published in Yearbook for Ethnomedicine 1997/98

Abstract

A summary is presented of human self-experiments or

psychonautic bioassays of pharmahuasca-capsules containing

crystalline N,N-dimethyltryptamine (DMT) plus harmine, as

well as combinations of other psychoactive tryptamines with

other ß-carbolines. The 1967 HOLMSTEDT-LINDGREN hypothesis

of the ayahuasca effect-oral psychoactivity of DMT

consequent to monoamine-oxidase [MAO] inhibition from

concomitant ingestion of ß -carbolines- has been confirmed

by 8 self-experimenters. Results of a total of some 70

bioassays are summarized and the literature on this subject

is reviewed. Discussion of ayahuasca analogues (anahuasca)

focuses on the contemporary non-traditional use of jurema

preta (Mimosa tenuiflora) and the ethnobotany and human

pharmacology of traditional vinho da jurema is also reviewed

[with 94 references and 1Table].

A 1967 analysis of a half-dozen South American snuffs used

in shamanic healing by the Tucano, Waiká , Araraibo, Piaroa

and Surára Indians, showed all but one of the powders to

contain tryptamines, mainly 5-methoxy-N,N-dimethyltryptamine

[5-MeO-DMT] and secondarily N,N-dimethyltryptamine [DMT]

(HOLMSTEDT & LINDGREN 1967). However, a paricá snuff of the

Piaroa Indians of the Venezuelan Orinoco region contained

tryptamines--5-OH-DMT [bufotenine], DMT and 5MeO-DMT

-together with the ß-carboline alkaloid harmine, while an

epéna snuff of the Surára Indians contained only

ß-carbolines, which had previously been reported in the same

Surára epéna (BERNAUER 1964) and in a Tucano Indian sample

of paricá snuff (BIOCCA et al. 1964). Since bufotenine and

DMT appeared to be non-psychoactive intranasally (TURNER &

MERLIS 1959), HOLMSTEDT and LINDGREN commented:

The occurrence of both tryptamines and ß-carbolines in the

South American snuffs is pharmacologically interesting. The

ß-carbolines are monoarnine-oxidase inhibitors, and could

potentiate the action of the simple indoles. The combination

of ß-carbolines and tryptamines would thus be advantageous.

(1967: 365)

The following year, HOLMSTEDT & LINDGREN, in collaboration

with AGURELL, found DMT in leaves of Diplopterys cabrerana

(Cuatrecasas) Gates (Malpighiaceæ] used as an admixture to

ayahuasca or yajé by Ecuadorian Kofan Indians (AGLTRELL et

al. 1968), a finding replicated by DER MARDEROSIAN and

colleagues (DER MARDEROSIAN et al, 1968). Ayahuasca is a

pan-Amazonian complex of shamanic potions based on aqueous

infusions of the stem of the ayahuasca liana, Banisteriopsis

caapi (Spruce ex Griseb.) Morton [Malpighiaceæ], to which

may be added visionary, stimulant or curative admixture-pl

ants, some 100 of which have been identified (OTT 1993,

1994, 1995c, 1999). Three years earlier, JACQUES POISSON had

isolated DMT from dried leaves of D. cabrerana [in all three

of these reports the synonym Banisteriopsis rusbyana

(Niedenzu) Morton was used] added to natem[a] or ayahuasca

by Ecuadorian Shuar Indians (POISSON 1965). Later research

documented widespread use of DMT-rich leaves of Psychotria

viridis Ruiz et Pavan [Rubiaceæl in ayahuasca potions (OTT

1994, 1999). HOLMSTEDT, LINDGREN and AGURELL extended the

earlier observation regarding the snuffs to ayahuasca,

noting: "The combination in yajé of monoamine oxidase

inhibiting harman alkaloids with N,N-dimethyltryptamine

might result in specific pharmacological effects" (AGURELL

et al. 1968: 148), an observation echoed by the DER

MARDEROSIAN group (1968: 146).

This hypothesis of DMT/ß-carboline synergy was proposed to

account for presumed oral activity of DMT in ayahuasca

potions. Although this idea was first suggested in relation

to the snuffs, this has been all but forgotten, and there

have been only rudimentary attempts to model the

psychoactivity of the snuffs with pure compounds, although

such studies are underway and will be reported in due

course. For DMT, first synthesized in 1931 (MANSKE 193 1)

and first isolated from Anadenanthera peregrina (L.) Speg.

