Despite their presence in the human pharmacopeia for millennia, we have yet to resolve the biochemical mechanisms by which the hallucinogens (psychedelics) so dramatically alter perception and consciousness. It is the only class of compounds that efficiently and specifically does so. For that matter, we do not fully understand the biochemistry of perception itself or how we live such a vivid and complex internal life in the absence of external stimulation. We do not understand the basic biochemical mechanisms of some of our most common experiences, such as the many human aspects of creativity, imagination or dream states. This is also true for extraordinary states of consciousness such as “visions” or spontaneous hallucinations or phenomena such as near-death experiences (NDE). And it is troubling that we have not sufficiently turned the scientific method on these latter subjects despite the profound role they have played in the evolution of our science, philosophy, psychology and culture.

The experiences derived from the administration of hallucinogens are often compared to dream states. However, the experience of administered hallucinogenic substances is far more intense, robust and overwhelming than the subtlety of mere dreams. By comparison, the natural biochemical processes for our related “hallucinatory” experiences are obviously far more highly regulated, occurring as an orchestrated and inherent function of the “normal” brain. Nonetheless, it is conceivable that attaining an explanation for these related natural human phenomena may lie in resolving the biochemical mechanisms involved in the more dramatic pharmacology of hallucinogens, recognizing that the complexities and intensity of the “administered” experience are, essentially, an overdose relative to corresponding natural regulatory controls. Given their status as “psychedelics” (mind-manifesting substances), increased study of the hallucinogens, particularly with advanced brain imaging and molecular biology approaches, may provide a better understanding of the “common” biochemistry that creates mind.

Perhaps the science behind the discovery of endogenous opioids offers us a corollary. We came to better understand the common human experience of pain through examining the pharmacology of administered opiates and the subsequent discovery of endogenous opioid ligands, receptors and pathways that are predominantly responsible for and regulate the experience and perception of pain. Such may also be the case for understanding perception and consciousness. With the discovery of the endogenous hallucinogen N, N-dimethyltryptamine (DMT, 1, Figure 1), perhaps, as with the endogenous opioids, we have a similar opportunity to understand perception and consciousness. Recent research has stimulated a renewed interest in further study of this compound as a neuro-regulatory substance and, thus, a potential neuro-pharmacological target. Taking results from these and more classical studies of DMT biochemistry and pharmacology together, this report examines some of the past and current data in the field and proposes several new directions and experiments to ascertain the role of endogenous DMT.

In terms of Western culture, DMT was first synthesized by a Canadian chemist, Richard Manske, in 1931 (Manske, 1931) but was, at the time, not assessed for human pharmacological effects. In 1946 the microbiologist Oswaldo Gonçalves de Lima discovered DMT's natural occurrence in plants (Goncalves de Lima, 1946). DMT's hallucinogenic properties were not discovered until 1956 when Stephen Szara, a pioneering Hungarian chemist and psychiatrist, extracted DMT from the Mimosa hostilis plant and administered the extract to himself intramuscularly (Szára, 1956). This sequence of events formed the link between modern science and the historical use of many DMT-containing plants as a cultural and religious ritual sacrament (McKenna et al., 1998), their effect on the psyche and the chemical structure of N, N-dimethyltryptamine.

The discovery of a number of hallucinogens in the 1950's and observations of their effects on perception, affect and behavior prompted hypotheses that the syndrome known as schizophrenia might be caused by an error in metabolism that produced such hallucinogens in the human brain, forming a schizo- or psycho-toxin (Osmond and Smythies, 1952). The presence of endogenous hallucinogenic compounds, related mainly to those resembling dopamine (mescaline) or serotonin (DMT), were subsequently sought. Although several interesting new compounds were found, the only known hallucinogens isolated were those derived from tryptophan (DMT, and 5-methoxy-DMT). Data were subsequently developed illustrating pathways for their endogenous synthesis in mammalian species, including humans. Over 60 studies were eventually undertaken in an attempt to correlate the presence or concentration of these compounds in blood and/or urine with a particular psychiatric diagnosis (for a review see Barker et al., 2012). However, there has yet to be any clear-cut or repeatable correlation of the presence or level of DMT in peripheral body fluids with any psychiatric diagnosis. Nonetheless, the discovery of endogenous hallucinogens and the possibilities rendered in various hypotheses surrounding their role and function in mental illness, normal and “extraordinary” brain function spurred further research into the mechanisms for their biosynthesis, metabolism and mode of action as well as for their known and profound effects on consciousness (Mishor et al., 2011; Araújo et al., 2015).

