Psychedelic mushrooms (e.g., Psilocybe azurescens or Psilocybe mexicana) and their associated mind- and consciousness-altering effects have been explored for over 3,000 years (Van Court et al., 2022). Most recently, psilocybin has been investigated for its therapeutic properties in several affective disorders including anxiety, treatment-resistant major depression, and cluster headache (Moreno et al., 2006; Sewell et al., 2006; Grob et al., 2011; Griffiths et al., 2016; Ross et al., 2016; Carhart-Harris et al., 2017; Johnson et al., 2017; Bogenschutz et al., 2018; Becker et al., 2022).

As a psychoactive alkaloid, psilocybin undergoes first-pass metabolism and acts as a prodrug. Upon oral ingestion, its terminal phosphate group (PO₄) is rapidly cleaved by alkaline phosphates and non-specific esterases. This transforms psilocybin into psilocin (4-hydroxy-N,N-dimethyltryptamine), a more lipophilic molecule that can cross the blood-brain barrier more readily (Horita and Weber, 1961a; Horita and Weber, 1962; Eivindvik et al., 1989; Hasler et al., 1997) to produce its psychedelic effects via serotonin 5-HT_2A receptors (Vollenweider et al., 1998; Rickli et al., 2016). Psilocin is therefore the active agent that produces psilocybin’s mind-altering effects. Psilocin concentration peaks in plasma around 2 h after oral administration and the subjective effects last for approximately 6 h (Griffiths et al., 2016; Johnson et al., 2017; Holze et al., 2022a). In humans, psilocin plasma concentrations of approximately 15–20 ng/mL and an elimination half-life (t_1/2) of 2–3 h were observed after a single oral dose of 25 mg psilocybin (Kolaczynska et al., 2021; Holze et al., 2022a). Psilocin is primarily excreted in urine, although only a minor fraction of administered psilocybin (1.5%) is eliminated as unconjugated psilocin in the first 24 h (Holze et al., 2022a). Psilocin is extensively glucuronidated to psilocin-O-glucuronide, a major urine metabolite. Approximately 20% of orally administered psilocybin is excreted as glucuronidated psilocin in humans within 24 h (Holze et al., 2022a). This conjugation is mainly catalyzed by UDP-glucuronosyltransferase (UGT) 1A10 and UGT1A9 which are highly expressed in the small intestine and liver, respectively (Sticht and Käferstein, 2000; Grieshaber et al., 2001; Hasler et al., 2002; Manevski et al., 2010; Kolaczynska et al., 2021). Concurrently, psilocin can undergo demethylation and oxidative deamination to form the intermediate metabolite 4-hydroxyindole-3-acetaldehyde (4-HIA), presumably catalyzed by monoamine oxidase (MAO). 4-HIA undergoes either reduction to 4-hydroxytryptophol (4-HTP) or oxidation to 4-hydroxindole-3-acetic acid (4-HIAA) (Kalberer et al., 1962; Hasler et al., 1997; Lindenblatt et al., 1998; Kolaczynska et al., 2021). 4-HIAA is, along with psilocin-O-glucuronide, another major urine metabolite. Around 33% of a psilocybin dose is renally excreted as 4-HIAA (Holze et al., 2022a). It is suggested that similar to the metabolic pathway of serotonin (5-hydroxytryptamine, 5-HT), aldehyde (ALDH) and alcohol dehydrogenase (ADH) might play a key role in transforming 4-HIA to 4-HIAA and 4-HTP, respectively (Svensson et al., 1999; Dinis-Oliveira, 2017). A further minor metabolic route is proposed to be the hydroxylation and oxidation of the indole moiety of psilocin to form an iminoquinone or o-quinone structure. It is hypothesized that this reaction is exerted by oxidative enzymes such as ceruloplasmin (copper carrying oxidase in human blood), cytochrome oxidases, or non-enzymatically by ferric oxide (Fe³⁺) but not MAO (Blaschko and Levine, 1960; Horita and Weber, 1961b; Dinis-Oliveira, 2017). However, the exact structure of the oxidized metabolite has not been experimentally elucidated. Moreover, to our best knowledge, the role of cytochrome P450 (CYP) enzymes in psilocybin’s metabolism has thus far not been demonstrated. The CYP superfamily is responsible for the oxidative metabolism of several psychoactive substances and has previously been shown to be involved in the metabolism of N,N-dimethyltryptamine (DMT), lysergic acid diethylamide (LSD), or 3,4-methylenedioxymethamphetamine (MDMA) (Vizeli et al., 2017; Luethi et al., 2019; Eckernäs et al., 2023).