[Leguminosæ] seeds used to prepare cohoba snuff 24 years

later (FISH et al. 1955), was found to be inactive orally,

in doses as high as 1.0 gram (ca. 13.0 mg/kg; SHULGIN 1976),

although it was dramatically psychoactive via intramuscular

[i.m.] injection in doses of 30-150 mg (0.4-2.0 mg/kg; SZARA

1957), seems to be quite as active when inhaled as vaporized

freebase (0.4-0.5 mg/kg highly active; OTT 1993), and more

active still when injected intravenously (i.v.; 0.2-0.4

mg/kg 'hallucinogenic'; STRASSMAN &QUALLS 1994; STRASSMAN et

al. 1994). According to the ingenious HOLMSTEDT-LINDGREN

hypothesis, the ß-carbolines present in ayahuasca potions

were serving to inhibit the catabolic enzyme monoamine

oxidase [MAO] -which would normally metabolize oral DMT

before it could get from the digestive system to the brain -

so allowing the DMT also present in the ayahuasca potions to

be absorbed and transported in the circulation to the brain,

there evoking visionary, psychotropic effects.

The HOLMSTEDT-LINDGREN theory — what we might call the

ayahuasca effect-won wide acceptance in the literature, for

it neatly explained the visionary effects of ayahuasca, that

could hardly have been due to the ß-carbolines alone, which

elicit rather sedative, Valium® [diazepam] - like

psychoactivity, and have a high threshold for oral activity,

8.0 mg/kg in the case of harmine, the main alkaloid of

ayahuasca plants and potions (NARANJO 1967). Plants rich in

ß-carbolines have found world-wide use as sedatives, a

property experimentally verified (MONARDES 1990; MOORE 1989;

OGA et al. 1984; SPERONI & MINGHETTI 1988). On the other

hand, 5 reported analyses of 17 ayahuasca potions showed an

average of 175 mg 0-carbolines per dose [range: 20-441

mg/dose; generally 3 parts harmine to I part d-leptaflorine

(R-1,2,3,4-tetrahydroharmine), with traces of harmaline

(3,4-dihydroharmine)], which would amount to just over 2

mg/kg [racemic leptaflorine was found to be even less active

than harmine, with an oral threshold of 12 mg/kg] (DER

MARDEROSIAN et al. 1970; LIWSZYC et al. 1992; McKENNA et al.

1984, 1998; NARANJO 1967; RIVIER & LINDGREN 1972). Since the

ß-carbolines per se could not explain the legendary

psychoptic activity of the jungle ambrosia, this had to be

due to its DMT-content, which amounted to an average of 29

mg/ dose in the 17 potions analyzed [range: 25-36 mg/dose].

Accordingly, the HOLMSTEDT-LINDGREN theory won the day (see

OTT 1999, for details and analysis of the phytochernistry of

ayahuasca plants and potions, summarized in Tables 11-A,

11-B and 11-C; pp. 36, 38 and 391.

Nevertheless this well-accepted theory had not been tested,

either in vitro or in vivo, and remained nothing but a

logical explanation of quite scanty phytochemical and

pharmacological data. Sixteen years passed before the group

of McKENNA showed that two Peruvian ayahuasca samples were

'extremely effective' MAOinhibitors in vitro, in a

"rat-liver cytosol fraction," as was an 'ayahuasca

analogue,' a solution of a mixture of 69% harmine, 26%

leptafforine [probably racemic) and 4.6% harmaline,

mimicking proportions that had been found in ayahuasca

potions (MCKENNA et al. 1984: 218). SO AYAHUASCA was

decidedly an MAO-inhibitor, but it remained to be seen

whether the sum total of all ß-carbolines present in a

typical dose of the potion, 175 mg, could render

psychoactive, in a human subject, the 25-36 mg [average: 29

mg] of DMT also present. Only psychonautic bioassays, known

as the 'Heffter Technique' (OTT 1995d) could establish this

with certainty, in the alembic of the human brain (MCKENNA

et al. 1984).

When I began to investigate this question in 1990, 1 was

able to build on the rudimentary experiments of Bigwood and

'Gracie and Zarkov.' Bigwood made a single bioassay of

pharmahuasca-a capsule containing 100 mg each of DMT

free-base [ 1. 16 mg/kg] and harmaline hydrochloride [= 86

mg free-base; 1.0 mg/ kg], noting: "DMT-like hallucinations

... very similar to ... a DMT- and harmaline-containing

ayahuasca brew that I had previously experimented with"

(BIGWOOD 1978). While this single experiment seemingly

confirmed the HOLMSTED -TLINDGREN theory, nonetheless it was

conducted with some 3-4 times the amount of DMT found in

typical doses of ayahuasca, and the ß-carboline chosen,

harmaline, "does not contribute significantly" to the

pharmacology of the potions (McKENNA et al. 1984: 221).

Subsequent 'underground' experiments by GRACIE & ZARKOV

found DMT active orally in combination with aqueous

infusions of ßcarboline-rich seeds of Peganum harmala L.

[Zygophyllaceæ; traditionally used as hypnotics; KIRTIKAR et

al. 19351, with a threshold level of 20 mg DMT; 30 or 40 mg

being a preferred dose (GRAcIE & ZARKOV 1986). Taken

together, these pioneering experiments provided tentative,

albeit fragmentary, confirmation of the hypothesized

ayahuasca effect, but it seemed to me desirable to conduct

more systematic psychonautic bioassays of pharmahuasca using

measured amounts of both pure DMT and ß-carbolines.