After the discovery of an indole-N-methyl transferase (INMT; Axelrod, 1961) in rat brain, researchers were soon examining whether the conversion of tryptophan (2, Figure 2) to tryptamine (TA; 3, Figure 2) could be converted to DMT in the brain and other tissues from several mammalian species. Numerous studies subsequently demonstrated the biosynthesis of DMT in mammalian tissue preparations in vitro and in vivo (Saavedra and Axelrod, 1972; Saavedra et al., 1973). In 1972, Juan Saavedra and Julius Axelrod reported that intracisternally administered TA was converted to N-methyltryptamine (NMT; 4, Figure 2) and DMT in the rat, the first demonstration of DMT's formation by brain tissue in vivo. Using dialyzed, centrifuged whole-brain homogenate supernatant from rats and humans, these same researchers determined that the rate of synthesis of DMT from TA was 350 and 450 pmol/g/hr and 250 and 360 pmol/g/h, using NMT as substrate, in these tissues, respectively. In 1973, Saavedra et al. characterized a nonspecific N-methyltransferase in rat and human brain, reporting a Km for the enzyme of 28 uM for TA as the substrate in rat brain. The highest enzyme activity in human brain was found in the subcortical layers of the fronto-parietal and temporal lobes and the cortical layers of the frontal parietal lobe. However, an INMT found in rabbit lung was shown to have a much higher Km (270 uM, Thompson and Weinshilboum, 1998; 340 uM, Raisanen and Karkkainen, 1978) than the brain enzyme in rats. This suggested that INMT may exist in several isoenzyme forms between species and possibly even within the same animal, each having different Km's and substrate affinities. INMT activity has subsequently been described in a variety of tissues and species. There have also been several reports of an endogenous inhibitor of INMT in vivo that may help regulate its activity and, thus, DMT biosynthesis (Wyatt et al., 1973a,b; Lin et al., 1974; Narasimhachari et al., 1974; Barker et al., 1981).

The combined data demonstrate that DMT is formed from tryptophan (2, Figure 2), a common dietary amino acid, via the enzyme aromatic L-amino acid decarboxylase (AADC) formation of TA (3, Figure 2) and its subsequent N, N-dimethylation. The enzyme indolethylamine-N-methyltransferase (INMT) uses S-adenosyl-l-methionine as the methyl source to produce N-methyltryptamine (NMT; 4, Figure 2) and then DMT (1, Figure 2). Both AADC and INMT act on other substrates as well. As a historical and research note regarding DMT, there was initial confusion and misidentification of the products formed when using 5-methyltetrahydrofolate (5-MTHF) as the methyl source in INMT studies due to formation of indole-ethylamine condensation products with formaldehyde (tetrahydro-beta-carbolines) (Barchas et al., 1974; Lin and Narasimhachari, 1974; Rosengarten and Friedhoff, 1976; Barker et al., 1981).

There has also been interest in the role of INMT and DMT biosynthesis in maturation and development. Relatively elevated levels of INMT activity have been found in the placenta from a variety of species, including humans (Thompson et al., 1999). INMT activity in rabbit lung was reported to be elevated in the fetus and to increase rapidly after birth, peaking at 15 days of age. It then declined to mature levels and remained constant through life (Lin et al., 1974). In this regard, Beaton and Morris (1984) have examined the ontogeny of DMT biosynthesis in the brain of neonatal rats and rats of various ages. Using gas chromatography-mass spectrometry with isotope dilution for their analyses, DMT was detected in the brain of neonatal rats from birth. DMT levels remained low (1–4 ng/g of whole brain tissue) until days 12 and 17 at which time they increased significantly and then returned to the initial low levels for all subsequent ages. There has yet to be any follow-on research as to the significance of this change in DMT concentrations during rat brain neurodevelopment or correlation with possible changes of INMT activity in other developing tissues, specifically during days 12–17. Nonetheless, these findings correlate well with the Lin et al. (1974) data for INMT changes in rabbits and deserve further inquiry.

There is a significant literature concerning INMT, particularly in peripheral tissues. INMT and its gene have been sequenced (Thompson et al., 1999), commercial antibodies for its detection have been developed and commercial probes exist for monitoring its mRNA and gene expression. A study using Northern blot detection of the INMT mRNA conducted by Thompson et al. (1999) in the rabbit suggested that INMT was present in significant quantities in the periphery, and particularly the lung, but that it was almost non-existent (low to absent) in the brain. These data became the foundation for several hypotheses that any neuropharmacological effects of endogenous DMT must lie in its formation in the periphery and its subsequent transport into the brain. This idea was strengthened by the fact that DMT has been shown to be readily, and perhaps actively, transported into the brain (Cozzi et al., 2009). However, the data concerning the apparent absence of INMT in brain would appear to be in conflict with the many earlier studies that demonstrated both in vivo and in vitro biosynthesis of DMT in the brain. Indeed, several studies had identified INMT activity or the enzyme itself in the central nervous system (CNS) including the medulla, the amygdala, uncus, and frontal cortex (Morgan and Mandell, 1969), the fronto-parietal and temporal lobes (Saavedra et al., 1973) and, more recently, the anterior horn of the spinal cord as well as the pineal gland (Cozzi et al., 2011).

