Molecular design of a therapeutic LSD analogue with reduced hallucinogenic potential

Abstract

Decreased dendritic spine density in the cortex is a key pathological feature of neuropsychiatric diseases including depression, addiction, and schizophrenia (SCZ). Psychedelics possess a remarkable ability to promote cortical neuron growth and increase spine density; however, these compounds are contraindicated for patients with SCZ or a family history of psychosis. Here, we report the molecular design and de novo total synthesis of (+)-JRT, a structural analogue of lysergic acid diethylamide (LSD) with lower hallucinogenic potential and potent neuroplasticity-promoting properties. In addition to promoting spinogenesis in the cortex, (+)-JRT produces therapeutic effects in behavioral assays relevant to depression and cognition without exacerbating behavioral and gene expression signatures relevant to psychosis. This work underscores the potential of nonhallucinogenic psychoplastogens for treating diseases where the use of psychedelics presents significant safety concerns.

Keywords: LSD; neuroplasticity; neuropsychiatric disease; nonhallucinogenic; psychedelic.

Fig. 1.

Structural basis for the rational design of JRT. (A) Comparison between the structures of N,N-dimethyltryptamine and LSD with N,N-dimethylisotryptamine (isoDMT) and JRT. (B) Structure of LSD with key 5-HT2AR contacts indicated. Key hydrogen bonding interactions between LSD and the 5-HT2AR are highlighted in red. These interactions are not possible for JRT due to its lack of an indole N–H bond. (C and D) Molecular docking of JRT into the crystal structures of the 5-HT2AR bound to LSD (PDB: 6WGT (Left) and 7WC6 (Right)) demonstrates that JRT is predicted to adopt similar binding poses as LSD with the exception that the indole nitrogen of JRT is predicted to be more removed from S242^5.46 or G238^5.42, respectively. The binding modes with the lowest RMSD score relative to native bound (+)-LSD are shown. Transmembrane helix 4 was removed for clarity. DMT = N,N-dimethyltryptamine; LSD = lysergic acid diethylamide. See also SI Appendix, Fig. S1.

Fig. 2.

Total synthesis of JRT. (A) Retrosynthetic analysis indicates that JRT can be accessed from a 7-substituted indole and a 3-substituted nicotinic acid. (B) The total synthesis of (±)-JRT was completed in 12 steps and 11% overall yield. (C) X-ray crystal structure of methylated (+)-JRT (16). The counterion (iodide) and a solvent molecule (DCM) in the crystal structure were removed for clarity. red = oxygen, blue = nitrogen, gray = carbon, white = hydrogen. See also SI Appendix, Fig. S2.

Fig. 3.

(+)-JRT is highly selective for serotonin receptors. (A) Radioligand binding (RLB) profiles of (+)-JRT and (–)-JRT compared to (+)-LSD. Orange and white cells indicate K_i values that are less than or greater than 10 μM, respectively. Data for LSD were obtained from literature values (see SI Appendix and Dataset S1). (B) BRET-based assay of G_q activation indicates that (+)-JRT is a potent partial agonist of 5-HT2A receptors. For antagonist mode (open circles), compounds were cotreated with 10 μM serotonin. (C and D) PsychLight assays indicate that (+)-JRT is a lower potency 5-HT2A receptor partial agonist than LSD. Differences in potency/efficacy between LSD and (+)-JRT are greater for the wild type human (C) than the murinized variant (D). For antagonist mode (open circles), compounds were cotreated with 10 μM serotonin. (E) Comparison of E_max/EC₅₀ values reveals that LSD is a stronger activator of the wild type human 5-HT2A receptor than the S242A^5.46 mutant. Data for concentration–response curves represent the mean ± SEM of 3 to 4 biological replicates (performed in duplicate). *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. 5-HT = serotonin; ns = not significant; ND = not determined for incomplete curves.

Fig. 4.

