ENX‐101, a GABA_A receptor α2,3,5‐selective positive allosteric modulator, displays antiseizure effects in rodent seizure and epilepsy models

Abstract

Objective

γ‐Aminobutyric acid type A (GABA_A) receptor positive allosteric modulators (PAMs) that lack α‐subunit selectivity, including benzodiazepines such as diazepam, exhibit antiseizure actions in animal models and in humans. ENX‐101 is a deuterated analog of the ⍺2,3,5‐selective GABA_A receptor PAM L‐838,417. The purpose of this study was to characterize the α‐subunit selectivity of ENX‐101 and evaluate its antiseizure potential in preclinical seizure and epilepsy models.

Methods

ENX‐101 potentiation of GABA chloride current responses in cells expressing recombinant GABA_A receptors were evaluated using an automated patch clamp assay. Antiseizure effects of ENX‐101 were examined in the mouse 6 Hz test at 32 and 44 mA, amygdala kindled rats, and Genetic Absence Epilepsy Rat from Strasbourg (GAERS).

Results

ENX‐101 displayed partial PAM activity with respect to diazepam at GABA_A receptors containing α2, α3, or α5 subunits but did not enhance GABA responses of GABA_A receptors containing α1 subunits. ENX‐101 (30, 100, and 300 mg/kg, i.p.) and diazepam protected most animals in the 6 Hz model at 32 mA but was less effective at 44 mA. In amygdala kindled rats, ENX‐101 (1–100 mg/kg, p.o.) reduced behavioral seizure severity and afterdischarge duration in a dose‐dependent manner. ENX‐101 (0.075–100 mg/kg, p.o.) caused dose‐dependent, persistent (>130 min) inhibition of spontaneous spike‐and‐wave discharges (SWDs) in GAERS, whereas diazepam transiently inhibited discharges. ENX‐101 did not cause motor impairment, as measured by performance in the rotarod assay.

Significance

ENX‐101 is an α2,α3,α5‐selective GABA_A receptor PAM that has high potency and partial efficacy. The drug is highly effective in rodent seizure and epilepsy models. ENX‐101 is most potent in the GAERS model of absence epilepsy, and active in the 6 Hz model and amygdala kindled rats. These results demonstrate that a partial, subtype‐selective GABA_A receptor PAM has activity in translationally validated preclinical epilepsy screening models. Clinical evaluation of ENX‐101 as a treatment for focal and generalized epilepsies is warranted.

Key points.

• ENX‐101 is a α2,α3,α5‐selective γ‐aminobutyric acid type A (GABA_A) receptor partial positive allosteric modulator (PAM).

• ENX‐101 showed dose‐related antiseizure effects in clinically validated rodent seizure models at doses not causing motor impairment.

• ENX‐101 was most potent at suppressing spike‐and‐wave discharges (SWDs) in Genetic Absence Epilepsy Rat from Strasbourg (GAERS), a model of generalized absence epilepsy.

• ENX‐101 was also effective in the mouse 6 Hz test and amygdala kindled rats, indicating utility in treating focal‐onset seizures.

1. INTRODUCTION

γ‐Aminobutyric acid type A (GABA_A) receptors, members of the pentameric ligand‐gated ion channel superfamily, are chloride channels that serve as the main mediators of inhibitory neurotransmission in the brain. ¹ The five protein subunits that compose GABA_A receptors commonly include two α‐subunits, two β‐subunits, and a γ2‐subunit, which can exist in long (γ2L) and short (γ2S) variants. ² , ³ Genetic or acquired defects in GABA‐mediated neurotransmission are believed to underlie some forms of epilepsy. ⁴ Moreover, a variety of drugs that act on brain GABA mechanisms are useful in the acute treatment of seizures and in the prevention of seizures in people with epilepsy. ⁵ , ⁶ Among the various ways in which antiseizure medications can influence GABAergic neurotransmission, positive modulation of GABA_A receptors is a key mechanism. ⁷ , ⁸ Benzodiazepines are the most widely used GABA_A receptor positive allosteric modulators (PAMs) for the treatment of seizures and epilepsy. Benzodiazepines in clinical use today, such as 1,4‐benzodiazepines including diazepam, lorazepam, midazolam, and clonazepam, and the 1,5‐benzodiazepine clobazam, act as PAMs on GABA_A receptors containing α1, α2, α3, or α5 subunits, and can be characterized as “non‐selective.” Such nonselective GABA_A PAMs have broad‐spectrum antiseizure activity but, except for clobazam, are not used widely in the chronic treatment of epilepsy due to dose‐limiting side effects, including sedation and ataxia, and are subject to tolerance. ⁹ Genetic manipulation studies in mice have revealed that GABA_A receptors with different α‐subunits have diverse roles in mediating the effects of benzodiazepines. ⁸ Receptors containing α2, α3, or α5 subunits contribute to the anxiolytic effects of benzodiazepines, whereas α1 subunit–containing receptors are thought to be responsible primarily for the sedative and motor‐impairing effects. Receptors containing the α1 subunit were formerly thought to mediate the antiseizure effects of benzodiazepines, but recent animal studies with drugs that are selective for α2, α3, or α5 have provided evidence that GABA_A receptors containing these subunits can be targets for antiseizure compounds Therefore it is not necessary to positively modulate receptors with α1 subunits to obtain an antiseizure effect. ⁸ GABA_A receptors containing α2 subunits that mediate phasic synaptic inhibition and are expressed on excitatory neurons in the cortex and hippocampus are a particularly relevant target. ⁴ Consequently, in recent years, there has been an interest in investigating the potential of selective α2,3,5 PAMs for the treatment of epilepsy, with the possibility that such compounds would have reduced liability for sedation and ataxia due to lack of activity on receptors that contain α1 subunits.