Thus, in 2011, Cozzi et al. sought to determine why earlier studies (Thompson et al., 1999) had not detected significant INMT in brain using Northern blots despite several reports that brain tissue had been shown to synthesize DMT from TA. One possibility was that INMT was “expressed in nervous tissue but that in some situations, INMT mRNA is not detectable by Northern analysis (e.g., the INMT gene is inducible, INMT expression is limited to specific brain nuclei, or INMT mRNA in brain is short-lived).” Examining primate nervous system tissues (Rhesus macaque spinal cord, pineal gland, and retina) probed with rabbit polyclonal antibodies to human INMT, all three tissues tested positive. INMT immunoreactivity in spinal cord was found to be localized in ventral horn motoneurons. The study also showed that INMT response was “robust and punctuate” in the pineal gland. Further, intense INMT immunoreactivity was detected in retinal ganglion neurons and at synapses in the inner and outer plexiform layers (Cozzi et al., 2011). In 2012, Mavlyutov et al. reported that INMT is also localized in postsynaptic sites of C-terminals of rat motoneurons in close proximity to sigma-1 receptors, which have been linked to control of the activities of ion channels and G-protein-coupled receptors. It was proposed that the close association of INMT and sigma-1 receptors suggests that DMT is synthesized locally to effectively activate sigma-1 in motoneurons. It has been further proposed that DMT is an endogenous sigma-1 receptor regulator (Fontanilla et al., 2009; Su et al., 2009).

Taking these newer data together with historical in vitro and in vivo results regarding INMT enzyme activity in the brain and CNS, it is now clear that the work of Thompson and Weinshilboum (1998) is not the final word on DMT biosynthesis in the brain.

The metabolism of DMT has been thoroughly studied, with a great deal of newer data being provided from studies of ayahuasca administration (McIlhenny et al., 2012; Riba et al., 2012). All of the in vivo metabolism studies have shown that exogenously administered (IV, IM, smoking, etc.) DMT is rapidly metabolized and cleared, with only a small fraction of IV or IM administered DMT subsequently being found in urine. For example, 0.16% of an intramuscular dose of DMT was recovered as the parent compound following a 24 h urine collection (Kaplan et al., 1974). DMT administered in this manner reached a peak concentration in blood within 10–15 min and was below the limits of detection within 1 h. It was estimated that only 1.8% of an injected dose was present in blood at any one time. Due to rapid metabolism in the periphery, DMT is not orally active, being converted to inactive metabolites before sufficient penetration to the brain can occur (low bioavailability). DMT is only orally active if co-administered with a monoamine oxidase inhibitor (MAOI). DMT is pharmacologically active following administration by injection (intravenous or intramuscular routes) or smoking (vaporization and inhalation), pathways which can avoid first-pass metabolism by the liver to some degree (Riba et al., 2015). The time to onset of effects is rapid (seconds to minutes) by these routes and short lived (15–60 min depending on dose and route).

The primary route of metabolism for DMT (1, Figure 2) is via monoamine oxidase A (MAO-A), yielding indoleacetic acid (IAA; auxin; 5, Figure 2). The other metabolites formed include DMT-N-oxide (DMT-NO; 6, Figure 2), the second most abundant metabolite, and lesser amounts of N-methyltryptamine (NMT; 4, Figure 2), which, along with TA, is also a substrate for MAO-A (Fish et al., 1955; Szara and Axelrod, 1959; Barker et al., 1980, 1981; Figure 2), with both yielding IAA. Inhibition of MAO leads to a shift in favor of the amounts of DMT-NO and NMT formed (Riba et al., 2003). Other metabolites have been reported, such as 6-hydroxy-DMT (6-OH-DMT), (Szára, 1961) as well as products from a peroxidase pathway, reported to yield N, N-dimethyl-N-formyl-kynuramine, and N, N-dimethyl-kynuramine (Tourino et al., 2013; Gomes et al., 2014). However, these latter metabolites have yet to be identified in vivo. Metabolites also result from the cyclization of an intermediate iminium ion that forms during demethylation of DMT, yielding 2-methyl- 1,2,3,4- tetrahydro-beta-carboline (MTHBC; 7, Figure 2) and THBC (8, Figure 2; Barker et al., 1980, 1981).

The primary role of MAO-A in the metabolism of DMT has been further confirmed by pretreatment of experimental subjects with the MAO inhibitor (MAOI) iproniazid as well as other MAOIs (Lu and Domino, 1974; Moore et al., 1975; Shah and Hedden, 1978; Barker et al., 1980, 1981), the ability of the MAO-inhibiting harmala alkaloids of ayahuasca to make DMT orally active (McKenna, 2004) and the increased half-life and extended effects of an α, α, β, β-tetradeutero-DMT (D₄DMT; 9, Figure 3), which is less susceptible to MAO-A metabolism due to the kinetic isotope effect (Barker et al., 1982, 1984; Beaton et al., 1982).

