Psychedelics in Psychiatry: Neuroplastic, Immunomodulatory, and Neurotransmitter Mechanisms

1. Ketamine

Subanesthetic doses of ketamine elicit rapid, robust, and sustained antidepressant effects (Berman et al., 2000; Price et al., 2009; Murrough et al., 2013a) at least partially via activating mTOR (Li et al., 2010), a key intracellular signaling pathway altered in MDD and other psychiatric disorders, such as ASD and SCZ (Li et al., 2010; Jernigan et al., 2011; Costa-Mattioli and Monteggia, 2013). In turn, this leads to the BDNF-mediated stimulation of dendritic, synaptic, and neuronal plasticity, which may well translate to network plasticity (Murrough et al., 2013b; Choi et al., 2017; Ly et al., 2018). Ketamine treatment increases BDNF levels in the prefrontal cortex (PFC) and CA3 and dentate gyrus regions of the hippocampus, and the latter region is intimately involved with adult neurogenesis (Dong et al., 2017). Moreover, ketamine treatment during stress exposure prevents the dendritic spine density loss observed in the CA3 region and dentate gyrus of the hippocampus of controls subjects (Dong et al., 2017). Similarly, ketamine increases spine density and synaptic plasticity markers such as synapsin 1, postsynaptic density protein 95, and mGluR1 in rats exposed to stress (Sarkar and Kabbaj, 2016).

Ketamine also interacts with σ1 receptor (S1R, discussed in the immunomodulation section) and S2R, key signal transducers of neuroplastic and neurotrophic pathways that at least partially mediate the therapeutic improvements elicited by currently approved antidepressant medications (Yagasaki et al., 2006; Dhir and Kulkarni, 2007; Fishback et al., 2010; Fujimoto et al., 2012; Krystal et al., 2019). For this reason, S1R and S2R may be involved in ketamine’s antidepressant action. (R)-Ketamine possess higher affinity for S1R than its enantiomer (S)-ketamine, and S1R antagonism does not decrease the antidepressant effects of ketamine (Robson et al., 2012). However, S1R antagonism does decrease the nerve growth factor (NGF)-mediated neuroplastic effects of ketamine. These results demonstrate that the ketamine-initiated, S1R-mediated modulation of NGF might be involved in the antidepressant effects elicited by ketamine (Fujimoto et al., 2012; Yagasaki et al., 2006; Krystal et al., 2019). Lastly, low-dose ketamine increases hippocampal AMPA/NMDA receptor density ratio in a preclinical MDD model (Tizabi et al., 2012). Therefore, decreased AMPA/NMDA receptor density ratio could be involved in the pathogenesis of MDD (Aleksandrova et al., 2017), and ketamine might alter this ratio to ultimately enhance synaptogenesis via upregulating mTOR/BDNF signaling (Tizabi et al., 2012; Aleksandrova et al., 2017).

2. N,N-Dimethyltryptamine, 5-Methoxy-N,N-dimethyltryptamine, and Ayahuasca

Neuroplastic effects have been observed in response to ayahuasca. For example, a single dose of ayahuasca increases circulating BDNF levels in healthy controls and patients with treatment-resistant depression (de Almeida et al., 2019). Preclinical findings support that neuroplasticity-enhancing effects might mediate therapeutic improvements in response to psychedelics. For instance, repeated administration of ayahuasca increases hippocampal BDNF levels in female rats (Colaço et al., 2020). Moreover, a single dose of 5-MeO-DMT increases neurogenesis in the dentate gyrus of the adult mouse (Lima da Cruz et al., 2018). This is of particular relevance given that antidepressant-free patients with depression have decreased BDNF levels (Diniz et al., 2010), and antidepressants increase BDNF (Aydemir et al., 2005), thereby enhancing hippocampal neurogenesis (Anacker et al., 2011). Similarly, other nonpsychedelic S1R agonists suppress neurodegeneration and attenuate disease progression in preclinical models via boosting neuronal proliferation and maturation (Ono et al., 2014), increasing hippocampal BDNF, and activating antioxidant pathways (Kikuchi-Utsumi and Nakaki, 2008; Pal et al., 2012).

S1R modulation might represent a synergistic mechanism to 5-HT_2A receptor modulation by psychedelics, which mediates the upregulation of neurotrophic factors. In fact, S1Rs are enriched in brain areas involved with cognition, stress, and psychiatric disorders, and their stimulation induces adaptive neuroplasticity (Hindmarch and Hashimoto, 2010); psychedelic and nonpsychedelic S1R agonists ameliorate psychiatric symptoms (Hindmarch and Hashimoto, 2010; Davis et al., 2019; Palhano-Fontes et al., 2019). The other active compounds contained in ayahuasca are the β-carbolines harmol, harmine, harmaline, and tetrahydroharmine (found in B. caapi), which inhibit the metabolism of DMT to render it orally active. These compounds also elicit neuroplastic effects. In vitro studies found that these molecules induce differentiation, proliferation, and migration of neural precursors (Morales-Garcia et al., 2017). This strengthens the notion that ayahuasca alkaloids might be useful for neurodegenerative states. Contrasting preclinical evidence suggests that repeated DMT might hinder synaptic plasticity (a sex-specific effect in females) (Cameron et al., 2019). Although preliminary, these results highlight the need for further preclinical studies with greater statistical power to identify potential neuroplasticity-related side effects.

3. Lysergic Acid Diethylamide

It was recently reported that microdoses of LSD increase circulating BDNF levels in healthy volunteers, suggesting that LSD might have neurotrophic effects in humans (Hutten et al., 2020). Another study investigated whether LSD produces acute gene expression changes in peripheral blood mononuclear cells 1.5 and 24 hours after administration, with no detected changes in 5-HT2A receptor or EGR-1, -2, and -3 transcript levels (Dolder et al., 2017a). In rodents, LSD upregulates CCAAT/enhancer-binding protein-β in the prefrontal cortex (Nichols and Sanders-Bush, 2004). This transcription factor affects synaptic scaling, an essential “housekeeping” neuronal process that modulates synaptic function (Turrigiano, 2008). Consistent with neuroplastic and neuroprotective outcomes, LSD increases cFOS, Egr-1, and Egr-2 in murine primary neuronal cultures (González-Maeso et al., 2003, 2007; Li et al., 2005). Moreover, LSD increases dehydroepiandrosterone (DHEA), the most abundant neurosteroid in the central nervous system (Strajhar et al., 2016). Similar to DMT (Fontanilla et al., 2009), DHEA is an S1R agonist, and signaling of DHEA and pharmacological compounds at S1R stimulates synaptic activity and neurogenesis and ameliorates drug-induced cognitive impairments (Meunier and Maurice, 2004; Moriguchi et al., 2013). This indirect action of LSD on S1R might be involved in the clinical improvements elicited by LSD (Moriguchi et al., 2013; Szabo et al., 2014; Schmidt et al., 2016). S1R forms heterodimers with D2 receptors, and such interaction may also boost neurogenesis (Beggiato et al., 2017). Lastly, DOI, a close relative of LSD, rapidly increases spine density growth in cortical cells, activating a synaptogenic pathway mediated by kalirin 7, while increasing Egr-1 in the mouse somatosensory cortex (Jones et al., 2009). Given that the EGR family of transcription factors is involved in synaptic plasticity, neurogenesis, and the pathologic processes underlying psychiatric symptoms, the effects of psychedelics on EGR-mediated transcription should be further elucidated (González-Maeso et al., 2003, 2007; Clark et al., 2010; Duclot and Kabbaj, 2017).

4. Psilocybin

A recent study reported that psilocybin modulates neurotrophic-related gene expression in the PFC and hippocampus, with a preferential effect on the PFC, but inducing changes which are also appreciable in the hippocampus (Jefsen et al., 2020). More specifically, in the PFC psilocybin increased the expression of CEBPB, c-Fos, dual specificity protein phosphatase 1 (Dusp1), transcription factor jun-β, NF-kappa-β inhibitor-α (Iκβα), nuclear receptor subfamily 4 group A member 1, while the serine/threonine-protein kinase 1 (Sgk1), protein fosB, protein S100-A10, and postsynaptic density protein 95 were increased only response to some doses of psilocybin. Dual specificity protein kinase (CLK1) was dose dependently decreased in response to psilocybin (Jefsen et al., 2020). In the hippocampus, Dusp1, Iκβα, and Sgk1 transcripts were similarly increased by acute psilocybin and Clk1 expression was strongly decreased (Jefsen et al., 2020). Psilocybin affects hippocampal neurogenesis in a biphasic fashion (Catlow et al., 2013). Although at lower doses (0.1 mg/kg) a nonstatistically significant trend was observed toward increased neurogenesis, higher doses (1–5 mg/kg) decreased neurogenesis 2 weeks after treatment (Catlow et al., 2013). These neuroplastic changes were associated with a facilitation of fear extinction at low doses, hypothetically mediated by a psilocybin-induced, 5-HT–mediated DA enhancement, an effect known to facilitate fear-extinction learning (Borowski and Kokkinidis, 1998; Catlow et al., 2013).

5. 3,4-Methylenedioxymethamphetamine

The debate on the neurotoxicity of MDMA creates a long-standing divide (Mithoefer et al., 2003; Parrott, 2013, 2014; Doblin et al., 2014; Pantoni and Anagnostaras, 2019; Ricaurte et al., 2002 [retracted]). Most of the available early preclinical research is focused on the neurotoxic effects of MDMA, which may explain cognitive impairments and psychiatric sequelae in MDMA abusers (Parrott, 2001, 2013). To simulate binge abuse and the resulting neurotoxic effects, relatively high doses (often in a chronic-administration design) were administered in these studies, with considerable neurotoxicity (Pantoni and Anagnostaras, 2019). Indeed, human studies indicate damaging effects on SERT homeostasis in heavy MDMA users (Baumann et al., 2007; Müller et al., 2019). However, although these studies represent a valid paradigm for MDMA abuse, they do not seem to adequately model the sparing use of relatively low doses employed in MDMA-augmented psychotherapy (Amoroso, 2019; Pantoni and Anagnostaras, 2019), which elicits notable improvements in treatment-refractory PTSD symptoms (Mithoefer et al., 2011, 2013; Mithoefer et al., 2018; Feduccia et al., 2019; Bahji et al., 2020).

Importantly, metabolites from MDMA metabolism seem to be responsible for the neurotoxic effects of MDMA given that direct intracerebroventricular administration of MDMA does not elicit neurotoxicity (Green et al., 2003). Accordingly, intrastriatal administration of 2,4,5-trihydroxymethamphetamine significantly depletes both 5-HT and DA, intracortical administration decreases 5-HT, and intracerebroventricular administration moderately depletes striatal DA without affecting 5-HT levels (Johnson et al., 1992; Zhao et al., 1992). Other metabolites, such as 2-hydroxy-4,5(methylenedioxy)methamphetamine (6-HO-MDMA), appear to be nontoxic given that intrastriatal and intracerebroventricular administration does not affect 5-HT or DA levels (Zhao et al., 1992). Such metabolites are quinone-thioethers, orto-quinones, and the glutathione conjugates 5-(glutathion-Syl)-α-methyldopamine (5-GSyl-α-MeDA) and 2,5-bis-(glutathion-S-yl)-α-methyldopamine (2,5-bis-(glutathione-S-yl)-α-MeDA) (Miller et al., 1996; Bai et al., 1999; Monks et al., 2001; Green et al., 2003).