An example of one such α2,3,5‐selective PAM is the non‐benzodiazepine L‐838,417, ¹⁰ , ¹¹ , ¹² which confers partial positive modulation at α2‐, α3‐, and α5‐containing GABA_A receptors but is an antagonist at α1 subunit‐containing receptors. ¹⁰ Benzodiazepines are considered “full” PAMs as the degree of positive modulation they confer is equal to the most efficacious agents, with diazepam being considered as a prototypical full PAM. Studies in animals suggest that partial PAMs, which have less efficacy than diazepam, may have reduced liability for tolerance. ¹³ , ¹⁴ , ¹⁵ The fact that L‐838,417 is a partial positive modulator could confer a favorable characteristic. However, due to its poor pharmacokinetic profile in preclinical models, ¹¹ L‐838,417 was not advanced into clinical trials so its side‐effect profile and liability for tolerance in humans are unknown. ENX‐101 is structurally similar to L‐838,417, except that several key hydrogen atoms are substituted with deuterium. ¹⁶ , ¹⁷ , ¹⁸ Deuterium substitution can improve the metabolic and pharmacokinetic performance of a compound while preserving its pharmacological activity. Herein, using patch clamp recording, we demonstrate that ENX‐101 is a GABA_A receptor PAM with selectivity for α2, α3, and α5, and that it possesses antiseizure activity in several animal models.

2. MATERIALS AND METHODS

2.1. Automated patch clamp assay

Human GABA_A receptor subunit combinations α1β3γ2L, α2β2γ2L, α2β3γ2L, α3β3γ2L, and α5β3γ2L were stably expressed in Chinese hamster ovary (CHO) cells that were cultured using standard techniques. GABA_A receptor chloride currents were assessed using the SynchroPatch automated platform at room temperature. To test for PAM activity, GABA was applied at its EC₂₀ (drug concentration causing 20% of maximal effect) and then GABA is applied again in the presence of the test compound, either ENX‐101 or diazepam, both in 0.2% DMSO (dimethyl sulfoxide). ENX‐101 (Figure 1) was tested at 10 concentrations (0.01 nM, 0.1 nM, 1 nM, 3 nM, 10 nM, 30 nM, 100 nM, 300 nM, 1 μM, and 10 μM) in duplicates. Diazepam was tested at five concentrations (0.01 μM, 0.1 μM, 1 μM, 10 μM, and 100 μM) in duplicates. The fold change over GABA alone was determined with the formula (I_comp/I_GABA) − 1, where I_comp is the current amplitude in the presence of the test compound and I_GABA is the current amplitude in the presence of GABA EC₂₀ alone. The compound EC₅₀ (half maximal effective concentration) is determined from a plot of the fold value vs. concentration. The % activation relative to agonist (GABA EC₂₀) was determined with the formula [(I_comp/I_GABA) − 1] × 100, where I_comp is the current amplitude in the presence of the compound and I_GABA is the current amplitude in the presence of GABA alone. E_max (maximal efficacious concentration) is the largest % activation value obtained with any compound concentration. Further details on methods are included in Supplemental Information.

FIGURE 1.

(A) Chemical structure of ENX‐101 (1,2,4‐triazolo[4,3‐b]pyridazine, 3‐(2,5‐difluorophenyl)‐7‐[1,1‐di(methyl‐d3)ethyl‐2,2,2‐d3]‐6‐[(1‐methyl‐1H‐1,2,4‐triazol‐5‐yl)methoxy] hydrate), a novel GABA_A receptor α2,3,5 subtype‐selective PAM. Trideuteromethyl groups (CD₃) are present in key metabolic hotspots. (B) Determination of the concentration dependence of the positive modulation of recombinant GABA_A receptors containing the α1, α2, α3, or α5 subunits stably expressed in heterologous cells using the SynchroPatch assay. The graph plots fold‐increases in chloride current with respect to GABA EC_5‐20 values produced by ENX‐101 at concentrations from 0.01 nM to 1 μM. Points indicate mean ± SEM. (C) SynchroPatch determination of the potency and efficacy of ENX‐101 and diazepam for positive modulation of recombinant GABA_A receptor subtypes stably expressed in heterologous cells. E_max (%) values are the maximum percent increase in chloride current with respect to GABA EC_5–20 values. EC_5–20‐% of maximal effective concentration, EC₅₀‐ half maximal effective concentration, Emax ‐ maximal effective concentration, SEM‐ standard error of mean.