Accordingly, for such psychonautic bioassays, I isolated and

purified DMT [as free-base, mp 45,' thin-layer

chromatographic (TLC) comparison with reference sample] from

roots of Desmanthus illinoensis (Michaux) MacM.

[Leguminosæl, and harmine [as the hydrochloride salt, mp

262,' TLC comparison with authentic reference] from seeds of

Peganum harmala, using standard alkaloid-purification

techniques as outlined in the literature (McKENNA et al.

1984; MANSKE 1952). Both plants were obtained commercially

on the U.S. herbal market. All bioassays were conducted

outside of the United States, with standard

'double-conscious' procedure, as described by SHULGIN &

SHULGIN (199 1: XXVII). 'Double-conscious,' a term

introduced by GORDON ALLES, means simply that the

human-bioassay subject be informed both regarding the

identity [as well as the dosage] of the drug being tested,

and also as to the nature of the effects which might be

anticipated.

In a total of some three dozen experiments [most of which

are detailed in OTT 1994], I was able to confirm in vivo the

HOLMSTEDT-LINDGREN ayahuasca effect, in my own body. I found

that DMT was indeed rendered psychoactive orally in

combination with harmine hydrochloride taken simultaneously

in a single gelatine-capsule. Starting with quantities near

the lowest levels found in ayahuasca portions [20 mg DMT and

40 mg harmine], I systematically tested increasing doses. I

found 120 mg of harmine [expressed as the free-base; 1.5

mg/kgl to be the threshold for the ayahuasca effect, whereas

in a control experiment with this amount absent DMT,

barely-perceptible sedative effects resulted- harmine

hydrochloride has been characterized as a 'stupefying' agent

(FONT QUER 1993: 424). Although I could feet 20 mg DMT [0.25

mg/kgl combined with 120 mg harmine, for me the visionary or

psychoptic threshold-level was 30 mg DMT [0.38 mg/kg]. I

have tested doses as high as 160 mg DMT [2.0 mg/kgl,

experiencing progressively more intense psychotropic

effects, but always with the same approximate

pharmacodynamics, quite similar to what I have enjoyed with

genuine Amazonian ayahuasca potions in Brasil, Ecuador and

Perú - 45 minutes to an hour incubation period; the effects

quickly building to a peak by 1: 15 and maintaining a

plateau for 45 minutes to an hour; followed by about an hour

of diminishing effects; the experience usually all but over

around the 3 hour point. In no case have I ever experienced

nausea in pharmahuasca experiments, although I have

weathered nausea and episodes of vomiting provoked by

genuine ayahuasca in Amazonia. In any case, I generally eat

little or nothing on the day of ingestion. During the

experimental series, I always allowed roughly a minimum of a

week to elapse between the individual experiments.

I have been able to extend these observations, based on the

experiences of eight psychonauts in all, involving a total

of about 70 self-experiments. As some of the experimenters

wished to remain anonymous, I will merely cite the sole

published account and one personal communication (CALLAWAY

1992; MARKUS 1989). In all cases 'double-conscious'

self-experiments were involved; in no case were the

compounds administered to anybody else; and no non-human

animal-experimentation of any kind was conducted.

It was found that both harmaline and 6-methoxy-harmalan

(MARKUS 1989) could substitute for harmine in pharmahuasca,

at approximately commensurate doses-CALLAWAY found 70 mg

harmaline [as free-base; 1.2 mg/kg] to activate tryptamines

in pharmahuasca, close to the level BIGWOOD had found active

[1.0 mg/kg] (BIGWOOD 1978; CALLAWAY 1992). Another

psychonaut found 175 mg harmaline hydrochloride [146 mg

base; 2.25 mg/kg] alone to be a mild sedative. Chemical

analysis showed the aged commercial sample of harmaline used

had partially oxidized to harmine, being in reality a

mixture of some two thirds harmaline, one-third harmine

(SHULGIN 1993). Three different dose-levels of this

'harmaline' were then tested in combinations with relatively

high doses of tryptamines. Whereas 50 mg [43 mg base; 0.66

mg/kgl was not effective as tryptamine-activator, doses of

100 mg [86 mg base; 1.32 mg/kgj and 150 mg [ 130 mg base;

2.0 mg/kg] definitely were. Thus it would appear that for

harmaline, too, there is a threshold for activity in

pharmahuasca, somewhere around 0.75 - 1.0 mg/kg. Doses of

6-methoxy-harmalan found to be effective were not divulged

to us in the sketchy second-hand report (MARKUS 1989).