Barker et al. (2012) have published a thorough overview of the 69 published studies examining blood, urine and cerebrospinal fluid detection of endogenous N, N-dimethylated tryptamines [N, N-dimethyltryptamine (DMT), 5-hydroxy-DMT (bufotenine, HDMT), and 5-methoxy-DMT (MDMT)]. Nearly all of the studies were directed at establishing a relationship between the presence and/or level of these compounds and a psychiatric diagnosis. In total, the 69 studies examined DMT in thousands of subjects. A critical review of these data determined, however, that most early studies reporting rather high concentrations of these compounds in blood and/or urine were most likely in error and any correlations based on these data were likewise probably incorrect. The reasons for this conclusion were: (1) Based on current analytical requirements for unequivocal structure identification, it is highly probable that many of these studies misidentified the target analyte. (2) If properly identified, the studies showed that a psychiatric diagnosis was not a necessary or sufficient criterion for finding one or more of these hallucinogens in various body fluids; “normal” controls were also positive (and sometimes higher) for these compounds. Nevertheless, it was also concluded that, particularly where mass spectral evidence was provided, DMT and HDMT are endogenous and can often be successfully measured in human body fluids. The evidence was less compelling for MDMT where the only two MS-based positive studies—in CSF—were performed by the same research group. There was no mass spectral data demonstrating detection of MDMT in blood or urine. There was also no study that attempted a determination of HDMT in CSF.

In conducting studies to determine the natural occurrence of a compound as being endogenous, it is also necessary to eliminate other possible dietary or environmental sources. Of the 69 studies reviewed, many addressed the possible source of DMT as being from diet or gut bacteria (Barker et al., 2012) by using special diets. Of those conducted, it was determined that neither was a source but additional research in this area using more modern technology and a more standard diet across studies is a necessity. There have also been only a few efforts to examine the many variables that may influence the levels of these compounds, such as circadian or diurnal variations, sleep stages and gender-age-related differences. Indeed, most of the studies collected only a single time point or were from 24 h collections (urine). Such infrequent sampling makes it impossible to assess central DMT production from peripheral measurements and suggests, perhaps incorrectly, that DMT only appears intermittently or not at all. In trying to compare the results, interpretations and correlation of the data were hampered by variability in sampling methods, amount of sample assayed, type of sample (plasma, serum and/or whole blood), divergent techniques and analytical methodology that also had highly variable or unspecified limits of detection.

There is a significant literature correlating the binding affinity of DMT and related hallucinogens for the 5HT2A receptor and its subset of receptors with other hallucinogens and their subsequent behavioral effects (Glennon et al., 1978; Nichols, 2004, 2016; Blough et al., 2014; Carbonaro and Gatch, 2016). However, DMT has been shown to interact with a variety of ionotropic and metabotropic receptors. While the subjective behavioral effects of exogenously administered DMT appear to be primarily acting via 5-HT2A receptors, the interaction of other receptors, such as other serotonergic and glutaminergic receptors, may also play a synergistic and confounding role. Indeed, the activation of frontocortical glutamate receptors, secondary to serotonin 5-HT2A receptor-mediated glutamate release, appears to be a controlling mechanism of serotonergic hallucinogens (dos Santos et al., 2016a,b). However, although this type of receptor research is quite mature, these findings have yet to define and accurately correlate what makes a compound hallucinogenic vs. compounds that have similar binding characteristics that are not hallucinogenic. Clearly, we are missing some pieces to the hallucinogen receptor/mode-of-action puzzle.

For example, Keiser et al. (2009) have shown that DMT binds to a variety of 5-HT receptors and that such binding does have physiological relevance. In their study, the role of 5-HT2A agonism in DMT-induced cellular and behavioral effects was examined in both cell-based and 5-HT2A knock-out mouse models. It was reported that “DMT binds to 5-HT1A, 5-HT1B, 5-HT1D, 5-HT2A, 5-HT2B, 5-HT2C, 5-HT5A, 5-HT6, and 5-HT7 receptors with affinities from 39 nM to 2.1 μM” (Keiser et al., 2009). Nonetheless, it was observed that DMT was not only a potent partial agonist at 5-HT2A but also that the DMT-induced head twitch response, a common measure of hallucinogenic activity, occurred only in wild-type mice but not in 5-HT2A knockout mice. However, it has been shown (Strassman, 1996) that the mixed 5-HT1A/1B antagonist pindolol markedly potentiates the subjective effects of DMT in humans. Furthermore, DMT-enhanced inositol trisphosphate production has been shown to persist even in the presence of the 5-HT2A antagonist ketanserin (Deliganis et al., 1991), suggesting other receptor sites for DMT's effects. Of interest is the finding of Urban et al. (2007) that receptors, such as the 5-HT family, can couple to multiple effectors, which allows receptor agonists to produce different pharmacological endpoints. Thus, certain compounds may selectively activate a specific subset of effectors producing a functional selectivity that complicates the interpretation of observed psychopharmacological or biochemical effects. In this regard, Carhart-Harris and Nutt (2017) have recently offered a novel bipartite model of serotonin neurotransmission involving co-modulation of the 5-HT1A and 5-HT2A receptors. This bipartite model purports to explain how different serotonergic drugs (including psychedelics) modulate the serotonergic system in different ways to achieve their observed pharmacology.

Clearly the 5-HT2A receptor is involved in the mode of action of DMT and other hallucinogens, but is it also clear that this is not the sole receptor on which we should rely for an overall explanation (Ray, 2010; Halberstadt and Geyer, 2011; Nichols, 2012, 2016).

