A multiorganism pipeline for antiseizure drug discovery: Identification of chlorothymol as a novel γ-aminobutyric acidergic anticonvulsant

Objective: Current medicines are ineffective in approximately one-third of people with epilepsy. Therefore, new antiseizure drugs are urgently needed to address this problem of pharmacoresistance. However, traditional rodent seizure and epilepsy models are poorly suited to high-throughput compound screening. Furthermore, testing in a single species increases the chance that therapeutic compounds act on molecular targets that may not be conserved in humans. To address these issues, we developed a pipeline approach using four different organisms. Methods: We sequentially employed compound library screening in the zebrafish, Danio rerio, chemical genetics in the worm, Caenorhabditis elegans, electrophysiological analysis in mouse and human brain slices, and preclinical validation in mouse seizure models to identify novel antiseizure drugs and their molecular mechanism of action. Results: Initially, a library of 1690 compounds was screened in an acute pentylenetetrazol seizure model using D rerio. From this screen, the compound chlorothymol was identified as an effective anticonvulsant not only in fish, but also in worms. A subsequent genetic screen in C elegans revealed the molecular target of chlorothymol to be LGC-37, a worm γ-aminobutyric acid type A (GABAA) receptor subunit. This GABAergic effect was confirmed using in vitro brain slice preparations from both mice and humans, as chlorothymol was shown to enhance tonic and phasic inhibition and this action was reversed by the GABAA receptor antagonist, bicuculline. Finally, chlorothymol exhibited in vivo anticonvulsant efficacy in several mouse seizure assays, including the 6-Hz 44-mA model of pharmacoresistant seizures. Significance: These findings establish a multiorganism approach that can identify compounds with evolutionarily conserved molecular targets and translational potential, and so may be useful in drug discovery for epilepsy and possibly other conditions.

labs were blinded to each other doing the work, although experiments in each individual lab were not conducted blind.For the mouse studies, with the exception of the corneal kindled mouse model, all in vivo seizure tests were assessed with a binary "yes/no" scoring system using well-established evaluation criteriathey do not appear on a gradient that could lead to subjective interpretation and so blinding was not deemed necessary.Acute seizure testing was conducted with secondary, simultaneous verification of assigned seizure score by a trained investigator (MB-H) or other technical staff, whenever possible.For the corneal kindling model, the ED50 was calculated by two independent investigators (AJ and MB-H) using a binary "protected/not protected" scoring system based on the modified Racine seizure score.

Materials
All materials were obtained from Sigma Aldrich (Poole UK) unless stated otherwise in the text.
The Johns Hopkins Clinical Compound Library (version 1.0) was kindly supplied by Professor David Sullivan, of the Johns Hopkins School of Medicine (Baltimore, USA).

Animal maintenance
Zebrafish were maintained according to the zebrafish standards of care 1 at 28°C on a 14 hours light/10 hours dark cycle.Zebrafish eggs were collected at the beginning of each light cycle.
Male albino CF-1 mice (18-25g, ~4-5 weeks old; Envigo, Harlan) were used as experimental animals.All animals were allowed free access to food (irradiated Picolab 20 5053, LabDiets) and filtered tap water, except during experimental procedures, and were used only once.In vivo antiseizure activity was established through both electrical and chemoconvulsant seizure paradigms.Vehicle-treated mice were also tested to confirm no anticonvulsant activity of the formulation vehicle.All rodent studies were performed at and approved by the University of Washington Institutional Animal Care and Use Committee and conformed to the ARRIVE Guidelines 3 .

Compound Library Screening in zebrafish
The Johns Hopkins Clinical Compound Library (JHCCL; 4 ) was stored with each compound at a concentration of 5 mM in DMSO in 25 v-bottomed 96-well microtitre plates (Matrix) at -80°C.Assay plates contained compounds diluted to 25 M in E3 media for drug screening.For the anticonvulsant assay, embryos were raised to 50 hpf and treated with Pronase (Sigma) to remove the chorions.Embryos were aliquoted at three or four embryos per well into Multiscreen mesh-bottomed plates (100 mm; Millipore) and transferred to Multiscreen 96-well culture receiver trays (Millipore) containing the JHCCL compounds at 25 M in columns 2-11, with control wells containing either 2.5 mM VPA, E3 or DMSO only in columns 1 and 12.
Assay plates were incubated in the dark at 28°C for 90 minutes followed by addition of PTZ to a final concentration of 20 mM to all the compound wells and half of the control wells.Assay plates were incubated for 1 hour at 28°C and the embryos were then transferred to fixative containing 4% PFA and stored at 4°C overnight.Embryos were bleached according to the standard protocol 5 and stored at -20°C in methanol until required for in situ hybridization.In order to facilitate screening and eliminate the need to transfer embryos between plates at any stage of the process, samples were maintained in the same 96-well mesh-bottomed Multiscreen plates (Millipore) during drug treatment, in situ hybridization and hit detection.Chlorothymol was one of the hits identified in the primary screen and rescreened using the method above.
Selected compound stocks were then purchased separately from Sigma and tested in concentration-response assays to further confirm the identity of the selected hit compound.

