Abstract
Approximately one-third of neonatal seizures do not respond to first-line anticonvulsants, including phenobarbital, which enhances phasic inhibition. Whether enhancing tonic inhibition decreases seizure-like activity in the neonate when GABA is mainly depolarizing at this age is unknown. We evaluated if increasing tonic inhibition using THIP [4,5,6,7-tetrahydroisoxazolo(5,4-c)pyridin-3-ol, gaboxadol], a δ-subunit–selective GABAA receptor agonist, decreases seizure-like activity in neonatal C57BL/6J mice (postnatal day P5–8, both sexes) using acute brain slices. Whole-cell patch-clamp recordings showed that THIP enhanced GABAergic tonic inhibitory conductances in layer V neocortical and CA1 pyramidal neurons and increased their rheobase without altering sEPSC characteristics. Two-photon calcium imaging demonstrated that enhancing the activity of extrasynaptic GABAARs decreased neuronal firing in both brain regions. In the 4-aminopyridine and the low-Mg2+ model of pharmacoresistant seizures, THIP reduced epileptiform activity in the neocortex and CA1 hippocampal region of neonatal and adult brain slices in a dose-dependent manner. We conclude that neocortical layer V and CA1 pyramidal neurons have tonic inhibitory conductances, and when enhanced, they reduce neuronal firing and decrease seizure-like activity. Therefore, augmenting tonic inhibition could be a viable approach for treating neonatal seizures.
Significance Statement
Neonatal seizures occur in roughly 1–3/1,000 term births and 10-fold more frequently in preterm neonates. Approximately 30–50% of neonatal seizures resist current treatments, and drug resistance often develops if diagnosis and treatment are delayed, for example, in underserved areas. There has been no significant improvement in treating neonatal seizures for nearly 50 years, especially for patients in whom pharmacoresistance develops. Our results show that enhancing tonic inhibition in neonatal mice, a developmental age when GABA mainly depolarizes, decreases neuronal firing and epileptiform activity in vitro. Our results suggest that augmenting tonic inhibition could be an alternative treatment for neonatal seizures, especially in pharmacy-resistant ones.
Introduction
Neonatal seizures occur in 1–3/1,000 term births, with a 10-fold increase in preterm neonates (Vasudevan and Levene, 2013). Hypoxic–ischemic encephalopathy and stroke are common causes of neonatal seizures, and if they have a high electrographic seizure burden, seizures can lead to developmental delay, epilepsy, and intellectual disability (Pisani and Spagnoli, 2015; Kharoshankaya et al., 2016). Despite new options for adults, treating neonatal seizures still relies on phenobarbital as the first-line drug (Pressler and Mangum, 2013; Pressler et al., 2023). However, 30–50% of neonatal seizure cases do not respond to phenobarbital due to treatment delays and biological factors (Donovan et al., 2016; Sharpe et al., 2020; Burman et al., 2022). Therefore, new approaches are needed, especially in cases where treatment is delayed and pharmacoresistance develops.
Gamma-aminobutyric acid (GABA) is the brain's primary inhibitory neurotransmitter, yet in neonates, GABA can have a depolarizing action (Ben-Ari et al., 2007). Whether GABA is inhibitory or depolarizing depends on the relationship between the neuron's membrane potential and GABA's reversal potential (EGABA), determined mainly by the neuronal chloride concentration ([Cl−]i). During early development, neurons have a wide range of [Cl−]i, but on average, it is higher than adults (Rheims et al., 2008; Glykys et al., 2009; Kirmse et al., 2015; Sulis Sato et al., 2017), resulting in more depolarizing GABAergic signaling. Also, during prolonged seizures, including status epilepticus, high neuronal activity increases [Cl−]i, and EGABA is further depolarized (Dzhala et al., 2010; Nardou et al., 2011). Therefore, GABAergic anticonvulsants like phenobarbital and diazepam (DZP), which act on synaptic receptors, may be ineffective or proconvulsant in neonates (Staley, 1992; Glykys et al., 2009; Glykys and Staley, 2015).
