Syrian Rue

Syrian rue, Peganum harmala, is a shrub of the arid steppes whose seeds contain beta-carbolines, harmine and harmaline foremost, which inhibit monoamine oxidase type A reversibly and competitively. It is almost nothing on its own: the cleanest trial that exists on its principal alkaloid finds no hallucination at any dose. It is dangerous because it lifts a metabolic barrier the body uses to protect itself from everything else. Its principal risk does not belong to it.

Neurochemical action

The seed contains two chemical families unrelated to each other, and confusing them is the first mistake one can make. The beta-carbolines (harmine, harmaline, harmalol, harmol, tetrahydroharmine) carry the MAO inhibition. The quinazoline alkaloids (vasicine, vasicinone, peganine) carry the uterotonic and abortifacient effect, which is the traditional use of the plant that is genuinely documented.

The seed extract inhibits human MAO-A reversibly and competitively, IC50 27 µg/L, and inhibits MAO-B poorly. On the isolated molecules, the published values span more than two orders of magnitude: keep only the order of magnitude, nanomolar to submicromolar at MAO-A, inactive at MAO-B.

The blockade is predominantly peripheral at ayahuasca doses, gastrointestinal and hepatic, which is enough to prevent the first pass of DMT: Riba et al., against placebo in 18 volunteers, measure negligible plasma harmine concentrations and an increased excretion of normetanephrine without the fall in deaminated metabolites that would be expected of an MAOI. A common over-reading concludes from this that harmine does not enter the brain, and that is false: [11C]-harmine is a validated PET radioligand of human brain MAO-A. Harmine crosses the barrier; it is its exposure that is low at these doses, and the distinction matters as soon as the dose rises.

The beta-carbolines are not classic psychedelics, and their own authors say so: Glennon et al. find a modest affinity at 5-HT2A, with no agonist action, and conclude that it is difficult to make them resemble the serotonergic hallucinogens. The only constituent with a direct serotonergic action is tetrahydroharmine, which weakly inhibits reuptake. It is not harmaline: the text that circulates under the heading of harmaline's SSRI effect has no support in the indexed literature. A second layer is finally added, independent of MAO: the beta-carbolines inhibit CYP3A4 and CYP2D6 in vitro, which concerns any narrow-margin medicine passing through these enzymes, and three substances of this wiki by name. 25I-NBOMe has CYP3A4 as its major enzyme and is active below the milligram: this is an inference drawn from two sources read, with no measurement and no published case, but it bears on the narrowest margin of the set. Dextromethorphan is a substrate of CYP2D6, which stacks a metabolic layer on top of the serotonergic risk treated below. 5-MeO-DiPT has CYP2D6 as its only O-demethylation route, with no documented backup route; but the direction of the clinical consequence is not established there, it is not known whether its dominant metabolite is active, and one must not write that closing this route makes things worse.

What it does on its own, and what no one has measured

The only object proper to it is recent: a phase 1 trial, 25 healthy volunteers, pure pharmaceutical-grade harmine up to 500 mg by the oral route. No frank hallucination, at any dose, neither reported nor observed. Marked somnolence, impaired concentration, a dose-dependent sensation of being high, and vomiting which becomes the dose-limiting toxicity between 100 and 200 mg and resolves by around 3 h whatever the antiemetic. No one, among those who felt effects, described them as pleasant. The authors conclude that harmine is not the principal psychoactive component of ayahuasca. Nothing equivalent exists for harmaline, which is nonetheless the majority constituent of the seed: the ranges that circulate for it come from work of 1967 and 1977 available as PDFs only, which we have not read.

At poisoning doses, the picture is neurological, digestive, then cardiovascular: visual hallucinations, agitation, diffuse tremor, ataxia, vomiting, bradycardia and hypotension. The most solid series is Moroccan, 200 cases over 24 years, with a lethality of 6.2% among the 114 cases whose outcome is specified; it bears on ingestions of several tens of grams, not on psychonautic use. We have moreover found no documented traditional psychedelic use of this plant, only sedative, emmenagogue, abortifacient and anthelmintic uses: its employment as a substitute for the ayahuasca vine is recent, and it is better to say so than to manufacture an antiquity for it.