The existing lack of clear knowledge on potential neurotoxic effects of MDMA could be filled by employing realistic preclinical models that simulate clinically relevant pharmacodynamics (Pantoni and Anagnostaras, 2019; Vollenweider et al., 1999a). If further studies will determine that neurotoxic effects are elicited by MDMA administration in humans at clinically relevant doses, strategies could be implemented in an effort to protect against these effects (Tourino et al., 2010).

1. N,N-Dimethyltryptamine, 5-Methoxy-N,N-dimethyltryptamine, and Ayahuasca

Strong anti-inflammatory effects mediated by psychedelic-induced S1R activation have been described (Szabo et al., 2014, 2016). These effects closely resemble the anti-inflammatory action of some SSRIs, SNRIs, and tricyclic antidepressants, (Kenis and Maes, 2002; Köhler et al., 2014; Strawbridge et al., 2015) and the antidepressant outcome of some anti-inflammatory therapies (Kappelmann et al., 2018). For example, DMT and its analog 5-MeO-DMT decrease IL1β, IL6, IL8, and TNF-α while increasing IL10 in immune-challenged, human monocyte–derived dendritic cells (Szabo et al., 2014). Similarly, in hypoxic stress–challenged neuronal and microglial cultures, DMT administration enhances cell survival and hypoxic stress–related neuroprotective signaling (Szabo et al., 2016). Accordingly, a single inhalation of 5-MeO-DMT decreases the levels of circulating IL6 (Uthaug et al., 2020). Such fast onset of anti-inflammatory effects is of particular interest and could be harnessed in psychiatry and other areas of emergency medicine—for example, in the treatment of systemic inflammatory response syndrome, also called the “cytokine storm” (Mastronardi et al., 2007; Tisoncik et al., 2012; Szabo et al., 2014, 2016; Mastronardi et al., 2015; Inserra et al., 2019; Mehta et al., 2020; Uthaug et al., 2020). Given that in psychiatric disorders there is an imbalance in immune system equilibria, which are shifted toward proinflammatory responses, therapeutic approaches with the ability to shift this balance in favor of an enhancement of immunomodulatory and anti-inflammatory responses have long been sought in psychiatry (Pulendran, 2004; Myint et al., 2005; Gola et al., 2013; Alcocer-Gómez et al., 2014). Whether psychedelics produce a decrease of proinflammatory cytokines in microglial cells similar to SSRIs and SNRIs, which decrease TNF-α and nitric oxide production and microglial activation, is an intriguing possibility that has been partially demonstrated and that warrants further investigation (Tynan et al., 2012; Szabo et al., 2016).

A study investigating proteomics changes in brain organoids in response to 5-MeO-DMT reported a downregulation of NF-ĸB and nuclear factor of activated T cells via toll-like signaling and Gq-coupled receptors (Dakic et al., 2017). Importantly, those molecules represent critical checkpoints of immune homeostasis interfacing inflammatory responses, mood, and psychiatric disorders (Keller et al., 2017; Zorn et al., 2017). Similarly, ayahuasca and DMT modulate cortisol release (Strassman and Qualls, 1994; Dos Santos et al., 2012; Galvao et al., 2018), and ayahuasca normalizes the blunted awakening cortisol responses observed in patients with treatment-resistant depression (Galvao et al., 2018) while decreasing the circulating levels of C-reactive protein (CRP) in these patients (Galvão-Coelho et al., 2020). Such an HPA axis modulation is quite remarkable given the well characterized dysregulation in psychiatric disorders and the long-sought possibility of modulating the HPA axis to ameliorate psychiatric symptoms (Carvalho and Pariante, 2008; Pariante and Lightman, 2008; Heim et al., 2008; Holsen et al., 2013; Keller et al., 2017).

2. Psilocybin

Psilocybin was shown to acutely increase the circulating levels of the stress-responsive adrenocorticotropic hormone (ACTH) and cortisol, as well as prolactin and thyroid-stimulating hormone (Hasler et al., 2004). The increase in the levels of these hormones was not correlated with stress-induced symptoms, such as anxiety, leading the authors to speculate that the observed temporary increase was likely due to a transient 5-HT_2A receptor–mediated activation of the HPA axis, resulting in an increase in the release of ACTH and circulating cortisol (Van de Kar et al., 2001; Hasler et al., 2004).

3. Lysergic Acid Diethylamide

LSD possesses anti-inflammatory activity in vitro. Specifically, LSD downregulates IL2, IL4, and IL6 and upregulates mitogen-activated protein kinase phosphatase-1 (House et al., 1994). Moreover, LSD administration increases circulating cortisol, cortisol, cortisone, corticosterone, prolactin, oxytocin, and epinephrine in humans (Schmid et al., 2015a; Strajhar et al., 2016). No reports are available so far on potential long-term immunomodulatory outcomes of LSD. A recent study investigated the effects of LSD on retinal cells and suggested that LSD might be toxic to retinal cells via releasing proinflammatory cytokines (Hu et al., 2018). However, retinal cells do not adequately model central and peripheral neural networks, and whether the concentrations of LSD employed in this study are reached in retinal cells in vivo remains to be determined. Therefore, these results should be interpreted with caution. Nevertheless, more studies are required to investigate whether LSD administration might be damaging to retinal cells in vivo.

4. 2,5-Dimethoxy-4-iodoamphetamine

DOI has been widely studied for its selectivity as a 5-HT_2A/2C receptor agonist and for its powerful anti-inflammatory effects, which can be elicited with very low doses of this compound. DOI binding to the 5-HT_2A receptor triggers anti-inflammatory signaling, which is different from 5-HT binding to the same receptor, which elicits proinflammatory signaling (Flanagan and Nichols, 2018). DOI powerfully downregulates inflammation in vitro and in vivo via decreasing TNF-α, IL1β and IL6, intracellular adhesion molecule-1, vascular cell adhesion molecule-1, monocyte chemoattractant protein 1 (MCP1), and fractalkine, ameliorating parameters of immune-induced gut pathology (Yu et al., 2008; Nau et al., 2013; Flanagan et al., 2019). Importantly, DOI blocks NF-ĸB activity and nuclear translocation while decreasing iNOS activity (Miller and Gonzalez, 1998; Yu et al., 2008; Nau et al., 2013). DOI was also shown to be protective in a mouse model of asthma via inhibiting IL3, IL5, IL6, IL13, MCP1, and granulocyte-macrophage colony-stimulating factor (Nau et al., 2015). A recent study suggests that the anti-inflammatory effects of DOI could also be harnessed for the treatment of systemic inflammatory conditions, such as high-fat diet–induced cardiovascular dysfunction (Flanagan et al., 2019b). In this study, the authors observed that DOI treatment during diet-induced cardiovascular inflammation in apolipoprotein-E knockout mice lowers total and low-density lipoprotein cholesterol. Moreover, this treatment decreases IL6, TNF-α, and vascular cell adhesion molecule-1 in the aortic tissue. Furthermore, DOI decreases the circulating levels of the C-X-C motif chemokine 10 (Flanagan et al., 2019b). Together, these findings suggest that DOI and other serotonergic psychedelics might prove valuable for systemic inflammatory diseases, which are often comorbid with psychiatric symptoms.

5. 3,4-Methylenedioxymethamphetamine

MDMA also possesses immunomodulatory properties that influence both the innate and adaptive arms of the immune system (Boyle and Connor, 2010; Mithoefer et al., 2018). Acute administration in humans increases cortisol and prolactin levels (Gouzoulis-Mayfrank et al., 1999) as well as the inflammatory mediators hydroxyeicosatetraenoic acid, dihydroxyeicosatetraenoic acid, and octadecadienoic acid (Boxler et al., 2018) while decreasing circulating CD4⁺ helper T cells and increasing natural killer cells, potentially as a result of increased HPA axis activation (Pacifici et al., 1999, 2000, 2007; Young, 2004). Increases in the production of immunosuppressive Th2 cytokines, such as IL4 and IL10, and a decrease of Th1 cytokines, such as IL2 and IFN-γ, were also reported (Pacifici et al., 2001). Interestingly, repeated exposure to MDMA decreases the number of PV+ GABA neurons in the hippocampal dentate gyrus, and this effect is attenuated by pretreatment with the anti-inflammatory ketoprofen, suggesting that it might be mediated by cyclooxygenase-mediated pathways (Anneken et al., 2013).

Given that PTSD is underlined by a low-grade chronic inflammatory state and a shift of cytokine production from Th2 to Th1 coupled to chronic lymphocyte activation, the immunomodulatory and anti-inflammatory effects of MDMA might relate to PTSD symptom reduction after MDMA-assisted psychotherapy (Wilson et al., 1999; Boyle and Connor, 2010; Yuan et al., 2016; Mithoefer et al., 2018). MDMA also decreases the production of the proinflammatory cytokines IL1β, IL6, TNF-α, and IFN-γ while stimulating the production of the anti-inflammatory IL10 and oxytocin (OT) (Boyle and Connor, 2010; Yuan et al., 2016). Similarly, MDMA inhibits lipopolysaccharide-induced TNF-α secretion (Connor et al., 2000). These lines of evidence suggest that MDMA might decrease the immune reactivity, and therefore the damaging potential, of immune cells in patients with PTSD (which are hyper-reactive in this disorder), and such an effect might be involved in reduction of clinical symptoms (Wilson et al., 1999; Mithoefer et al., 2018; Bahji et al., 2020).

6. Ketamine

Ketamine displays a biphasic action on immune responses. Acutely, it upregulates IL1β, IL6, and TNF-α. Chronically, it decreases TNF-α (Blandino-Rosano et al., 2017). Mice exposed to highly immunogenic stressors, such as the bacterial toxin lipopolysaccharide, were shown to have higher survival rates when administered ketamine and decreased levels of the proinflammatory TNF-α and IFN-γ, which might be responsible for their improved survival rates (Takahashi et al., 2010). Given the direct role of inflammation in psychiatric symptoms, the anti-inflammatory action of ketamine could relate to its beneficial and fast-onset antidepressant properties (Dantzer et al., 2008; Hotamisligil, 2017). In fact, TNF-α is depressogenic, and TNF-α antagonism is antidepressant (Raison et al., 2013). The short-term proinflammatory effects elicited by ketamine might activate necroptotic pathways, which are required for brain regeneration, a process that may (hypothetically) mediate fear extinction and memory reconsolidation (Feduccia and Mithoefer, 2018; Lloyd et al., 2019).

Together, this body of evidence suggests that the anti-inflammatory and immunomodulatory action of psychedelics might be involved in the antidepressant and anxiolytic effects observed in psychiatric populations via pushing the system toward a homeostatic state and, overall, ameliorating psychiatric symptoms. Future studies should explore whether these immunologic outcomes observed in vitro and in preclinical models translate to human physiology. Safety concerns arising from immunosuppression, even though partial, need to be addressed in immunocompetent and immunodeficient models (Boyle and Connor, 2010).