(+)-JRT promotes structural plasticity in vitro and in vivo. (A) Representative images of embryonic day 18 (E18) rat cortical neurons (DIV6) treated with compounds (1 μM) demonstrate that (+)-JRT promotes dendritogenesis (white, MAP2). (B) Maximum numbers of crossings (N_max) of the Sholl plots in SI Appendix, Fig. S8A. (C) Representative images of E18 rat embryonic cortical neurons (DIV18) treated with compounds (1 μM) for 24 h demonstrate that (+)-JRT promotes spinogenesis (white, F-actin). (D) Quantification of spine density in F. (E) The 5-HT2 antagonist ketanserin (10 μM) blocks dendritic spine density induced by LSD or (+)-JRT (1 μM). (F) Schematic depicting the design of in vivo experiments. Scanning electron microscopy of serial sections demonstrates that a single dose of (+)-JRT (1 mg/kg) increases structural plasticity in the PFC 24 h after administration. (G) Representative reconstruction from EM data of dendritic segments (blue) and associated dendritic spines (magenta). (H) (+)-JRT promotes spinogenesis in vivo (N = 3 female animals, n = 8 × 27 μm³ volumes each per condition). (I) Representative S3EM image of mPFC neuropil from animals treated with (+)-JRT 24 h prior to tissue collection. Synapses are highlighted in magenta. (J) (+)-JRT promotes synaptogenesis in vivo (3 female animals, 6 fields each per condition). (K) (+)-JRT does not impact synapse size in vivo (L) Schematic depicting the design of cortical atrophy rescue experiments. (M) Representative images of cortical neurons from the PFC of Thy1-EGFP mice exposed to chronic CORT and/or (+)-JRT. (N) Confocal microscopy demonstrates that a single dose of (+)-JRT (1 mg/kg) rescues chronic CORT-induced dendritic spine loss in layer 2/3 of the mPFC 24 h after administration (2-way ANOVA). Black asterisks with black bars indicate comparisons within treatment groups (i.e., VEH or JRT. White asterisks indicate comparisons within stress groups (i.e., GH or CORT). VEH = vehicle; LSD = lysergic acid diethylamide; CLZ = clozapine; S3EM = scanning electron microscopy of serial sections; GH = gentle handling; CORT = corticosterone; ns = not significant. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, as compared to VEH control or indicated comparator. See Materials and Methods and SI Appendix, Table S1 for full details on statistics.

Fig. 5.

(+)-JRT does not exacerbate hallucinogen-like activity in vivo. (A) Mouse HTR assays in male and female animals demonstrate that (+)-JRT has low hallucinogenic potential when the assay is run in agonist mode (head-twitches quantified over 20 min). Furthermore, (+)-JRT (1 mg/kg) demonstrates antipsychotic properties by antagonizing a HTR induced by LSD (0.2 mg/kg). (B) Pretreating mice with (+)-JRT (1 mg/kg) does not induce a deficit in prepulse inhibition or exacerbate MK-801-induced deficits. AMPH = (+)-amphetamine; KET = ketamine; PCP = phencyclidine; VEH = vehicle. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, as compared to VEH controls or the comparator(s) indicated by a horizontal bar for A. For B, a 3-way ANOVA indicated a significant main effect of sound level and MK-801 treatment, but not (+)-JRT treatment, nor any significant interaction effects. ns = not significant. See Materials and Methods and SI Appendix, Table S1 for full details on statistics.

Fig. 6.

(+)-JRT exhibits antidepressant effects in vivo. (A) Schematic depicting a rat FST conducted in males 24 h after compound administration. (B) A single dose of (+)-JRT (IP) produces antidepressant-like effects at substantially lower doses than ketamine. Doses (mg/kg) are indicated within the bars representing various treatment groups. (C) Schematic depicting CORT-induced anhedonia and pharmacological rescue assessed with the SPT in mice. (D) CORT-sensitive animals (+) were defined as sucrose-preferring animals that lost that preference (i.e., score of ≤ 0.5) following 20 d of CORT administration (i.e., Day 23). (E) Treatment with (+)-JRT (1 mg/kg, IP) rescued CORT-induced deficits in the SPT (2-way ANOVA). (F) Schematic depicting chronic cold-water stress-induced anhedonia and pharmacological rescue assessed with the PRT in rats. (G) Chronic cold-water stress leads to deficits in response bias (log b) that is rescued by both KET (10 mg/kg, IP) and (+)-JRT (1 mg/kg). KET = ketamine; VEH = vehicle; CORT = corticosterone; SPT = sucrose preference test; PRT = probabilistic reward task. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, as compared to VEH controls (A), the comparator indicated by horizontal/vertical bars (D and E), or Day 1 log b values (G) for the same treatment (indicated by the color of the asterisks). See Materials and Methods and SI Appendix, Table S1 for full details on statistics.

Fig. 7.

(+)-JRT exhibits procognitive effects in vivo. (A) Schematic depicting UMS-induced impairments in cognitive flexibility and pharmacological rescue assessed with a 4-odor discrimination and reversal assay. (B) A single dose of (+)-JRT (1 mg/mL, IP) does not impact stimulus discrimination but rescues cognitive deficits induced by UMS. UMS = unpredictable mild stress; VEH = vehicle; ns = not significant. *P < 0.05, **P < 0.01, as compared to the comparator indicated by a horizontal bar. ns = not significant. See Materials and Methods and SI Appendix, Table S1 for full details on statistics.