The seed is not ayahuasca, and the gap runs the wrong way

Two published assays of the dried seed give, as total beta-carboline load, 10.6% and 6.15%, that is a factor close to two and not the factor of five that circulates: that factor of five is obtained by forgetting harmalol, itself also an MAO inhibitor. One gram of seeds therefore delivers between 62 and 106 mg of beta-carbolines, and the split between harmine, harmaline and harmalol is not stable from one method to another. No clinical trial of the plant in humans exists, moreover, and the bioavailability of harmine from a crushed seed has never been measured. Two sources converge on 2 to 4 g as the reference dose; at the highest contents, 3 to 4 g carry 320 to 420 mg of beta-carbolines, of the same order as an ayahuasca dose, which three sources place at around 180 to 250 mg.

The comparison is nonetheless not reassuring, and this is the central fact of this page. In ayahuasca, harmaline is a minor constituent, of the order of 3.5% of the load, behind harmine and tetrahydroharmine. In the seed it is the majority constituent, about 53%. At equal total load, the seed delivers of the order of fifteen times more harmaline. And harmaline is precisely the alkaloid that the most recent systematic review designates as the most problematic in preclinical work: it is harmaline that produces the lethal convulsive myoclonic syndrome with 5-HTP in the rat, harmaline that potentiates the hyperthermia of 5-MeO-DMT in the mouse, harmaline that carries the Purkinje cell degeneration observed in the rat and absent in the mouse. Some ritual groups do in fact boil their decoction longer on purpose, in order to convert harmaline into tetrahydroharmine. A crushed seed allows no adjustment of that kind.

Duration and tolerance

The plasma half-life of harmine is given as between 1 and 3 h, that of harmaline as around 2 h. Vomiting occurs 60 to 90 min after dosing and resolves by around 3 h. The only subjective effects properly timed, 4 to 5 h, were timed by the buccal route with pure harmine, which does not transpose to swallowed seeds.

The figure that is missing is the one that counts: the real duration of MAO-A inhibition in humans has never been measured, and a plasma half-life does not date the lifting of an enzymatic blockade. There is therefore no quantified safety window, and no one can say from what point on a combination goes back to being what it would be without the plant. The order of magnitude known in the other class gives the measure of the problem: after stopping tranylcypromine, 30 days were needed for the pressor response to tyramine to normalise. Tolerance in humans has never been studied.

Harm reduction

The plant on its own. The published severe cases show renal failure requiring haemodialysis, hepatic cytolysis, raised troponin and CPK-MB with a normal electrocardiogram. Prolonged neurological sequelae have been described in two cases, a severe peripheral polyneuropathy in both, a cerebellar ataxia in one only; the authors of the first themselves call them unusual and recall that the ataxia was reversible elsewhere. There is no antidote. Signs that should send someone to the emergency department: bradycardia or marked hypotension, consciousness altered beyond the expected somnolence, convulsions, ataxia that does not resolve, a fall in urine output, persistent tremor, raised temperature. Hyperthermia deserves a separate word: it marks serotonergic toxicity, not simple harmala poisoning, where hypothermia or a slight elevation is described instead.

Pregnancy: a contraindication, and it is the best supported point of this page. The plant is used as an abortifacient in several traditions, abortion is the reason for taking it in 13.5% of the Moroccan series, and two published cases associate foetal death with severe maternal sequelae. This contraindication rests on the quinazoline alkaloids and not on the beta-carbolines: the reassuring data on ayahuasca can neither ground it nor lift it.

Combinations. Syrian rue makes nothing stronger by an effect of its own: it closes the exit route. From this, three consequences that must not be confused. Overdose by blocked clearance, for any molecule eliminated by MAO, which makes the dose unpredictable. Serotonergic toxicity, which requires two different mechanisms, typically an MAOI plus a reuptake inhibitor. And interaction through the cytochromes, independent of MAO.

• Absolute contraindication: pharmaceutical MAOIs, including those not recognised as such, linezolid, methylene blue, moclobemide. Stacking two blockades on the same enzyme does not average them out.