1. Lysergic Acid Diethylamide

LSD is a 5-HT_2A receptor partial agonist (Marek and Aghajanian, 1996; Rickli et al., 2015; De Gregorio et al., 2016a). Similarly, LSD is an agonist at the 5-HT_1A receptor (Norman et al., 1985; Reissig et al., 2005; Rickli et al., 2016). Chronic administration of the widely used antidepressant and anxiolytic SSRI medications desensitizes the postsynaptic 5-HT_1A receptor, increasing serotonergic firing and ultimately eliciting antidepressant effects (Blier et al., 1993; Artigas et al., 1996; Haddjeri et al., 1998). The effects of LSD are strikingly similar to the mechanism of action of SSRIs in terms of 5-HT_1A receptor dynamics. In fact, acutely, LSD decreases the firing and burst activity of 5-HT/DRN neurons via 5-HT_1A receptor activation, and this effect can be blocked by 5-HT_1A and 5-HT_2A receptor antagonism (De Gregorio et al., 2016b). On the contrary, repeated LSD administration increases firing and burst activity in the DRN via a mechanism mediated by 5-HT_1A, D2, and TAAR receptors, resembling classic antidepressants (De Gregorio et al., 2016b). LSD is also an agonist at the 5-HT_2B receptor and a partial agonist at the 5-HT_2C receptor (Fiorella et al., 1995; Egan et al., 1998; Rickli et al., 2016; Wacker et al., 2017). Interactions at the 5-HT_2C receptor are thought to be responsible for the lack of addictive properties of psychedelic compounds (Canal and Murnane, 2017) and to partially mediate their anxiolytic and antidepressant effects (Millan, 2005; Nichols and Nichols, 2008). Agonist action has been described at the 5-HT_1B, 5-HT_1D, 5-HT_1E, 5-HT₆, and 5-HT₇ receptors (Hoyer, 1988; Erlander et al., 1993; Lovenberg et al., 1993; Blaho et al., 1997; Passie et al., 2008; Wacker et al., 2017). Evidence for LSD interaction at SERT remains controversial. In fact, although an in vitro study reported no interaction with SERT (Rickli et al., 2015), in vivo investigations reported decreased LSD effects in Sert null mice (Krall et al., 2008; Kyzar et al., 2016). Given that 1) suicidal patients with depression and PTSD have increased incidence of the low-expressing SERT polymorphisms (Austin et al., 2002; Oquendo et al., 2007), 2) long-term psychedelic users have increased SERT binding (Callaway et al., 1994), 3) intact SERT signaling is necessary for fear extinction (Young et al., 2017), and 4) there is lower incidence of psychological distress and suicidality among psychedelics users (Hendricks et al., 2015; Argento et al., 2017), it cannot be excluded that SERT modulation is a common pharmacological denominator mediating the therapeutic effects of psychedelics in psychiatric disorders. Aside from serotonergic interaction, the pharmacological and behavioral effects of LSD, and likely of other psychedelics, require activation of the 5-HT_2A-mGluR2 complex in glutamatergic pyramidal neurons (discussed below in the Biased Signaling, Biased Phosphoproteomics, and Psychedelic Compounds section) (Moreno et al., 2011). Recently, LSD was reported to increase the social adaptation to opinions similar to one’s own through stimulation of 5-HT_2A receptors, which increase neuronal activity in the mPFC in response to social feedback processing (Duerler et al., 2020).

2. 3,4-Methylenedioxymethamphetamine

Upon ingestion, MDMA is rapidly absorbed in the intestinal tract (Kalant, 2001; Green et al., 2003). Peak plasmatic concentrations are reached after about 2 hours (Kalant, 2001). MDMA is metabolized by several pathways, such as N-demethylation, O-dealkylation, conjugation, and O-sulfation, into 14 metabolites [reviewed in Green et al. (2003))]. Initially, N-demethylation gives rise to MDA, demethylenation gives rise to 3,4-dihydroxymethamphetamine and N-methyl-α-methyldopamine, and ring hydroxylation gives rise to 6-HO-MDMA (Green et al., 2003). Given that both MDMA and its main metabolite, MDA, elicit psychoactive effects, and given that both interact (although to different extents) with the 5-HT system, it is difficult to separate the effects of one from the other. Therefore, the behavioral and neurobiological effects observed should be considered in light of this duality (Green et al., 2003).

MDMA is a 5-HT_2A receptor agonist, and this effect is thought to be responsible for the MDMA-induced mesolimbic DA release (Teitler et al., 1990; Orejarena et al., 2011). Acute MDMA administration induces a transient, MDA-mediated, dose-related increase in extracellular 5-HT in the mPFC, striatum, NAc, and hippocampus (Gough et al., 1991; Gudelsky and Nash, 1996; Kankaanpää et al., 1998; O’Loinsigh et al., 2001; Mechan et al., 2002), and repeated MDMA decreases 5-HT concentration in the striatum (Nash and Yamamoto, 1992). MDMA is a weak 5-HT_1A receptor agonist, and MDMA administration results in a postsynaptic upregulation of this receptor in the cortex and hypothalamus that is noticeable 1 week after acute administration (Aguirre et al., 1998; Battaglia et al., 1988). Indirect 5-HT_1B receptor activation is involved in MDMA-induced hyperlocomotion when the substance is administered at low doses, whereas 5-HT_2A receptor activation is involved at high doses (Rempel et al., 1993; McCreary et al., 1999; Vollenweider et al., 2002). The acute MDMA-elicited 5-HT increase is disrupted by the administration of a 5-HT reuptake inhibitor, suggesting an inhibitory effect of MDMA on SERT (Gudelsky and Nash, 1996; Liechti et al., 2000; Mechan et al., 2002). Indeed, MDMA increases the available concentration of 5-HT via inhibiting SERT, reversing the direction of the membrane transporter and ultimately resulting in the accumulation of 5-HT in the synaptic cleft (Battaglia et al., 1988; Teitler et al., 1990; Koch and Galloway, 1997; Cole and Sumnall, 2003; Green et al., 2003; Verrico et al., 2007). MDMA use has been associated with decreased SERT density, and abstinence time and SERT density correlate positively, suggesting that these alterations are reversible to extents to be determined (Müller et al., 2019). The increase in available 5-HT after MDMA administration is only partially responsible for the enhanced release of DA given that the SSRI fluoxetine does not completely block the increase in 5-HT (Koch and Galloway, 1997). Preclinical studies corroborate the crucial involvement of SERT in the effects of MDMA given that MDMA has no psychostimulant effects in SERT^−/− mice (Bengel et al., 1998). Remarkably, the effects of MDMA on SERT within the NAc are sufficient to elicit the prosocial effects of MDMA (Heifets et al., 2019).

The modulation of SERT by MDMA has implications in psychiatry given that SERT abundance has a dramatic impact on synaptic 5-HT and behavior. For example, genetic SERT variants and environmental exposure eliciting epigenetic modifications (Caspi et al., 2003) downregulate SERT expression, leading to increased psychiatric susceptibility (Karg et al., 2011; Lee et al., 2005). Accordingly, patients with MDD and BD, as well as suicide completers, have decreased SERT expression (Austin et al., 2002; Oquendo et al., 2007). On the other side of the spectrum, high-expressing SERT polymorphisms are associated with obsessive-compulsive disorder (Hu et al., 2006). Therefore, MDMA-induced SERT modulation could be exploited therapeutically—for example, in the treatment of ASD (Danforth et al., 2016, 2018). Future studies should assess acute and long-term MDMA-induced SERT transcriptional dynamics to clarify the effects of trauma, MDMA, and psychotherapy on SERT modulation.

MDMA has an inhibitory effect on the production of 5-HT via decreasing tryptophan hydroxylase (the rate-limiting enzyme of 5-HT biosynthesis) activity within 15 minutes from administration (Johnson et al., 1992). This effect might be mediated by the peripheral generation of metabolites given that it is not appreciable in vitro (Schmidt et al., 1987; Che et al., 1995; Colado et al., 1999). The decrease in tryptophan hydroxylase has been observed for up to 2 weeks after a single acute administration of MDMA (Schmidt and Taylor, 1987). A further effect on the 5-HT system is exerted via the inhibitory effect of MDMA on monoamine oxidase A and monoamine oxidase B, which are also involved in the extracellular monoaminergic increase after MDMA administration (Leonardi and Azmitia, 1994). Repeated MDA administration was shown to cause long-lasting depletion of cortical, hippocampal, and striatal 5-HT (Miller et al., 1997).

3. N,N-Dimethyltryptamine, 5-Methoxy-N,N-dimethyltryptamine, and Ayahuasca

The psychedelic tryptamine DMT, contained in the shrub Psychotria viridis (among other plants), is a 5-HT_2A/2C receptor agonist, although its action at 5-HT_2C receptors appears less relevant than its effects at 5-HT_2A receptors (Smith et al., 1998; Carbonaro et al., 2015). DMT stimulates 5-HT release and inhibits its reuptake via interacting with SERT, in which it shows a high binding-to-uptake ratio, meaning that DMT interacts with different substrate and inhibitor SERT sites and that it might be actively taken up and stored by cells (Cozzi et al., 2009; Blough et al., 2014; Rickli et al., 2016). DMT is also both an inhibitor and substrate of the vesicular monoamine transporter 2 (VMAT2), which is involved in the reuptake of monoamines. Much like the effects at SERT, DMT also shows a high binding-to-uptake ratio, suggesting that VMAT2 might be another carrier involved in the uptake of DMT by cells (Cozzi et al., 2009). DMT also interacts with the 5-HT_1A, 5-HT_1D, 5-HT_1E, 5-HT_2B, 5-HT_5A, 5-HT₆, and 5-HT₇ receptors (Deliganis et al., 1991; Smith et al., 1998; Bunzow et al., 2001; Ray, 2010). The 5-HT_1A receptor antagonism potentiates the hallucinogenic effects of DMT, suggesting that 5-HT_1A receptor blockade can enhance 5-HT_2A receptor–mediated hallucinogenic effects (Strassman, 1996). Like other serotonergic psychedelics, DMT inhibits 5-HT firing in the DRN (Aghajanian et al., 1970).

The β-carbolines harmol, harmine, harmaline, and tetrahydroharmine are obtained from B. caapi (ayahuasca, or “vine of the souls”) and function as MAOIs to block the metabolism of DMT, rendering it orally active (Pähkla et al., 2000; Yritia et al., 2002; Gambelunghe et al., 2008; Carbonaro et al., 2015). These compounds are tricyclic indole alkaloids that resemble tryptamines (Hamill et al., 2019). The presence of an endogenous counterpart in the pineal gland and retina, 6-methoxytetrahydro-β-carbolin (pinoline), has been reported (Langer et al., 1984; Leino, 1984). Harmine and harmaline act as reversible monoamine oxidase A inhibitors, whereas tetrahydroharmine binds 5-HT_2A and 5-HT_2C receptors but not the presynaptic 5-HT_1A receptor as previously hypothesized (Udenfriend et al., 1958; Callaway et al., 1999). This effect results in an increase in available extracellular 5-HT and might be involved in the fast-onset antidepressant effects of ayahuasca in treatment-resistant MDD (Palhano-Fontes et al., 2019; Jiménez-Garrido et al., 2020) and potentially in the long-term neuromorphologic changes observed in long-term ayahuasca users (Bouso et al., 2015).