Despite the failure of serotonin receptor binding theory to completely explain hallucinogenic activity, these observations support the 5-HT2A receptor as being a possible primary target for DMT's hallucinogenic effects (Keiser et al., 2009). While DMT has been shown to bind to the 5-HT2A receptor with relative high affinity (IC50 75 ± 1 nM; McKenna et al., 1990), many other compounds that lack DMT's visual effects have a higher affinity for the 5-HT2A receptor (McKenna et al., 1990).

In examining the possible complex interaction of multiple systems that may be necessary to explain the effects of compounds such as DMT, attention has also turned toward additional possible binding sites. Another set of functionally relevant binding sites for DMT is the family of trace amine-associated receptors (TAARs) (Burchett and Hicks, 2006; Wallach, 2009). DMT has been shown to be an agonist in binding to TAAR-1 with high affinity, causing activation of adenylyl cyclase and cAMP accumulation in TAAR1 transfected HEK293 cells. However, as is the case with the 5-HT2A receptor, other psychedelics and non-psychedelics also stimulate cAMP production following binding at TAAR1. There has yet to be sufficient research of TAAR to determine what role, if any, this class of receptors plays in the pharmacology or endogenous function of DMT. Thus, the research to date regarding the role of TAAR receptors suffers from the same lack of explanation for the mode of action of the hallucinogens as the 5-HT2A but may comprise a piece of what is obviously a complex set of interactions.

Another receptor family has also been implicated; the sigma-1 receptor. One of the possible roles of the sigma-1 receptor appears to be to act as an intracellular chaperone between the endoplasmic reticulum (ER) and mitochondria. In this role, it is involved in the transmission of ER stress to the nucleus (Carbonaro and Gatch, 2016). This process would be expected to result in the enhanced production of anti-stress and antioxidant proteins, with the activation of sigma-1 mitigating the possible damage done by hypoxia or oxidative stress (Szabo et al., 2014, 2016; Szabo and Frecska, 2016). Using in vitro cultured human cortical neurons (derived from induced pluripotent stem cells), monocyte-derived macrophages, and dendritic cells, Szabo et al. (2014) have shown that DMT greatly increases the survival of these cell types in severe hypoxia (0.5% O₂), apparently via its interaction with sigma-1 receptors. A decreased expression and function of the alpha subunit of the hypoxia-inducible factor (HIF-1) was also observed, suggesting that DMT-mediated sigma-1 activation may alleviate hypoxia-induced cellular stress and increase survival via decreased expression and function of the stress factor HIF-1α in severe hypoxia. Such a mechanism has relevance to stroke, myocardial infarct or similar arterial occlusive disorders, cardiac arrest, and perinatal asphyxia, all conditions associated with hypoxic consequences (Carbonaro and Gatch, 2016). Szabo et al. (2016) and Szabo and Frecska (2016) have speculated that DMT may also contribute to neuroregenerative and neurorestorative processes by modulating the survival of microglia-like cells.

These sigma-1 associated effects may also be related to findings that DMT affects the rate of genetic transcription associated with synaptic plasticity (O'Donovan et al., 1999; González-Maeso et al., 2007), increased expression of brain-derived neurotrophic factor (BDNF) expression associated with synaptic plasticity (O'Donovan et al., 1999), cognitive processes such as memory and attention (DeSteno and Schmauss, 2008), and modulation of efficacy and plasticity of synapses (Soulé et al., 2006).

The sigma-1 receptor has been implicated in several neurobiological disorders and conditions and is found widely distributed though out the body, including in the CNS. However, both hallucinogens and non-hallucinogens bind to sigma-1 receptors, again complicating an attribution to this receptor as the primary site of DMT's action. Further, DMT binds to sigma-1 receptors at what should be considered as a high concentration (EC50 = 14 μM vs. about 75 nM for 5-HT2A) but does, nonetheless, have agonist activity. INMT has been shown to be co-localized with sigma-1 receptors in C-terminals of motor neurons (Mavlyutov et al., 2012) and such intracellular synthesis would allow for DMT accumulation and storage, producing the necessary μM concentrations for its action. It is also important to consider that the role of endogenous DMT is not necessarily to produce the same effects as observed from exogenous administration and such a “normal” role may be one of its biological assets.

It has also been observed that sigma-1 receptor agonists are potentially neuroprotective (Frecska et al., 2013). DMT has been shown to reduce neuronal inflammation via the sigma-1 receptor (Szabo et al., 2014) and can also induce neuronal plasticity, a long-term recuperative process that goes beyond neuroprotection (Tsai et al., 2009; Ruscher et al., 2011; Kourrich et al., 2012). Sigma-1 receptors can also influence cell survival and proliferation (Collina et al., 2013) and Frecska et al. (2013) have suggested that DMT is protective during cardiac arrest and perinatal development. With respect to the ontogeny of DMT, Lin et al. (1974) and Beaton and Morris (1984) have examined changes in INMT activity and DMT biosynthesis, respectively, with age in the rat. Taken together, changes in INMT levels consequently yielded increased DMT synthesis. It is possible that DMT-mediated sigma-1 receptor activity is also increased during this period to induce neuronal changes in newborns. Several selective sigma-1 receptor agonists have been shown to be protective against excitotoxic perinatal brain injury (Griesmaier et al., 2012) and ischemic neurodegeneration in neonatal striatum (Yang et al., 2010). In addition, it has been suggested that adequate expression of placental INMT may be necessary for pregnancy success (Nuno-Ayala et al., 2012).

