Abstract
γ-Aminobutyric acid (GABA) is the principal inhibitory neurotransmitter in the adult brain which mediates its rapid effects on neuronal excitability via ionotropic GABAA receptors. GABA levels in the brain are critically dependent upon GABA-aminotransferase (GABA-AT) which promotes its degradation. Vigabatrin, a low-affinity GABA-AT inhibitor, exhibits anticonvulsant efficacy, but its use is limited due to cumulative ocular toxicity. OV329 is a rationally designed, next-generation GABA-AT inhibitor with enhanced potency. We demonstrate that sustained exposure to OV329 in mice reduces GABA-AT activity and subsequently elevates GABA levels in the brain. Parallel increases in the efficacy of GABAergic inhibition were evident, together with elevations in electroencephalographic delta power. Consistent with this, OV329 exposure reduced the severity of status epilepticus and the development of benzodiazepine refractory seizures. Thus, OV329 may be of utility in treating seizure disorders and associated pathologies that result from neuronal hyperexcitability.
Significance Statement
Enhancing inhibitory control over neurons to reduce excitability is a common strategy in treating seizure disorders. Here, we describe a novel compound, OV329, which acts on a common pathway to vigabatrin to increase inhibitory signaling following a low repeated dose paradigm. In vivo application of OV329 exhibited enhanced tonic γ-aminobutyric acid signaling in mice at the synaptic level in the hippocampus and at the network level reduced seizure severity and the development of benzodiazepine refractory seizures. This suggests OV329 may be of clinical use in the treatment of seizure disorders.
Introduction
Ionotropic GABAA receptors (GABAARs) are the primary mediators of inhibition in the central nervous system and shape neural function by constraining the excitability of neurons. These ligand-gated ion channels mediate Cl−-dependent neuronal hyperpolarization, the unitary events that underpin fast synaptic or phasic inhibitory neurotransmission and sustained or tonic inhibition (Brickley and Mody, 2012; Braat and Kooy, 2015). Consistent with their essential roles in limiting neuronal excitability, deficits in fast inhibitory neurotransmission result in epilepsy and associated pathologies arising from aberrant neuronal hyperexcitability (Brickley and Mody, 2012; Braat and Kooy, 2015).
The efficacy of GABAergic transmission is critically dependent upon the rates of synthesis and degradation of γ-aminobutyric acid (GABA). GABA is catabolized to succinic semialdehyde by GABA-aminotransferase (GABA-AT), an enzyme expressed at high levels by both neurons and glia. Accordingly, inhibitors of GABA-AT, such as vigabatrin (VGB), are used to treat infantile spasms and focal seizures with impaired awareness in adults (French, 1999; Wheless et al., 2007). However, the therapeutic use of VGB is severely limited due to retinal degeneration (Wild et al., 2009; Maguire et al., 2010; Yang et al., 2012; Foroozan, 2018). The mechanisms underlying this toxicity may result from the low potency of VGB and its irreversible effects on GABA-AT activity and possible off-target effects that include inhibition of GABA uptake mechanisms (Schousboe et al., 1986; Eckstein-Ludwig et al., 1999; Walters et al., 2019).
Recently, a rationally designed mechanism-based inactivator of GABA-AT, OV329, has been synthesized (Juncosa et al., 2018). This compound exhibits a marked increase (200–1,000-fold) in potency for GABA-AT compared with VGB (Juncosa et al., 2018; Shen et al., 2020; Feja et al., 2021). OV329 itself is an unreactive compound that undergoes activation when used as a substrate by the GABA-AT enzyme; the metabolite of OV329 inhibits enzymatic function by forming a noncovalent tight-binding complex with a residue key to the catalytic function of GABA-AT (Shen et al., 2020; Weerawarna et al., 2021). In this study, we examined the effects of sustained exposure of mice to OV329 on the efficacy of GABAergic inhibition, on the development of benzodiazepine refractory status epileptics (SE), and on basal electroencephalographic (EEG) power. Sustained exposure of mice to OV329 reduced GABA-AT activity, increased steady-state GABA levels in the brain, and induced elevations in phasic and tonic inhibition. These modifications paralleled a significant increase in EEG power selectively in the delta frequency. OV329 reduced the severity of kainate-induced SE and prevented the development of benzodiazepine refractory seizures. Collectively, our results demonstrate that OV329 acts on GABA-AT to increase GABA accumulation in the brain, elevate GABAergic inhibition, and exert potent anticonvulsant efficacy.
Materials and Methods
Animals and drug treatments
Five- to seven-week-old male C57Bl6/J mice (group housed and kept on regular day/night cycle) were weighed and labeled with tail markings for identification purposes and received up to 6 d of repeated i.p. injections with volumes of 0.01 ml/g of vehicle (pH 7.4 PBS) or 5 mg/kg OV329 (sustained exposure/repeated dose), or a single injection of 5 mg/kg OV329 at the same volume. Cohorts of cohoused mice received i.p. injections between 12:00 and 2:00 P.M. Injections were staggered so that individual mice within the same cohort finished their final scheduled injection on successive days. Upon completing their injection schedule, mice were killed for experiments the following day within a 24 h postfinal injection window.