Although data on the acute monoaminergic effects of DMT and ayahuasca in humans are scarce, preclinical studies suggest potent and region-specific modulation of neurotransmitter release. For example, a study quantifying monoaminergic changes in the rat brain found that ayahuasca increases 5-HT in the hippocampus at high doses and in the amygdala at all doses tested. Hippocampal 5-HT turnover was significantly decreased only at high doses. Amygdalar 5-HT turnover was decreased at all doses tested (de Castro-Neto et al., 2013). Similarly, a study investigating the effects of repeated ayahuasca administration on whole-brain neurotransmitter levels found an increase in whole-brain 5-HT levels in females receiving repeated ayahuasca at the highest does tested without significant changes in 5-HT turnover (Colaço et al., 2020). A study investigating potential anticancer activity of tryptamine and DMT found that these compounds inhibit IDO in vitro, potentiating antitumor activity (Tourino et al., 2013). Although this study was not psychiatry-oriented, the fact that compounds that modulate the kynurenine pathway to decrease IDO activity are useful in psychiatry to decrease IDO emphasizes the relevance of this finding for psychiatry (Savitz, 2020). Consistent with the presence of naturally occurring MAOIs in ayahuasca, neurotransmitter turnover has been reported to be decreased upon ayahuasca administration in rodents (de Castro-Neto et al., 2013). Clinical imaging and biochemical studies should aim at closing the research gap on the monoaminergic effects of DMT and ayahuasca in humans to better elucidate its mechanism of action and maximize its therapeutic applications.

DMT is a selective SERT releaser (Blough et al., 2014; Rickli et al., 2016), whereas 5-MeO-DMT inhibits synaptosomal 5-HT reuptake (Berge et al., 1983). A study observed platelet SERT upregulation in humans after repeated ayahuasca ingestion, potentially as a result of repeated MAOI ingestion and/or repeated exposure to high levels of DMT, which affect SERT pharmacodynamics (Callaway et al., 1994; Blough et al., 2014). Tentatively, SERT polymorphisms could be investigated as predictors of positive/negative outcomes and experience intensity given that low- and high-expressing SERT genotypes have different psychopharmacological profiles (Pezawas et al., 2005; Möller et al., 2019). A close relative of DMT, 5-MeO-DMT, is found in several plants (such as A. peregrine) and in certain species of toads (such as I. alvarius, which also produces bufotenine) (Glennon and Rosecrans, 1982; Shen et al., 2010). 5-MeO-DMT acts as a 5-HT_2A, 5-HT_2C, 5-HT_1A, S1R, and TAAR1 receptor agonist (Krebs-Thomson et al., 2006; Ray, 2010; Szabo et al., 2014). 5-MeO-DMT is a weak 5-HT uptake inhibitor (Blough et al., 2014). The affinity of 5-MeO-DMT for 5-HT_1A receptor is lower compared with that of DMT (McKenna et al., 1990).

4. Ketamine

Acute ketamine does not affect DRN/5-HT firing (El Iskandrani et al., 2015) and the release of 5-HT in the DRN (López-Gil et al., 2019), but it does enhance the stimulated release of 5-HT in the DRN (Tso et al., 2004). Moreover, ketamine enhances the release of 5-HT in the mPFC (López-Gil et al., 2019), strongly activating the 5-HT/PFC system through an AMPA receptor–independent mechanism (Ago et al., 2019). Supporting an enhancement of 5-HT release, 5-hydroxy-indoleacetic acid levels are increased after both acute and repeated ketamine administration (Lindefors et al., 1997). The antidepressant effects of ketamine appear to be elicited via a 5-HT–dependent mechanism given that 5-HT depletion in mice blocks its antidepressant effects (Gigliucci et al., 2013). Similarly to SSRIs, ketamine reduces 5-HT reuptake, although the exact mechanism underlying this effect have not been fully elucidated yet (Martin et al., 1982; Nishimura et al., 1998; Can et al., 2016; Yamamoto et al., 2013). Another known serotonergic interaction occurs at 5-HT_1B receptors (an action mediated by AMPA receptors) (Yamanaka et al., 2014), and this effect has been suggested to be involved in ketamine’s antidepressant effects (du Jardin et al., 2018). Antagonist action has been reported at 5-HT₃ receptors at higher than clinically relevant concentrations (Appadu and Lambert, 1996). The analgesic action of ketamine is thought to be mediated by 5-HT_2B and 5-HT_2C receptor activation, among others (Crisp et al., 1991). Increased 5-HT and NE efflux to the mPFC seems to be the essential step to produce ketamine’s antidepressant effects, and such a process is thought to be mediated by ketamine-mediated glutamate release in the mPFC, which in turn excites DRN neurons both directly, via AMPA receptors on serotonergic terminals, and indirectly, via DRN efferents (Amargós-Bosch et al., 2006; Pham et al., 2017; López-Gil et al., 2019).

5. Psilocybin

Similarly to LSD, psilocybin decreases DRN/5-HT firing (Aghajanian and Hailgler, 1975). Serotonergic and dopaminergic activity are mutually regulated. Indeed, the selective lesion of DA neurons decreases the spontaneous firing activity of DRN/5-HT neurons by 60%, whereas the selective lesion of 5-HT neurons enhances the firing activity of VTA/DA neurons by 36%, indicating an inhibitory effect of the 5-HT input on DA neurons and an excitatory effect of DA on 5-HT neurons (Guiard et al., 2008; Dremencov et al., 2009). The 5-HT–mediated regulation of striatal DA neurotransmission via serotonergic afferents is involved in modulating reward, and serotonergic antidepressants improve mood, cognition, and hedonic behavior partially via 5-HT/DA interactions (Shuto et al., 2020). Employing radioligand binding assays, psilocybin has been demonstrated to display a high affinity at human serotonin receptors 5-HT_1A, 5-HT_2A, and 5-HT_2B (Chadeayne et al., 2020).

5-HT_2A receptors located on DA neurons within the striatum and NAc are thought to be involved in the modulation of psychedelic-induced, 5-HT–mediated dopaminergic effects (Vollenweider et al., 1999b). Psilocybin is thought to exert its psychedelic effects mostly through 5-HT_2A receptor agonism, occupying up to 72% of all human brain 5-HT_2A receptors at clinically relevant doses (Vollenweider et al., 1998; Winter et al., 2007; Quednow et al., 2012; Kometer et al., 2013; Madsen et al., 2019). Psilocybin also interacts with the 5-HT_1A receptor in vivo without producing behavioral effects and 5-HT_1D, 5-HT_1E, 5-HT_2B, 5-HT_2C, 5-HT₅, 5-HT₆, and 5-HT₇ receptors in vitro, although it is not known whether these interactions produce a clinically relevant effect (McKenna et al., 1990; Passie et al., 2002; Winter et al., 2007). Some authors have suggested that the 5-HT_1A receptor mediates the deficits in attentional performance induced by psilocybin (Carter et al., 2005). Although the modulatory action of psilocybin on mood states and emotional face recognition are mediated by the 5-HT_2A receptor, the psilocybin-induced bias toward the processing of positive emotions is mediated by other receptors, possibly the 5-HT_2B, 5-HT_2C, or 5-HT_1A receptor (Kometer et al., 2012). Also, psilocybin decreases 5-HT reuptake by inhibiting SERT (Rickli et al., 2016).

1. Ketamine

The direct involvement of DA receptors in the pharmacology of ketamine is contradictory. In fact, although earlier studies reported high affinity and partial agonistic action at D2 receptors (Breier et al., 1998; Vollenweider et al., 2000; Kapur and Seeman, 2002; Seeman and Kapur, 2003) and ketamine-induced blockade of striatal DAT (Tsukada et al., 2001), later reports failed to replicate these findings, potentially because of the high doses necessary to initiate DA interactions (Aalto et al., 2002; Kegeles et al., 2002; Can et al., 2016). Nevertheless, ketamine is thought to restore DA activity and synaptic plasticity, at least partially, via D1 activation in the NAc (Belujon and Grace, 2014). Given that patients with MDD present NAc dysfunction with impaired dopamine-mediated reward circuits, a protective effect of ketamine on the DA system might relate to its therapeutic effects in patients with treatment-resistant depression (Feder et al., 2014; Kim et al., 2019; Krystal et al., 2019; Pizzagalli et al., 2009; Phillips et al., 2019). Acute ketamine induces a 5-fold increase in DA release in the mPFC and enhances striatal DA release, whereas repeated (7 days) ketamine increases basal dopamine to almost double that of preketamine values and attenuates ketamine-induced DA release in the mPFC (Lindefors et al., 1997; Breier et al., 1998; El Iskandrani et al., 2015; Can et al., 2016). Moreover, ketamine antagonizes DA reuptake in the striatum (Tso et al., 2004). Although ketamine does not affect firing activity in the VTA (El Iskandrani et al., 2015), it restores the stress-induced dopaminergic dysfunction in the VTA (Rincón-Cortés and Grace, 2017). Nevertheless, ketamine restores the amphetamine-induced decrease of VTA activity (Belujon et al., 2016).

2. Lysergic Acid Diethylamide

LSD is a DA D1 (Schindler et al., 2012), D2 (Seeman et al., 2005), and D4 (Marona-Lewicka et al., 2009) receptor agonist. LSD modulates DA neurotransmission in a biphasic, dose-dependent fashion. Although at low doses no effects are observed in VTA/DA activity, the latter is decreased at higher doses via a multireceptorial mechanism involving D2, 5-HT_1A, and TAAR1 receptors (Marona-Lewicka et al., 2005; De Gregorio et al., 2016b). Confirming a dopaminergic involvement in LSD pharmacology, in vitro studies confirmed that LSD binds the human and murine D1 and D2 receptors (Pieri et al., 1974; Watts et al., 1995; Marona-Lewicka et al., 2009; Rickli et al., 2015). Moreover, the D4 receptor is involved in the discriminative stimulus effects in rats of LSD (Marona-Lewicka et al., 2009).

3. 3,4-Methylenedioxymethamphetamine

MDMA stimulates dopaminergic neurotransmission via 1) reversing the direction of the DA transporter; 2) a 5-HT_2A receptor–induced DA release, which results in an overall increase in synaptic DA availability; and 3) an inhibition of DA nigrostriatal and mesolimbocortical reuptake in a time-, dose-, and region-dependent manner (Yamamoto and Spanos, 1988; Gudelsky et al., 1994; Bankson and Cunningham, 2001; Mayerhofer et al., 2001; Verrico et al., 2007). Acute MDMA induces DA release, and repeated administration decreases DA concentration in the striatum (Nash and Yamamoto, 1992). At high doses (i.e., 10 mg/kg), the effects of MDMA on DA release are more marked in the caudate nucleus (∼400%) compared with the NAc (∼200%), whereas at lower doses (i.e., 2.5 and 5 mg/kg), this increase is almost equivalent in the two regions and is not so remarkable (∼20%) (Yamamoto and Spanos, 1988). Interaction of MDMA with the DA transporter is relatively low, and it is thought to be partially responsible for the increase in extracellular DA after MDMA administration (Hagino et al., 2011). Decreased DAT binding in the striatum has been reported in rodents 1 week after acute MDMA (Biezonski et al., 2013). In humans, a decrease in striatal DAT was observed only in recreational MDMA users who also used amphetamines (Reneman et al., 2002a). Lastly, MDMA possesses weak D1 and D2 receptor affinity (Battaglia et al., 1988), and these effects are involved in MDMA-induced hyperlocomotion (Risbrough et al., 2006), while D2 receptors are involved in the dopaminergic toxicity induced by high doses of MDMA (Granado et al., 2011).