• Serotonergic antidepressants: the best documented risk. There exists at least one published case of serotonergic toxicity attributed to P. harmala in a patient on fluoxetine and quetiapine, qualified as such by a systematic review read in full. A reservation not to be kept quiet: this case is inaccessible by any consultable route, we have not read it, and its clinical detail remains to be confirmed. For ayahuasca, by contrast, two independent systematic reviews conclude that no clear case of serotonin syndrome was found. Distinguish the two.

• Dextromethorphan is the most serious partner, and the only one with a case. A 46-year-old man, after three days of ayahuasca on a retreat, took a DXM cough syrup and diphenhydramine: severe serotonin syndrome, 38.9 °C, CK at 1,432 IU/L, cyproheptadine, midazolam, emergency intubation, discharge on day 3. Two reservations inseparable from the case: three products are present together in it, and the urine toxicology was in addition positive for benzodiazepines. Only the conference abstract has been read, not the article. DXM is moreover a SERT inhibitor as potent as fluoxetine at the concentrations reached in recreational use, and Gillman names it among the molecules involved in fatal serotonergic toxicities with MAOIs.

• Opioids are not to be treated as a block, and the distinction is actionable for someone on substitution. Implicated, deaths included: pethidine, tramadol, methadone, dextromethorphan, propoxyphene. Known not to be: morphine, codeine, oxycodone, buprenorphine. This second half is not an absolution: oxycodone figures among the opioids most frequently reported in serotonin syndrome in the WHO database, by mechanisms independent of the transporter.

• The other substances of this wiki: none has a published case with this plant, and it is necessary to say which is which. A list where everything is in red is no longer read. The six entries that follow are given at their real level, from the most solid mechanism to the most open, down to the one that has none.

• 5-MeO-DiPT, the most concerning combination of the set, and yet without a single case. Four converging supports, not to be confused with one another. A full 5-HT2A agonism at nanomolar affinity. A 5-HT1A brake, which is the strong argument because it is sourced: extracellular serotonin is held there only by its own autoreceptors, and the authors of that measurement conclude that anything inhibiting their function would produce greater effects. A minor MAO elimination route, blocking which closes a way out. And the precedent of the closest analogue, 5-MeO-DMT, which has one published death and one severe case with harmala alkaloids, and whose hyperthermia is very strongly potentiated by harmaline in the mouse, a fact this page already carries higher up. No human case of 5-MeO-DiPT plus an MAOI exists. To this is added a field fact that changes the reach of the warning: of eleven Japanese products sold as Amanita muscaria, four liquid extracts contained 5-MeO-DiPT and 5-MeO-DMT together with harmine and harmaline, plus atropine and scopolamine, none of these compounds being declared and nothing being quantified. One can therefore absorb the combination without having chosen it, and this is the only place in the wiki where that is documented.

• AMT, the only molecule of the set that already carries its own MAOI. It is both a serotonin releaser and a reversible inhibitor of MAO-A: adding an exogenous MAOI to it amounts to stacking the same risk twice. No published case exists, and the search was redone in order to be able to see the answer. Two inseparable reservations: this is an extrapolation and it must be given as such; and the inhibitory potency of AMT is not quantified, the only range read spanning three orders of magnitude, so do not write that AMT would be a potent MAOI.

• The three 2Cs of the set, 2C-I, 2C-E and 2C-T-7: the mechanism is clear, and the right word is not serotonin syndrome. MAO-A and MAO-B are their major elimination routes; blocking them closes the exit and exposure rises at unchanged quantity. The foreseeable risk is an overdose by blocked clearance, on molecules active at a few tens of milligrams: these three are direct 5-HT2A agonists, neither releasers nor reuptake inhibitors. No human case exists with any MAOI whatsoever: this is a class extension, and it must be written that way. Two internal nuances, already carried by their own pages. 2C-I is the most solid case, CYP2D6 not being involved for this precise compound, hence with no documented backup route. 2C-T-7 is on the contrary weakened: the only published direct test is negative, a team having co-incubated the 2C-Ts with an MAO-A inhibitor and an MAO-B inhibitor with no significant effect on cytotoxicity, whereas the same manipulation does aggravate that of mescaline and 2C-B. It is a cell model, which does not refute plasma accumulation, but it is the only test that has looked. Caution, never an established fact.