Synaptic and extrasynaptic GABAA receptors (GABAARs) have different subunit compositions, and these differences affect their kinetics. Traditional GABAergic anticonvulsants like phenobarbital target synaptic GABAARs, mainly containing γ subunits and mediating phasic inhibition. In contrast, extrasynaptic GABAARs primarily contain δ or α5 subunits and mediate tonic inhibition (Glykys et al., 2008). Tonic inhibition mediates 75% of the inhibitory conductance of neurons compared to just 25% mediated by phasic inhibition (Mody and Pearce, 2004). Importantly, tonic inhibition has a sustained influence on membrane potential (Chandra et al., 2006; Glykys and Mody, 2007a), and it can provide effective inhibition by shunting excitatory currents (Staley and Mody, 1992; Chance et al., 2002). However, as GABA is depolarizing in the neonatal age, and neuronal [Cl−]i increases with repetitive seizures, it is unknown if enhancing tonic “inhibition” will worsen neonatal seizures.
Here, using mice of both sexes, we evaluated if the δ-selective agonist THIP [4,5,6,7-tetrahydroisoxazolo(5,4-c)pyridin-3-ol, gaboxadol] can enhance tonic inhibition in neonatal cortical and CA1 pyramidal cells (PCs) and if it affects neuronal activity. We also evaluated whether increasing tonic inhibition decreases seizure-like activity in two in vitro neonatal seizure models. We demonstrate that boosting tonic inhibition in the mouse hippocampus and neocortex reduces seizure-like activity in these models by reducing neuronal excitability. Thus, enhancing tonic inhibition may be a viable treatment for neonatal seizures, especially in cases where pharmacoresistance has developed.
Materials and Methods
Animals
C57BL/6J (postnatal day P5–8, both sexes) and adult (1.5 months old, male) mice (Jackson Laboratory, strain #000664) and transgenic P5–8 Thy1-GCaMP6s mice (C57BL/6J-Tg(Thy1-GCaMP6s)GP4.3Dkim/J, Jackson Laboratory, strain #024275) were used for the experiments. The mice were housed in The University of Iowa Animal Facility. Experiments were approved by the Institutional Animal Care and Use Committee of The University of Iowa.
Preparation of acute brain slices
Mice were anesthetized with isoflurane and decapitated per approved protocol. The brain was placed in ice-cold artificial cerebrospinal fluid (aCSF) containing (in mM) NaCl (120), KCl (3.3), CaCl2 (1.3), MgCl2 (2), NaH2PO4 (1.25), NaHCO3 (25), and D-glucose (10) with pH 7.3–7.4 when bubbled with 95% O2 and 5% CO2. For current clamp and sEPSC recordings, the cutting and recording solution had NaCl 116 instead of 120 mM and mannitol 20 mM. Coronal brain slices 450 µm (P5–8) and 350 µm (adults) thick were cut using a vibratome (Leica VT1000S) while submerged in aCSF containing 2 mM kynurenic acid and 2 mM MgCl2 to block glutamate receptors. Brain slices were then placed in an interface holding chamber containing aCSF (1.3 mM MgCl2) at room temperature for 30 min. Next, the temperature was slowly increased and held at 30°C until the time of the experiment. Slices were stored for at least 1 h at 30°C before experimental use.
Field electrophysiology
Brain slices of C57Bl/6J mice were placed in an interface recording chamber and perfused with aCSF containing 100 µM of 4-aminopyridine (4-AP) or low-Mg2+ (0 mM MgCl2 aCSF) to induce seizure-like events (SLEs). Slices were held at 32–34°C and oxygenated with 95% O2 and 5% CO2. aCSF-filled glass electrodes were placed in the neocortex (motor and somatosensory regions, up to the anterior hippocampus, layer IV/V) or the hippocampal CA1 stratum pyramidale. A stereomicroscope (AmScope) was used to visualize electrode placement. Extracellular field potentials were recorded using a low-noise differential amplifier (100× gain, DP-311, Warner Instruments) and digitized at 2 kHz with an analog-to-digital converter (IX/408, iWorx Systems Incorporated). SLE were measured by fast Fourier transform (FFT) power during baseline and subsequent perfusion of THIP or DZP using a custom-written macro programmed in Igor Pro v8.03 (WaveMetrics) as previously described (Glykys et al., 2009). SLE characteristics were analyzed using a custom-written macro in Igor Pro (Sturgeon et al., 2021) using the same timeframes (900 s) for baseline and drug. SLE analysis parameters were as follows: event threshold 3× baseline, minimum event length 0.04 s, and dead time 1 s. Events lasting ≥10 s were considered ictal, and events lasting <10 s were considered interictal events (White et al., 2006).