The MDMA-induced increase in available 5-HT is partially responsible for the increased release of DA given that fluoxetine (SSRI) only partially blocks the MDMA-induced increase in extracellular DA (Koch and Galloway, 1997). These dopaminergic effects elicited by MDMA are the basis for concern around potential abuse liability and addiction. However, at clinically relevant doses (i.e., 1.5 mg/kg), the increase in DA release is minimal, suggesting that abuse concerns are not so relevant at clinical doses (Yamamoto and Spanos, 1988; Pantoni and Anagnostaras, 2019). As discussed earlier, the hepatic metabolism of MDMA to MDA and other catechols with substantial DA activity is likely involved in the potential neurotoxic effects of MDMA (Green et al., 2003).

4. N,N-Dimethyltryptamine, 5-Methoxy-N,N-dimethyltryptamine, and Ayahuasca

Acute Ayahuasca administration increases amygdalar DA levels and decreases its turnover (de Castro-Neto et al., 2013). A recent study reported that repeated ayahuasca administration increases the concentration of the DA metabolites 3,4-dihydroxyphenylacetic acid in whole-brain homogenates in rodents (Colaço et al., 2020). Earlier studies reported that DMT increases striatal (Smith, 1977) and whole-brain DA synthesis (Waldmeier and Maitre, 1977), although it lacks direct dopaminergic activity (Hungen et al., 1975). DMT was also reported to increase the levels of the extraneuronal DA metabolite 3-methoxytyramine (Waldmeier et al., 1976) acutely and after 1 month of treatment. This suggests that repeated DMT might accelerate DA turnover, potentially as a mechanism to compensate the increased DA synthesis (Waldmeier et al., 1976; Smith, 1977; Colaço et al., 2020). One study reported decreased levels of forebrain DA in response to DMT (Haubrich and Wang, 1977). Given that S1R agonists modulate DA neurotransmission, important effects of ayahuasca, DMT, and 5-MeO-DMT over the dopaminergic system are plausible, which might be involved in the amelioration of suicidality and addiction behaviors by modulating reward-related circuitries. For example, S1R ligands can interact with the DA transporter, preventing DA efflux and nullifying drug-induced reward (Sambo et al., 2017, 2018). Therefore, future studies should investigate whether these compounds could address reward circuitry dysfunction in addiction (Sambo et al., 2017).

Together, a dose-dependent involvement of the DA system is recognized in the action of psychedelics and entactogens. Given the role of the DA system in mediating reward responses and developing addiction, future studies are required to clarify whether 1) acute and repeated psychedelic administration leads to changes in homeostasis of reward-related pathways 2) and, if so, at which doses; 3) whether this represents a concrete danger for abuse; and 4) whether adjunctive therapies can pharmacologically counteract these effects without blocking the desirable effects.

1. Lysergic Acid Diethylamide

LSD elevates prefrontal-limbic glutamate via stimulating postsynaptic 5-HT_2A receptors on pyramidal cells in deep cortical layers (Aghajanian and Marek, 1997, 1999; Scruggs et al., 2003; Muschamp et al., 2004), and this effect is reversed by 5-HT_2A receptor antagonists (Vollenweider et al., 1998), AMPA receptor antagonists (Benneyworth et al., 2007), selective antagonists of the NR2B subunit of NMDA receptors (Zhang and Marek, 2008), and positive mGluR2 allosteric modulators (Lambe and Aghajanian, 2006). Electrophysiological studies show that antagonism of the NMDA receptor subunit NR2B suppresses the glutamate release induced by LSD, suggesting an essential role for glutamate in the downstream effects of LSD (Lambe and Aghajanian, 2006). Accordingly, neurodegeneration, aging, and inflammation associate with layer 5 synaptic and dendritic dysfunction (de Brabander et al., 1998; Fogarty et al., 2015). LSD also needs a functional mGluR2 or mGluR3 given that mGluR2/3 antagonists block LSD-induced gene transcriptional changes (Moreno et al., 2013). Therefore, it seems plausible that psychedelics might have protective effects on the neurodegenerative aspect of MDD and other psychiatric disorders via their glutamatergic-enhancing effects in corticolimbic areas.

2. Psilocybin

Multimodal neuroimaging studies after psilocybin administration reported a region-specific hypermetabolic state, especially in the frontolateral and frontomedial cortices. Given that these areas signal primarily via glutamatergic projections and are rich in 5-HT_2A-mGluR2 heteroreceptors, it is possible that glutamate might mediate these effects (González-Maeso et al., 2008; Carhart-Harris et al., 2016c). Interestingly, in a preclinical study, although administration of psilocybin or ketamine alone did not ameliorate depressive-like behavior, the concomitant administration of these compounds elicited antidepressant-like effects comparable to those of fluoxetine (Martin-Ruiz et al., 2001). This finding suggests that subthreshold doses of multiple psychedelic compounds might have synergistic antidepressant effects, potentially via activating pathways that converge on the modulation of corticolimbic serotonergic and glutamatergic systems (Martin-Ruiz et al., 2001).

3. Ketamine

The main effects of ketamine and its metabolite norketamine were first attributed to NMDA receptor antagonism (Lodge et al., 1982; Anis et al., 1983; MacDonald et al., 1987; Franks and Lieb, 1994; Ebert et al., 1997). Subsequent work found that mTOR activation is a step required for ketamine to elicit antidepressant effects and to enhance synaptogenesis in the PFC (Li et al., 2010). It seems likely that these effects are not mutually exclusive and that the enhancement of AMPA receptors over NMDA receptor throughput in cortical circuits, leading to increased plasticity and neurogenesis, might be responsible for the onset of antidepressant effects (Maeng et al., 2008). The antidepressant effects of ketamine are elicited via NMDA receptor antagonism, which results in an increase of AMPA receptor activation compared with NMDA receptor activation (Maeng et al., 2008; Sanacora et al., 2008; Zarate and Manji, 2008; Autry et al., 2011). Ketamine rapidly increases extracellular cortical and striatal glutamate (Stone et al., 2012; López-Gil et al., 2019), enhancing cortical excitability (Cornwell et al., 2012), and this action correlates with improvement of psychiatric symptoms (Stone et al., 2012; Abdallah et al., 2015; Lisek et al., 2017). Moreover, ketamine increases glutamate release in the DRN (López-Gil et al., 2019). The metabolism of ketamine to hydroxynorketamine may be another key necessary step to elicit antidepressant effects, which are mediated by early and sustained activation of AMPA receptor, independently from NMDA receptor (Maeng et al., 2008). The therapeutic effects of ketamine may also be driven, at least partially, by a downregulation of excitatory amino acid transporter (EAAT) 2, which delays glutamate reuptake (Lisek et al., 2017). Ketamine administration 1) potentiates firing but decreases burst activity in glutamatergic neurons in the PFC (Jackson et al., 2004), 2) decreases firing activity in PFC GABAergic interneurons (ultimately diminishing the inhibitory influence on the cortex) (Homayoun and Moghaddam, 2007), 3) enhances striatal dopamine release (Breier et al., 1998), and 4) modulates reticular thalamus (RT) activity (Troyano-Rodriguez et al., 2014). These effects resemble the schizophrenic brain and can be partly restored by antipsychotic treatment (Kargieman et al., 2007). AMPA receptor blockade prevents the behavioral effects of ketamine and the increase in firing and bursting in the locus coeruleus and VTA activity elicited by ketamine (Moghaddam et al., 1997; El Iskandrani et al., 2015). Moreover, AMPA receptor potentiation of CA3 pyramidal neurons was also observed after ketamine (El Iskandrani et al., 2015). These findings suggest that an enhancement of catecholaminergic neurotransmission via AMPA receptor–mediated neurotransmission is required for ketamine to elicit antidepressant effects (Maeng et al., 2008; Sanacora et al., 2008; El Iskandrani et al., 2015). Importantly, ketamine blocks NMDA receptor–mediated burst firing in the lateral habenula, the “antireward” center, disinhibiting monoaminergic reward centers and rapidly relieving depression (Yang et al., 2018). A novel mechanism has recently been suggested that might at least partially explain the psychotomimetic and/or the antidepressant effects of ketamine: although physiologically α₂-adrenoceptors and GABA_B receptors selectively inhibit AMPA and NMDA receptors, respectively, ketamine reduces the levels and activity of the neuromodulatory regulator of G protein signaling 4, leading to selectivity loss and broad glutamatergic inhibition (Lur et al., 2019).

4. 3,4-Methylenedioxymethamphetamine

Acute MDMA increases glutamate release in the anteromedial striatum and dorsal hippocampus (Nash and Yamamoto, 1992; Anneken et al., 2013), whereas repeated MDMA does not affect glutamate efflux in the striatum (Nash and Yamamoto, 1992). This effect might be mediated by the stimulation of 5-HT_2A/2C receptors on non-neuronal cells given that pretreatment with the 5-HT_2A/2C receptor antagonist ketanserin prevents the increase in hippocampal glutamate release (Anneken and Gudelsky, 2012). Microiontophoretic investigations found that MDMA inhibits neuronal firing in the NAc similarly to 5-HT and DA, and this effect might be due to the MDMA-induced increase in 5-HT and DA release (White et al., 1994, 1995).

Repeated administration of MDMA during adolescence leads to important neuroadaptive changes in glutamatergic-related gene expression in corticolimbic structures (Kindlundh-Hogberg et al., 2008). For example, GluR2, mGluR1, mGluR5, NR1, NR2A, NR2B, EAAT1, and EAAT2 were increased in the adult cortex, whereas GluR3, NR2A, and NR2B receptor subunits were increased in the caudate putamen after repeated adolescent MDMA (Kindlundh-Hogberg et al., 2008). Moreover, transcription levels of GluR1 were reduced in the hippocampus, whereas GluR1, GluR3, mGluR1, and mGluR3 were increased in the hypothalamus (Kindlundh-Hogberg et al., 2008). These effects might be involved in the neurotoxic effects of MDMA binge use and repeated use in individuals of young age and potentially in the MDMA-induced, 5-HT–mediated loss of parvalbumin interneurons in the hippocampal dentate gyrus (Collins et al., 2015). NMDA receptors take part in the acquisition of conditioned rewarding effects of MDMA (García-Pardo et al., 2015), whereas NMDA and AMPA receptor antagonists block these, as well as striatal DA efflux and GABAergic toxicity in the hippocampus, suggesting potential in the treatment of MDMA and methamphetamine abuse (Finnegan and Taraska, 1996; Johnson and Kotermanski, 2006; Huff et al., 2016; García-Pardo et al., 2018, 2019).