Furthermore, the pharmacological activity or relevance of psilocybin’s metabolites needs to be assessed. Psilocin’s 5-HT receptor binding and receptor activation potency have been determined previously with relevant activity at human 5-HT_1A, 5-HT_2A, 5-HT_2B, and 5-HT_2C receptors (Rickli et al., 2016). In addition, psilocin moderately inhibits the serotonin transporter (SERT) but not dopamine (DAT) or norepinephrine transporters (NET) (Rickli et al., 2016). However, it is not clear if any of psilocybin’s remaining metabolites display relevant pharmacological activity at 5-HT receptors or the monoamine transporters.

As psilocybin transforms from a recreational substance to a potential therapeutic agent, a comprehensive understanding of its pharmacological characteristics and metabolic breakdown is necessary. In this study, we aimed to characterize the metabolic pathways of psilocin in vitro, focusing on the involvement of CYP enzymes. Furthermore, we compared the in vitro findings to in vivo findings in both humans and mice.

In the present study, we elucidated psilocybin’s metabolic pathways in vitro by incubating psilocin with a range of human metabolic enzymes, as well as in vivo in mice and humans. In addition, we tested the 5-HT receptor binding affinity and activation potency of psilocin and some of its major metabolites. We observed that psilocin is metabolized by HLM and MAO-A to 4-HIAA and 4-HTP. Minor formation of psilocin-O-glucuronide was observed in HIM, but not with recombinant UGT1A10. The putative minor metabolites, norpsilocin, and oxidized psilocin, were formed by human recombinant CYP2D6. In contrast to in vitro observations, humans and mice showed significant psilocin glucuronidation and formation of 4-HIAA but not 4-HTP. We confirmed that putative oxidized psilocin was present in both species, while putative norpsilocin was observed solely in mice. Psilocin interacted with 5-HT_1A, 5-HT_2A, 5-HT_2B, and 5-HT_2C receptors in the nanomolar range, while the metabolites 4-HIAA and 4-HTP showed no relevant interaction.

Based on our findings and the research of others, we propose a metabolic pathway for psilocybin as outlined in Figure 5. Psilocybin is rapidly dephosphorylated to psilocin by alkaline phosphatases and non-specific esterases in the intestines, kidneys, and probably also in the blood circulation (Hasler et al., 1997; Dinis-Oliveira, 2017). In mice, a competitive alkaline phosphatase substrate can occupy most of the enzyme and prevent the conversion of the prodrug psilocybin to psilocin (Horita, 1963). Studies in rats have shown that psilocybin probably undergoes a complete transformation to psilocin before being absorbed and distributed in the blood circulation (Eivindvik et al., 1989). In accordance with these findings, we did not detect psilocybin in the plasma of humans or mice. The plasma elimination half-life of psilocin in humans is approximately 2 h, while mice metabolize psilocin at double this rate, resulting in a half-life of around 0.9 h. Therefore, psilocin is extensively metabolized and can undergo several metabolic reactions.

A major part of psilocin is glucuronidated to psilocin-O-glucuronide and is excreted via urine in the first 24 h. In clinical studies, plasma concentrations of conjugated psilocin were around 4-fold higher than concentrations of free psilocin (Holze et al., 2022a). The half-life of psilocin-O-glucuronide in humans is around 4 h, while in mice it was approximately 4-fold lower (∼1 h) (Holze et al., 2022a). Manevski et al. reported that UGT1A10 has the highest glucuronidation activity for psilocin, while UGT1A9 may also be a main contributor due to its high expression in the liver (Manevski et al., 2010). Contrary to those findings, our in vitro glucuronidation systems were not able to recreate the extensive in vivo glucuronidation of psilocin. Although the positive control OH-efavirenz was readily glucuronidated, only minor amounts of psilocin were transformed into the conjugated compound by HIM and no visible transformation was observed with recombinant UGT1A10. Therefore, using more complex systems like primary hepatocytes or human spheroid three-dimensional models might be more suitable to investigate in vivo phase II metabolism of psilocybin (Ooka et al., 2020).

Psilocin undergoes another major oxidative metabolic pathway similar to the structurally related monoamine neurotransmitter 5-HT (Svensson et al., 1999). MAO enzymes metabolize the tertiary amino group of psilocin to yield the intermediate aldehyde 4-HIA (Hasler et al., 1997). 4-HIA is then rapidly metabolized to either 4-HIAA or 4-HTP (Hasler et al., 1997; Becker et al., 2022). Recombinant MAO-A enzymes as well as HLM were able to produce minor amounts of 4-HIAA and 4-HTP in vitro, while selective MAO-A inhibitors reduced this reaction. Neither MAO-B nor any of the assessed CYP enzymes produced 4-HIAA or 4-HTP, suggesting that only MAO-A is involved. However, since MAO-A likely only catalyzes the production of the intermediate 4-HIA, major amounts of 4-HIAA and 4-HTP might only be produced when functional ALDH or ADH enzymes are also present, similar to 5-HT’s metabolism (Svensson et al., 1999; Dinis-Oliveira, 2017). Moreover, careful interpretation of the produced amounts is advised, considering that DMT, a substance well-documented to be rapidly metabolized by MAO in vivo, exhibited only slow metabolism in vitro, thus resulting in only minor amounts of its metabolite IAA (Luethi et al., 2022; Vogt et al., 2023). Interestingly, although reported before in other studies (Horita, 1963; Hasler et al., 1997), 4-HTP was neither detected in humans nor in mice but only in vitro with HLM. Most likely 4-HTP is either rapidly metabolized in humans and mice or not produced at all.