Whole-cell patch-clamp recording
Brain slices were transferred to a submerged recording chamber and perfused with aCSF (32°C–34°C, 95% O2/5% CO2) containing kynurenic acid (2 mM) to block glutamatergic transmission. PCs in the neocortex (somatosensory and motor, layer V) or hippocampal CA1 PCs were identified by IR-Dodt Gradient Contrast video microscopy (BX51WIF, 40× water immersion objective, Olympus with an infracontrast DGC, Luigs & Neumann), imaged with a CS505MU CMOS camera (Thorlabs), and recorded with an integrated patch-clamp amplifier (Double IPA, Sutter Instruments). For tonic and phasic inhibitory current measurements, micropipettes (2–5 MΩ) contained the following internal solution (in mM): CsCl (140), MgCl2 (1), HEPES (10), EGTA (0.15), NaCl (4), Mg-ATP (2), Na2-GTP (0.3), QX-314 (5), and pH ∼7.25, ∼280 mOsm. For current clamp and spontaneous excitatory currents, micropipettes were filled with a potassium-based intracellular solution containing (in mM) K-gluconate (120), EGTA (0.15), HEPES (10), KCl (20), GTP (0.3), ATP (4), phosphocreatine (10, pH = 7.2, 289 mOsm). Whole-cell voltage-clamp recordings were performed at a holding potential of −70 mV. Series resistance and whole-cell capacitance were estimated and compensated. Recordings were discontinued if series resistance increased by >25% during the experiment. Recordings were acquired at 10 kHz and low-pass filtered at 3 kHz. Drugs were perfused after obtaining a stable control recording period of at least 1 min. Bicuculline methiodide (BMI, 10 mM stock) 20 μl was injected into the perfusion chamber to block GABAARs. Recordings were analyzed post-acquisition using SutterPatch, Igor Pro 8, and GraphPad Prism.
Tonic inhibition analysis
The tonic inhibitory current was measured using a custom-written IgorPro macro (Glykys and Mody, 2007a). Briefly, the tonic current was obtained by fitting a Gaussian to the all-points histogram over a specific time epoch (baseline, 20–30 s; plus BMI, 10–15 s). The Gaussian was fitted to the part of the distribution not skewed by synaptic events during the baseline condition. We obtained the magnitude of the tonic current by subtracting the current recorded in the presence of BMI from the baseline. The tonic current was expressed as a conductance normalized to the membrane capacitance (pS/pF) to account for cell-to-cell variability.
Spontaneous inhibitory and excitatory postsynaptic current measurement
Spontaneous inhibitory postsynaptic currents (sIPSCs) were detected using the SutterPatch synaptic event analysis template (rise = 1.5 ms, decay = 15 ms) and selected for analysis based on a 4 pA threshold amplitude. Spontaneous sEPSCs were recorded at baseline and after 10 min of THIP perfusion. sEPSCs were detected using an event template (rise = 1.5 ms, decay = 15 ms) and selected for analysis based on a 5 pA threshold amplitude.