5. N,N-Dimethyltryptamine, Ayahuasca, 5-Methoxy-N,N-dimethyltryptamine

Acute ayahuasca administration (up to 800 mg/kg) did not affect glutamate levels in the hippocampus or amygdala in rats (de Castro-Neto et al., 2013). mGluR5 is downregulated by 5-MeO-DMT in brain organoids, and this receptor is involved in addiction; therefore, the lower incidence of alcohol and substance use disorder among people who ingest psychedelics multiple times during their lifetime could be mediated by the influence of this receptor (Bird et al., 2008; Stoker et al., 2012; Dakic et al., 2017). Future studies should investigate the glutamatergic effects elicited by repeated administration of ayahuasca, DMT, and 5-MeO-DMT, which might mediate the activation of antidepressant and antiaddiction corticolimbic circuits (Bouso et al., 2013; Riga et al., 2014; Osório et al., 2015; Sanches et al., 2016; Morales-Garcia et al., 2017; Inserra, 2018; da Silva et al., 2019; de Almeida et al., 2019; Palhano-Fontes et al., 2019; Zeifman et al., 2019; Murphy-Beiner and Soar, 2020).

1. Ketamine

The main body of work investigating the effects of psychedelics on the GABAergic system probably comes from studying the effects of ketamine on GABAergic corticolimbic interneurons (Behrens et al., 2007; Quirk et al., 2009; Belforte et al., 2010). The effects of ketamine in modulating GABAergic populations within the CSTC circuit have been described (Behrens et al., 2007; Höflich et al., 2015; Rivolta et al., 2015; Jeevakumar and Kroener, 2016), including a disruption of RT activity (Liu et al., 2016; Mahdavi et al., 2020). The antidepressant effects of ketamine require the modulation of GABA_A and GABA_B receptors in discrete networks of cortical PV+ GABAergic interneurons (Quirk et al., 2009; Rosa et al., 2016; Yang et al., 2016; Wang et al., 2017a; Gerhard et al., 2020) and a downregulation of GABA reuptake (Mantz et al., 1995; Flood and Krasowski, 2000). Further strengthening the case for a modulatory effect of ketamine on the GABAergic system, ketamine administration in young animals leads to alterations in adult GABA circuits (Jeevakumar and Kroener, 2016). Both corticothalamic (Dawson et al., 2013; Anderson et al., 2017) and thalamocortical (Höflich et al., 2015) neurotransmission is modulated by ketamine via NMDA receptor blockade on GABAergic neurons in the RT, which leads to a disinhibition of DA neurons and increased DA release (Liu et al., 2016). Such effects on the RT are of particular interest given that NMDA receptor antagonism in the RT disrupts T-type calcium channel–mediated bursting in GABAergic neurons, leading to decreases in burst activity (Zhang et al., 2009) that are reminiscent of SCZ (Ferrarelli and Tononi, 2011; Hunt et al., 2017). Given that the RT (Min, 2010) and, more broadly, the CSTC circuit (Herrera et al., 2016) are 1) implicated in the generation of consciousness (Ward, 2011), 2) dysfunctional in psychiatric conditions (Swerdlow and Koob, 1987; Brown et al., 2017a; Posner et al., 2017), and 3) targeted by psychedelics (Scruggs et al., 2000; Preller et al., 2019), it is plausible that a modulatory action of this circuit is at the therapeutic core of psychedelics in psychiatric disorders (discussed in the Going Beyond Receptors: Neuronal Circuits Activated by Psychedelic Drugs section).

2. Lysergic Acid Diethylamide

LSD and DOI are partial agonists at 5-HT_2A receptors in cortical interneurons, and this effect is involved in the inhibition of specific cortical pyramidal networks (Marek and Aghajanian, 1996). DOI was shown to activate cortical GABAergic interneurons and to dose-dependently increase cortical GABA levels (Abi-Saab et al., 1999; Wischhof and Koch, 2012). DOI directly depolarizes interneurons (Marek and Aghajanian, 1994), and this effect might be mediated by direct presynaptic 5-HTR binding by DOI or by psychedelic-induced, 5-HT–mediated presynaptic stimulation (Willins et al., 1997; Jakab and Goldman-Rakic, 1998). Moreover, DOI upregulates PV expression in cFOS+ GABAergic interneurons in the mPFC and somatosensory cortex (Martin and Nichols, 2016).

3. 3,4-Methylenedioxymethamphetamine

MDMA-induced DA release is mediated at least partially by GABAergic interactions. For example, MDMA enhances GABA efflux in the VTA, which dampens the MDMA-mediated DA release in the NAc shell (Bankson and Yamamoto, 2004) and decreases GABA efflux in the substantia nigra (Yamamoto et al., 1995). MDMA also modulates GABA-related gene expression in a specific spatiotemporal fashion. For example, MDMA increased GABA transporter (GAT) 1 and GAT4, but not GAT2 transcription, in the PFC and midbrain (Peng and Simantov, 2003). These results suggest a pivotal effect of MDMA on GABAergic gene expression (Simantov and Peng, 2004). Accordingly, a binge-like MDMA administration decreased PV+ GABAergic interneurons in the dorsal hippocampus (Anneken et al., 2013). Interestingly, repeated exposure to MDMA decreases the number of PV+ GABA neurons in the dentate gyrus, and this effect is attenuated by pretreatment with the anti-inflammatory ketoprofen, suggesting that the cyclooxygenase-mediated pathway might be involved (Anneken et al., 2013).

4. Psilocybin

5-HT_2A receptors located on GABAergic interneurons within the striatum and NAc might be involved in the modulation of psychedelic-induced, 5-HT–mediated dopaminergic effects (Vollenweider et al., 1999b). Given that psilocybin, and psychedelics in general, decrease neuronal activity in discrete brain areas, such as the default mode network (DMN) (Carhart-Harris et al., 2012, 2016c; Palhano-Fontes et al., 2015; Tagliazucchi et al., 2016), these effects could be mediated by the serotonergic action of psilocybin or psilocybin-induced 5-HT release on GABAergic interneurons in these areas, enhancing local inhibition, as is the case for DOI (Cumming-Hood et al., 1993; Marek and Aghajanian, 1996) and LSD (Marek and Aghajanian, 1996).

5. N,N-Dimethyltryptamine, 5-Methoxy-N,N-dimethyltryptamine, Ayahuasca

Ayahuasca increases GABA levels in the hippocampus and in the amygdala only at high doses in preclinical models (de Castro-Neto et al., 2013). This mechanism could be involved in the retrieval of repressed traumatic memories experienced by some individuals who ingest ayahuasca given that the GABAergic network synchronization in the hippocampus can create the prerequisites for the synaptic changes (Paulsen and Moser, 1998; Inserra, 2018) required for memory reconsolidation (Tronson and Taylor, 2007). Such effects could be also exploited for the treatment of substance abuse (Liester and Prickett, 2012; Thomas et al., 2013) given the high comorbidity rates with PTSD (Brown and Wolfe, 1994).

1. Ketamine

The dissociative anesthetics ketamine and PCP have been shown to increase NE release and to inhibit its reuptake. Ketamine increases NE efflux in the mPFC in vivo (Kubota et al., 1999; López-Gil et al., 2019 and NE efflux in the ventral bed nucleus of the stria terminalis in vitro (Tso et al., 2004). No changes in the DRN efflux of NE have been reported (López-Gil et al., 2019). Early studies suggested that ketamine potentiates the contraction-inducing effects of NE on vascular adrenergic neurons (Nedergaard, 1973). α-Adrenergic, but not β-adrenergic, blockade attenuated most of the cardiorespiratory responses to ketamine, suggesting an involvement of α-adrenergic neurotransmission in the effects of ketamine (Traber et al., 1970, 1971). More recent studies reported that ketamine interacts both with peripheral α₁- and β₂-adrenoceptor binding sites, eliciting vasoconstriction and vasodilation, respectively (Bevan et al., 1997). Supporting an involvement of α₂-adrenoceptors in the mechanism of action of ketamine, α₂-adrenoceptor antagonism attenuates the cardiostimulatory and psychotomimetic effects of ketamine anesthesia (Doak and Duke, 1993; Tanaka and Nishikawa, 1994; Levanen et al., 1995). Further, ketamine attenuates the cardiodepressant effects of the α₂-adrenoceptor agonist dexmedetomidine (Char et al., 2013).

Ketamine and PCP were shown to inhibit NE reuptake in NE neurons of the rat cortex (Taube et al., 1975; Mandela and Ordway, 2006). However, although the inhibitory activity of PCP was high (to extents similar to those of cocaine), the inhibitory activity of ketamine was much lower (Taube et al., 1975). Similarly, in the bed nucleus of the stria terminalis, ketamine decreased the rate of NE reuptake (Tso et al., 2004). Other studies reported that ketamine inhibits NE reuptake in the heart tissue (Miletich et al., 1973; Aronson and Hanno, 1978; Salt et al., 1979) and that the antagonistic effects of ketamine at NET are responsible for the arrhythmias generated by the drug at high doses (Koehntop et al., 1977). However, different timing of ketamine exposure appears to have different effects over NE reuptake. For example, acute (30 minutes) exposure of bovine adrenal medullary cells to ketamine decreases NE uptake (Hara et al., 1998), whereas exposure for longer than 3 hours increases it (Hara et al., 2000). Importantly, 24-hour incubation with ketamine increases NET mRNA (Hara et al., 2000, 2002). The exact mechanisms through which ketamine affects NET remain unknown. It has been hypothesized that the inhibitory effects of ketamine on NE reuptake might be mediated by a direct action at the desipramine binding site of NET (Hara et al., 1998) or by an indirect effect of NMDA receptors over NET (Mandela and Ordway, 2006).

2. 3,4-Methylenedioxymethamphetamine

MDMA is an α2-adrenergic receptor agonist (Battaglia et al., 1988; Bexis and Docherty, 2005, 2009; Hysek et al., 2012a, 2013). Lower affinity has been described for β-adrenoceptors (Battaglia et al., 1988). Agonism at β3-adrenergic receptor, together with agonism at α1-adrenergic receptor, is thought to be involved in hyperthermia-induced rhabdomyolysis given that concomitant antagonism at β3- and α1-adrenergic receptor attenuates hyperthermia-induced rhabdomyolysis (Sprague et al., 2004). The empathogen MDMA increases NE release (Rothman et al., 2001; Verrico et al., 2007; Mithoefer et al., 2018). Moreover, MDMA interacts with NET, inhibiting NE reuptake (Rothman et al., 2001; Verrico et al., 2007). Accordingly, the NET inhibitors reboxetine and desipramine, respectively, decrease the biological (such as plasmatic NE release and cardiostimulant action) and subjective (such as subjective drug high, psychostimulation, and emotional excitation) effects of MDMA in humans (Hysek et al., 2011) and the acute MDMA-induced goal-oriented task impairments observed in other primates (Verrico et al., 2008). Further suggesting an adrenergic involvement in the biological and cardiostimulatory effects of MDMA, the β-adrenergic antagonist propranolol, but not the α-adrenergic antagonist prazosin, suppresses the MDMA-induced glycogenolysis and consequent glucose release in the striatum (Pachmerhiwala et al., 2010), whereas the nonselective β-blocker pindolol and the α1- and β-adrenergic antagonist carvedilol attenuate the cardiostimulatory effects of MDMA (Hysek et al., 2010, 2012a,b).