2.2. 6 Hz Seizure test

All mouse experimentation was approved by the University of Washington Institutional Animal Care and Use Committee (protocol 4387‐01). Adult male CF‐1 mice (20–35 g) were obtained from Envigo (Haslett, MI). ENX‐101 was formulated in 0.5% methylcellulose 4000 (Sigma Aldrich M0512) and administered by the intraperitoneal (i.p.) route 0.25 h prior to challenge with a 6 Hz (hertz) 32 mA (milliampere) or 44 mA equivalent current for 3 s (second) delivered through corneal electrodes to elicit a psychomotor seizure characterized by unilateral or bilateral forelimb clonus, vibrissae twitching, and Straub tail. For the 6 Hz test, the i.p. route of administration was chosen deliberately to ensure sufficient exposure for efficacy in the mouse 6 Hz model due to prior pharmacokinetic studies with the parent compound demonstrating that i.p. administration in mice gives superior and desired exposure profile as compared to oral administration. ¹¹ The experimenter was blinded to the treatment condition. Mice not displaying all these behaviors within the immediate (5–10 s) period after stimulation were considered “protected.” Ten mice were tested in each treatment group. Diazepam was dosed at 1 mg/kg, i.p., as supplied from Hospira. Further details on methods are included in Supplemental Information.

2.3. Rotarod tests

Rotarod test of minimal motor impairment (MMI) in mice was established by mouse performance on a fixed‐speed rotarod, which was assessed at the University of Washington in all mice immediately prior to the 6 Hz seizure stimulation for each current intensity. ¹⁹ When a normal mouse is placed on a rod that rotates at a speed of 6 rpm, the animal can maintain its equilibrium for long periods of time. If a mouse falls off the rotarod three times during the 60 s trial period, it is considered “impaired.” Each time a mouse falls off the rotarod in the 60 s trial period, it is immediately returned to the rotarod until the 60 s trial period has elapsed. Evaluation on the rotarod is a binary outcome measure (Yes/No fall) and scoring is conducted by an experimenter who is blinded to treatment condition. The number of mice that exhibit minimal motor impairment on the rotarod assay is reported, with a total of n = 10 mice per treatment group from each current intensity in the 6 Hz study (Table 1).

TABLE 1.

Comparison of ENX‐101 and diazepam in the mouse 6 Hz seizure test.

Treatment	Dose (mg/kg, i.p.)	Seizure protection		Rotarod impairment
		32 mA	44 mA	32 mA	44 mA
Vehicle		0/10	0/10	0/10	0/10
Diazepam	1	10/10	3/10	0/10	0/10
ENX‐101	30	8/10	3/10	0/10	0/10
ENX‐101	100	10/10	6/10	0/10	0/10
ENX‐101	300	9/10	7/10	0/10	0/10

Note: Each treatment was tested in separate groups of 10 mice with stimulation intensities of 32 and 44 mA. Seizure protection column entries indicate the number of mice protected (out of a total of 10 mice tested) under the different treatment conditions: vehicle, diazepam (1 mg/kg), or ENX‐101 (30, 100, or 300 mg/kg). Minimal motor impairment in mice was established by performance on a fixed‐speed rotarod immediately prior to in‐life seizure testing. Each mouse was considered impaired if it fell off the rotarod 3 times during a 1‐min period. i.p.‐ intraperitoneal, mA‐ milliampere, min‐ minute, mg/kg‐ milligram per kilogram.

Performance of rats in the rotarod assay was examined 30 min after oral administration of ENX‐101. During the training session, rats were placed on the rotarod, with the rotation speed accelerating from 4 r/m (rotations/minute) to 40 r/m within 5 min. Rats were trained twice a day for 3 days. The latency to fall off the rotarod in the last trial was recorded and considered as baseline. On the fourth day, during the testing phase, rats were randomly assigned to test groups and 30 min after drug administration, placed on the rotarod with the same settings as that of the training. The duration of time that each rat remained on the rotarod was recorded. Dosing and rotarod measurement were conducted by two scientists separately. Pentobarbital was used as a positive control for assay sensitivity. Data were analyzed using one‐way analyses of variance (ANOVAs) with post hoc analysis using Tukey's test and significance considered at p < 0.05.

2.4. Amygdala kindled rat seizure model

Male Wistar rats (Janvier Labs) were anesthetized with isoflurane and injected with carprofen (5 mg/kg, s.c., subcutaneous). This work was performed at Porsolt Research Laboratory in compliance with animal health regulations (Council Directive No. 2010/63/UE of Sept. 22, 2010 on the protection of animals used for scientific purposes and French decree No. 2013‐118) and in accordance with the Porsolt facility accreditation for experimentation (E 53 1031) and recommendations of the Association for Assessment and Accreditation of Laboratory Animal Care. Screw electrodes were placed over the fronto‐parietal cortex, a depth electrode was placed in the basolateral amygdala, a screw electrode was placed over the right occipital cortex to serve as ground, and an anchor screw was placed over the left occipital cortex. The animals were allowed at least 10 days to recover from surgery. Rats were stimulated (A‐M Systems stimulator, Model 2100) via the amygdala electrode twice daily, 5 days per week until they displayed four consecutive seizures with scores of 4 or 5 on the Racine scale. ²⁰ Each rat received the four treatments according to an incomplete Latin Square design, by investigators blinded to the test substance (n = 12–16 rats per treatment group, except 100 mg/kg group with four rats). Testing of each treatment required 2 days, with a washout time of 2 days between treatments. On day 1 (baseline session) the animal received the vehicle orally and was stimulated 60 or 120 min later. On day 2 (test session) ~24 h after the baseline session, the test substance was administered orally, and the animal was stimulated 60 or 120 min later. After each stimulation, the seizure behavior was scored using the Racine scale (0 to 5 point scale). ²⁰ This work was performed at Porsolt Research Laboratory (France) in accordance with ethical standards of approved animal protocols. In addition, cortical and amygdala electroencephalography (EEG) recordings were collected from the amygdala and cortex electrodes and were analyzed off‐line to determine the afterdischarge duration. Mean seizure severity scores in the baseline and test sessions were compared with the Wilcoxon test; mean afterdischarge durations were compared using the paired t test. Vehicle used was 0.5% methylcellulose solution. Further details on methods are included in Supplemental Information.