GABA-AT assay
GABA-AT assay kit was procured from Biomedical Research Service, SUNY Buffalo, catalog #E-134. The assay was done following the manufacturers’ protocol using a whole-brain homogenate. Briefly, net tissue weights were noted accurately after tissue collection, and the volume of 1× lysis buffer (1:3 ratio) was determined according to the tissue weight. A handheld tissue homogenizer was used to grind the tissue, and to avoid heating during this step, samples were kept on ice. One hour after homogenization, samples were centrifuged at 15,000× g for 20 min. Supernatant was collected and stored immediately at −80°C for the enzyme assay. Protein concentrations were determined using the BCA assay kit per manufacturer's recommendations (Thermo Fisher Scientific/Pierce, catalog #23225). The enzyme assay was carried out in a 96-well optical bottom, black plate (Thermo Fisher Scientific, catalog #265301) following the supplier's protocol. Each sample had control and reaction wells which contained either water or GABA-AT substrate. Twenty micrograms of protein was used for this assay, and each sample was run in duplicate. Upon addition of the assay buffer, plates were incubated at 37°C for 60 min. To protect from light, plates were covered with aluminum foil and incubated at 37°C for another 60 min. Absorbance at 492 nm was measured using a plate reader (GloMax Discover, Promega). For each sample, the optical density (OD) value of the control well was subtracted from the reaction well, and the Δ-OD values were used to calculate the enzyme activity. GABA-AT activity was presented as % control, as vehicle-treated samples were considered as 100% and OV329-treated samples were normalized to control.
Western blotting
Western blotting was done following the standard procedure using the Invitrogen NuPAGE system. Twenty micrograms of total protein was loaded in each well and transferred to a PVDF membrane and probed with rabbit monoclonal GABA-AT (1 µg/ml, Abcam, catalog #ab216465) and b-actin (Cell Signaling Technology, catalog #4970, 1:2,000) antibodies (Hines et al., 2013; Nathanson et al., 2019; Lee et al., 2022). Membranes were probed with goat anti-rabbit secondary antibodies (Jackson ImmunoResearch) and subsequently developed with an enhanced chemiluminescence kit (Pierce, Thermo Fisher scientific). Images were captured using a Bio-Rad image analyzer and quantified using ImageJ/Fiji software.
GABA measurement by liquid chromatography/mass spectrometry (LC/MS/MS)
GABA concentration was measured using whole-brain homogenates prepared for the GABA-AT activity assay. Twenty-five microliters of tissue homogenate was treated with 300 μl of acetonitrile. After centrifugation, the supernatant was completely dried under nitrogen and reconstituted with 200 µl of water prior to LC/MS/MS detection. All samples were then analyzed with an API7500 equipped with Phenomenex C6-Phenyl HPLC column. For internal standard, nonlabeled GABA (Sigma-Aldrich: 43811) and isotopic GABA, d6-GABA (Sigma-Aldrich: 615587) were used. GABA was detected using a LC/MS/MS, Sciex API 5500 with an electrospray ionization (ESI) interface (fee for service, Harvard Taplin Facility). A GABA calibration curve was prepared from 1.56 to 200 µg/ml. In the mobile phase of HPLC, acetonitrile was used in the organic phase, and the aqueous phase was constituted with 0.02% formic acid and 2.5 mM ammonium formate in water. For the MS detection, ESI-positive mode was used. Transition ions for GABA were Q1 m/z 104.1 and Q3 m/z 87. Transition ions for GABA-d6 were Q1 m/z 110.1 and Q3 m/z 93.
Slice preparation
Six- to eight-week-old male C57Bl6/J mice that received either 1 or 6 d of vehicle or OV329 injections were anesthetized with isoflurane before decapitation. The brain was rapidly extracted and placed in an ice-cold, oxygenated (5% CO2/95%O2) slicing solution containing the following (in mM): 126 NaCl, 2.5 KCl, 1.25 NaH2PO4, 2 MgCl2, 0.5 CaCl2, 10 glucose, and 26 NaHCO3, adjusted to pH 7.4 by NaOH and osmolarity 305–315 mmol/kg and allowed to cool before being blocked and glued to a vibratome stage. In an ice-cold bubbled slicing solution, coronal (350 μm) sections containing the hippocampus were prepared with a vibratome (Leica VT1200). Sections were then hemisected through the midline, then allowed to recover in an oxygenated aCSF at 32°C in an incubator for 1 h, and then kept bubbled at room temperature.
Patch recordings
The recording chamber was perfused with an oxygenated aCSF containing the following (in mM): 126 NaCl, 2.5 KCl, 1.25 NaH2PO4, 2 MgCl2, 0.5 CaCl2, 1 glutamine, 1.5 Na pyruvate, 10 glucose, 26 NaHCO3, and 3 kynurenic acid and the GABABR antagonist CGP54626 (1 μM in DMSO; CGP), adjusted to pH 7.4 by NaOH and osmolarity 305–315 mmol/kg; this solution was heated to 32°C via an inline heater (Warner Instruments). Cells were visualized with differential interference contrast optics, and whole-cell patch recordings were made using borosilicate glass pipettes (3–5 MΩ, World Precision Instruments) containing the following CsCl internal solution (in mM): 140 CsCl, 4 NaCl, 1 MgCl2, 0.1 EGTA, 10 HEPES, 2 Mg-ATP, and 0.3 Na2-GTP, adjusted to pH 7.25 by CsOH and osmolarity 290 mmol/kg. Recordings were amplified (MultiClamp 700B, Molecular Devices), digitized at 10 kHz (1440A digitizer, Molecular Devices), low-pass filtered at 2 kHz, and recorded (pClamp 10.7, Molecular Devices) for offline analysis (Clampfit 10.7, Molecular Devices; Mini Analysis, Synaptosoft). Access resistance was monitored throughout the recordings, and cells with an access resistance larger than 20 MΩ or a change greater than 20% were excluded from analysis.
Dentate gyrus granule cells (DGGCs) were voltage clamped to −60 mV and gap-free spontaneous inhibitory postsynaptic current (sIPSC) recordings of 5–8 min duration at baseline. The perfusate was then switched to aCSF containing the GABAAR antagonist picrotoxin (PTX; 100 µM). PTX was applied for 4–8 min until both phasic and tonic GABAAR-mediated activities in the slice had ceased, and the holding current had reached steady state (Colmers and Bains, 2018; Nathanson et al., 2019; Lee et al., 2022).