Multiphoton laser scanning microscopy
Neuronal Ca2+ transients were imaged using a two-photon microscope (Bruker Ultima galvo-resonant system) mounted on an Olympus BX51WIF upright microscope with a 20× immersion objective (NA 1.00). A TI:sapphire mode-locked laser (Mai Tai HPDS; Spectra-Physics) generated two-photon fluorescence at 920 nm. Time-series (T-series) recordings were taken at a frequency of 2.21 Hz at a resolution of 256 × 256 pixels with 2× digital magnification. Acute brain slices from transgenic mice expressing neuronal GCaMP6s were imaged in a submerged chamber perfused with aCSF at 32–34°C with 95% O2 and 5% CO2 with or without 4-AP (10 µM). Two 2 min time series were recorded 2 min apart, and then the slices were perfused with THIP (50 µM). After 5 and 10 min of THIP perfusion, two additional time series were taken. As a control, a separate group of slices was perfused with aCSF plus 4-AP (10 µM) for the whole experiment. Evoked Ca2+ transients: a tungsten monopolar stimulating electrode was placed in the neocortical layer IV/V or the CA3 region of the hippocampus. T-series recordings were obtained in the neocortical layer II/III and hippocampal CA1 stratum pyramidale at a distance from the electrode to record synaptic activity. T-series recordings were 40 s long and recorded in galvo-resonant mode at 6.98 Hz. First, a 10 s baseline was recorded with sequentially longer stimulations from 0.5 ms to 20 ms at 3.9 mV, with a 1 min interval between T-series to allow for a return to baseline. Next, this protocol was repeated using the same slice under THIP 50 µM without moving the stimulating electrode.
Multiphoton image processing
Using custom-written ImageJ macros, raw videos were motion-corrected, background-subtracted, smoothed (median filter, radius = 2), and compressed into maximum intensity projections. Active neurons were selected manually, and an automated algorithm created predefined regions of interest (ROI). All ROI pixels were averaged, and the ΔF/F was calculated over time with an ImageJ macro. Peak Ca2+ transients were identified using the FindPeaks function in MATLAB, where the ΔF/F values for each ROI had to rise and fall by 0.15 (ΔF/F) before the next event could be detected. For evoked Ca2+ transients, ΔF/F values for each ROI had to rise and fall by 0.5 (ΔF/F). Areas under the curve (AUC) were fitted by a Boltzmann sigmoidal equation: f(W) = (Max)/1 + exp (W50-X/slope), where W is the stimulus duration (ms), Max is the maximum AUC, W50 is the stimulus duration that elicits 50% of Max, and slope is a steepness factor.
Statistical analysis
The normality of distributions was evaluated with the Shapiro-Wilk test and QQ plots. Gaussian distributed data are presented as mean ± 95% confidence interval [CI] while non-Gaussian are presented as median ± interquartile range (IQR). Lognormal distributions were log-transformed before applying parametric tests and presented as geometric mean ± 95% CI. A one-sample t test was used for a single-sample comparison. Comparisons of two paired groups used the paired t test (parametric) or Wilcoxon matched-pair signed-rank test (WSRT, nonparametric). Comparison of two unpaired groups used the unpaired t test (parametric). Repeated measures (RM) ANOVA was used for multiple comparisons of parametric data, and regular ANOVA for unpaired parametric data with the Tukey post hoc test. For multiple comparisons with unequal variance, the Brown–Forsythe ANOVA tests were used, followed by the Dunnett T3 post hoc test. Comparisons of nonparametric paired data used the Friedman test, while unpaired data used the Kruskal–Wallis test with the Dunn post hoc test. To analyze the cumulative incidence curves, the stimulus duration firing thresholds for all ROIs in each slice in aCSF and THIP were used to generate the curves, followed by the log-rank (Mantel–Cox) test and hazard ratios for aCSF versus THIP. Statistical significance was defined as p < 0.05. Data analysis used Igor Pro v8.03 and Prism 9 (GraphPad software).
Coding accessibility
Custom-written codes and macros will be provided upon request.
Reagents
4-AP, THIP [4,5,6,7-tetrahydroisoxazolo[5,4-c]pyridin-3-ol], BMI, and kynurenic acid (4-hydroxyquinoline-2-carboxylic acid) were obtained from Sigma-Aldrich. QX-314 was obtained from Tocris Bioscience. All drugs (except kynurenic acid) were prepared as stock in deionized water and diluted to the appropriate concentration in aCSF. DZP was prepared as stock in 100% ethanol and diluted to 2 µM with a final ethanol concentration of 0.04%.