2.5. GAERS model

Fourteen adult male rats (age 4–8 months) of the Genetic Absence Epilepsy Rat from Strasbourg (GAERS) strain were obtained from Dr. Antoine Depaulis (INSERM, Grenoble Institute of Neurosciences, Grenoble, France) under exclusive license to SynapCell. This experiment was performed at SynapCell's facility, approved by the ethical committee of the Grenoble Institute of Neurosciences, University Grenoble Alpes, France, and performed in accordance with the European Committee Council directive of September 22, 2010 (2010/63/EU; Facility accreditation number‐ A 38 397 10 001). All efforts were made to minimize animal suffering and reduce the number of animals used. EEG recordings were obtained from right and left, frontal and parietal cortices of freely moving animals using SystemPlus Evolution software (Micromed). One week after electrode implantation, EEG was recorded for 1 h and 12 rats that demonstrated ≥20 spike‐and‐wave discharges (SWDs) during the recording were selected for the study. A crossover design was used in which each animal received treatments in a randomized fashion including vehicle (0.5% methylcellulose), diazepam (3 mg/kg), and doses of ENX‐101 from 0.075 mg/kg to 100 mg/kg on different days by oral gavage (n = 4–10 rats/dose group). SWDs were counted during a 20 min baseline period (from 25 to 5 min before treatment administration), and for successive 20 min epochs up to 130 min starting 10 min after administration. The initial 10 min was ignored because dosing disturbs the occurrence of SWDs. When required, animals were returned to a state of quiet wakefulness by gentle stimulation. SWD counts per epoch were expressed as mean ± standard error of the mean (SEM). Statistical analysis was performed by ANOVA using a mixed model with repeated measures, followed by paired comparisons (Dunnett) vs. vehicle (Graphpad Prism v9.1). ED₅₀ (median effective dose) values were calculated using the dose–response curve fitting of Prism v9.1, with a variable slope model and an R‐squared method. Further details on methods included in Supplemental Information.

2.6. Plasma ENX‐101 collection and assay

Following completion of in‐life testing, blood was collected from a subset of animals (n = 3/dose group) post testing in the 6 Hz, amygdala kindled rats, and GAERS model (n = 2–3/dose group) to determine plasma ENX‐101 concentrations. Whole blood samples were collected 0.25 h after dosing in mice in the 6 Hz study and 2–2.5 h after dosing in the amygdala kindled rats and GAERS experiments. Blood samples were either from the trunk following decapitation or through retroorbital sinus puncture, in accordance with institutionally approved methods, and isolated plasma was frozen and stored at −80°C (with anticoagulant K2EDTA (dipotassium ethylenediamine tetraacetic acid) until assayed by LC/MS/MS (liquid chromatography tandem mass spectrometry) using an internal standard (IS). After addition of the IS, plasma samples were extracted by protein precipitation using acetonitrile. Reversed‐phase HPLC (high performance liquid chromatography) separation was achieved with a Phenomenex, Synergi Max‐RP column (50 × 2.0 mm, 4 micron). MS/MS (tandem mass spectrometry) detection was conducted in Thermo‐Ion Spray (TIS) positive mode. Quantitation was based on linear regression of analyte/IS area ratio vs. a weighting factor. The dynamic range of detection for both the mouse and rat methods was 5.00–5000 ng/mL for ENX‐101. Further details on methods are included in Supplemental Information.

3. RESULTS

3.1. ENX‐101 is a selective GABA_A receptor α2,3,5 PAM

To demonstrate the positive modulatory activity and selectivity of ENX‐101 on GABA_A receptors with different α‐subunits, single cell electrophysiological studies using an automated patch clamp system applied to cells that expressed recombinant GABA_A receptors was conducted. ENX‐101 (Figure 1A) potently enhanced currents generated by GABA in receptors containing α2, α3, and α5 subunits but caused negligible potentiation of α1 subunit–containing receptors (Figure 1B). In contrast, diazepam was more efficacious on α2, α3, and α5 subunit–containing receptors and was an effective PAM of α1 subunit–containing receptors. Despite its greater functional activity (E_max values), diazepam had weaker potency (i.e., exhibited larger EC₅₀ values) than ENX‐101 on receptors with α2, α3, and α5 subunits (Figure 1C). Thus, on α2, α3, and α5 subunit–containing receptors, ENX‐101 displayed potent but partial PAM activity relative to diazepam.