CALLAWAY found 10 mg of 5-methoxy-N,N-dimethyltryptamine

[5-MeODMT] to be psychoactive in pharmahuasca [expressed as

free-base; 0. 17 mg/kg] (CALLAWAY 1992); MARKUS also found

this compound psychoactive but we know not the dose (MARKUS

1989). Thus it would appear that 5-MeO-DMT is several times

as active as DMT in pharmahuasca; mirroring the higher

activity of this compound by other routes. SHULGIN found

5-10 mg doses psychoactive by inhaling the vapor of the

free-base (0.07-0.13 mg/kg; SHULGIN 1970,1983). The

artificial compound N,N-diethyltryptamine [DET or T-91 was

likewise found to be psychoactive orally in pharmahuasca

capsules by two psychonauts, who employed doses of 60 mg

free-base [0.7 mg/kg1 and 150 mg free-base [2.3 mg/kg]

respectively. The latter quantity was characterised as

"definitely an overdosage." We lack enough data to speculate

on oral threshold-levels of DET -it is likely this compound

is roughly equipotent with DMT in pharmahuasca, much as it

is via intramuscular injection. SZARA found 60 mg DET 10.8

mg/kg] psychoactive i.m. (SZARA 1957); the BÖSZÖRMENYI group

found it active in doses of from 0.650.85 mg/kg i.m.

(BOSZORMENYI et a]. 1959). A recent book (SHULGIN & SHULGIN

1997) gives full details of the chemistry and human

pharmacology of the ß-carbolines and tryptamines discussed

here, along with some further pharmahuasca data. Whereas 20

and 50 mg of harmaline were insufficient orally to activate

55 and 60 mg of DMT respectively; 80, 100, 150 and 150 mg

harmaline did activate 40, 120, 35 and 80 mg DMT

respectively. Moreover, 70, 80 and 150 mg harmaline sufficed

orally to activate 10, 10 and 25 mg of 5-MeO-DMT

respectively. On the other hand, it was noted that harmine

[as HCI? 141 mg=120 mg base] could activate 3540 mg DMT

orally, in doses in the 140-190 mg range; whereas doses of

120-140 mg harmine were ineffective when taken with 30 mg of

DMT.

It has lately been alleged that the ayahuasca effect

constitutes 'potentiation' of tryptamines by ß-carbolines,

as originally suggested with regard to the snuffs (HOLMSTEDT

& LINDGREN 1967). In a review article, for example, CALLAWAY

noted: "It is well known that ßCs potentiate the activity of

methylated tryptamines" (CALLAWAY 1995: 25). However,

orally-active DMT in pharmahuasca seems to be weaker than

via other routes of administration. It would appear that the

descending order of potency via distinct routes is: i.v.

injection > inhalation of vapor > i.m. injection >

subcutaneous injection > orally in pharmahuasca. Intravenous

injection as the fumarate salt appears to be the most

effective route; 0.2-0.4 mg/kg was described as

'hallucinogenic,' with the higher quantity seemingly

representing the maximum effects of the drug (STRASSMAN &

QUALLS 1994: 86). Inhalation of the vaporized free-base has

a threshold of activity in the 0.2-0.4 mg/kg range (BIGWOOD

& OTT 1977), and 40-50 mg was described as a 'large dose'

(0.5-0.7 mg/kg; MEYER 1992: 154); whereas SHULGIN noted 30

mg (0.4 mg/ kg) evoked a "complete psychedelic experience"

(SHULGIN 1976: 167). While psychoactivity was observed with

intramuscular injection of 30 mg or 0.4 mg/kg [as

hydrochloride salt; this dose was misstated as 0.2 mg/kgl,

0.7-1.0 mg/kg was described as the 'optimum' i.m. dose

(SZkRA 1957: 461); and experienced users found 1.0 mglkg

i.m. (as fumarate salt) "significantly less ...

hallucinogenic than ... previous experience with the smoked

[sic] drug" (STRASSMAN & QUALLS 1994: 86). Another

researcher characterized the i.m. dose-range as 0.75-1.0 mg/

kg, fixing the threshold level at 0.60 mg/kg: "there are no

symptoms at all on administering only 0.50-0.55 mg/kg"

(SAI-HALASZ 1962: 137; SAI-HALASZ et al. 1958); although

SHULGIN pegged an "abrupt threshold of activity" at 30 mg

(0.4 mg/kg) and estimated 50-70 mg i.m. (0.7-0.9 mg/kg)

provoked the "complete psychedelic experience." SHULGIN gave

75 mg (1.0 mg/kg) as the equivalent 'complete' dose via

subcutaneous injection (SHULGIN 1976: 167). We have seen

that the oral DMT-threshold in pharmahuasca is about 0.

3-0.4 mg/kg, and I would estimate that maximum effects would

require doses between 1.5-2.0 mg/kg, perhaps more.

Intranasal DMT free-base was inactive in doses of 5-20 mg

(0.07-0.28 mg/kg; TURNER & MERLIS 1959); likewise inactive

were doses of up to 125 mg DMT intrarectally (as a solution

of 185 mg bioxalate salt; 1.7 mg/kg; DE SMET 1983). These

psychonautic bioassay data are summarized in Table 1.