A study investigating a potential causal role of genetic variants at the NET gene, which could explain biological and subjective variability in the effects of MDMA, found a weak correlation between NET polymorphisms and the cardiostimulatory effects of MDMA (Vizeli et al., 2018). Specifically, 1) individuals carrying the GG genotype of the NET polymorphism rs1861647 presented greater cardiostimulant effects compared with subjects with one or no G alleles, 2) subjects carrying a C allele of the NET polymorphism rs2242446 presented greater MDMA-induced heart rate elevations compared with the TT genotype, and 3) G allele carriers of the polymorphism rs36029 displayed decreased elevations of arterial pressure compared with the AA genotype (Vizeli et al., 2018). Lastly, NE is involved in the MDMA-induced prolongation of the latency and light-induced miosis (pupil constriction) (Hysek and Liechti, 2012).

3. Lysergic Acid Diethylamide

LSD is an α2-adrenoceptor agonist (Marona-Lewicka and Nichols, 1995). Microiontophoretically applied LSD was reported to slightly increase the firing rate of some, but not all, LC/NE neurons, whereas intravenous LSD induces a delayed partial suppression of some others (Svensson et al., 1975; Rogawski and Aghajanian, 1979). Interestingly, aside from its direct effects on firing rate, LSD increases the reactivity of LC/NE neurons to stimuli that normally do not activate these neurons during anesthesia, such as fur stroking or air puffs, and it potentiates the response to stimuli that normally excite LC/NE neurons, such as skin pinching (Aghajanian, 1980). Further, LSD was reported to decrease the NE-induced cAMP increase in vivo in the rat hypothalamus and brainstem and in vitro in the hippocampus and to attenuate the vascular and cardiac responses mediated by α- and β-adrenoceptor activation, suggesting that LSD possesses adrenergic activity via α- and β-adrenoceptor interactions (Palmer and Burks, 1971; Hungen et al., 1975; Tang and Seeman, 1980). Several earlier preclinical studies reported no effects of LSD administration over brain NE levels or turnover (Katz and Kopin, 1969; Leonard and Tonge, 1969; Peters, 1974; Stolk et al., 1974; McCall and Aghajanian, 1980), whereas others reported a slight (∼20%) decrease in nonstressed and stressed animals (Barchas and Freedman, 1963; Diaz et al., 1968; McGrath and Olverman, 1978). In humans, LSD was reported not to alter the urinary excretion of NE (Hollister and Moore, 1967). Acutely, LSD was shown to facilitate neuronal excitation in the rat facial motor nucleus, potentiating the response to microiontophoretically applied NE (McCall and Aghajanian, 1980). Moreover, NE-depleted rats were reported to have an attenuated behavioral phenotype in response to LSD, suggesting that NE projections from the LC may indeed be relevant for the behavioral effects of LSD (Geyer et al., 1985). One study investigating the effects of repeated LSD administration (14 days) reported a decrease of brainstem NE levels of 20% compared with controls (Peters, 1974). LSD was reported not to inhibit NE reuptake (Dengler et al., 1961).

4. Psilocybin

Psilocybin is an α2-adrenoceptor agonist (Marona-Lewicka and Nichols, 1995). Psilocybin decreases available NE levels by up to 25% in the rat brain in the 4 hours after administration (Stolk et al., 1974). Interestingly, NE synthesis in the brain was transiently increased after psilocybin administration, whereas normetanephrine, a product of NE turnover, was increased by over 2-fold 1 hour after psilocybin administration, suggesting that aside from affecting the levels of extrasynaptic NE, psilocybin might affect NE turnover (Stolk et al., 1974). Earlier in vitro studies reported that psilocybin increases NE reuptake (Herblin and O’Brien, 1968), whereas recent studies suggest that psilocybin binds NET, likely inhibiting NE reuptake (Rickli et al., 2016).

5. N,N-Dimethyltryptamine, 5-Methoxy-N,N-dimethyltryptamine, Ayahuasca

Similar to LSD, mescaline, and psilocin, DMT was reported to potentiate the effects of 5-HT and NE in the facial nucleus in rats (McCall and Aghajanian, 1980). One early report found that DMT slightly increases NE synthesis in the rat forebrain (Waldmeier and Maitre, 1977), whereas another found no effect (Smith, 1977). Neither report found evidence of effects on NE turnover (Smith, 1977; Waldmeier and Maitre, 1977). Ayahuasca administration increases NE levels in the amygdala but not the hippocampus of rats (de Castro-Neto et al., 2013). Moreover, the level of 4-hydroxy-3-methoxy mandelic acid, a product of NE metabolism, was decreased in the amygdala at all doses tested, and in the hippocampus, it was decreased only at the highest doses tested (de Castro-Neto et al., 2013). On the contrary, another study found no changes in the levels of the NE metabolite 3-methoxy-4-hydrohyphenylglycol after repeated ayahuasca administration (Colaço et al., 2020). In humans, ayahuasca administration was reported to increase the urinary excretion of normetanephrine (Riba et al., 2003). Given that harmaline binds adrenoceptors, it is possible that the effects observed in response to ayahuasca administration over NEergic neurotransmission might arise from such interaction (Miralles et al., 2005). Depletion of NE was shown to block and reverse the analgesic effects of 5-MeO-DMT, suggesting that 5-MeO-DMT might elicit analgesic effects via modulating the NEergic system (Archer et al., 1985; Alhaider et al., 1993). Lastly, 5-MeO-DMT has been suggested to suppress NE reuptake and to increase its turnover (Fuxe et al., 1972; Nagai et al., 2007).

1. Acute Physical Side Effects Encountered in Clinical Trials

Mild to moderate acute physical side effects have been reported by individuals receiving human-grade psychedelic compounds in controlled clinical settings. The most common include 1) headaches; 2) nausea and vomiting (especially when ayahuasca or psilocybin is administered); and increased 5) blood pressure, 6) heart rate, and 7) body temperature (Vollenweider et al., 1999b; Carhart-Harris et al., 2012; Johnson et al., 2012; Murrough et al., 2013b; Feder et al., 2014; Palhano-Fontes et al., 2015, 2019; Schmid et al., 2015a; Griffiths et al., 2016; Nichols, 2016; Mithoefer et al., 2018). The cardiovascular effects elicited by psychedelic compounds are likely mediated by their serotonergic action and should not be overlooked, especially in patients with pre-existing cardiac conditions or a family history of cardiovascular disease.

2. Acute Psychological Side Effects Encountered in Clinical Trials

So far, psychedelics have been considered physiologically and psychologically safe when administered acutely in controlled clinical settings. In healthy patients, the administration of medical-grade psychedelics in controlled clinical settings induces a transient psychosis-like state, which can be accompanied by hallucinations, sense of unity and transcendence, mystical experiences, feelings of bliss and boundlessness, dissociation, derealization, revelations, and the re-experiencing of traumatic memories, which wear off as the compound is metabolized (Nichols, 2016; De Gregorio et al., 2018). Such acute psychological effects might be mediated by the activation of the serotonergic system and HPA axis, effects shared by almost all psychedelics (Vollenweider et al., 1999b; Carhart-Harris et al., 2012; Johnson et al., 2012; Murrough et al., 2013b; Feder et al., 2014; Palhano-Fontes et al., 2015, 2019; Schmid et al., 2015a; Griffiths et al., 2016; Nichols, 2016; Mithoefer et al., 2018).

Given that the safety concerns largely revolve around the psychological rather than physiological effects of psychedelics, interpersonal support should be available during clinical administration to address psychological distress that might arise. Crucially, since psychedelics can facilitate trauma access, psychological guidance to facilitate integration should be available before, during, and after administration (Nutt and Carhart-Harris, 2020). This could maximize the positive outcomes while minimizing the negative outcomes. For example, integration support might create the prerequisites for fear extinction and memory reconsolidation, thus erasing the fear memory (Feduccia and Mithoefer, 2018; Inserra, 2018; Nutt and Carhart-Harris, 2020). However, absence of guidance after trauma access may also lead to the reinstauration of trauma and significant distress, highlighting the importance of adequate follow-up.

1. Risk of Switch to Mania in Patients with Bipolar Disorder

Another potential side effect that could be encountered after the administration of psychedelic compounds is the risk of switch to mania in patients with BD (Szmulewicz et al., 2015). At least two such case reports are available in the literature. One is of a patient with BD with a current depressive episode who switched to a manic episode after a 4-day ayahuasca ritual (Szmulewicz et al., 2015). The patient had previously experienced hypomanic episodes, with the most recent occurring 10 days prior to ritual. The man was admitted to hospital and treated with the antipsychotic risperidone and the benzodiazepine clonazepam for 1 month, after which he became asymptomatic and was discharged (Szmulewicz et al., 2015). Another case report is of a male BD type 1 treatment-refractory psychiatrist who had attempted to self-medicate with smoked DMT (up to 1 g daily), phenelzine (a MAOI inhibitor), and clonazepam and who was admitted to hospital after developing a hypomanic psychotic episode with highly disturbed and agitated behavior 2 to 3 days after interrupting his self-medication schedule (Brown et al., 2017c). These reports suggest the need to assess the likelihood that patients with BD might switch to mania if receiving psychedelic therapy and to identify the most suited antimanic pharmacological approaches to treat such patients should the switch take place.

2. Risk of Psychosis

One of the most significant potential psychological side effects after the ingestion of psychedelic compounds in uncontrolled settings is the onset of a psychotic episode, which can last for several hours or days after the drug has worn off. The most vulnerable patients to the onset of long-lasting psychoses appear to be those who have previously received a diagnosis of SCZ, BD, or personality disorder or patients more inclined to mistrustfulness, fearfulness, and susceptibility on projection as defense (Anastasopoulos and Photiades, 1962; De Gregorio et al., 2016a, 2018; Dos Santos et al., 2017). Other risk factors include mixing other psychoactive substances, such as cannabis. For example, it was reported that an individual with a history of cannabis consumption and hypomania started experiencing psychotic symptoms associated to his hypomanic condition after the inhalation of a cannabis/DMT mix (Umut et al., 2011). Psychotic episodes have also been reported in individuals affiliated with religious groups who ingest ayahuasca for ritualistic and religious purposes, although the incidence appears to overlap that of the general population (dos Santos et al., 2017).

3. Acute Severe Physical Side Effects Encountered in Uncontrolled Settings

From a physiological point of view, psychedelic compounds appear to have a considerable margin of safety. For example, even after the accidental ingestion of very high doses (up to 550 doses) of LSD, no major adverse physiological side effects have been reported in a recent case report of three individuals (Haden and Woods, 2020). On the contrary, positive mental health changes were reported in two out of three patients, such as a 20-year-long reduction of mania with psychotic features in one subject and a reduction in physical pain and morphine addiction in another (Haden and Woods, 2020).

Although rare, conspicuous physiological adverse reactions such as vasoconstriction, coronary artery spasms, and rhabdomyolysis have been reported after the consumption of high doses of psychedelic compounds in unsupervised settings. This may be attributable to the serotonergic action of psychedelics, which affects cardiovascular function and can lead to serotonin syndrome and cardiovascular complications (Berrens et al., 2010; Cogen et al., 1978; Nichols, 2016).