Patch analysis
Spontaneous IPSCs recorded under all relevant conditions were analyzed with automated parameters in Mini Analysis (Synaptosoft) and confirmed manually (parameters: threshold, 20 pA; period to search a local maximum, 10,000 ms; time before a peak for baseline, 5,000 ms; period to search a decay time, 20,000 ms; fraction of peak to find a decay time, 0.37; period to average a baseline, 1,000 ms; area threshold, 10; number of points to average, 3; direction of peak, negative). The first 200 events observed, starting 120 s after recordings began, were analyzed for amplitude and interevent interval (IEI); cumulative distributions of both measures were tested for significant differences using via the Kolmogorov–Smirnov (KS) statistics. The decay kinetics of sIPSCs were estimated by calculating the weighted time constant based on a single or double exponential fit as described (Nathanson et al., 2019).
Tonic current shifts were analyzed in ClampFit; 30 s regions taken from stable baseline and PTX treatment conditions were selected and used to generate all-points histograms, which were fit with Gaussian curves to determine the peak value of the holding current (Colmers and Bains, 2018; Nathanson et al., 2019). The shift from baseline for their parameters was calculated for individual cells, and the mean difference between vehicle- and OV329-treated groups was determined using these values, normalized to cell capacitance, to calculate current density. To visualize the shift in the holding current, the all-points histograms were plotted normalized to baseline, with the PTX-induced shift centered at their group's mean to show differences in noise distribution in addition to the overall shift in the holding current. Data were entered into Prism (GraphPad Software) for data manipulation, normality testing, statistical analysis, and graphing.
EEG surgeries and recording
Adult male C57Bl/6J mice (11 weeks old) were surgically implanted with head mounts to permit EEG recordings. Mice were anesthetized via isoflurane inhalation, and stereotaxically positioned three-channel EEG/electromyography (EMG) head mounts (Pinnacle Technology, catalog #8201) were superglued to the skull in alignment with lambda. For cortical recordings, four screws were placed in the skull above the frontal and parietal lobes; silver epoxy was applied to each screw head to improve electrical connectivity between the electrodes and the head mount. After the surgery, the mice recovered for 7 d in their home cages before experimentation. On the day of recording, mice were connected to the preamps in the recording chambers for an hour to acclimatize them to the chamber prior to any recording. Data were recorded with a three-channel system (Pinnacle Technology) using LabChart. For the sustained dose experiment, a 4-h-long baseline recording of freely behaving mice was obtained. Following this, mice were injected with 0.5 mg/kg OV329 (i.p.) for 6 consecutive days. Twenty-four hours after the sixth, and final, dose of OV329, mice received 20 mg/kg kainate (i.p.), and 2 h after the dose of kainate, mice received a single dose of 5 mg/kg diazepam (DZ; i.p.). Recordings were terminated 3 h after the kainite injection to study the effectiveness of OV329 pretreatment in terminating kainate-induced seizures. EEG recordings were sampled at 2 kHz, and a 1 V range was used throughout the experiment (Moore et al., 2018; Jarvis et al., 2023).
EEG analysis
To evaluate the potential impact of a sustained dose of OV329 on baseline EEG power, a 10 min silent period was analyzed where there was no muscular movement detected in the EMG channel. This represented the pretreatment measurement. All mice were administered a dose of 0.5 mg/kg OV329 for 6 consecutive days. The effects of OV329 were examined 1 h after the injections during Days 2, 4, and 6 for another 10 min silent epoch and statistically compared with the pretreatment baseline EEG epoch. To compare the EEG signals, they were converted into the frequency domain via the fast Fourier transform (FFT) resulting in power spectral plots using the LabChart software, with an 8 K FFT size, Hann (cosine-bell), and 87.5% window overlap parameters. The EEG frequency analysis was performed by binning the total signal into various frequencies such as delta (0–4 Hz), theta (4–8 Hz), alpha (8–13 Hz), beta (13–30 Hz), and gamma (30–100 Hz; Lee et al., 2022; Jarvis et al., 2023). The contribution of each frequency band to total power was then determined and compared between the treatment groups.
EEG recordings were further analyzed to determine the onset to first seizure and SE. Seizure activity was defined as EEG activity 2.5 times the standard deviation of the preceding 1 min of activity, which persisted for at least 20 s. SE was defined as seizure activity lasting longer than 30 min or continuous events that were separated by <30 s return to baseline. To examine the effectiveness of OV329 in preventing the development of DZ-resistant SE, we waited 20 min after administering DZ to allow the animals to return to SE and then compared the remaining 40-min-long EEG epochs between the two treatment groups. One mouse in the vehicle group and two mice in the OV329-treated group did not respond to kainate; these mice were excluded from comparisons evaluating the anticonvulsant efficacy of OV329. Statistical tests were performed using GraphPad Prism. Values are expressed as mean ± standard error of the mean (SEM). The Shapiro–Wilk test was performed to find normal distribution of the data sets. For all comparisons, parametric and nonparametric t tests were performed as appropriate.
Results
Sustained exposure of mice to OV329 inhibits GABA-AT activity and increases GABA accumulation in the brain
To examine the cumulative effects of OV329 on GABA-AT activity in the brain, mice were treated daily for 6 d with 5 mg/kg (i.p.; Fig. 1A). Sustained treatment with OV329 significantly reduced GABA-AT activity in the brain to 62.6 ± 4.4% of control (Fig. 1B; vehicle, 100.0 ± 2.7%; n = 6 mice per group; two-tailed unpaired t test). However, as measured by immunoblotting (Fig. 1C), OV329 did not modify the expression levels of GABA-AT (vehicle, 99.9 ± 8.2%; OV329, 105.0 ± 10.3%; n = 6 mice per group; two-tailed unpaired t test). Next, the effects of OV329 exposure on brain GABA levels were examined using liquid chromatography coupled with mass spectroscopy (Fig. 1D). OV329 treatment significantly increased GABA levels to 134.0 ± 7.2% of control (vehicle, 100.0 ± 5.9%; n = 6 mice per group; two-tailed unpaired t test). Thus, sustained exposure of mice to OV329 reduces GABA-AT activity without modifying its expression levels, which parallels increased GABA levels in the brain.