Minor interspecies differences were detected when comparing the results of human with mouse samples. In humans, 4–5-fold higher maximal concentrations of 4-HIAA were found compared to psilocin (Holze et al., 2022a; Becker et al., 2022), while the concentrations in mice were 2-fold lower compared to psilocin. The elimination half-life of 4-HIAA in mice (0.75 h) was more than 2-fold shorter than in humans. Moreover, 4-HIAA is most likely glucuronidated in mice whilst in human samples there was no evidence of glucuronidated 4-HIAA. Although 4-HIAA possesses several glucuronidation sites, we suggest that it undergoes a similar metabolic reaction as 4-hydroxyindole and psilocin in humans to form the O-glucuronide (Manevski et al., 2010). The O-glucuronidation of 4-hydroxyindole in humans is mainly catalyzed by UGT1A6 and UGT1A9 which also could play an important role in this pathway (Manevski et al., 2010). Alternatively, 4-HIAA could undergo acyl-glucuronidation, to form an unstable acyl-glucuronide, or N-glucuronidation (Regan et al., 2010; Kaivosaari et al., 2011).

Out of the eight assessed CYP enzymes, only CYP2D6 and CYP3A4 showed a relevant metabolic activity, suggesting that CYP1A2, 2B6, 2C8, 2C9, 2C19, and 2E1 are unlikely to be involved in psilocybin’s metabolism. Neither 4-HIAA nor 4-HTP formation could be observed with any of the CYP enzymes, confirming that CYPs are not involved in these metabolic pathways. Psilocin can, however, continue to follow another minor metabolic pathway that may be of relevance in the presence of MAO inhibitors (e.g., antidepressants). Several studies suggested a minor oxidative metabolite with an o-quinone or iminoquinone structure produced by, e.g., ceruloplasmin, cytochrome oxidases, or non-enzymatically by Fe³⁺ (Blaschko and Levine, 1960; Horita and Weber, 1961b; Horita, 1963; Kovacic, 2009). We hypothesized that CYP enzymes might be involved in this step and investigated this in vitro. CYP3A4 showed minor activity in metabolizing psilocin, however, no previously described metabolite could be identified. Estimating the extent of the contribution of CYP3A4 in psilocin’s metabolism is challenging. Although our data implies CYP3A4 involvement, additional investigation is necessary for a more accurate assessment. CYP2D6 was more active than CYP3A4 as it rapidly and completely metabolized psilocin. Subsequently, oxidized psilocin, potentially a quinone-type structure like psilocin iminoquinone (e.g., 4-hydroxy-5-oxo-N,N-dimethyltryptamine) or psilocin hydroquinone (e.g., 4,5-hydroxy-N,N-dimethyltryptamine) has been detected following incubation of psilocin with CYP2D6. Afterward, the compound has also been detected in human and mouse plasma. This metabolite might have been described as a psilocybin metabolite in mammalian tissue homogenates, but never before in humans (Horita and Weber, 1961b; Horita, 1963; Dinis-Oliveira, 2017). Based on the parent mass of 220.0 Da and the observed fragmentation pattern in the mass spectrometer (m/z 221.0 → 176.0 Da), we can postulate that this metabolite possesses an additional oxygen atom on the benzene ring of the indole moiety. However, the exact position of the oxygen cannot be determined solely with tandem mass spectrometry. We can also not exclude that there are several different metabolites present with different hydroxylation or oxidation sites on the indole ring. Similar metabolic findings have been made for structurally related synthetic tryptamines (Bergh et al., 2024). An o-quinone structure as proposed by Dinis-Oliveira et al. can be ruled out since this metabolite would have a different molecular weight (218.0 Da) than the metabolite observed in our experiments (Dinis-Oliveira, 2017). Further structural analysis is needed to elucidate the elemental formula and exact position of the oxygen atom.

