Structure of clemizole and clemalogues 1–28. Chemical structure of clemizole and 28 clemizole analogues synthesized as part of SAR studies detailed herein. Sites of modification to the clemizole structure are highlighted in blue for each clemalogue subtype shown. Full chemical structures are provided in Supplementary Table S1.

Synthesis of clemalogues. Summary of synthetic routes used to prepare clemalogues 1–28. (a) cyclopentylacetic acid, PPA, μw, 80°C, 20%; (b) ClCH₂(4-ClPh), K₂CO₃, DMF, 60°C, 67%; (c) ethyl 4-Cl-3-oxobutanoate, SnCl₂, EtOH, 80°C; (d) pyrrolidine, EtOH, 95°C; 83% for two steps (e) Br-CH₂R, NaH, THF, 0°C to rt, 15–60% or Br-CH₂R, NaH, TBAI, THF, 0°C to rt, 13–54% or propyl iodide, NaH, THF, 0°C to rt, 47%;(f) p-Cl-benzylchloride, NaH, DMF, rt, 54%; (g) (i) LiOH, MeOH/water, rt, (ii) pyrrolidine, HATU, DIEA, DMF, rt, 43% over 2 steps; (h) p-Cl-benzylchloride, K₂CO₃, CH₃CN, 45°C, 85% or cyclohexylmethyl bromide, K₂CO₃, TBAI, CH3CN, 55°C, 30%; (i) pyrrolidine, Na(OAc)₃BH, CH₂Cl₂; 30–46%; (j) ClCH₂Ar, K₂CO₃, CH₃CN, 45°C, 44%; (k) pyrrolidine, Na(OAc)₃BH, CH₂Cl₂, rt, 65%; (l) HNR₂, Na(OAc)₃BH, CH₂Cl₂, rt, 47–85%; (m) H₂, Pd/C, MeOH, rt, 55%; (n) 4 M HCl in dioxane, rt.

Phenotypic screening of clemizole analogues. Twenty-eight clemizole analogues were screened for efficacy in suppressing the high-velocity seizure-like swim behaviour observed in scn1lab mutant zebrafish. Plots show the change in mean swim velocity of 5 dpf larvae screened at (a) 100 µM, or (b) 250 µM. Threshold for inhibition of seizure activity (positive hits—yellow data points) was determined as a reduction in mean swim velocity of ≥40% (red dashed line). The red data points represent compounds that were classified as toxic after 90-min exposure. The heat map shows the % change in velocity for the six individual larva from the first trial (1–6). Mean velocity change from six individual fish is shown for trial 1 and 2. Clemizole analogue 25 (*) failed to go into solution at 250 µM so it was not considered for further testing.

Evaluation of clemizole analogues that reduce seizure-like swim behaviour in scn1lab mutant zebrafish. Clemizole analogues identified as positive from the in vivo screen were freshly synthesized and retested for efficacy in suppressing the seizure-like swimming behaviour of 5 dpf scn1Lab mutant zebrafish. Graphs show the change in mean velocity over four concentrations of (a) compound 4, (b) compound 6 and (c) compound 20. Each bar represents the mean change in velocity ± SEM from three independent experiments (six individual larva per experiment). Toxicity is indicated by dashed bars. The threshold for a decrease in velocity is ≥40% (red line). Locomotion of larvae was recorded for 10 min after an exposure of 30 min (blue bars) and 90 min (yellow bars). A representative raw 10 min tracking plot is shown for a single experiment of six individual scn1Lab zebrafish. The chemical structure for each clemizole analogue is shown (d–f). In vitro radioligand binding analyses of (g) compound 4, (h) compound 6 and (i) compound 20 revealed specificity for 5-HT_2BR over 5-HT₂R subtypes. SB206553 was used as a positive control for 5-HT_2BR binding (black). The binding affinity for the other clemizole analogues is given in Supplementary Table S2.

Dose response evaluation of 5HT2BR agonists in scn1lab mutant zebrafish. 5HT2BR agonists were tested for efficacy in reducing the high-speed seizure-like behaviour in 5 dpf scn1lab mutant zebrafish. Graphs show the change in mean velocity over three concentrations of (a) methylergonovine, (b) 6-APB, (c) norfenfluamine, (d) BW-723C86, (e) RO-60-0175, (f) TL-99, (g) m-CPP and (h) CP-809, 101. Larvae locomotion was recorded for 10 min after an exposure of 30 min (blue bars) and 90 min (yellow bars). Each bar represents the mean change in velocity ± SEM from three independent experiments (six individual larva per experiment). The threshold for a decrease in velocity is ≥40% (red line). Representative tracking plots of a 10 min recording are shown for six individual 5 dpf scn1lab zebrafish at baseline, and following 30 min and 90 min exposure of 100 µM of each compound.

Electrophysiological assay to identify drugs that rescue the scn1lab mutant epilepsy phenotype. (a) Electrophysiology recording were obtained with an electrode placed in the forebrain of 5 dpf agar-immobilized scn1lab larvae that had previously showed suppressed seizure-like behaviour in the locomotion assay. (b) Bar graphs show the frequency of epileptiform events in a 10 min recording epoch for scn1lab larvae exposed to clemizole analogues 4 (n = 15), 6 (n = 12), 20 (n = 9), 6-APB (n = 6), norfenfluramine (NorFA) (n = 8), methylergonovine (MeERGO) (n = 7) or scn1lab mutants (n = 15). The graph represents mean ± SEM and individual data points are shown. Kruskal–Wallis one-way analysis of variance was used to test for significance. *P < 0.05; **P < 0.01; ***P < 0.001. (c) Representative field electrode recording epochs (10 min) are shown for clemizole analogues 4, 6, 20, methylergonovine (MeERGO) and 6-APB. These compounds showed significant changes in the frequency of events compared to untreated scn1lab mutant zebrafish (red).