Examining the effects of OV329 on GABA-AT activity in vitro and in vivo. A, The line diagram represents the i.p. dosing employed in our experiments. B, GABA-AT activity in the brains of mice dosed with 5 mg/kg OV329 for 6 d was compared with vehicle-treated controls. C, Brain extracts from mice dosed with OV329 were immunoblotted with GABA-AT and actin antibodies. The levels of GABA-AT in mice treated with OV329 were normalized to vehicle. D, GABA levels were measured in the brains of mice treated with OV329 and values were normalized to vehicle. Data shown as mean ± SEM.
Sustained exposure of mice to OV329 enhances tonic inhibition
To determine the functional effect of GABA-AT enzyme inhibition on inhibitory neurotransmission in the brain, electrophysiological recordings were taken following a 6 d drug injection paradigm. Mice (males, 6–8 weeks old at the time of experiment) received daily injections of either 6 d of vehicle, 6 d of 5 mg/kg OV329, or a single day of 5 mg/kg OV329 (Fig. 1A). About 24 h after the last injection, brain slices containing the hippocampus were prepared in which whole-cell patch-clamp recordings of DGGCs were held in voltage clamp at −60 mV (Modgil et al., 2017; Nathanson et al., 2019). We recorded from stable cells for 5–8 min and then applied the noncompetitive GABAAR antagonist PTX (100 µM) to the slice for 4–8 min. As expected, both the fast, phasic inhibition mediated by synaptic GABAAR and slow, tonic inhibition mediated by extrasynaptic GABAAR were blocked by PTX (Modgil et al., 2017; Nathanson et al., 2019).
The degree of tonic inhibition was calculated by examining the holding current of cells prior to and following PTX application (Fig. 2A). The basal level of tonic inhibition present in vehicle-treated mice (9.7 ± 2.2 pA; n = 16; N = 5 mice) was significantly enhanced in DGGCs from mice after 6 d of sustained treatment with 5 mg/kg OV329 (60.8 ± 2.2 pA; n = 15; N = 4 mice). However, a single dose of 5 mg/kg OV329 24 h prior was unable to significantly enhance the tonic inhibitory tone (13.2 ± 2.9 pA; n = 8; N = 5 mice; Fig. 2B; one-way repeated measures (RM) ANOVA: F(2,36) = 34.44). To control for the density of extrasynaptic GABAAR expression across cells, tonic current amplitudes were normalized to cell capacitance (Saliba et al., 2012; Fig. 2C; one-way RM ANOVA: F(2,36) = 30.7). Repeated dosing with 5 mg/kg OV329 significantly enhanced current density (2.4 ± 0.2 pA/pF; n = 15; N = 4 mice) compared with repeated doses with vehicle (0.3 ± 0.08 pA/pF; n = 16; N = 5 mice) or a single dose of 5 mg/kg OV329 (0.1 ± 0.08 pA/pF; n = 8; N = 5 mice). Despite the enhanced tonic current seen following sustained OV329 treatment, the basal holding current was unaltered across all conditions (Fig. 2D; vehicle: −121.9 ± 12.1 pA, n = 16; 6 d 5 mg/kg: −148.5 ± 13.0 pA, n = 15; 1 d 5 mg/kg: −38.7 ± 27.6 pA, n = 8; N = 4–5 mice per group; one-way RM ANOVA: F(2,36) = 0.89).
Repeated doses of 5 mg/kg OV329 significantly enhances tonic and phasic inhibition of DGGCs. Mice were given 6 d of repeated injections of vehicle (blue) or 5 mg/kg OV329 (green) or a single dose of 5 mg/kg OV329 (orange) prior to collection of brain slices for whole-cell patch-clamp recordings on DGGCs. Representative recordings in hippocampal slices of mice treated with 6 d repeated vehicle (A, top) or 5 mg/kg OV329 (A, bottom) showing a shift in the whole-cell holding current of DGGCs at baseline (gray-shaded region) following application of the noncompetitive GABAAR antagonist PTX (colored-shaded region) resulting from the blockade of extrasynaptic GABAARs. B, All-points histogram showing the mean shift in holding current distribution normalized to baseline (unshaded color-coded traces) following blockade of GABAARs (shaded color-coded traces). C, Tonic GABA current shift normalized to cell capacitance shows significant difference in current density following repeated dosing of OV329 at 5 mg/kg. D, Despite differences in the magnitude of tonic current, the holding current at baseline was unchanged. E, Representative spontaneous IPSC traces in mice given repeated doses of vehicle (left) or 5 mg/kg OV329 (right). Cumulative distribution of sIPSC amplitude (F) showing no significant differences in the mean amplitude (F, inset) following single or repeated doses of 5 mg/kg OV329. Cumulative distribution of sIPSC IEI (G) showing no significant differences in the mean frequency (G, inset) following single or repeated doses of 5 mg/kg OV329. H, Weighted decay of sIPSCs showing no significant changes following repeated 5 mg/kg OV329 administration. Data shown as mean ± SEM.