While the ß-carbolines clearly render DMT active orally, we

can hardly characterise this as potentiation. Indeed, it was

demonstrated more than 30 years ago that the artificial

MAO-inhibitor iproniazid markedly inhibited psychoactive

effects of DMT. In subjects given 0.35-0.83 mg/kg DMT

i.m.,greatly reduced psychoactivity was experienced when the

injections were repeated two days after having received 100

mg iproniazid daily for four days: "the DMT psychosis [sic]

... was less pronounced: there were illusions and

hallucinations, but without colours, or only with a few of

them, and only having the eyes closed" (SAI-HALASZ 1963:

386). The following year, pretreatment with the

MAO-inhibitor isocarboxazide "very markedly attenuated" or

entirely blocked effects of oral LSD-tartrate (subjects each

received LSD, 40 and 75 mcg; then both doses after 2 weeks

of isocarboxazide, 30 mg/day; both again after 5 weeks of

isocaboxazide treatment; RESNICK et al. 1964: 1211). The

MAOI nialamide also 'blocked' effects of LSD (GROF & DYTRYCH

1965). A survey conducted by researchers at the U.S.

National Institute of Mental Health found: "decrease in

response to LSD... in those people who had been taking an

MAO inhibitor" (BONSON 1994: 9; BONSON et al. 1996). It is

interesting to note that in a single experiment I found

pretreatment with isocarboxazide [Marplang], 3 doses of 10

mg in a single day, to render psychoactive 30 mg DMT

free-base ingested an hour after the final dose of this

artificial MAOinhibitor (OTT 1994,1999), and the reversible

inhibitor of monoamine oxidase, moclobemide, at 75-300 mg

oral dosage, is also effective at catalyzing the ayahuasca

effect in combination with an appropriate oral dose of a

visionary tryptamine (TORSTEN 1998).

It would thus appear that the locus of the ayahuasca effect

is peripheral and that the MAO-inhibitors which catalyze

oral activity of DMT may exert a sort of DMT blocking effect

in the brain. While MAOI can render DMT and other

tryptamines orally-active, they appear to render it far less

potent than when administered via other routes; serving as

activators, but not potentiators. We have seen that

long-term, daily administration of medicinal MAO-inhibitors

(which theoretically elevates brain-serotonine levels) can

partially or completely block the effects of both DMT and

LSD. This has been documented experimentally and also in

surveys of patients undergoing daily MAOI-therapy.

Conversely, the potent serotonine-antagonist methysergide or

UNIL-491 (Sanserte had "a very strong potentiating effect"

on i.m. DMT (SAI-HALASZ 1962: 138), at oral (1-2 mg) or i.m.

(0.5 mg) doses well below its own threshold for

psychoactivity (4.3 mg; ABRAMSON & ROLLO 1967).

Possible DMT-attenuating actions of MAOI might have some

bearing on the fact that DMT orally in pharmahuasca appears

to be significantly weaker than via other routes of

administration. Strangely, and quite at odds with the

limited data at our disposal, ß-carbolines and Peganum

harmala seeds have acquired the reputation of all-purpose

pan-potentiators of shamanic inebriants, and have been

combined by avid 'basement shamans' [contemporary,

non-traditional aficionados of shamanic inebriants] with

psilocine[4-OH-DMT]-containing mushrooms, LSD, even leaves

of Salvia divinorum Epling et Jativa [Labiatæl, the

visionary principle of which, the diterpenoid salvinorin A,

is not even an amine (OTT 1995b, 1996; SIEBERT 1994)!

Nevertheless, this ingenious discovery by South American

Indians of the ayahuasca effect- conceivably the most

sophisticated pharmacognostical discovery ever made in the

archaic world-bids fair to revolutionize contemporary,

non-traditional entheobotany of visionary shamanic

inebriants (OTT 1997).

These results with pharmahuasca have also been extended to

so-called 'ayahuasca analogues' or anahuasca-the use of

non-traditional source plants for either or both tryptamines

and ß-carbolines. Since well over 100 plant species in 27

families are known to contain simple ß-carbolines (ALLEN &

HOLMSTEDT 1980), some 70 of which contain known

MAO-inhibitors (OTT 1993,1994), and some 75 species in 14

families are reported to contain DMT and/or 5-MeO-DMT (OTT

1994; SMFM 1977), there are theoretically several thousand

combinations of two plants which could provoke the ayahuasca

effect. Indeed, such ayahuasca analogues have lately been

made from a variety of plants, although possibly only in

Amazonia was the ayahuasca effect exploited in archaic

ethnomedicine. Always used as source of ß-carbolines are the

seeds of harmel, Syrian rue or Peganum harmala, which are

sold worldwide for use as an incense, and the plant is

naturalized in North America and Europe (GRACIE & ZARKOV

1986; HASSAN 1967; OTT 1994). Since these seeds contain much

higher levels of ß-carbolines than do stems of

Banisteriopsis caapi ordinarily used in ayahuasca (2-7%

alkaloids, as opposed to an average of 0.45%; McKENNA et al.