3.2. Evaluation of ENX‐101 in the mouse 6 Hz seizure test and rotarod test

ENX‐101 was evaluated in the mouse 6 Hz model at stimulation intensities of 32 mA and 44 mA, and compared to parallel evaluation of diazepam, which is known to be effective in this test, ²¹ as a positive control. As shown in Table 1, all vehicle‐pretreated animals at either stimulation intensity exhibited seizures. Diazepam pretreatment protected all mice stimulated with the lower intensity (32 mA) current from seizures but only partially protected the group stimulated with the higher intensity (44 mA) current. The reduced protection conferred by diazepam with higher intensity current is consistent with the results of a prior study in the same mouse strain. ²² Similarly, ENX‐101 protected most animals stimulated at 32 mA but fewer animals at 44 mA. It is noteworthy that ENX‐101 at 30–100 mg/kg is equally or more effective than diazepam at 1 mg/kg. It is important to note that none of the mice receiving ENX‐101 exhibited impairment in the rotarod test at any of the doses tested (Table 1).

3.3. Antiseizure effects of ENX‐101 in the amygdala kindled seizure model

ENX‐101 was evaluated for its ability to inhibit behavioral seizures and EEG afterdischarge in amygdala kindled rats. Fully kindled rats were administered either vehicle, ENX‐101 at four escalating test doses (1, 6, 30, and 100 mg/kg), or diazepam (16 mg/kg) as a positive control. With vehicle treatment (baseline), stage 5 seizures were observed in all animals (Figure 2A). Diazepam (16 mg/kg), administered orally 60 min prior to the electrical stimulation, substantially reduced the mean severity seizure score (−3.7, p < .001, Figure 2A), and mean afterdischarge duration in the amygdala (−69%, p < .001, Figure 2B) and the cortex (−75%, p < .001, Figure 2C). ENX‐101, administered 2 h prior to electrical stimulation, also significantly reduced the mean seizure severity score and the mean afterdischarge duration in both the amygdala and the cortex. At doses of 1, 6, and 30 mg/kg, ENX‐101 significantly reduced the mean seizure severity score by −1.8 (p < .01), −3.2 (p < .001), and −3.8 (p < .01), respectively (Figure 2A). Compared to baseline, ENX‐101 at 1, 6, and 30 mg/kg also reduced mean afterdischarge duration in the amygdala by −26% (p < .01), −46% (p < .01), and −64% (p < .001), respectively (Figure 2B); and the afterdischarge duration in the cortex (−30%, p < .01; −53%, p < .01; and −66%, p < .001) (Figure 2C). The 100 mg/kg ENX‐101 group showed the greatest reductions in seizure severity score (−4.0) and afterdischarge duration in the cortex (−76%, p < .05) and in the amygdala (−78%) but because there were fewer animals in the group it was not possible to demonstrate statistical significance with respect to baseline for two of the measures. Overall, ENX‐101 produced a dose‐dependent and significant inhibition of the behavioral seizure response and the EEG afterdischarge in the amygdala and the cortex over the dose range of 1 to 100 mg/kg. The maximal effect obtained with ENX‐101 was comparable to that produced by diazepam at 16 mg/kg. Average plasma exposures 2 h after dosing in a separate set of rats treated with 1, 6, and 30 mg/kg oral doses of ENX‐101 were 191, 487, and 1102 ng/mL (mean, n = 3 rats per dose), respectively.

FIGURE 2.

Comparison of oral ENX‐101 and diazepam in amygdala kindled rats. ENX‐101 was tested in kindled rats at doses of 1, 6, 30, and 100 mg/kg, orally. Diazepam was tested at a dose of 16 mg/kg, orally. (A) Prior to drug treatment, all animals exhibited stage 5 seizures. ENX‐101 treatment caused a dose‐dependent reduction in the seizure severity score, with maximal reduction comparable to that produced by diazepam. (B, C) ENX‐101 treatment caused dose‐dependent reductions in afterdischarge duration as assessed by the EEG recorded from depth electrodes in the amygdala (B) and cortical surface recording (C). The number of animals in each group ranged from 12 to 16 except in the 100 mg/kg ENX‐101 group, which consisted of four animals. Bars indicate mean ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001. SEM‐ standard error of mean.

3.4. Antiseizure effects of ENX‐101 in the Genetic Absence Epilepsy Rat from Strasbourg (GAERS)

ENX‐101 and diazepam were assessed for their ability to reduce the occurrence of spontaneous SWDs in GAERS. ENX‐101 was evaluated at six oral doses of 0.075 mg/kg, 0.75 mg/kg, 2 mg/kg, 6 mg/kg, 30 mg/kg, 100 mg/kg; diazepam was evaluated at 3 mg/kg oral dose. As shown in Figure 3A, following treatment with the lowest tested ENX‐101 dose of 0.075 mg/kg, there was a numerical decrease in the mean SWD counts in each epoch compared to the vehicle group, but the reductions were not statistically significant. However, the overall SWD rate during the post‐treatment observation period (10–130 min) was significantly reduced even at this low dose (Figure 3B). At doses of 0.75, 2, 6, 30, and 100 mg/kg, ENX‐101 significantly reduced the SWD counts per epoch and the overall SWD rate when compared to the vehicle group values. The reduction was observed even in the first epoch after dosing (10–30 min). For each of these doses there was a monotonic reduction in mean SWD count in successive epochs with complete suppression of the discharges occurring more rapidly at the higher doses. Once suppressed there was no recovery of activity during the observation period. A dose–response curve constructed from the overall SWD rate data in Figure 3B provided a calculated ED₅₀ value for ENX‐101 of 0.10 mg/kg (95% confidence interval [CI] 0.05–0.21).