Results
THIP enhances neonatal tonic inhibition in neocortical and CA1 PCs
We measured tonic inhibitory conductances using whole-cell voltage-clamp recordings in neonatal (P5–P8) neocortical (layer V) and CA1 PCs (Fig. 1A–C). All recordings were done in aCSF without GABA and contained 2 mM kynurenic acid to block ionotropic glutamate receptors with a high CsCl internal pipette solution to enhance Cl−-mediated currents. The tonic inhibitory conductance of neocortical PCs was 6.68 pS/pF [95% CI (4.10, 9.26), n = 7] and 3.92 pS/pF [(2.54, 5.30), n = 5] for CA1 PCs. This tonic conductance was significantly enhanced by 10 and 50 µM of THIP, a GABAAR agonist with a preference for δ- compared to γ2-containing GABAARs (Brown et al., 2002). We chose these concentrations based on the EC50 of 44 µM measured in slices (Drasbek and Jensen, 2006) and 51.8 µM in oocytes (Stórustovu and Ebert, 2006). While these concentrations may not be entirely specific to δ-GABAARs, they primarily enhance tonic inhibition (Meera, 2011). Neocortical PC tonic inhibitory conductance significantly increased with THIP 10 µM to 62.4 pS/pF [(40.5, 84.3), n = 5] and 177 pS/pF [(89.9, 263), n = 6] with 50 µM (Brown–Forsythe ANOVA, F(2, 5.5) = 19.9, p = 0.003; Fig. 1C). In CA1 PCs, the conductance also increased with THIP 10 µM to 82.3 pS/pF [(56.0, 109), n = 5] and 220 pS/pF [(141, 299), n = 7] with 50 µM (Brown–Forsythe ANOVA, F(2, 6.87) = 31.3, p = 0.0004; Fig. 1C). We compared the sIPSCs in control and THIP 10 µM, as they were not clearly disernible in 50 µM THIP due to the increase in baseline noise (Fig. 1D, Table 1). There was a significant increase in peak amplitude and rise time in the neocortex PC and a decrease in the frequency of sIPSCs in CA1 PCs under THIP. Part of these observed differences could be due to small sIPSCs being buried in the high baseline noise under THIP. When we repeated the analysis by selecting control sIPSCs larger than the THIP baseline variance, this eliminated the significant differences except in neocortical PC rise time. Thus, our results show that neocortical layer V and CA1 PCs have a THIP-sensitive inhibitory tonic conductance during the neonatal period.
The neocortical and CA1 pyramidal neurons have tonic currents in the neonatal period, which THIP enhances. A, Voltage-clamp recordings of a neocortical pyramidal cell (layer V) and (B) a CA1 pyramidal cell during control (top), 10 µM (middle), and 50 µM THIP perfusion (bottom; Vh = −70 mV). Top line indicates bicuculline perfusion. Right, fits to the all-point histograms under each condition. Dashed lines: control, THIP, and bicuculline holding currents. Time scale bar: 30 s. C, Tonic inhibitory conductance calculated in the neocortex (left) and CA1 region (right). Neocortex: Brown–Forsythe ANOVA, F(2, 5.5) = 19.9, p = 0.003, Dunnett's T3 post hoc test. CA1: Brown–Forsythe ANOVA, F(2, 6.87) = 31.3, p = 0.0004, Dunnett's T3 post hoc test. D, Example of sIPSC traces in aCSF (top), 10 µM (middle), and 50 µM (bottom). Asterisks: putative sIPSCs were not detected due to high baseline noise. White circles: individual slices. Black lines: mean ± 95% CI.