Induction and analysis of PTZ-induced convulsive behavior in zebrafish and C. elegans.
D. rerio: 3 dpf fish were transferred to 48-well plates for measurement in E3 solution.20 mM PTZ was added immediately before measurement to the appropriate wells.Fish were preincubated in treatment compounds for 90 min before measurements were taken.Fish movements were measured using the Zebrabox/Zebralab (Zebrabox Viewpoint) automated locomotion tracking system 6 .Measurements were taken over a 1-hour period with light/dark intervals every 2 min (light driver intensity: 10%).For each embryo group incubated with compounds, the mean distance moved over the total time analysed and at individual time intervals were calculated.
C. elegans were age synchronised using the bleaching technique.Worms were washed off of the plate using 3.5 ml sterile H2O.1.5 ml of commercial bleach diluted with 5 M NaOH (2:1) was then added.Following repeated vortexing over a 10-min period, this solution was centrifuged at 1300 g and aspirated to 0.1 ml.Sterile H2O was used to resuspend the pellet and a second centrifuge/aspiration step performed.Eggs were pipetted onto a freshly seeded NGM plate.
All C. elegans PTZ assays were conducted in Dent's solution: 0.81 g NaCl, 0.45 g KCL, 0.20 g MgCl2, 0.12 g Hepes, 1 ml of 1 M CaCl2 and made up to 100 ml with dH2O (pH to 7.4).BSA was added at 0.06% to prevent worm adherence to the assay plate.Depending on the measurements being conducted, various concentrations of PTZ were dissolved into Dent's solution.For the majority of the experiments conducted, 7 mg/ml of PTZ (50 mM) was used and worms were incubated in this solution for 15 mins before 30 seconds of observation.
Both qualitative and quantitative measurements were performed as previously described 2 .For qualitative data, a yes/no seizing score was determined by if the measured C. elegans showed three repetitive convulsions at any time during a 30-second measuring period.
Also included in the qualitative data was scoring if worms were paralysed by the treatment (no movement for 30 seconds).A convulsion was defined as an extension and retraction of the head of the worm (an atypical worm behaviour when not presented with PTZ).The number of convulsions was also scored over the 30 seconds to provide a quantitative measure.Worms were excluded from the data if abnormal behaviour was seen such as stiffness or minimal movement determined using touch responsiveness post-recording.This was classified as a potential fatality during the experiment.

RNA extraction
Immediately following locomotion assays, 3 dpf fish were isolated for qPCR analysis and snapfrozen on dry ice.All 8 fish were taken per treatment for each repeat creating 3 samples per group.RNA extraction was conducted using the Trizol/chloroform method.Following RNA extraction samples were measured using nanodrop spectrophotometry for absorbance at 260 nm and 280 nm wavelengths to determine relative RNA concentration.cDNA synthesis cDNA was synthesised using Thermofisher superscript II reverse transcriptase.400 ng of RNA sample was combined with 0.5 µl of oligo dT (500 µg/ml), 0.5 µl dNTP (10 mM) which was topped up to 6.5 µl with sterile H2O.Using heat blocks; samples were heated to 65°C for 5 mins then chilled on ice.Samples were then centrifuged and an additional 2 µl of first strand buffer and 1 µl 0.1 M DTT was added.Following this samples were incubated at 42°C for 50 mins.Reaction was inactivated by heating to 70°C for 15 mins and stored at -70°C.

Quantitative PCR reaction assembly and thermal cycling
cDNA concentration and purity were measured using nanodrop spectrophotometry. 1 µl of 100 ng/µl cDNA template was loaded into each well with 5 µl Sybr green, 0.6 µl FW primer, 0.6 µl RV primer and 2.8 µl H2O, making a 10 µl total mixture.qPCR mixtures were loaded into Bio-Rad 96-well hard shell PCR plates.Thermo-cycling was conducted using the Bio-Rad CFX real-time PCR detection system.Temperature was held at 95°C for 3 min, then 40 cycles of the following temperatures; 95°C for 30 seconds, 60°C for 30 seconds and 72°C for 30 seconds.
Melting curves were measured through an increase from 65 to 95°C by 0.5°C every 5 seconds.

