A Caenorhabditis elegans assay of seizure-like activity optimised for identifying antiepileptic drugs and their mechanisms of action

Abstract

Background: Epilepsy affects around 1% of people, but existing antiepileptic drugs (AEDs) only offer symptomatic relief and are ineffective in approximately 30% of patients. Hence, new AEDs are sorely needed. However, a major bottleneck is the low-throughput nature of early-stage AED screens in conventional rodent models. This process could potentially be expedited by using simpler invertebrate systems, such as the nematode Caenorhabditis elegans.

New method: Head-bobbing convulsions were previously reported to be inducible by pentylenetetrazol (PTZ) in C. elegans with loss-of-function mutations in unc-49, which encodes a GABA_A receptor. Given that epilepsy-linked mutations in human GABA_A receptors are well documented, this could represent a clinically-relevant system for early-stage AED screens. However, the original agar plate-based assay is unsuited to large-scale screening and has not been validated for identifying AEDs. Therefore, we established an alternative streamlined, higher-throughput approach whereby mutants were treated with PTZ and AEDs via liquid-based incubation.

Results: Convulsions induced within minutes of PTZ exposure in unc-49 mutants were strongly inhibited by the established AED ethosuximide. This protective activity was independent of ethosuximide's suggested target, the T-type calcium channel, as a null mutation in the worm cca-1 ortholog did not affect ethosuximide's anticonvulsant action.

Comparison with existing method: Our streamlined assay is AED-validated, feasible for higher throughput compound screens, and can facilitate insights into AED mechanisms of action.

Conclusions: Based on an epilepsy-associated genetic background, this C. elegans unc-49 model of seizure-like activity presents an ethical, higher throughput alternative to conventional rodent seizure models for initial AED screens.

Keywords: Anticonvulsant; Caenorhabditis elegans; Calcium channel; Drug screens; Epilepsy; Ethosuximide; GABA receptor; Pentylenetetrazol.

Fig. 1

Optimization of the liquid-based assay. (A) Worms were exposed to the indicated concentrations of PTZ in solution for 15 min and then the proportion of worms experiencing head-bobbing convulsions (unc-25, unc-49) or full body convulsions (unc-43) in 30 s was scored. unc-49 mutants exhibited the highest incidence of seizure-like activity and were selected for further optimization. (B and C) unc-49 mutants were exposed to the indicated concentrations of PTZ for 15 min and then both the proportion of worms exhibiting head-bobbing convulsions (B) and the number of convulsions experienced per animal (C) was measured over 30 s. (D and E) unc-49 mutants were exposed to 7 mg/ml PTZ for the indicated times and then both the proportion of worms exhibiting head-bobbing convulsions (D) and the number of convulsions experienced per animal (E) was measured over 30 s. Data shown were pooled from three independent experiments (n = 27–30 worms in total per concentration or time point).

Fig. 2

Responses to PTZ are not age-dependent in younger worms. unc-49 mutants were age-synchronised by bleaching and grown to the indicated ages before being exposed to 7 mg/ml PTZ in solution for 15 min. At this point, the proportion of worms exhibiting head-bobbing convulsions (A) and the number of convulsions experienced per animal (B) were measured over 30 s. No age-associated difference in PTZ responsiveness was seen when the proportions of animals exhibiting convulsions were scored (A); but the mean number of convulsions per worm (B) was significantly reduced in 4-day-old adults as compared to younger ages of days 1 and 2. Data shown were pooled from three independent experiments (n = 27–30 worms in total per adult age point). Statistical analysis was performed with the Kruskal-Wallis test with Dunn's correction for multiple comparisons; *, p < 0.05.

Fig. 3

Ethosuximide reduces seizure-like activity in GABAergic mutant strains. (A) Worms were cultured on agar plates containing ethosuximide or vehicle control for 2 days before incubation in a 7 mg/ml PTZ solution for 15 min. The proportion of worms experiencing head-bobbing convulsions (unc-25, unc-49) or full body convulsions (unc-43) in 30 s was then scored. The incidence of PTZ-induced seizure-like activity was significantly reduced by ethosuximide treatment in unc-25 and unc-49 mutants, but not unc-43 mutants. Wild type N2 worms did not undergo convulsions in response to PTZ. (B) Worms were cultured on agar plates containing ethosuximide or vehicle control for 2 days before incubation in Dent’s Ringer solution for 10 min. The number of thrashing movements made by each worm in 60 s was then scored. Ethosuximide significantly reduced the thrashing frequency of all mutant strains, but had no effect on wild type N2 worms. Data shown were pooled from three independent experiments (n = 30 worms in total per strain per condition), with statistical comparisons made via the Mann-Whitney test; *, p < 0.05.

Fig. 4

Optimization of liquid-based AED screening conditions using ethosuximide in unc-49 mutants. (A and B) unc-49 mutants were pre-incubated in a solution containing 2 mg/ml ethosuximide for the indicated times; or (C and D) were pre-incubated for 2 h in the indicated concentrations of ethosuximide. The worms were then exposed to 7 mg/ml PTZ for 15 min and the proportion of worms exhibiting head-bobbing convulsions (A and C) and the number of convulsions experienced per animal (B and D) were measured over 30 s. Complete abolition of convulsions was achieved with a 2-hour incubation at 4 mg/ml ethosuximide. Data shown were pooled from three independent experiments (n = 10–15 worms in total per drug treatment time-point or concentration).

Fig. 5

Identification of non-toxic DMSO levels for drug screens in unc-49 mutants. Worms were exposed to the indicated concentrations of DMSO throughout a 2-hour pre-incubation period and a 15-minute PTZ exposure period. (A) The proportion of worms exhibiting convulsions and (B) the mean number of convulsions per animal were then scored over 30 s. PTZ responsiveness was unaffected up to 1% DMSO, but higher DMSO concentrations resulted in toxicity that reduced sensitivity to PTZ. Data shown were pooled from three independent experiments (n = 10–15 worms in total per DMSO concentration).

Fig. 6

Ethosuximide does not mediate anticonvulsant effects through the T-type calcium channel, CCA-1. Three unc-49 lines harbouring a loss-of-function mutation in the T-type voltage-gated calcium channel-encoding cca-1 gene (unc-49;cca-1) were generated. These and the parent unc-49 single mutant strain were pre-incubated for 2 h in a solution containing either 4 mg/ml ethosuximide or 4 mg/ml succinimide, before exposure to PTZ for 15 min. The loss of these channels did not affect protection from seizure-like activity by ethosuximide, as the drug (A) prevented overall convulsions and (B) significantly reduced the convulsion rates in all lines when compared to unc-49 worms treated with the inert succinimide control (p < 0.05). Furthermore, anticonvulsant activity in all lines were comparable to cca-1-intact ethosuximide-treated unc-49 controls (p > 0.05). Data shown were pooled from three independent experiments, with convulsion rates compared via the Kruskal-Wallis test with Dunn’s multiple comparison correction (n = 13–15 worms in total per strain per condition); *, p < 0.05.