• DPT: an evidentiary void, and in both directions. DPT combined with an inhibitor of MAO has never been studied, in any species. DPT being active by the oral route without an MAOI, adding one brings no clear pharmacological benefit and adds a real serotonergic load. The argument rests on the absence of benefit, not on a measured mechanism, since no one has ever measured DPT as a substrate of MAO. And one must not harden it in the other direction either: to write that DPT would be protected from MAO would go beyond everything that exists.

• Morning glory and Hawaiian baby woodrose: weak, and honestly open. No published case of interaction between LSA and an MAOI exists, and the usual caution comes from an analogy with the simple tryptamines that does not hold: LSA is an ergoline, and no one has established that it is a substrate of monoamine oxidase. That does not mean reassuring. The class evidence runs the other way and it is published: the ergot alkaloids are serotonergic agonists, dihydroergotamine can provoke a serotonin syndrome in humans, and the seed contains, as measured, methysergide and methylergometrine. The basis is mechanistic and not empirical, and the two symmetrical errors are both to be avoided.

• Nitrous oxide: the pharmacological orphan of the set, and its place is stated in the negative. It is the only substance of this wiki for which the question does not arise: neither SERT nor MAO has ever been tested for this gas, its own page already says so in that sense, and nothing therefore allows a pharmacological interaction with Syrian rue to be attributed to it. What links them is elsewhere, in the emetic chain treated below, where it figures as a sedative.

• 5-HTP is the most direct and the most neglected warning. In the rat, harmaline plus 5-HTP produces a lethal convulsive myoclonic syndrome that neither of the two provokes on its own. An animal datum, given as such, but it targets a real practice among users of psychedelics.

• The convulsive load is a cross-cutting fact of this wiki, and for this plant it never reads other than in pairs. Its own signal is weak and twofold: convulsions figure in the published clinical pictures of poisoning, with no denominator, and the lethal myoclonic syndrome above bears on a combination, not on the plant alone. Ten of the fourteen substances of this corpus carry this signal, six on counted human cases and four on weaker support, Syrian rue among them; but it is the only one whose intended use consists in taking something else with it, and adding together two lowered epileptogenic thresholds is here the ordinary situation rather than the accident. Outside the set, tramadol and lithium, the latter being the only medicine to which a death in this corpus has been imputed.

• MDMA, amphetamines, cocaine: no data, neither case nor study. What circulates is a class inference, and must be given in those terms.

• The objection that a reversible inhibitor would keep one safe does not hold: moclobemide, the only reversible MAO-A inhibitor in clinical use, figures by name in the class concerned in Gillman. The extrapolation is not there; it is in the leap from a medicine at a controlled dose to a seed of unknown content.

Tyramine. No challenge test has ever been done with this plant nor with harmine in humans, and two systematic reviews leave the question at the stage of class reasoning. The rare blood pressure measurements are rather reassuring, but two of the five cases of a published comparative table presented hypertensive, including a woman who had taken her seeds with coffee. The best available is what a clinical trial does: a low-tyramine diet, 72 h before dosing.

Vomiting under a sedative is the most commonplace and the most lethal chain. Syrian rue is a reliable emetic, 33 to 57% vomiting in an ayahuasca context and 69.9% in the largest survey available, and vomiting is even the dose-limiting toxicity of pure harmine in the only existing clinical trial. In the set of substances of this wiki, the sedatives are dextromethorphan, diphenhydramine and nitrous oxide; alcohol and the benzodiazepines belong to the same column, and the other reliable emetics are morning glory, Hawaiian baby woodrose and nutmeg. The case that establishes the chain is not in this page: it is the only published death on DPT, a 20-year-old man whose autopsy concludes that there was a deep aspiration of gastric contents and a cerebral oedema from oxygen deprivation. The published order there is collapse, convulsion, then vomiting, hence in a person already unconscious: supervision must not wait for the first vomit before it starts. Here vomiting is expected, not feared, and the risk it carries cannot be dealt with after the event: what protects is someone who takes nothing, posted before the dose, and who knows how to turn a body onto its side.

In the emergency department, no one will be identified. No routine test detects these alkaloids, and a negative screen rules out nothing. Saying what has been taken, and keeping the bag, materially changes the management: the treatment of the serotonergic side has a name, it is cyproheptadine, and it has been used in the published cases.