We next examined the effects of single and repeated daily doses of 5 mg/kg OV329 on phasic inhibition, in recordings of (action potential-dependent) sIPSCs from DGGCs held at −60 mV (Fig. 2E). The amplitude and duration between individual sIPSC events (IEI) were analyzed from the first 200 events recorded 2 min following the start of the recording, which allowed sufficient time for the pipette solution to equilibrate into the cell. Cumulative distribution plots were generated for the spontaneous amplitude (Fig. 2F) in which the distribution of amplitudes in both 1 d 5 mg/kg (p = 0.0027; KS test) and 6 d 5 mg/kg OV329 (p < 0.0001; KS test) show a significant right-shift compared with 6 d vehicle, consistent with a greater number of large events seen in DGGCs after OV329 treatment. Interestingly, however, the mean sIPSC amplitudes were not significantly changed by the drug (Fig. 2F, inset; 6 d vehicle: −30.5 ± 1.0 pA, n = 16; 1 d 5 mg/kg: −35.6 ± 2.9 pA, n = 8; 6 d 5 mg/kg: −32.2 ± 4.0 pA, n = 15; N = 4–5 mice per group; one-way RM ANOVA: F(2,36) = 0.55). Cumulative distribution plots generated for sIPSC IEIs (Fig. 2H) in mice treated with 5 mg/kg OV329 for 1 d (p = 0.0502; KS test) or 6 d (p = 0.0017; KS test) were left-shifted relative to those from 6 d vehicle-treated mice, consistent with a shorter duration between sIPSC events. However, the mean sIPSC frequency was not significantly altered (Fig. 2G, inset; 6 d vehicle: 2.8 ± 0.5 Hz, n = 16; 6 d 5 mg/kg: 2.5 ± 0.2 Hz, n = 15, 1 d 5 mg/kg: 2.8 ± 0.6 Hz, n = 8; N = 4–5 mice per group; one-way RM ANOVA: F(2,36) = 0.11). The mean decay kinetics of sIPSC decay (Fig. 2J) were also not significantly altered as measured by the weighted time constant (τw) in response to single or repeated doses of 5 mg/kg OV329 compared with vehicle (τw vehicle: 14.6 ± 1.5, n = 16 ms; τw 6 d 5 mg/kg: 10.5 ± 1.8 ms; n = 15; τw 1 d 5 mg/kg: 13.2 ± 1.8 ms, n = 8; N = 4–5 mice per group; one-way RM ANOVA: F(2,36) = 1.70).
Sustained exposure with 0.5 mg/kg OV329 enhances phasic and tonic inhibition
We next studied the effects of repeated treatment with a lower dose of OV329 on GABA-AT function. After a 6 d injection protocol with 0.5 mg/kg OV329 (Fig. 1A), we recorded tonic (Fig. 3A) and phasic (Fig. 3E) GABA currents in DGCCs held at −60 mV. In these animals, the PTX-mediated blockade of GABAARs unmasked a shift in the holding current (Fig. 3B; two-tailed unpaired t test: p = 0.018) in 0.5 mg/kg OV329-treated neurons (18.3 ± 2.7 pA; n = 16; N = 4 mice) significantly larger in OV329-treated than in vehicle-treated DGGCs (9.7 ± 2.2 pA; n = 16; N = 5 mice). Normalized to cell capacitance, the current density in mice given 6 d of 0.5 mg/kg OV329 (0.7 ± 0.1 pA/pF; n = 16; N = 4 mice) was significantly larger than in mice treated with 6 d of vehicle injections (0.3 ± 0.1 pA/pF; n = 16; N = 5 mice; Fig. 3C; two-tailed unpaired t test: p = 0.0041).
Repeated low doses of 0.5 mg/kg OV329 provide a significant enhancement to tonic and phasic inhibition onto DGGCs. Mice were given 6 d of repeated injections of vehicle (blue) or 0.5 mg/kg OV329 (purple) prior to experimentation. Representative recordings in slices from mice given 6 d repeated vehicle (A, top) or 0.5 mg/kg OV329 (A, bottom) showing a holding current shift from baseline (gray region) following PTX application (colored region). B, All-points histogram showing the mean tonic current shift from baseline (unshaded traces) following PTX application (shaded traces). C, Repeated dosing of OV329 at 0.5 mg/kg elicited a significant shift in the tonic current density with no significant change in the holding current (D). E, Representative spontaneous IPSC traces in mice given repeated doses of vehicle (left) or 0.5 mg/kg OV329 (right). Cumulative distribution of the sIPSC amplitude (F) showing no significant change in the mean amplitude (F, inset) in mice given repeated doses of 0.5 mg/kg OV329. Cumulative distribution of sIPSC IEI (G) with no significant change in the mean frequency (G, inset) following repeated doses of 0.5 mg/kg OV329. H, Weighted decay of sIPSCs showing no significant change following repeated 0.5 mg/kg OV329 administration. Data shown as mean ± SEM.
Following 6 d of 0.5 mg/kg treatment, the holding current for DGGCs at −60 mV was not significantly altered (Fig. 3D; two tailed unpaired t test: p = 0.055) following 0.5 mg/kg treatment (−93.0 ± 7.9 pA; n = 16; N = 4 mice) relative to vehicle treatment (−121.9 ± 12.1 pA; n = 16; N = 5 mice). sIPSC cumulative amplitude distribution (Fig. 3F) exhibited a right-shift (p < 0.0001; KS test) following 6 d of 0.5 mg/kg OV329 injections, consistent with a larger number of high-amplitude events, but no significant change to mean sIPSC amplitudes was observed (Fig. 3F, inset; 6 d vehicle: −30.5 ± 1.8 pA, n = 16, N = 5; 6 d 0.5 mg/kg: −36.3 ± 2.8 pA, n = 16, N = 4; two-tailed unpaired t test: p = 0.093). Similarly, the cumulative distribution of sIPSC IEIs (Fig. 3G) exhibited a left-shift (p < 0.0001; KS test) in mice treated with 0.5 mg/kg OV329, consistent with more frequent sIPSC events; however, the elevation observed in the mean frequency values did not reach significance (Fig. 3G, inset; 6 d vehicle: 2.79 ± 0.5 Hz, n = 16, N = 5; 6 d 0.5 mg/kg: 4.8 ± 0.9 Hz, n = 16, N = 4; two-tailed unpaired t test: p = 0.067). The increase in sIPSC decay kinetics following repeated dosing with 0.5 mg/kg OV329 (Fig. 3H) did not reach statistical significance (τw vehicle: 14.6 ± 1.5 ms, n = 16, N = 5; τw 6 d 0.5 mg/kg: 11.2 ± 0.8 ms; n = 16, N = 4; two-tailed unpaired t test: p = 0.058).