1984; OTT 1994; POISSON 1965; RIVIER & LINDGREN 1972; SHAMMA

& ABDUL-GHANY 1977), as little as 2-3 grams of harmel seeds

will suffice per dose of anahuasca. Although this dose-level

serves to activate pure DMT or tryptamines in plants added

to the ayahausca analogue, an aqueous infusion of 15 grams

of harmel seeds sans additives acted as a Valium®-like

sedative, with no visionary effects (OTT 1995c),

Various sources of tryptamines have been used in these

analogues, such as roots of Desmanthus illinoensis, Acacia

phlebophylla F. von Muell. leaves [Leguminosæ]; and halms of

various strains of Phalaris spp. [Gramineæ] (FESTI &

SAMORINI 1993[4]; OTT 1995a). Presently the most widely-used

source of anahuasca tryptamines is root-bark of Mimosa

tenuiflora (Willd.) Poir. [= M. hostilis (Mart.) Benth.;

Legurninosæ], source of the ayahuasca-like Brasilian

traditional entheogen vinho da jurema.

In 1946, Brasilian microbiologist OSWALDO GONÇALVES DE LIMA

reported the continuing shamanic use of ajucá or vinho da

jurema among the Pancarurú Indians of Brejo dos Padres near

Tacaratú in the valley of the Rio São Francisco in southern

Pernambuco. He described the preparation of the potion as a

manual, cold, aqueous infusion of pounded root-bark of

jurema preta [Mimosa tenuiflora; as M. hostilis], with no

additive-plants nor cooking (GONÇALVES DE LIMA 1946).

Although thought by SCHULTES to be extinct (SCHULTES &

HOFMANN, 1980), we now know that some forms of shamanic

ceremony involving vinho da jurema have survived into the

20th century at least among the following indigenous groups

in Brasil: Xucurú of Serra de Ararobá in northern Pernambuco

(HOHENTHAL 1952); Kariri-Shoko of Colegio near the mouth of

the Rio São Francisco which demarcates the Alagoas/Sergipe

border (DA MOTA 1987); the Atikum of the Serra do Umã in

western Pernambuco (DE AZEVEDo GRONEwALD 1995); the Truká

(BATISTA 1995) and numerous other groups scattered sparsely

over the immense caatinga of northeastern Brasil (PINTO

1995; TROMBONI 1995). Moreover, in this century, the

indigenous jurema ceremony has been adopted symbolically by

syncretic Umbandista churches along the Brasilian coast,

where jurema preta is not native, centered especially around

Alhandra in southern Paraíba (VANDEZANDE 1975).

On the other hand, our only report of contemporary

indigenous use of potions prepared from jurema preta

root-bark comes from the Atikum of the Serra do Umá region

of Pernambuco. Other indigenous groups rather employ one or

another type of jurema branca, of which some 10 species have

been reported from 4 genera, all but one in the family

Leguminosæ: Acacia jarnesiana Willd. (VANDEZANDE 1975); A.

piauhyensis Benth. (LEMOS DE ARRUDA CAMARGO 1988); Mimosa

burgonia Aubl. (DE ANDRADE MELLO 1955 [as jurema marginada])

M. pudica L. (PINTO 1995); M. verrucosa Benth. (DA MOTA 1987

[as jurema mansal; LEMOS DE ARRUDA CAMARGO 1988; SANGIRARDI

1983); Pithecellobium acacioides Ducke; P. diversifolium

Benth.; P. dumosum Benth.; P. tortum Mart. (BATISTA 1995; DE

ANDRADE MELLO 1955; LEMOS DE ARRUDA CAMARGO 1988; SANGIRARDI

1983); and Vitex agnus-castus L. [Verbenaceæl (DA MOTA

1987).

It has been reported that Mimosa verrucosa contains DMT, but

there is nothing in the chemical literature to support this

assertion, and in fact none of these jurema branca species

is known to contain visionary tryptamines, although several

other species of Acacia do contain DMT and/or 5-MeO-DMT (OTT

1994,1999), whereas bracatinga or Mimosa scabrella Benth.

contains low levels [0.03%] of DMT in stem-bark [root-bark

untested]. A common source of fuel-wood in southeastern

Brasil, this species is known especially for "honey of

bracatinga, used as a digestive stimulant and for

circulatory problems" (DE MORAES et al. 1990). For a

discussion of phytotoxins [especially psychoactive] in

honeys, see my recent review article (OTT 1998a).