Sources

• T. Herraiz et al. (2010), beta-carbolines of Peganum harmala and inhibition of human MAO, Food Chem Toxicol (seed contents, reversible and competitive inhibition).
• J.L. Ables et al. (2024), phase 1 single ascending dose trial of pure oral harmine, J Psychopharmacol (principal source for the dose).
• E. White et al. (2024), systematic review of the adverse events of ayahuasca and DMT, Int J Toxicol (78 articles; harmaline, ceremonial doses, preclinical data).
• A. Halman et al. (2024), drug interactions of the classic psychedelics, systematic review, J Psychopharmacol.
• P.K. Gillman (2005), MAOIs, opioid analgesics and serotonin toxicity, Br J Anaesth (the opioid-by-opioid distinction).
• J.P.M. Finberg and J.M. Rabey (2016), inhibitors of MAO-A and MAO-B, Front Pharmacol (the RIMA class, the 30-day delay after tranylcypromine).
• S. Achour et al. (2012), poisoning by Peganum harmala in Morocco, 200 cases, Therapie.
• M.A. Berdai et al. (2014), poisoning in a pregnant woman, Case Rep Emerg Med and G. El Aidouni et al. (2021), Ann Med Surg (neurological sequelae; note that the title of the second announces a maternal death that its text does not describe).
• J. Riba et al. (2003), human pharmacology of ayahuasca, placebo-controlled trial, JPET.
• M.R. Pranzatelli and S.R. Snodgrass (1987), harmala alkaloids and the myoclonic model, Exp Neurol (lethal syndrome with 5-HTP in the rat).
• X.L. Jiang, H.W. Shen, A.M. Yu (2015), potentiation by harmaline of the hyperthermia induced by 5-MeO-DMT, Neuropharmacology.
• T. Zhao et al. (2011), inhibition of CYP3A4 and CYP2D6 by the beta-carbolines, Phytother Res.
• L.M. Nielsen et al. (2017), cytochrome P450 enzymes involved in the metabolism of 25I-NBOMe and 25I-NBOH, Drug Testing and Analysis (source of the CYP3A4 warning, on the 25I-NBOMe side).
• Y. Hagino et al. (2021), dual actions of 5-MeO-DiPT at SERT and at the 5-HT1A receptor in the mouse, Neuropsychopharmacol Rep (the 5-HT1A brake, the sourced support of the 5-MeO-DiPT entry).
• K. Tsujikawa et al. (2006), analysis of Amanita mushrooms circulating in Japan, Forensic Sci Int (undeclared co-formulation with harmine and harmaline in the four liquid extracts, nothing quantified).
• E. Gil-Martins et al. (2024), neurotoxicity of the 2Cs and of the corresponding NBOMes, J Xenobiotics (the only published direct test of aggravation by MAO inhibition, negative for the 2C-Ts).
• M.A. Neukamm et al. (2024), fatal aspiration case after consumption of DPT, J Pharm Biomed Anal (anchor of the emetic chain; abstract read, full text not read).
• J.C. Bouso et al. (2022), adverse effects of ayahuasca, Global Ayahuasca Survey, PLOS Glob Public Health.
• ICEERS, Ayahuasca: Basic Info (harm reduction source, not academic; its formula on the absence of a documented case does not hold for the plant on its own).
• Erowid, Syrian Rue Vault (corpus of user reports, to be read on site; its licence forbids reproduction).
• B. Bakim, S. Sertçelik, O. Tankaya (2012), a case of serotonin syndrome under an antidepressant with concomitant use of Peganum harmala, Klinik Psikofarmakoloji Bülteni 22(4):359-361. Text not read: inaccessible by any consultable route. Authors and reference verified against the publisher's register; the existence and the qualification of the case are corroborated by Halman's systematic review, its clinical detail is not.
• D.E. Brush, S.B. Bird, E.W. Boyer (2004), MAOI poisoning resulting from online misinformation, J Toxicol Clin Toxicol 42(2):191-5. Primary report not read; details known through Halman. The dose in this case is unknown and it must not be presented as a case where three seeds would have sufficed.
• S. Pack, T. Ellett, J. Pham, Y. Ahmad (2026), Am J Respir Crit Care Med 212(Suppl 1), aamag162.4811. Conference abstract read, full article not read.