FIGURE 3.

Comparison of the effects of ENX‐101 and diazepam administered by oral gavage on SWDs in GAERS. (A) Time course of the effect of the treatments on SWDs counts in successive 20 min epochs beginning 10 min after treatment administration. (B) SWDs during the entire post‐treatment observation period from 10 to 130 min after treatment administration expressed as a percent of the baseline rate for all groups. Each treatment group consisted of 4 to 10 rats. *p < 0.05, ***p < 0.001, ****p < 0.0001. Error bars show SEM (standard error of mean).

Diazepam at 3 mg/kg caused a rapid and almost complete elimination of SWDs in the first epoch (10–30 min) but, the mean number of discharges increased progressively during the remainder of the observation period, in contrast to EN‐X‐101 where the suppression was persistent. This indicates an initial high efficacy of diazepam, followed by a “wearing off.”

3.5. ENX‐101 did not impact rat rotarod performance

ENX‐101 was administered orally to rats at four dose levels, 10, 30, 60, 100 mg/kg, prior to testing performance in a rotarod assay (Figure 4). Duration of time that the rat remained on an accelerating rotarod was recorded. ENX‐101‐treated animals, at all doses tested, remained on the rotarod for a similar amount of time as compared to vehicle‐treated animals. These data indicate that the ENX‐101 did not cause motor impairments at anti‐seizure doses, including the highest tested dose of 100 mg/kg.

FIGURE 4.

Effect of ENX‐101 on Rat rotarod performance. Impact of ENX‐101 (10, 30, 60, and 100 mg/kg, p.o.) on rotarod performance was tested in rats. During the training period, rats were placed on the rotarod (rotation speed accelerating from 4 rotations per minute (r/m) to 40 r/m within 5 min and trained twice a day for 3 days). Latency to fall off the rotarod was recorded. On the fourth day, rats were placed on the rotarod with the same setting of training, 30 min after drug dosing. Duration of time that the rat stayed on the rotarod was recorded. Data was analyzed using one‐way ANOVA with Tukey post hoc test. Significance was considered at *p < 0.05. Pentobarbital was administered i.p. at 15 mg/kg as a positive control for this assay. n = 12 rats per treatment group.

3.6. ENX‐101 pharmacokinetic–pharmacodynamic relationships

Blood was collected from selected mice in the 6 Hz test experiments and rats in the kindled and GAERS experiments after the completion of testing to allow assessment of the plasma levels associated with antiseizure activity. Figure 5A plots the extent of protection in the 6 Hz test at the two stimulation intensities, with the plasma levels of ENX‐101 achieved in mice receiving ENX‐101 at doses of 30, 100, and 300 mg/kg. At the lower stimulation intensity, nearly all animals were protected, so no concentration dependence was discernible. At the higher stimulation intensity, only a portion of the animals were protected and there is a clear concentration dependence to the extent of protection. With the 32 mA stimulation intensity, it is apparent that robust seizure protection is associated with plasma concentration of 100 ng/mL to 1000 ng/mL in mice.

FIGURE 5.

Pharmacokinetics and pharmacodynamic effects of ENX‐101 in mouse and rat models of seizure. (A) Pharmacokinetic–pharmacodynamic analysis of the results in the mouse 6 Hz seizure test. The mean ± SEM ENX‐101 plasma concentration values for the three doses of ENX‐101 (30, 100, and 300 mg/kg) at 0.25 h post‐dose are plotted against the efficacy values representing the fraction of mice protected in groups subjected to 32 mA (red circles) and 44 mA (blue squares) stimulation. (B) Plasma concentrations of ENX‐101 following oral administration in rats at doses of 0.075, 0.75, 1, 2, 6, and 30 mg/kg. Closed symbols are taken from the kindled rats and were collected 2 h after dosing (n = 3 rats/dose). Open symbols are from the GAERS experiment and were taken 0.5 h after the end of the last EEG recording of the study, about 2.5 h after dosing (three rats per dose, except the 0.75 mg dose, which was two rats). Symbols and error bars indicate mean ± SEM. Where error bars are not visible, they are smaller than the size of the symbols.

Figure 5B plots the plasma concentration values obtained in amygdala kindled rats and GAERS experiments. Efficacy in the amygdala kindled rats was associated with plasma ENX‐101 concentrations in the same range as those that were effective in mice in the 6 Hz test with 32 mA stimulation. In contrast, efficacy in GAERS occurs with much lower plasma exposures (< 200 ng/mL), suggesting that substantially less receptor occupancy with ENX‐101 may be required to treat absence seizures.