The taxonomy of jurema preta was recently systematized; its

distribution extending from the vast Brasilian caatinga

northward to the state of Oaxaca in southern Mexico (BARNEBY

1991). In Mexico the stem-bark is a well-known

ethnomedicine, tepescohuite, applied topically for burns and

as a vulnerary, but there is no evidence the ancient

Mesoamericans exploited the entheogenic virtues of jurema

preta (GRETHER 1988). Mexican material of tepescohuite stem

-bark was shown to contain low levels [0.03%] of DMT, but

there are no published analyses of corresponding root-bark,

although bioassays suggest it contains at least 1.0% DMT

(MECKES-LOZOYA et al. 1990; OTT 1994,1999).

GONÇALVES DE LIMA reported the isolation of 0.51% of an

alkaloid he called nigerina from root-bark of Mimosa

tenuiflora collected in Arcoverde, Pernambuco, giving the

melting-point as 45.8-46.8'C and the empirical formula C I

3HqNO (GONÇALVES DE LIMA 1946). Nine years later, 0.98%

nigerina was again isolated frorn Jurema preta root-bark

also collected in Arcoverde (DE MELO 1955). In 1959,

GONÇALVES DE LIMA supplied jurema preta root-material to

researchers at a U.S. pharmaceutical company, who isolated

0.57% DMT [m.p. 48-49'C, C12 H16 N21 (PACHTER et a]. 1959).

It is now thought that nigerina was an impure form of DMT,

perhaps contaminated with DMT-N-oxide [readily generated

from DMT under isolation conditions] and possibly other

compounds. Assuming that GONÇALVES DE LIMA's nigerina

consisted at least partially of DMT, his would represent the

first finding of DMT [originally synthesized in 1931 (MANSKE

193 1)] as a natural product, although priority must go to

M.S. FISH, who first isolated DMT from Anadenanthera

peregrina seeds and pods (FISH et al. 1955).

There are no published analyses of vinho da jurema potions,

but a 1983 collaboration between Karolinska Institutet and

Universidade Federal do Paraiba in João Pessoa led to two

independent analyses of jurema preta potions obtained from

an Umbandista juremeira in Alhandra, Paraíba. The first,

dated 12 November 1983, was analyzed in Sweden and reported

to contain 1-10 mg/mI DMT; while a more precise quantitative

analysis in João Pessoa of a potion said to be "identical to

the sample taken to Karolinska" but dated 5 December 1983,

found 7.46 mg/ml DMT, with the source root-bark containing

11% DMT (HOLMSTEDT 1983; SANCHEZ LEMUS 1984)! The Brasilian

group found DMT also in a jurema branca sample from

Alhandra, unfortunately unidentified. Thus we have a range

of reported DMT concentration in jurema preta root-bark from

1-11% [since PACHTER's group isolated 0.57% DMT, we can

assume a total content perhaps twice as high]; as compared

to 0.00-0.66% DMT in reported analyses [12 samples] of

Psychotria viridis or chacruna leaf, chief source of

tryptamines for ayahuasca potions. Even taking the highest

DMT level found in chacruna, jurema preta is 1.5-16.5 times

richer in DMT! As for the potions, we have no data on

amounts of vinho da jurema typically consumed, but 7.46

mg/ml DMT would correspond to 0.45-1.64 g DMT per dose,

taking the range of 60-220 ml reported for typical doses of

ayahuasca, analysis of which found 25 mg[220 ml]-36 mg[60

ml] DMT/dose, meaning that VINHO DA JUREMA may be 12.5-65

times higher in DMT than ayahuasca (OTT 1994, 1999)! The

vinho da jurema potions analyzed were said to be thick

residues or concentrates; even allowing for a 10-fold

concentration prior to analysis, vinho da jurema remains

1.25-6.5 times higher in DMT than ayahuasca. Inasmuch as the

36 mg[60 ml] DMT in ayahuasca doses was from Pucallpa and

Tarapoto, Perú, where the brews are also considerably

concentrated, we can assume the vinho da jurema analyzed to

be at least 2.5-3.0 times higher in DMT than typical

ayahuasca.

GONÇALVES DE LIMA and HOHENTHAL both described the formation

of foam atop the potions when the beaten jurema preta

root-bark was hand-squeezed in cold water, and analyses of

stem-bark of Mexican tepescohuite jurema preta have found

several novel triterpenoid saponins which could explain this

phenomenon (ANTON et al. 1993; JIANG et al. 1991 a, 1991 b).

Novel chalcone compounds called kukulkanins have also been

isolated from branches of Mexican tepescohuite (DOMINGUEZ

1989).