Szára (1956, 1961) originally reported that the effects of a medium dose (0.7 mg/kg) of DMT, given intramuscularly, were similar to those of mescaline and LSD, including visual illusions and hallucinations, distortion of body image, speech disturbances, mood changes and euphoria or anxiety (dependent on set and setting). Several other studies have replicated these findings using either IV or IM administrations (Turner and Merlis, 1959; Rosenberg et al., 1964; Gillin et al., 1976; Strassman et al., 1994a,b). Intramuscular effects of DMT at a reported dose of 0.2–1 mg/kg (Szára, 2007) generally had a rapid onset (2–5 min) and lasted 30–60 min. The IM effects are usually less intense than intravenous or inhalation-of-vapor routes of administration.

The subjective effects of DMT from ayahuasca administration (0.6–0.85 mg/kg DMT; Riba et al., 2003) usually appear within 60 min, peak at 90 min and can last for approximately 4 h (Cakic et al., 2010). The prolongation of effect is attributed to the MAOI effects of the constituent harmala alkaloids. Riba et al. (2015) have also reported the effects of oral and vaporized DMT alone. As expected, oral ingestion of pure DMT produced no psychotropic effects. Vaporized DMT was found to be quite psychoactive. This study also showed that smoked DMT caused a shift from the MAO-dependent route to the less active CYP-dependent route for DMT metabolism. Commonly used doses for vaporized or inhaled free-base DMT are 40–50 mg, although a dose may be as much as 100 mg (Shulgin and Shulgin, 1997). The onset of vaporized DMT is rapid, similar to that of i.v. administration, but lasts less than 30 min. It is of interest to note that intranasal free-base DMT is inactive (0.07–0.28 mg/kg; Turner and Merlis, 1959) as is DMT administered rectally (De Smet, 1983).

There is also additional significant literature concerning the administration of DMT via consumption of ayahuasca. While of great scientific interest, this subject is not reviewed here. This is mainly due to the complexity of composition of ayahuasca, especially the presence of significant MAOI effects.

Strassman et al. (1994a,b) have reported dose-response data for intravenously administered DMT fumarate's neuroendocrine, cardiovascular, autonomic, and subjective effects in a group of experienced hallucinogen users. DMT was administered at doses of 0.05, 0.1, 0.2, and 0.4 mg/kg to 11 experienced hallucinogen users. The results of these studies showed peak DMT blood levels and subjective effects were attained within 2 min after drug administration and were negligible at 30 min. DMT was also shown to dose-dependently elevate blood pressure, heart rate, pupil diameter, and rectal temperature, in addition to elevating blood concentrations of β-endorphin, corticotropin and cortisol. Prolactin and growth hormone levels rose equally at all doses of DMT. Levels of melatonin were unaffected. The lowest dose that produced statistically significant effects relative to placebo and that was also hallucinogenic was 0.2 mg/kg (Strassman, 1991, 1996).

The effects observed and the biochemical and physiological parameters measured in these studies add needed insight into the role and function of endogenous DMT. However, we must distinguish the effects of exogenously administered DMT from that which may be observed from its natural role as an endogenous substance. Exogenous administration of a bolus of DMT represents an “overdose” of a naturally occurring compound that may, when administered in this manner, exert a more complex pharmacology. However, this could also be true of any physiological change that produced a “normal” elevation in endogenous DMT, such as a response to stress or hypoxia, but with the entire process still remaining under a greater degree of biochemical control and response and the elevation possibly occurring in only certain brain areas or systems. For exogenously administered DMT we know plasma concentrations between 12 and 90 ng/ml (Callaway et al., 1999; Yritia et al., 2002; Riba et al., 2003) must be attained in order to produce hallucinogenic effects. The concentrations actually attained in whole brain or in specific areas required to produce hallucinogenic effects from such administrations are unknown.

While there have been several studies reporting neuroimaging data from volunteers consuming ayahuasca (Bouso et al., 2015; dos Santos et al., 2016a,b), there is minimal neuroimaging data for the administration of DMT alone (Daumann et al., 2008, 2010). Using functional magnetic resonance imagining (fMRI) techniques, administration of DMT “caused a decreased blood oxygenation level-dependent response during performance of an alertness task, particularly in the extrastriate regions during visual alerting and in temporal regions during auditory alerting.” (Daumann et al., 2010). It was concluded that the effects for the visual modality were more pronounced. Imaging data for other hallucinogens, such as psilocybin and LSD, have been generated (dos Santos et al., 2016b). dos Santos et al. (2016b) have concluded that “the acute effects of hallucinogen administration, as interpreted from imaging studies, included excitation of frontolateral/frontomedial cortex, medial temporal lobe, and occipital cortex, and inhibition of the default mode network.” For long-term use, the administration of hallucinogens was associated with “thinning of the posterior cingulate cortex, thickening of the anterior cingulate cortex, and decreased neocortical 5-HT2A receptor binding.” It was also suggested that hallucinogens “increase introspection and a positive mood by modulating brain activity in the fronto-temporo-parieto-occipital cortex” (dos Santos et al., 2016b).