In addition, another psilocin metabolite was identified in the CYP2D6 assay and subsequently in mouse plasma. The metabolite possesses a molecular mass of 190.0 Da and therefore most likely corresponds to norpsilocin (N-methyl-4-hydroxytryptamine). The fragments observed in the mass spectrometer (m/z 191.0 → 160.0 Da) suggest that the compound arises through demethylation of the tertiary amine of psilocin. Interestingly, norpsilocin could not be detected in human plasma samples. To the best of our knowledge, this metabolite has not been observed before as a product of psilocybin’s metabolism, but rather as a synthesis or metabolic product of baeocystin, a demethylated analog of psilocybin, present in hallucinogenic mushrooms of the Psilocybe genus (Lenz et al., 2017; Sherwood et al., 2020). Norpsilocin was previously suspected to exert psychoactive effects similar to psilocin. It is nearly a full agonist at the mouse and human 5-HT_2A receptor in G_q-dependent calcium flux assays (EC₅₀ < 10 nM) but interestingly did not show any psychoactive properties in head-twitch response experiments with mice (Sherwood et al., 2020). This may be due to the inability to cross the blood-brain barrier or rapid degradation of the secondary amine group (Sherwood et al., 2020; Sherwood et al., 2024). Nevertheless, further analysis with NMR or high-resolution mass spectrometry is needed to confirm the structure of this minor psilocin metabolite. Interestingly, the reduction in psilocin concentration observed with CYP2D6 can probably not be solely attributed to an oxidative metabolite or norpsilocin. This suggests the possibility of additional, unidentified metabolites being present.

Furthermore, we showed that genetically determined CYP2D6 activity did not change exposure to psilocin in humans. Thus, CYP2D6 might not be considered to be critically involved in the main metabolism of psilocin in humans and any between-individual differences in CYP2D6 activity are unlikely to influence the response to psilocin in humans. However, the significance of this data may be constrained by the limited number of samples examined. Nevertheless, these preliminary results could have direct clinical implications for psilocybin-assisted therapy and contrast with LSD and MDMA, wherein poor CYP2D6 function significantly increases exposure to these substances and enhances the acute effects (Schmid et al., 2016; Vizeli et al., 2021).

In general, there are only limited data available on the activity of psilocybin’s metabolites at the human 5-HT receptors. Here, psilocin showed sub-micromolar receptor binding affinity at human 5-HT_1A (128 ± 33 nM), 5-HT_2A (41.1 ± 8.9 nM), and 5-HT_2C (136 ± 35 nM) receptors, confirming the results of Rickli et al. (Rickli et al., 2016). Psilocin further showed sub-micromolar activation potencies at the 5-HT_1A, 5-HT_2A, and 5-HT_2B receptors (<74.5 ± 6.0 nM). However, neither 4-HIAA nor 4-HTP displayed relevant binding affinity at the 5-HT_1A, 5-HT_2A, and 5-HT_2C receptors (K _i > 10,000 nM) and also no receptor activation potency at the 5-HT_1A, 5-HT_2A, and 5-HT_2B receptors (EC ₅₀ > 10,000 nM). The inactivity of these metabolites at the human 5-HT receptors indicates that they are not involved in exerting the psychoactive effects in humans, which is in line with the time course of the subjective effects observed in clinical studies (Holze et al., 2022a). The herein-detected oxidative psilocin metabolite has so far not been investigated at human 5-HT receptors due to its uncertain structure.

In conclusion, this comprehensive study explored the metabolic pathways of psilocin both in vitro and in vivo and provides new evidence of involved enzymes. In total, we were able to detect six psilocin metabolites. While confirming the glucuronidation of psilocin in vivo, we also detected apparent interspecies differences with the glucuronidation of 4-HIAA and the presence of putative norpsilocin in mice compared with humans. While MAO-A was identified as a key enzyme responsible for psilocin’s oxidative transformation to 4-HIAA and 4-HTP, the additional roles of ALDH and ADH still have to be investigated. CYP2D6 and CYP3A4 seem to be involved to a minor extent in psilocin’s metabolism. CYP2D6 produced norpsilocin and a structurally unresolved oxidized metabolite. However, no metabolite was identified with CYP3A4, requiring further investigation into the extent of its role in psilocin’s metabolism. The herein-employed in vitro assays assisted in unraveling the metabolism of psilocin but were unable to closely reproduce phase II metabolic reactions of UGT and MAO as observed in humans and mice. Consequently, it is recommended to use and assess more complex hepatocellular assays to further investigate the metabolism of these tryptamines. The major metabolite 4-HIAA and 4-HTP were inactive at human 5-HT receptors but the activity of oxidized psilocin metabolites and norpsilocin remain to be assessed. Inhibition of psilocin inactivation by MAO could potentially augment the metabolic pathway catalyzed by CYP2D6, thereby altering the pharmacodynamics of psilocybin therapy. However, the CYP2D6 genotype did not influence psilocin concentrations in humans. Moreover, glucuronidation of psilocin would likely continue to be the predominant metabolic pathway, rendering MAO inhibition potentially less important.

Finally, our findings on psilocybin’s metabolism contribute to the safety and efficacy of psilocybin therapy by indicating potential drug-drug interactions and helping advance research on psilocybin as a therapeutic agent.