Cytochrome P450 enzymes contribute to the metabolism of LSD to nor-LSD and 2-oxo-3-hydroxy-LSD: Implications for clinical LSD use

Luethi and colleagues situate their work in the context of a revival of clinical and experimental research on lysergic acid diethylamide (LSD). They note that LSD undergoes rapid metabolism in humans and that several metabolites have been identified, with 2-oxo-3-hydroxy-LSD (O-H-LSD) described as the major urinary metabolite and 6-norlysergic acid diethylamide (nor-LSD) also detected in plasma. Despite prior demonstrations that human liver microsomes and primary hepatocytes metabolise LSD, the specific cytochrome P450 (CYP) isoforms responsible had not been fully resolved, and the potential effects of CYP polymorphisms or co-administered drugs on LSD pharmacokinetics remained unclear. This study aimed to identify which human CYP isoforms contribute to the formation of nor-LSD and O-H-LSD in vitro, to test whether CYP induction or inhibition alters metabolite formation, and to assess the serotonergic pharmacology of LSD and these two metabolites. The researchers used human liver microsomes (HLM), recombinant human CYPs (Supersomes), and primary hepatocyte-derived cells for metabolism experiments, and they measured receptor binding and functional activity at 5-HT1A, 5-HT2A, 5-HT2B and 5-HT2C receptors to characterise potential pharmacological activity of the metabolites.

The experimental programme combined in vitro metabolism assays with receptor-binding and functional pharmacology. For microsomal metabolism, LSD (1,000 nM) was incubated with pooled human liver microsomes (Corning UltraPool HLM) at 37 °C for up to 4 h. Reactions included an NADPH regenerating system and were sampled at 0, 0.5, 1, 2, 3 and 4 h. CYP-specific chemical inhibitors were added in parallel incubations to probe contributions of individual isoforms; control incubations using established CYP substrates confirmed inhibitor efficacy. To validate microsomal inhibition results, the authors performed analogous incubations using recombinant human CYP Supersomes (rhCYP1A2, 2C9, 2D6, 2E1, 3A4 and others) and compared formation rates of nor-LSD and O-H-LSD. CYP induction was assessed in HepatoCells (primary human hepatocyte-derived cells) seeded at 400,000 cells per well; cells were treated for 72 h with vehicle (0.1% DMSO), omeprazole (50 µM) to induce CYP1A2, or rifampicin (10 µM) to induce CYP3A4, followed by 4 h incubations with substrate (tizanidine, midazolam, or 1–10 µM LSD depending on the assay) and timed sampling. LSD and metabolite concentrations were quantified by UHPLC-tandem mass spectrometry (LC-MS/MS) with methods optimised for positive and negative ionisation modes. Calibration curves used internal standards including isotopically labelled compounds; multiple reaction monitoring (MRM) transitions for LSD, nor-LSD and O-H-LSD were summed to increase sensitivity. Microsomal stability experiments report mean ± SEM and used n = 3 replicates. Pharmacology at serotonergic receptors was assessed by radioligand binding using membrane preparations from HEK293 cells transfected with human 5-HT1A, 5-HT2A or 5-HT2C receptors; displacement of selective radioligands provided Ki values (binding affinity). Functional agonist activity at 5-HT2A and 5-HT2B receptors was measured in receptor-transfected NIH-3T3 or HEK293 cells using FLIPR calcium mobilisation assays to derive EC50 (concentration producing half-maximal effect) and Emax (maximum efficacy) values via concentration–response curves.