In 1976, Christian et al., published the accumulated evidence that DMT was a naturally occurring transmitter in mammalian brain, having met the criteria for such a designation at the time; “1) the synthetic enzymes and substrates are present in the CNS for the production of DMT, 2) a binding site is present to react with the compound and 3) the compound is found in human CSF and isolated synaptic vesicles from rat brain tissue” (Christian et al., 1976). Additional criteria have been added over the years, such as demonstration of electrophysiological activity. Indeed, DMT had also been shown to change the transepithelial and intracellular potentials of the blow-fly salivary gland and to increase the production of cyclic AMP (Berridge, 1972; Berridge and Prince, 1974) early on. Another added criterion is that a pathway for DMT's metabolism and removal must be demonstrated. Pathways of DMT metabolism in the brain are well understood (Barker et al., 1980) and newer data offers other mechanisms, such as uptake into synaptic vesicles and neurons, for controlling its synaptic levels (Cozzi et al., 2009; Blough et al., 2014). Like any neurotransmitter, uptake and storage can allow a reservoir of DMT to remain stored in vesicles, ready for release, and provide a mechanism for protecting and concentrating the compound (Frecska et al., 2013).

Christian et al. (1977) subsequently described a specific high-affinity (Kd = 30 nM) binding site for DMT on purified rat synaptosomal membranes that was also sensitive to LSD but not to serotonin. DMT was also shown to lead to the production of cAMP in synaptosomal membrane preparations as well as in rat brainstem slices and rat cerebrum in vivo (Christian et al., 1977). Unfortunately, no additional research on these findings has been reported. Other studies have also demonstrated that administered DMT becomes localized in the synaptosomal fraction of rat brain following administration and is detected in the vesicular fraction of such preparations (Barker et al., 1984). Further, the Mg²⁺ and ATP dependent uptake of DMT (Corbett et al., 1975) into rat brain vesicles has also been demonstrated as has apparent high and low affinity uptake sites for active transport of DMT in rat brain cortical slices.

The supporting data for DMT as a neurotransmitter have continued to accumulate. DMT has also been shown to be taken up into neuronal cells via serotonin uptake transporters (SERT) on neuronal plasma membrane (Berge et al., 1983; Whipple et al., 1983; Nagai et al., 2007; Cozzi et al., 2009) and Cozzi et al. (2009) have shown sequestration of DMT into synaptic vesicles from the cytoplasm by the neuronal vesicle monoamine transporter 2 (VMAT2). Blough et al. (2014) have also shown that DMT releases 5-HT via SERT with an EC50 in the low nM range. This indicates that DMT is a substrate for the SERT transporter and provides a further mechanism for the neuronal accumulation of DMT. Newer data concerning INMT in specific brain areas (Cozzi et al., 2011) and its presence in perfusates of the pineal gland of living rats (Barker et al., 2013) add additional evidence for DMT's potential role as a neurotransmitter. At a minimum, the anatomy, pharmacology and physiology of DMT have been sufficiently characterized and demonstrated to afford DMT the classification as a putative neurotransmitter.

The concentration of DMT into vesicles and its release at the synaptic cleft would permit elevated concentrations of DMT, perhaps sufficient to elicit its known pharmacological actions as well as other effects. It would also be protected from MAO degradation. Peripheral production of DMT would not be required. It may also be the case that brain DMT biosynthesis is inducible in response to specific physiological effects, causing an increase in concentration in specific cell types and areas. This being the case, the idea that a pharmacologically relevant blood level of DMT must be attained before such effects are observed (Nichols, 2017) from endogenous production of DMT would not be relevant.

There has been a renewed interest in using hallucinogenic drugs as therapeutics in clinical research to address depression (Berman et al., 2000; Aan het Rot et al., 2010; Buchborn et al., 2014), obsessive-compulsive disorder (Moreno and Delgado, 1997), the psychological impacts of terminal illness (Grof et al., 1973; Grob et al., 2011), prisoner recidivism (Hendricks et al., 2014), and substance abuse disorders, including alcohol (Bogenschutz et al., 2015) and tobacco (Mangini, 1998; Krebs and Johansen, 2012). Most of these studies have examined the use of LSD, psilocybin or ayahuasca instead of DMT alone.

In the history of use of DMT-containing “remedies,” ayahuasca has perhaps the longest record (Dos Santos et al., 2011; Alonso et al., 2015; Pic-Taylor et al., 2015). Long-term use of ayahuasca has been shown to produce measurable changes in the brain itself, such as differences in midline brain structures as determined from MRI studies (dos Santos et al., 2016a,b). While such effects may not appear to be of therapeutic value, long-term ayahuasca users (>10 years) have shown reduced ratings of hopelessness (Santos et al., 2007). Long-term ayahuasca use has also produced marked improvement in depressive symptoms with no concomitant mania or hypomania for up to 21 days after a single dose (Osório et al., 2015). These data suggest evidence for a potential antidepressant effect for DMT. However, ayahuasca is a complex mixture containing MAOIs (harmala alkaloids) which, as a class of drugs, have also been used alone to treat depression. Thus, it is impossible to say from such studies that DMT itself or the elevation of other brain neurotransmitters in combination is responsible for the perceived positive clinical effects or even if the hallucinations produced by DMT consumed under these conditions are themselves somehow cathartic.