4. DISCUSSION

In the present study, the antiseizure activity of ENX‐101 was characterized in a battery of well‐established rodent seizure and epilepsy models. The results indicate that ENX‐101 has robust antiseizure activity when administered parenterally and orally. The rodent models selected for the present study represent some of the most frequently used for discovery of antiseizure medication, and for which a high degree of predictive validity exists. The 6 Hz test is a well‐characterized mouse model of focal‐onset seizures that demonstrates not only pharmacoresistance at high stimulation intensities but also provides a valuable degree of differentiation between antiseizure medication standards of care. ²² , ²³ The amygdala kindled rat represents a validated rat model of secondarily generalized focal onset seizures that can identify clinically beneficial therapies (i.e., levetiracetam) and is also sensitive to benzodiazepines. ²³ , ²⁴ Finally, the GAERS model is highly reflective of SWDs associated with absence epilepsy, and informative for identifying compounds that may reduce or aggravate SWDs. ²⁵ Therefore, these models were selected to characterize and differentiate the preclinical profile of ENX‐101 and provide insight for predicting clinical potential.

At the highest dose tested (100mg/kg), ENX‐101 did not cause motor impairment as assessed with the rotarod assays, in both mice and rats (Figure 4, Table 1). Based on patch clamp recordings, in cells expressing recombinant GABA_A receptors with various α subunits, ENX‐101 displayed selective GABA_A receptor PAM activity, with partial functional activity at GABA_A receptor subtypes containing α2, α3, or α5 subunits but devoid of activity at receptors with the α1 subunit. The pharmacological profile of ENX‐101 differs substantially from that of diazepam, which is a fully active PAM at α1‐containing GABA_A receptors. The in vitro potencies of ENX‐101 at the various α subunit–containing receptors is 10‐ to 20‐fold greater than that of diazepam but the functional activity is substantially less, with diazepam maximally potentiating GABA_A receptor chloride currents as much as 5‐fold greater than ENX‐101.

There has been considerable interest in the potential of subtype‐selective GABA_A receptor PAMs, as well as partial agonists, for the treatment of seizures and epilepsy. High‐affinity ligands of the benzodiazepine binding site with partial efficacy, such as the beta‐carboline derivative abecarnil (ZK 112119), considered to be “anxioselective” due to anxiolytic‐like effects, also showed anticonvulsant effects in number of chemically and electrically induced seizure models, in both rodents and primates. ²⁶ It has long been recognized that GABA_A receptors containing the α1‐subunit contribute to the antiseizure actions of nonselective GABA_A receptor PAMs. ²⁷ However, α2 subunit–containing receptors have also been demonstrated to play a role. ²⁸ More recently, compounds with α2, α3 subunit selectivity, such as the imidazopyridazine derivative darigabat, have been reported to exhibit antiseizure effects in animal models such as the pentylenetetrazol seizure test and amygdala kindled rats. ²⁹ , ³⁰ The α2, α3‐selective PAM BAER‐101 (AZD7325), an N‐substituted cinnoline, has been reported to have activity in a mouse model of Dravet syndrome and in GAERS. ³¹ , ³² A large body of data is also available for the imidazodiazepine KRM‐II‐81, which is an α2,α3‐selective PAM. ³³ , ³⁴ , ³⁵ , ³⁶ , ³⁷ Surprisingly, KRM‐II‐81 demonstrated superior efficacy to diazepam in some models. ³⁶

Relative to diazepam, darigabat exhibits greater potency at α2 and α3 subunits, whereas KRM‐II‐81 has lower potency on these receptor configurations. ³⁸ It is important to note, however, that KRM‐II‐81 has comparable functional activity to diazepam on GABA_A receptors containing these subunits, whereas darigabat has lower functional activity than diazepam. In contrast, ENX‐101 shows lower functional activity but greater potency compared to diazepam, and displays robust antiseizure activity. Therefore, partial PAM activity is sufficient to confer antiseizure efficacy in animal seizure and epilepsy models relevant to absence seizures and focal‐onset seizures. BAER‐101 (AZD7325) has also been reported to be a partial α2, α3 PAM. ³² , ³⁹ , ⁴⁰ BAER‐101 has recently been reported to have efficacy in GAERS, consistent with the idea that partial PAM activity is sufficient to confer antiseizure efficacy, at least for SWD seizures. ³² A liability of GABA_A receptor PAMs for the acute treatment of seizures and the chronic treatment of epilepsy is the development of refractoriness or tolerance. There is evidence that partial GABA_A receptor PAMs may have reduced liability for tolerance. ⁴ It will be of interest to determine if ENX‐101 has reduced tolerance compared with PAMs that act in a benzodiazepine‐like fashion on GABA_A receptors and reduced tolerance in relation to subtype‐selective compounds such as darigabat and KRM‐II‐81.

Despite its partial PAM activity at the GABA_A receptor and lack of activity at α1 subunit–containing receptors, ENX‐101 demonstrated robust antiseizure activity in the various models tested. ENX‐101 was effective in the mouse 6 Hz model with stimulation intensity of 32 mA, and higher doses showed antiseizure efficacy with the higher 44 mA stimulation intensity. Notably, no signs of motor impairment in the rotarod were observed at any of the doses tested in mice. Diazepam (1 mg/kg, i.p.) protected all animals tested with the stimulation intensity of 32 mA, but only 3 of 10 animals at 44 mA. Most antiseizure drugs that are active in the 6 Hz test exhibit greater potency with 32 mA stimulation than with 44 mA stimulation. ²² Benzodiazepines, including diazepam and clonazepam, exhibit this property. ²¹ , ²² Thus, the selectivity of ENX‐101 does not cause it to act differently from established drugs in the 6 Hz test.