Since the ayahuasca effect depends on presence of

ß-carboline alkaloids from Banisteriopsis spp. or other

plants, which render DMT orally-active by inhibiting MAO,

there has been speculation concerning a lost or missing

ingredient to vinho da jurema, or regarding purported

content of ß-carbolines in jurema preta. However, no

ß-carbolines were found by HOLMSTEDT or SANCHEZ LEMUS, nor

in recent unpublished analyses of Mexican root-bark of

M..tenuiflora (CALLAWAY 1998). In the Serra do Umã, where

use of jurema preta potions survives, it was noted that

juice of maracuja was consumed freely during the jurema

ceremony (DE AZEVEDO GRUNEWALD 1995). Since maracuja juice,

from Passiflora spp. [Passifloraceae], contains ß-carbolines

(LUTOMSKI et al. 1975), it was suggested such might account

for oral activity of DMT in the potions. However, the

Passiflora spp. contain especially harman [or passiflorinel,

which was found not to be effective as a human MAOI in

pharmahuasca bioassays (OTT 1994,1999). None of the scanty

ethnographic reports support the notion of a lost or missing

additive-plant, although they do stress prodigal use of

smoked tobacco as adjunct to jurema ingestion. Recently it

was found that tobacco-smokers show 40% inhibition of

cerebral MAOB (FOWLER et al. 1996), which would seem to be

insufficient of itself to render DMT active orally, although

it could be a contributing factor. The MAOI effect of smoked

tobacco is not understood chemically, but low levels of

ß-carbolines are known from tobacco-smoke (JANIGER & DOBKIN

DE RIOS 1976).

Whereas a potion prepared from 10 g Mimosa tenuiflora

root-bark [3 times extracted in acidified hot water] was

inactive as to visionary effects, I recently found

psychoptic properties in a potion prepared from 25 g jurema

preta root-bark. The potion was prepared by the traditional

method, simply squeezing the beaten rootbark in cold water,

with no additives [2 times, 125 ml water each time]. The

vinho da jurema gave DMT-like effects commencing somewhat

sooner [20 min.] and lasting less time [> 2 h.] than is

typical for me with ayahuasca or pharmahuasca [ca. 45 min.;

lasting 2 h.+] (OTT 1998b). As there exist non-ß-carboline

MAOI, there is the possibility some unknown MAOI compound

exists in jurema preta. On the other hand, preliminary

analyses of Mexican root-bark suggest that, apart from free

DMT, jurema preta contains DMT bound or complexed to larger

molecules which might protect against deamination by MAO and

allow transport to the brain [where free DMT would probably

have to be generated by action of another enzyme] (CALLAWAY

1998). In any case, it is evident there is no lost or

missing ingredient, and vinho da jurema is potently

visionary by itself, prepared in the traditional manner, and

assuming an adequate dose. The apparent inactivity of simple

Atikurn jurema preta potions is thus likely due to

insufficient dosage or perhaps a weak strain, and not to the

lack of some lost ingredient. The same may hold true for the

reported lack of activity of a Kariri-Shoko jurema mansa

potion [Mimosa verrucosa]; or perhaps this species contains

no bioavailable DMT (DA MOTA 1987). Only further chemical

research linked to human psychonautic bioassays wilt resolve

the conundrum of the psychoptic pharmacology of vinho da

jurema. Meanwhile, jurema preta root-bark is being sold and

widely used in Europe and North America as a visionary

substrate for contemporary anahuasca potions.

Table 1

Human Pharmacology of Psychoptic Tryptamines

Compound/route Doses References

Range thresholdmaximal

Strassman &

DMT, i.v 0.05-0.40 0.2 0.4 Qualls 1994,

Strassman et

al. 1994

Meyer 1992;

DMT, vapor 0.06-1.00 0.2-0.4 1.0 Bigwood & Ott

1977

DMT, vapor 0.00-0.40 ? 0.4 Shulgin 1976

DMT, im 0.00-2.00 0.4 1.5 Szara 1957

Sai-Halasz et

DMT, i.m 0.50-1.00 0.6 1.0+ al. 1958;

Sai-Halasz

1962

DMT, i.m. ? 0.4 0.7-0.9Shulgin 1976

DMT, s.c. 0.00-1.00 ? 1.0 Shulgin 1976

DMT, p.o 0.00-13.00 - - Shulgin 1976

DMT, p.h. 0.25-2.00 0.4 2.0+ Ott 1993

(1996)

DMT, i.n 0.07-0.28 - - Turner &

Merlis 1959

DMT, i.r. 1.70-1.30 - - de Smet 1983

DET, i.m. 0.00-0.85 0.65-0.85? Boszormenyi et

al. 1959

5-MD 0.00-0.13 0.07-0.13? De Smet 1983

5-MD 0.17-0.17 0.17 ? Ott 1994

[i.v. = intravenous; i.m. = intramuscular; s.c.

subcutaneous; p.o. = peroral; p.h. = pharmahuasca; i.n. =

intranasal; i.r. = intrarectal; 5-MD = 5-MeO-DMT]

Acknowledgements

I am beholden to Drs. James C. Callaway, Mark S. Donnell,

Jochen Gartz, Robert Montgomery, Torsten and Alexander T.

Shulgin for discussions and advice. This paper is dedicated

to Prof. Richard Evans Schultes of Harvard University, in

grateful recognition of his pioneering role in elucidating

the botany and ethnopharmacognosy of ayahuasca potions.

---