4. Serotonin Syndrome

Another concern surrounding the use of psychedelic compounds in psychiatry is the onset of serotonin syndrome, a potentially life-threatening condition due to excessive stress on central and peripheral postsynaptic serotonin receptors that results in autonomic and neuromuscular aberrations that can prove fatal (Martin, 1996; Parrott, 2002; Boyer and Shannon, 2005; Scotton et al., 2019). Serotonin syndrome is characterized by changes in mental status including mild agitation; hypervigilance; delirium; autonomic hyperactivity including tachycardia, hyperthermia, hypertension, shivering, and hyperactive bowels; and neuromuscular abnormalities such as hyperreflexia and muscular rigidity (Martin, 1996; Boyer and Shannon, 2005; Scotton et al., 2019). If hyperthermia is not treated, it can lead to more severe side effects, such as rhabdomyolysis, metabolic acidosis, seizures, renal failure, intravascular coagulopathy, and death (Martin, 1996; Boyer and Shannon, 2005; Scotton et al., 2019).

Several approved serotonergic psychiatric drugs, including SSRIs, have been associated with serotonin syndrome, both at clinical doses and after accidental or voluntary overdose (Graudins et al., 1998; Kinzie and Meltzer-Brody, 2005; Paruchuri et al., 2006). A major concern revolves around the ingestion of serotonergic psychedelics by patients who medicate with SSRIs, a combination which can result in serotonin syndrome due to synergistic effects of SSRIs and serotonergic psychedelics or other compounds with MAOI activity such as Peganum harmala (Syrian rue) (Bakim et al., 2012). Concerning psychedelics, dissociative anesthetics, and empathogens, serotonin syndrome has been reported after the ingestion of LSD, MDMA, 5-methoxydiisopropyltryptamine (“foxy methoxy”), dextromethorphan, and P. harmala and by the combined use of pharmaceutical amphetamines and MDMA, mostly in patients abusing one or more of these substances, and in patients receiving MAOI or SSRI/SNRI therapies (Silbergeld and Hruska, 1979; Parrott, 2002; Ener et al., 2003; Boyer and Shannon, 2005; Arora and Kannikeswaran, 2010; Bakim et al., 2012; Davies et al., 2014; Tao et al., 2017; Singh et al., 2019).

Although no cases have been reported of serotonin syndrome after the administration of medical-grade psychedelic compounds in clinical settings, this side effect should be taken into consideration. Importantly, the use of SSRIs and SNRIs needs to be carefully taken into consideration, and if required, it should be suspended under medical supervision prior to psychedelic therapy (Bakim et al., 2012). Pharmacogenomic variability might also increase the likelihood of serotonin syndrome, as it was shown for carriers of polymorphisms at the T102C site of the 5-HT_2A receptor (Cooper et al., 2014). This raises the concept that pharmacogenomic screening tests could be implemented for the identification of patients who are suitable to receive psychedelic therapy and those more likely to undergo severe side effects.

5. 3,4-Methylenedioxymethamphetamine Neurotoxicity

Several studies have investigated whether MDMA is neurotoxic on neurotransmitter systems (Schmidt, 1987; Gudelsky et al., 1994; Armstrong and Noguchi, 2004; Granado et al., 2011; Shokry et al., 2019) given that this compound is also used to induce dopaminergic toxicity in mouse models (Blesa and Przedborski, 2014). Catechols derived from MDMA metabolism appear to be responsible for the neurotoxic effects of MDMA given that direct intracerebroventricular administration of MDMA does not elicit 5-HT neurotoxicity (Green et al., 2003). For example, intrastriatal administration of 2,4,5-trihydroxymethamphetamine significantly depletes both 5-HT and DA, intracortical administration decreases 5-HT, and intracerebroventricular administration moderately depletes striatal DA without affecting 5-HT levels (Johnson et al., 1992; Zhao et al., 1992). Other metabolites, such as 6-HO-MDMA, appear to be nontoxic given that intrastriatal and intracerebroventricular administration does not affect 5-HT or DA levels (Zhao et al., 1992). Further metabolites that might contribute to the neurotoxic properties of MDMA are quinone-thioethers, orto-quinones, and the glutathione conjugates 5-GSyl-α-MeDA and 2,5-bis-(glutathion-S-yl)-α-methyldopamine (2,5-bis-(glutathione-S-yl)-α-MeDA) (Monks et al., 2001; Green et al., 2003). Acute intracerebroventricular injections of 5-GSyl-α-MeDA were reported to acutely increase DA and 5-HT levels and to increase long-term DA (but not 5-HT) turnover (Miller et al., 1996). Multiple intrastriatal and intracortical administrations of this metabolite significantly deplete striatal and cortical (but not hippocampal) 5-HT (Bai et al., 1999). 2,5-Bis-(glutathion-S-yl)-α-MeDA administration was shown to reduce cortical 5-HT and hippocampal 5-HT and 5-hydroxy-indoleacetic acid, to modestly reduce striatal 5-HT concentrations, and to produce 5-HT neurotoxicity, which was limited to the terminal areas, without affecting DA concentrations or those of its metabolites (Miller et al., 1997). Similarly, intrastriatal and intracortical administration of 2,5-bis-(glutathion-S-yl)-α-MeDA significantly depletes striatal, cortical, and hippocampal 5-HT (Bai et al., 1999). However, in many toxicity studies, the administration is often repeated, and higher doses are employed (i.e., 5–20 mg/kg) compared with clinical trials (Pantoni and Anagnostaras, 2019). However, when taking into consideration clinical or recreational doses (i.e., 1–2 mg/kg), no cognitive impairments or long-term neurotoxicity is observable (Pantoni and Anagnostaras, 2019).

1. Codependence and Psychosis

Several psychoses induced by psychedelics are linked to the codependence with other substances, especially cannabis and alcohol. Even if psychosis induced by multiple substances is a frequent occurrence in the psychiatric emergency departments, there are only a few studies and case reports available (Warren et al., 2013; Paterson et al., 2015; Dos Santos et al., 2017).

2. Abuse Potential

The abuse potential profile of psychedelic compounds is weak, and so is the recurrence of significant psychiatric sequelae from the administration of psychedelic compounds in controlled clinical settings (Nichols, 2016). Studies in animals require more deep investigations (see below). A rare case of physical dependence from, and tolerance to, LSD has recently been reported in the literature (Modak et al., 2019). One recent study investigated the abuse liability of low- and high-dose LSD, psilocybin, and mescaline and repeated LSD administration via the intracranial self-stimulation paradigm in rats, which is widely adopted to determine the abuse potential of a putative drug (Sakloth et al., 2019). Interestingly, the acute administration of LSD, psilocybin, and mescaline fails to produce reliable evidence of abuse (Sakloth et al., 2019). Moreover, the authors found that repeated LSD (at the highest doses tested, but not at low doses) tends to attenuate the rewarding effects of the psychostimulant methamphetamine and the depressant κ-opioid receptor agonist U69,593, although only findings for the latter were statistically significant (Sakloth et al., 2019). In another study, no preference for the self-administration of DMT or psilocybin was observed in comparison with saline (aside from one subject showing transient higher rates of DMT and psilocybin self-administration), and none self-administered DOI (Fantegrossi et al., 2004). On the other hand, in nonhuman primates, minimal rates of LSD self-injection were observed after a daily access procedure, but a more substantial self-administration was observed after an intermittent access procedure (Goodwin, 2016). Therefore, although it seems that these compounds present low abuse liability, more adequately powered studies are required to confirm these findings, especially those using different doses and different regimens of psychedelics.

3. Hallucinogen-Persisting Perception Disorder

Hallucinogen-persisting perception disorder is the reoccurrence of drug-like effects long after the substance has worn out (Orsolini et al., 2017; Halpern et al., 2018; Knuijver et al., 2018; Skryabin et al., 2018). Although the incidence of this condition is low and seldom impairs life quality significantly, it has been reported in patients with alcoholism (Batzer et al., 1999), military personnel (Stanton and Bardoni, 1972), and individuals with a long-term history of psychedelic use/abuse proportionally to the number of times the substance was ingested but sometimes after one administration (Batzer et al., 1999; Orsolini et al., 2017; Anderson et al., 2018; Halpern et al., 2018; Knuijver et al., 2018; Martinotti et al., 2018; Kurtom et al., 2019; Skryabin et al., 2018; Goldman et al., 2007). Successful treatment of hallucinogen-persisting perception disorder has been reported with high-potency serotonergic benzodiazepines such as clonazepam (Lerner et al., 2003).

1. Cardiovascular Side Effects

Other potential severe side effects that might arise after microdosing concern the direct repeated stimulation of cardiac 5-HT_2B receptors by psychedelics, as well as the indirect 5-HT_2B receptors stimulation due to increased circulating 5-HT levels (Connolly et al., 1997; Fishman, 1999). Cardiac 5-HT_2B receptor overstimulation can lead to cardiovascular dysfunction after significant histologic and functional changes to the heart valves (Gustafsson et al., 2005; Elangbam et al., 2008; Hutcheson et al., 2011). For example, valvular fibrosis leading to valvular thickening and valvular regurgitation were observed in patients receiving anorectic (appetite-depressing) drugs with 5-HT_2A/2B receptor action, such as the fenfluramine-phentermine combination (Connolly et al., 1997; Curfman, 1997; Hutcheson et al., 2011). Animal models of repeated 5-HT administration confirmed those findings and added that these effects are accompanied by increased cardiac 5-HT_2B receptor and decreased SERT expression (Gustafsson et al., 2005; Elangbam et al., 2008; Hutcheson et al., 2011). Similarly, the ergot alkaloids ergotamine and methysergide, which are used for the treatment of migraine and cluster headaches, were found to damage valvular function to extents that may require valve replacement due to pathologic fibrosis and thickening (Graham et al., 1966; Graham, 1967; Mason et al., 1977; Redfield et al., 1992). Pulmonary hypertension is another risk that should be assessed in response to the repeated administration of serotonergic psychedelics. This severe side effect was observed after the repeated administration of fenfluramine and dexfenfluramine, a postsynaptic 5-HT releaser and presynaptic SERT inhibitor, respectively, and aminorex, which is an NE/catecholaminergic releaser (Gurtner, 1985; Fishman, 1999; Gaine et al., 2000).

Severe cardiovascular side effects (cardiopulmonary arrest) were reported in one individual after frequent ingestion of Psilocybe semilanceata mushrooms for 1 month (Borowiak et al., 1998). Although this represents an extremely rare case report, future efforts should be directed toward identifying potential cardiac effects that might result from the repeated administration of psychedelic compounds, even in microdoses. Further efforts should be devoted to identifying biological biomarkers and psychological and neuroimaging predictors of psychedelic-induced subjective side effects. Further efforts should be devoted to identifying biological biomarkers and psychological predictors of psychedelic-induced subjective effects (such as rostral ACC thickness), as well as dissociation, negative experiences, and psychoses. This could help identify patients more likely to encounter challenging experiences (such as those with high neuroticism, emotional instability, and family history of psychosis, bipolar disorders, and childhood trauma) and implement adequate strategies to maximize therapeutic improvement while minimizing potential biological and psychological short- and long-term side effects (Barrett et al., 2017; Thal et al., 2019; Lewis et al., 2020).