Gene
Sequence FW Sequence RV

C. elegans paralysis assay
C. elegans were incubated in a 300 µM chlorothymol, 0.1% DMSO and 0.6% BSA solution for 15 min prior to observation for 30 seconds.Worms were scored as having a paralysis phenotype if they exhibited a 10 second period of limited movement.This was defined as an inability of the worm to move either head or tail across the head-tail axis.Worms were given a mechano-sensory stimulus to ascertain survival following observation.For the addition of bicuculline methiodide, it was first dissolved in a stock concentration in 100% DMSO, following which it was diluted to 10 mM for incubation treatment and the assay conducted as above.

C. elegans microinjection
The UBC_f80M224Q and CBGtg9050C11145D fosmid constructs were obtained from Source-Bioscience, with CBGtg9050C11145D having been described previously by Gendrel et al 7 .All injections were conducted as follows.Young-adult C. elegans were injected with combinations of 15 ng/µl fosmid construct, 5 ng/µl myo-2::mCherry and made up to 120 ng/µl with empty vector filler DNA (pBluescript).For each set of injections, three independently derived lines of each transgenic strain were analysed.

Mouse brain slice preparation and electrophysiology.
Mice were euthanised by cervical dislocation in accordance with Schedule Chlorothymol was prepared as a concentrated 10 mg/ml stock solution (1000x) in DMSO and diluted to the final concentration (54 µM) in ECS.The final maximum DMSO concentration (0.1 % v/v) had no effect on mIPSCs, or the tonic current.Chlorothymol was applied via the perfusion system (2-4 ml/min) and allowed to infiltrate the slice for a minimum of 10 min while recordings were acquired.Salts were obtained from VWR and all antagonists from HelloBio (UK).
Whole-cell recordings were analysed offline using the Strathclyde Electrophysiology Software, WinEDR/WinWCP (J.Dempster, University of Strathclyde, UK) software package.
Miniature IPSCs were analysed with respect to peak amplitude, 10-90% rise time, charge transfer, and decay time course.The decay time course of mIPSCs was described adequately in all cases by fitting a double exponential function (y(t) = A1e (-t/1) + A2e (-t/2) ) using the least squares method, where A is amplitude, t is time and  is the decay time constant.From the initial bi-exponential fit, a weighted decay time constant (w) was also calculated according to Given the low intracellular Cl -concentration and holding current of +10 mV, inhibitory currents are recorded as positive deflections in the traces (Figure 4B,C).This method allows for recording inhibitory currents experienced by a neuron without altering the activity of the network.
In order to isolate and calculate the individual contribution of phasic and tonic inhibitory currents (IPha and ITon) we used a method previously developed and described by Glykys and Mody 10 .In brief, the recordings were split into 30 seconds long epochs, an allpoint histogram was constructed for each 30 seconds epoch for the distribution of inhibitory currents, and a Savitzky-Golay filter was used to smooth the histogram and determine its peak.
A Gaussian was then fitted to the side of the histogram more negative than the peak, which is un-affected by the phasic events.All points within the Gaussian fit were classified as part of the tonic current (ITon), while points more positive than a threshold (set at the peak of the Gaussian + 3 SDs), were classified as part of the phasic current (IPha).We used 2-minute periods before and after different drug manipulations to test the effects of drugs on IPha and ITon.The peak of the Gausian was considered as the mean holding current and ITon was calculated as the difference in holding current before and after drugs manipulation 10 .IPha was calculated as the mean of all the recording points after subtraction of the holding current from the recording, which brings the mean of the Gaussian to 0 and thus captures only the contribution of IPha 10 .
We tested the effect of chlorothymol on IPha and ITon by sequentially applying chlorothymol (54 µM).We initially applied chlorothymol and recorded the change in holding current and the change in IPha induced by chlorothymol.We next washed in bicuculine, which blocks both phasic and tonic GABAergic currents.This allowed us to calculate the absolute ITon as the differences in holding current between baseline conditions and during bicuculline treatment.The effect of chlorothymol on ITon was calculated as percentage change relative to the absolute ITon.
Data was analysed using custom made scripts in Matlab R2017B and presented as mean ± standard error of the mean (SEM).