Sustained dosing of mice with 0.5 mg/kg OV329 selectively enhances EEG δ power
Administering a sustained dose of 0.5 mg/kg OV329 has been found to increase both tonic and phasic inhibition in the mice. We further examined whether this low sustained dose of OV329 had any impact on the baseline EEG in freely moving mice (Fig. 4A; Hines et al., 2013; Nathanson et al., 2019). A 4-h-long baseline recording was conducted (Fig. 4B), after which a single dose of 0.5 mg/kg was administered daily for 6 consecutive days, with recordings continuing for an additional 4 h after the final dose (Fig. 4C).
Sustained exposure to 0.5 mg/kg OV329 alters baseline EEG power. A, Line diagram shows the experimental paradigm. B, Representative EEG trace and its spectrogram show distribution of power across different frequency bands before OV329 treatment. C, Representative EEG trace and its spectrogram show distribution of power across different frequency bands after treatment with 0.5 mg/kg OV329 for 6 consecutive days. D, EEG recordings from Day 2 were subjected to FFT, and spectral plot is shown for frequencies between 0 and 100 Hz. The total EEG power was significantly increased following 2 d of OV329 treatment. Percentage change in EEG power was observed across delta, theta, beta, and gamma following 2 d of OV329 treatment. E, EEG recordings from Day 4 were subjected to FFT, and spectral plot is shown for frequencies between 0 and 100 Hz. The total EEG power was not significantly elevated following 4 d of OV329 treatment; however, EEG power remained significantly altered across delta and gamma frequency bands. F, EEG recordings from Day 6 were subjected to FFT, and spectral plot is shown for frequencies between 0 and 100 Hz. No change in the total EEG power following 6 d of OV329 treatment; EEG power remained significantly altered across delta and gamma frequency bands.
To study the effect of repeated doses of OV329 on EEG power, we compared 10 min segments during which there was no EMG activity from post-treatment timepoints on Days 2, 4, and 6 (Fig. 4C) against a pretreatment timepoint (Fig. 4B). To quantify the effects of OV329 on the EEG, recordings were converted into the frequency domain via FFT, generating a power spectral density plot for frequencies between 0 and 100 Hz (Fig. 4D, left; Nathanson et al., 2019; Jarvis et al., 2023). We normalized the post-treatment values in mice to their paired pretreatment values to calculate the percentage change and observed that OV329 significantly increased the total EEG power (Fig. 4D, middle; 193.5 ± 39.38% of pretreatment, two-tailed Wilcoxon matched-pairs signed-rank test (Wilcoxon t test): p = 0.03, N = 8). To assess possible effects on the EEG frequency bands (delta; 0–4 Hz, theta; 4–8 Hz, alpha; 8–13 Hz, beta; 13–30 Hz, and gamma; 30–100 Hz; Jarvis et al., 2023), we observed that OV329 treatment resulted in a significant power increase in the lower frequency domains, especially in the delta, theta, and beta power bands (Fig. 4D, right; delta: 405.0 ± 140.5%, p = 0.01; theta: 175.5 ± 40.55%, p = 0.02; beta: 177.9 ± 23.67%, p = 0.01; significance derived from Wilcoxon t test: N = 8). There was a concomitant decrease in the gamma power (Fig. 4D, right; 67.49 ± 9.07%, Wilcoxon t test: p = 0.04, N = 8 mice). On Day 4, power spectral density plots for frequencies between 0 and 100 Hz (Fig. 4E, left) showed no difference in the total EEG power (Fig. 4E, middle). However, we again observed a significant increase in the delta power band (Fig. 4E, right; 321.1 ± 79.21%, Wilcoxon t test: p = 0.04, N = 8) along with a reduction in the EEG power of the gamma frequency band (Fig. 4E, right; 46.38 ± 6.76%, Wilcoxon t test: p = 0.01, N = 8). By Day 6, spectral density analysis (Fig. 4F, left) again revealed no difference in the total EEG power (Fig. 4F, middle), but confirmed a persistent increase in the delta band power (Fig. 4F, right; 203.6 ± 36.5%, Wilcoxon t test: p = 0.02, N = 8) and the previous reduction in the gamma band power (Fig. 4F, right; 47.7 ± 6.2%, Wilcoxon t test: p = 0.007, N = 8).
Sustained dosing of mice with OV329 reduces the severity of status epilepticus and prevents the development of benzodiazepine resistance
To explore the anticonvulsant properties of OV329, its effects on the development of kainic acid (KA)-induced SE in mice was examined using EEG recordings. This model was chosen because of the similarities with patients undergoing SE, as KA-induced seizures become refractory to DZ within minutes (Reddy and Kuruba, 2013; Lee et al., 2022; Vien et al., 2022; Jarvis et al., 2023). Mice implanted with EEG/EMG electrodes were injected with a single dose of 0.5 mg/kg OV329 for 6 consecutive days. Twenty-four hours after the final dose of OV329, mice received KA (20 mg/kg i.p.) to induce SE (Puttachary et al., 2015). EEG recordings captured the response of vehicle- and OV329-treated mice to KA injection for 2 h before a saturating concentration of DZ (5 mg/kg i.p.) was administered; recordings were terminated 1 h following DZ administration (Lee et al., 2022; Vien et al., 2022; Jarvis et al., 2023; Fig. 5A–C). The effects of OV329 on the development of SE were first examined, revealing no effects on latency to first seizure (Fig. 5D) or SE compared with vehicle (Fig. 5E).