Microsomal stability experiments showed a significant decrease in measured LSD concentration after 4 h with HLM: mean LSD fell from 1015 ± 5 nM to 735 ± 47 nM (P < 0.01, n = 3). Concurrently, formation of nor-LSD and O-H-LSD was detected but at low absolute amounts: 11.3 ± 1.6 nM nor-LSD (≈1% of parent) and 2.1 ± 0.7 nM O-H-LSD (≈0.2%) after 4 h. No metabolite formation occurred in incubations lacking microsomes. CYP inhibition in HLM implicated several enzymes. Quinidine (CYP2D6 inhibitor), 4-methylpyrazole (CYP2E1 inhibitor) and ketoconazole (CYP3A4 inhibitor) significantly reduced the formation rate of nor-LSD. For O-H-LSD formation, inhibition of CYP1A2, CYP2C9, CYP2E1 and CYP3A4 led to significant decreases. Control experiments using established CYP substrates confirmed inhibitor activity under the assay conditions. Incubations with recombinant human CYPs reproduced these patterns: inhibition experiments and formation-rate comparisons indicated that rhCYP2D6, 2E1 and 3A4 contributed significantly to nor-LSD formation, while rhCYP1A2, 2C9, 2E1 and 3A4 were involved in O-H-LSD formation (significance reported as P < 0.05). Analysis of formation-rate slopes across three experiments used analysis of covariance. CYP induction in HepatoCells modestly increased LSD metabolism. Over 4 h, non-induced HepatoCells reduced LSD by about 26% on average; omeprazole-treated (CYP1A2-induced) cells showed a 27% decrease and rifampicin-treated (CYP3A4-induced) cells a 33% decrease. Metabolite concentrations after 4 h were: nor-LSD 0.5 nM (non-induced), 1.4 nM (CYP1A2 induced), and 5.3 nM (CYP3A4 induced); O-H-LSD 0.6 nM (non-induced and CYP1A2 induced) and 2.5 nM (CYP3A4 induced). Positive control substrates (tizanidine for CYP1A2 and midazolam for CYP3A4) showed expected metabolism in HepatoCells. Binding and functional assays showed that LSD had high affinity for human 5-HT1A, 5-HT2A and 5-HT2C receptors (Ki values in the 1–10 nM range). Nor-LSD retained substantial affinity at 5-HT2A but had reduced affinity at 5-HT2C relative to LSD. O-H-LSD displayed markedly weaker binding: Ki at 5-HT2C exceeded 14 µM and binding at 5-HT2A was roughly three orders of magnitude weaker than LSD. Functionally, O-H-LSD produced partial agonism at 5-HT2A and 5-HT2B only at high concentrations (> 10 µM). By contrast, nor-LSD showed higher activity and efficacy at both 5-HT2A and 5-HT2B in the in vitro assays compared with LSD, although the authors caution that activation potency in these assays does not directly translate to clinical psychedelic potency.

The investigators interpret their results as evidence that multiple CYP enzymes contribute to the in vitro metabolism of LSD to nor-LSD and O-H-LSD, and that both inhibition and induction of these enzymes produce minor but statistically significant changes in LSD stability and metabolite formation. CYP3A4 emerged as quantitatively important for formation of both metabolites, with additional contributions from CYP2D6, CYP1A2, CYP2C9 and CYP2E1 depending on the pathway. Induction with rifampicin (a CYP3A4 inducer) substantially increased metabolite formation, supporting a role for CYP3A4, while omeprazole (CYP1A2 inducer) had only a modest effect. With respect to pharmacology, nor-LSD retained significant 5-HT2A affinity and displayed substantial functional activity in vitro, whereas O-H-LSD had substantially lower affinity and activity at serotonergic receptors, consistent with being pharmacologically inactive at clinically relevant concentrations. The authors note that N-demethylation (generating nor-LSD) only marginally reduced 5-HT2A affinity but produced a larger reduction at 5-HT2C, which could imply a role for 5-HT2C in LSD’s effects if nor-LSD contributes to in vivo activity. The authors acknowledge several limitations and uncertainties. First, the metabolism experiments used an LSD concentration (1,000 nM) that exceeds typical human peak plasma concentrations (low nanomolar) and was chosen to allow detection of small metabolite pools. Second, only nor-LSD and O-H-LSD were quantified because other metabolites were not commercially available for calibration; therefore the observed decrease in parent LSD exceeded the accounted-for amounts of these two metabolites, suggesting formation of additional, unmeasured metabolites or involvement of non-microsomal pathways such as peroxidase-mediated reactions reported previously. Third, rifampicin can induce multiple CYPs beyond CYP3A4, complicating interpretation of induction data. The authors further note discrepancies with a recent study reporting CYP2C19 involvement in N-demethylation; such differences may reflect methodological variation between studies. Clinically, the findings imply that genetic polymorphisms and drug–drug interactions affecting CYP3A4, CYP2D6, CYP1A2, CYP2C9 and CYP2E1 could alter LSD pharmacokinetics and potentially pharmacodynamics, although the in vivo relevance remains to be established. They highlight practical examples: St John’s wort as a CYP3A4 inducer, smoking-induced CYP1A2 induction, CYP1A2 inhibitors such as fluvoxamine or ciprofloxacin, ethanol induction of CYP2E1, and the sizeable roles of CYP2C9 and CYP2D6 in drug metabolism generally. The study concludes that further human drug–drug interaction studies and expanded metabolite profiling are required to determine the clinical significance of these in vitro observations.