Mouse in-vivo seizure and motor impairment tests.
Mice were prepared for electrically-evoked seizures with application of a drop of anaesthetic/electrolyte solution (0.5% tetracaine hydrochloride in 0.9% saline) to the eyes before placement of bilateral corneal electrodes.A chlorothymol stock (25 mg/ml) was prepared in 40% hydroxypropyl-β-cyclodextrin (HPBCD) solution for IP administration.For MES test, the electrical stimulus was 50 mA, 60 Hz for 0.2 sec delivered using equipment similar to that described by Woodbury and Davenport 11 .Absence of tonic hindlimb extension was considered protected.Mice were challenged with 6 Hz stimulation used to induce seizures at 32 mA and 44mA for a duration of 3 sec via corneal electrodes.32 mA and 44 mA are 1.5 times and 2 times the CC97 of male CF-1 mice (current intensity that evokes seizure in 97% of the population tested 12 ).Typically, 6 Hz seizures are characterised by an initial momentary stun followed immediately by forelimb clonus, twitching of vibrissae, and Straub tail.
Animals not displaying this behaviour were considered protected.The subcutaneous PTZ chemoconvulsant seizure assay was conducted in mice with a dose of 85 mg/kg dissolved in 0.9% saline, the convulsant dose of 97% of male CF-1 mice 12 .Animals were observed for 30 mins for the presence of a 3-sec clonic episode.Absence of a seizure in the 30-min observation period was scored as protection.
For corneal kindling, mice were kindled to a criterion of 5 consecutive secondarily generalized seizures (Stage 4 or 5, as described by Racine 13 ) according to the corneal kindling protocol previously described 14; 15 .Twice daily, 0.5% tetracaine was applied to each eye and the optic nerve was stimulated through bilateral corneal electrodes (3 mA, 60 Hz, 3 seconds).
Each stimulation session was separated by at least 4 hours.After receiving twice daily corneal stimulations, CF-1 mice typically reached the first Stage 5 seizure within approximately 8-10 days.Twice daily stimulations continued for each mouse until that animal had achieved the criterion of 5 consecutive Stage 5 seizures, whereby it was considered "fully kindled", which typically occurs after 10-14 days.Fully kindled mice were then stimulated once every other day until all other mice within the group achieved the criterion of 5 consecutive Stage 5 seizures.Testing of chlorothymol commenced at least 5-7 days after the last stimulation necessary to achieve kindling criterion for all mice in the group 14 .Quantitative differentiation studies were performed 1 hour after drug dosing (at the TPE).Mice displaying a seizure score < 5 were considered protected.Each corneal kindled mouse was allowed at least 3-4 days between tests to washout any investigational compound after testing.
The fixed-speed rotarod test was used to establish minimal motor impairment (MMI) 16 .
When a mouse is placed on a 1-inch knurled rod that rotates at a speed of 6 r.p.m., the animal can maintain its equilibrium for sustained periods of time.Compounds were considered toxic if treated mice fell off this rotating rod three times during a 1-minute period.

Statistical analysis
Phenotype analysis: Statistical analysis was generally performed using GraphPad Prism version 6 (Graphpad Software Inc, California USA) using Student's t tests or ANOVA with appropriate corrections for multiple testing.For mouse phenotyping; time of peak efficacy (TPE) of chlorothymol was first established in n = 4 mice for all acute seizure tests.All quantitative in vivo antiseizure/behavioral impairment studies were then conducted at the TPE of that seizure test.For all acute seizure tests, groups of at least n = 8 mice were tested with various doses of chlorothymol until at least 2 points were established between the limits of 100% protection or MMI and 0% protection or MMI.The dose of drug required to produce the desired endpoint in 50% of animals (ED50 or TD50) in each test, the 95% confidence interval and the slope of the regression line were then calculated by Probit analysis 17 .
qPCR analysis: The fold change in gene expression for qPCR experiments was calculated using the delta-delta CT quantification method.c-fos fold change was normalised against actin as a housekeeping gene.Larvae from each treatment group were pooled, giving 3 samples of 8 larvae.Data shown are the mean of the 3 samples with SEM.Seizure-prone unc-49(e407) mutant worms were pre-incubated for 2 hours in the presence of the indicated concentra ons of VPA.They were then transferred to droplets containing 50 mM PTZ and the same concentra on of VPA.The total number of head-bobbing convulsions was measured over a 30-second period and displayed as a % of PTZ-treated controls ± SEM (n=10 worms per experiment; N=3 independent experiments).
on of the op mal an convulsant VPA concentra on in C. elegans.