In kindled rats, ENX‐101 produced dose‐dependent (1–100 mg/kg, oral) inhibition of behavioral seizures as assessed with the Racine scale. There was a corresponding attenuation of the afterdischarge duration. Benzodiazepines are well recognized to inhibit behavioral seizures and the afterdischarge duration in amygdala kindled rats. ⁴¹ , ⁴² We confirmed this effect with diazepam at an oral dose of 16 mg/kg. The magnitude of the antiseizure action of ENX‐101 in amygdala kindled rats was similar to that obtained with diazepam. Given that higher doses of ENX‐101 reduced seizure severity to under stage 2 but did not fully block it, it appears that ENX‐101, like diazepam, does not completely block focal seizures but may substantially reduce secondary or generalized seizures. Darigabat has been reported to have antiseizure effects in the kindled rats, but insufficient data were provided to allow a direct comparison. ²⁹

ENX‐101 effectively reduced SWDs in GAERS, which is a model of absence epilepsy. Seizures in the GAERS model are highly sensitive to diazepam, ⁴³ , ⁴⁴ and also to the α2, α3 selective compounds BAER‐101 and darigabat. ³² , ⁴⁴ Although diazepam cannot be used in the prolonged treatment of absence seizures in humans due to the rapid development of tolerance, the benzodiazepine clonazepam is approved by the U.S. Food and Drug Administration (FDA) for that indication but is used only rarely because of side effects, such as drowsiness and the liability for tolerance. ⁴⁵ ENX‐101 was effective in GAERS at non‐sedating doses that were associated with plasma exposures substantially lower than those that were required in the other seizure paradigms. The remarkable efficacy of these GABA_A receptor PAMs in GAERS is not fully understood. Thalamocortical mechanisms are well recognized to play a role in absence seizure in GAERS and other absence epilepsy models. ⁴⁶ Dysfunction of GABAergic neurons in the nucleus reticularis thalami (nRT) that project to thalamic nuclei, may lead to SWDs, including in GAERS. ⁴⁷ More specifically, it has been demonstrated that cortico‐thalamic and nRT neurons are hyperexcitable in the GAERS. ⁴ , ⁴⁷ , ⁴⁸ , ⁴⁹ GABA_A receptors containing α3 subunits are highly enriched in the nRT. ⁵⁰ GABA_A receptors containing α3 subunits are therefore positioned to suppress the pathological hyperexcitability of these neurons, which could be an important factor in the therapeutic activity of α3‐targeting PAMs. Thus, although the α2‐driven activity of α2,α3,α5‐selective PAMs such as ENX‐101 may exceed α3 driven activity in treating focal‐onset seizures, it can be speculated that α3 PAM activity may be of particular significance in the anti‐absence seizure activity in GAERS. Support for this concept comes from studies in mice showing that a point mutation in α3 renders the receptors benzodiazepine insensitive and benzodiazepines fail to suppress absence seizures in these anaimals. ⁵¹ Consistent with the role of α3‐subunit containing GABA_A receptors in regulating absence seizures is the observation that certain individuals with rare loss‐of‐function variants in the GABRA3 gene encoding the α3 subunit exhibit absence seizures. ⁵² Loss‐of‐function gene variants in humans GABRA2 and GABRA5, encoding α2 and α5 subunits, respectively, have been associated with non‐absence seizures. ⁵³ It will be of interest to determine if ENX‐101 is an effective treatment for such genetic epilepsies.

5. CONCLUSION

The pharmacological studies presented herein show that ENX‐101 is a partial GABA_A receptor PAM, active at receptors containing α2, α3, and α5 subunits but devoid of activity at receptors with the α1 subunit. ENX‐101 displayed efficacy in multiple rodent seizure and epilepsy models with predictive validity for focal‐onset seizures, and was particularly efficacious in GAERS, a validated model of absence epilepsy, the most common idiopathic generalized epilepsy. These results support the evaluation of ENX‐101 as an antiseizure treatment in patients with focal and generalized epilepsies.

CONFLICT OF INTEREST STATEMENT

J.S., K.C.V., W.B., E.T., K.E.V., S.C., and V.S. are/were employees of Engrail Therapeutics at the time of this research and may own stock and/or stock options in the company. A.E. and C.R. are employees of SynapCell SAS. M.B.H. has no conflicts to disclose. H.S.W. has served on the Scientific Advisory Board of Otsuka Pharmaceuticals; has served as an advisor to Biogen Pharmaceuticals, JAZZ Pharmaceuticals, Neurelis, Inc., and Takeda Pharmaceuticals; is a member of the UCB Pharma Speakers Bureau; and is a scientific co‐founder of NeuroAdjuvants, Inc., Salt Lake City, UT. M.A.R. has served as an advisor to Engrail Therapeutics but received no compensation in relation to this work. We confirm that we have read the Journal's position on issues involved in ethical publication and affirm that this report is consistent with those guidelines.

Supporting information

Data S1.

Serrats J, Vadodaria KC, Brubaker W, Barker‐Haliski M, White HS, Evrard A, et al. ENX‐101, a GABA_A receptor α2,3,5‐selective positive allosteric modulator, displays antiseizure effects in rodent seizure and epilepsy models. Epilepsia. 2025;66:2124–2136. 10.1111/epi.18340

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the corresponding author upon reasonable request.