While other classic hallucinogens (LSD, psilocybin, etc.) are beginning to show promise in the treatment of addictions (drugs, alcohol, etc.) as well as post-traumatic stress (PTSD) and other mental disorders (Bogenschutz and Pommy, 2012) there has yet to be generated conclusive evidence regarding the efficacy of DMT in any of them. DMT has been shown to exert anti-anxiety/anti-psychotic properties at the trace amino acid receptor (TAAR) and others have suggested that the possible positive symptoms observed in schizophrenia may be mediated by the effects of endogenous DMT (Cakic et al., 2010; Grammenos and Barker, 2015). These findings do not necessarily support the conclusion that DMT is useful for treatment of anxiety or mental illness, however. The possible use of DMT as an adjunct to psychiatric therapy has been proposed by numerous investigators, a proposal that contravenes the tenets of the transmethylation hypothesis.

Frecska et al. (2013) have suggested that DMT may be involved in significant adaptive mechanisms that can also serve as a promising tool in the development of future medical therapies and there have been proposals that DMT might be useful to treat substance abuse, inflammation, or even cancer. However, at this point, the necessary data to support such proposals have not been presented and it would be premature to propose that DMT will become commonly used for clinical purposes. If it is a neurotransmitter, then understanding its role and function in normal or disease states could provide pharmacological targets to alter these functions, however.

It has been 86 years since DMT's first synthesis by Manske and 61 years since Szara discovered its hallucinogenic properties. It has been 41 years since Christian et al. characterized DMT as a neurotransmitter. Further research has better defined the latter's characteristics such that a compelling case can be made, at a minimum, to consider DMT as a putative neurotransmitter.

Over time, the observations of the hallucinogenic phenomena experienced following the administration of DMT have led to speculation that endogenous DMT is possibly involved in psychosis, normal attributes and experiences such as creativity, imagination and dream states, maintenance of waking reality, altered states of consciousness including religious and/or spiritual phenomena, and NDEs. Even more far reaching and “other worldly” hypotheses have also been offered, suggesting that DMT, as well as other hallucinogens, may provide actual proof of and/or philosophical insights into many of our unanswered questions regarding extraordinary states of consciousness. Regardless of the level and cause of such speculation and hypotheses, it is only scientific research that can inform or refute such thinking. There is no doubt that hallucinogen research has been a forbidden fruit long ripening on the tree of knowledge.

Recent research has demonstrated that DMT is present in and is released from the pineal gland of live, freely-moving rodents. Although older data suggested that DMT might not be synthesized to any great extent in brain, studies have now shown that the necessary enzymatic components for the biosynthesis of DMT are present in discreet brain cell types and areas as well as other tissues not previously examined. New receptors for DMT have been identified and a potential role for DMT as a neuroprotectant and/or neuroregenerative agent has been suggested. Hallucinogens have been shown to produce brain patterning resembling dream states, apparently mediated through 5-HT2A receptor activation. DMT's effect in this regard has yet to be examined, but raises speculation as to one of the possible roles of endogenous DMT.

As discussed and delineated above, more research is needed on DMT's natural role and function and interaction with other neurotransmitter systems. This will require the recommended future research into DMT biosynthesis, metabolism and binding, new methods for peripheral and central detection and data from administration, imaging and therapeutic trial studies. The data derived from the areas of research addressed above will no doubt suggest several possible new avenues for additional future research on DMT. In order to advance, however, regulatory blockades to hallucinogen research must be removed. Progress in hallucinogen research in these areas has been slowed due to over-regulation. For at least the last 50 years, research on DMT and other hallucinogens has been impeded in the United States by passage of the Congressional Amendment of 1965 and the Controlled Substances Act of 1970 by the United States Congress that classified DMT and other major hallucinogens as Schedule-I substances. Given the endogenous nature of DMT, it deserves a special status for future research.

It is evident that we have too long ignored the field of hallucinogen research, in all of its potential aspects. This is especially true if continuing research demonstrates a clear role for one of its more prominent members, DMT, as an endogenous regulator of brain function. It is my opinion that these and many other possible approaches and hypotheses regarding DMT and other psychedelics are research endeavors that have great potential and are worthy of attention and support. Turning the newest technologies to this work, in genetics, analytical chemistry, molecular biology, imaging and others, we will no doubt acquire both new knowledge and ask new questions. If the politics of any one nation forbid it, perhaps others will take up the challenge to further the knowledge of our own potential and the further development and understanding of what we prize as our most unique human characteristic; the untapped possibilities of the mind.

The work presented is the sole effort of SB who agrees to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.