Sustained dosing with OV329 reduces the severity of SE. A, Line diagram shows an experimental timeline we utilized. Sustained dosing of mice with 0.5 mg/kg OV329 enhances phasic and tonic inhibition in the DG. We further explored the effects of GABA-AT inhibition on the basal EEG power by performing subcortical EEG surgeries. Mice were implanted with EEG/EMG electrode to record a 4-h-long baseline (pretreatment period), and then mice received a single dose of 0.5 mg/kg OV329 for 6 consecutive days. Twenty-four hours after the final dose of OV329, mice received KA (20 mg/kg i.p.) to induce SE. Two hours following KA injection, mice were dosed i.p. with 5 mg/kg DZ, and EEG recordings were extended for a further 1 h. B, Representative EEG trace and its spectrogram show distribution of power across different frequency bands for mice injected with vehicle (i.p.). C, Representative EEG trace and its spectrogram show distribution of power across different frequency bands for mice injected with OV329. D, Graph shows a comparison of latency to the first seizure between the two treatment groups. E, A comparison of the onset of SE between the two treatment groups. F, To provide the quantitative insights of effects of sustained dosage of OV329 on the basal EEG power, recordings were subjected to FFT, and EEG signal was converted from the time to the frequency domain to get power spectral density plot for frequencies between 0 and 100 Hz for both groups of mice experiencing SE. G, Graph shows a comparison of SE power between the two treatment groups. H, EEG recordings from mice pretreated with vehicle or OV329, 20 min following the DZ injection were subjected to FFT, and spectral plots are shown for frequencies between 0 and 100 Hz. I, Graph shows a comparison of the total EEG power post DZ treatment between the two groups of mice. J, Graph shows a comparison of EEG power before and after DZ treatment from the vehicle-treated mice. K, Graph shows no difference in percentage change in the total EEG power after DZ treatment in vehicle-treated mice. L, Vehicle treatment does not modify EEG power of any individual frequency band. M, A comparison of EEG power before and after DZ treatment from OV329-treated mice. N, Percentage change in the total EEG power was observed after DZ treatment in OV329-treated mice. O, Percentage change in EEG power was observed across delta, theta, and alpha was observed upon OV329 treatment.
In mice pretreated with OV329, the total EEG power of SE was significantly reduced compared with vehicle (Fig. 5F,G; vehicle, 2.0 ± 0.3 × 10−6 V2, and OV329, 0.9 ± 0.2 × 10−6 V2; two-tailed Welch's t test: p = 0.005; N = 6). Two hours after KA injection, mice were injected with DZ, and the ability of OV329 pretreatment to suppress EEG power was compared between the two groups 20 min after DZ injections. After DZ treatment, EEG power postictally was significantly reduced in OV329-treated mice compared with vehicle-treated mice (Fig. 5H,I; vehicle, 1.0 ± 0.2 × 10−6 V2, and OV329, 0.4 ± 0.04 × 10−6 V2; two-tailed Welch's t test: p = 0.02; N = 6). As in previous studies, DZ did not modify the total EEG power or any individual frequency band in mice pretreated with vehicle (Fig. 5J–L), as characterized by the resurgence in EEG power seen in Figure 5C (Goodkin et al., 2003; Silayeva et al., 2015; Bertoglio et al., 2017; Moore et al., 2018; Jarvis et al., 2023), consistent with the DZ insensitivity reported in the KA model (Reddy and Kuruba, 2013). By contrast, DZ significantly reduced EEG power in mice pretreated with OV329 (Fig. 5M; −DZ, 0.9 ± 0.2 × 10−6 V2, and +DZ, 0.4 ± 0.05 × 10−6 V2; Wilcoxon t test: p = 0.02, N = 6; Reddy and Kuruba, 2013; Lee et al., 2022; Vien et al., 2022; Jarvis et al., 2023). We found a significant reduction in the total EEG power following DZ treatment in mice pretreated with OV329 (Fig. 5N; 48.85 ± 9.10%; Wilcoxon t test: p = 0.03, N = 6). Examining specific frequency bands, this reduction in EEG power was observed across the delta (Fig. 5O; 43.95 ± 13.69%; Wilcoxon t test: p = 0.03, N = 6), theta (Fig. 5O; 33.91 ± 4.9%; Wilcoxon t test: p = 0.005, N = 6), and alpha (Fig. 5O; 37.14 ± 8.07%; Wilcoxon t test: p = 0.008, N = 6) frequency bands following OV329 pretreatment. Together, these findings suggest that increasing GABA-mediated inhibition with OV329 enhances the efficacy by which DZ terminates kainate-induced SE.
Discussion
OV329 is a mechanism-based inactivator of GABA-AT, rationally designed for greater potency and selectivity (Juncosa et al., 2018; Feja et al., 2021) compared with VGB, an established GABA-AT mechanism-based inactivator with clinical application limited by its cumulative toxicity. Here we have examined the effects of sustained low dosing with OV329 in mice on the efficacy of GABAAR-mediated neuronal inhibition in the hippocampus. We found that 6 d of OV329 treatment at 5 mg/kg significantly enhanced tonic GABAAR-mediated current in DGGCs, while the mean sIPSC amplitude, decay, and frequency were not significantly changed. Consistent with this, we observed a significant elevation in brain GABA levels in treated mice that correlated with reductions in GABA-AT activity, without changes in its expression levels. Thus, the reduced GABA-AT activity and subsequent rise in brain GABA content are likely to underpin the increased tonic current measured in hippocampal slices following in vivo exposure to OV329. When mice were treated with just 0.5 mg/kg OV329 for 6 d, we also observed significant elevations in tonic inhibition—as with the higher dose. These results demonstrate that exposure of mice to OV329 increases the magnitude of neuronal inhibition mediated by extrasynaptic GABAARs, consistent with published studies showing VGB increases tonic inhibition in vitro (Jackson et al., 1994; Engel et al., 2001; Overstreet and Westbrook, 2001; Wu et al., 2001, 2003), but with greater potency (Wild et al., 2009; Maguire et al., 2010; Yang et al., 2012; Foroozan, 2018). The therapeutic effects of OV329 are time and dose dependent, as an acute dose of 40 mg/kg was able to reduce the severity of pentylenetetrazol-induced seizures and suppress generalized seizures in kindled rats, while an acute dose of 5 mg/kg had no effect (Feja et al., 2021). Accordingly, we show that an acute dose of 5 mg/kg is unable to significantly alter the tonic current in DGGCs; however, with sustained doses at 0.5 or 5 mg/kg, the cumulative inhibition of GABA-AT over multiple days was able to significantly enhance tonic GABA signaling at lower doses. The mechanism for VGB-associated visual field loss is unclear; however, it is thought that the high daily dose, poor blood–brain barrier permeability, and low inactivation efficiency may play a role (Wild et al., 2009; Maguire et al., 2010; Yang et al., 2012; Foroozan, 2018; Shen et al., 2020). Recent evidence suggests a role for off-target or downstream alterations in neuroactive amino acids (Jammoul et al., 2009; Walters et al., 2019). CPP-115, a cyclopentane-based precursor of OV329 with similar structure, was tested for retinal damage in a model for infantile spasms in rats. CPP-115 was given at a dose roughly 20-fold its effective dose for 45 d and showed a substantially lower retinal loss compared with VGB at its effective dose over the same duration (Silverman, 2012). This suggests OV329 may have a similarly enhanced safety profile; however, further examination is necessary.
To examine if these modifications to inhibition impact brain activity, we measured their effects on EEG power. Six days of treatment with 0.5 mg/kg OV329 significantly increased the power of delta frequency (0–4 Hz) and a decrease in gamma frequency (30–100 Hz). Similar elevations in delta power are seen with lorazepam and zolpidem, benzodiazepines that enhance phasic inhibition via their efficacy as GABAAR-positive allosteric modulators (Christian et al., 2015). Moreover, delta frequency EEG power also increases upon treatment with GABA reuptake inhibitor-based antiepileptic drugs (Frohlich et al., 2023). This suggests that OV329 treatment increases inhibition by regulating delta oscillations, which may suppress seizures.
To evaluate the anticonvulsant efficacy of OV329, we tested the effects of 6 d exposure to 5 mg/kg OV329 on the development and severity of SE induced by KA using EEG recording. While OV329 did not modify the latency to SE, the power of KA-induced seizures was reduced, suggesting this compound exhibits anticonvulsant activity. In agreement with our results, a single 40 mg/kg dose of OV329 has been reported to reduce the severity of pentylenetetrazol-induced convulsions as measured on the Racine scale (Feja et al., 2021).
Significantly, pretreatment of mice with OV329 prevented the development of benzodiazepine resistance following entrance into KA-induced SE. In humans, as in rodents, SE (Reddy and Kuruba, 2013; Lee et al., 2022; Jarvis et al., 2023) becomes resistant to benzodiazepines resulting in increased mortality of this trauma and severe brain damage in surviving patients and significantly increases the probability of developing chronic epilepsy (Walker, 2018; Crawshaw and Cock, 2020). Sustained elevations in neuronal excitability led to profound modifications in the efficacy of synaptic inhibition which include reductions in GABA levels, internalization of GABAARs, and modifications in their subunit composition, in addition to increased neuronal Cl− accumulation (Walker, 2018; Burman et al., 2019, 2022). While further studies are required to understand the mechanism underlying the effects of OV329, presumably its ability to enhance the efficacy of GABAergic inhibition is of significance in preventing the development of benzodiazepine refractory seizures in SE.
In summary, our results demonstrate that OV329 treatment of mice induces sustained elevations in the magnitude of GABAergic inhibition and selective elevations in EEG delta power. These sustained effects on neuronal excitability suggest that OV329 may be of use as an anticonvulsant and for treatment of disorders that result from aberrant elevations in neuronal excitability.
Footnotes
S.J.M. serves as a consultant for AstraZeneca, Ovid Therapeutics, and Sage Therapeutics, relationships that are regulated by Tufts University. S.J.M. holds equity in Sage Therapeutics. P.L.W.C., S.F.J.N., and S.J.M. hold equity in Ovid Therapeutics. J.M., S.L., and P.S. are employees of Ovid Therapeutics and hold equity.
S.J.M. is supported by the National Institutes of Health (NIH) – National Institute of Neurological Disorders and Stroke Grants NS087662, NS081986, NS108378, NS101888, NS103865, and NS111338 and NIH – National Institute of Mental Health Grant MH118263 and a sponsored research agreement with Ovid Therapeutics.
↵*P.L.W.C. and M.N.A. contributed equally to this work.
This is an open-access article distributed under the terms of the Creative Commons Attribution 4.0 International license, which permits unrestricted use, distribution and reproduction in any medium provided that the original work is properly attributed.
References
In this issue