THIP (10 µM) actions on spontaneous inhibitory postsynaptic currents (sIPSCs) in the neonatal neocortex layer V and CA1 PCs
Augmenting tonic inhibition decreases the excitability of neonatal neocortical layer V pyramidal neurons without abolishing excitatory synaptic currents
Because the tonic inhibitory conductance was large using a CsCl internal solution, we evaluated whether a high THIP dose (50 µM) has a significant effect on the intrinsic neuronal properties and spontaneous excitatory synaptic currents (sEPSCs) using a more physiological internal pipette solution (K-gluconate). We chose to evaluate the intrinsic cellular properties of layer V PCs by current clamp in aCSF (baseline) followed by THIP (50 µM) perfusion, as the effect of THIP was similar between CA1 and Layer V PCs (Fig. 2A). THIP hyperpolarized the resting membrane potential, increased the rheobase, and shifted the input–output curves to the right (Fig. 2B–D). Additionally, THIP reduced the slope of the IV relationship and reduced the input resistance (Fig. 2E,F). THIP did not alter the amplitudes or frequencies of sEPSCs recorded in layer V PCs under whole-cell voltage-clamp (p = 0.303 and p = 0.416; Fig. 2G–I). Thus, our data demonstrate that THIP 50 µM causes neocortical PCs to be less excitable without significantly altering synaptic communication via sEPSCs.
Enhancing tonic inhibition decreases the excitability of neonatal neocortical PCs without altering spontaneous excitatory postsynaptic currents (sEPSCs). A, Representative traces of voltage responses to depolarizing or hyperpolarizing current injection steps (from −100 to 200 pA, 10 pA step), recorded in current clamp mode. The first action potential discharge at rheobase (minimum amount of current that evoked an action potential) is depicted in black. B, Resting membrane potential [baseline, −63 mV (−70.6, −55.4), THIP, −70.1 mV (−76.9, −63.3), paired t test, n = 8]. C, Rheobase [baseline, 36.1 pA (20.2, 52.3), THIP, 56.3 pA (30.2, 82.3), paired t test, n = 8]. D, Action potential firing frequency per input current (F/I) curves in baseline and THIP, mean ± SEM. E, IV relationship curve, mean ± SEM. F, Input resistance [baseline, 412 MOhms (246, 579), THIP, 340 MOhms (185, 496), paired t test, n = 8]. G, Representative traces showing sEPSCs (top) and magnification of individual events (bottom) recorded in aCSF. H, sEPSCs amplitudes. Baseline, 13.7 pA (8.04, 19.4); THIP, 10.3 pA (3.09, 17.6), p = 0.303, unpaired t test. I, Frequency of sEPSCs. Baseline, 0.34 Hz (0.12, 0.56); THIP = 0.23 (−0.01, 0.54), p = 0.416, unpaired t test. White circles: matched individual cells. Black lines: mean ± 95% CI.
Increasing tonic inhibition dampens Ca2+ transients in the neonatal neocortex and CA1 region
During early development, GABA can be depolarizing in the neocortex and hippocampus due to elevated neuronal [Cl−]i (Owens et al., 1996; Dzhala and Staley, 2003; Tyzio et al., 2007; Rheims et al., 2008; Glykys et al., 2009). However, instead of depolarizing neurons with high neuronal [Cl−]i, THIP may enhance inhibition by augmenting extrasynaptic GABAAR conductance and shunting inhibition (Chance et al., 2002). To evaluate if neurons decrease their firing rate when tonic inhibition is enhanced, we used two-photon Ca2+ imaging (Thy1-GCaMP6s) to determine the effect of THIP on Ca2+ transients in the neocortex (layer IV/V) and CA1 hippocampal region at P5–8 (Figs. 3 and 4). During baseline, there were infrequent but stable Ca2+ transients in both areas (Fig. 3). Pefusion of THIP (50 µM) significantly decreased but did not abolish spontaneous neuronal Ca2+ transients in the neocortex (p < 0.0001, Friedman test, n = 9 slices; Fig. 3A–D) and in the CA1 region (p = 0.0003, Friedman test, n = 10 slices; Fig. 3E–H). There were still Ca2+ transients under THIP in 3/9 slices at 5 min and 2/9 at 10 min in the neocortex and 5/10 slices at 5 min and 3/10 at 10 min in the CA1 region (6 mice). As there were infrequent neuronal Ca2+ transients in regular aCSF, we repeated these experiments in the presence of 4-AP (10 µM) to increase neuronal activity (Fig. 4). Similarly, THIP (50 µM) significantly reduced but did not abolish neuronal Ca2+ transients in the neocortex (p = 0.0002, Friedman test, n = 8 slices; Fig. 4A–D) and CA1 region (p < 0.0001, Friedman test, n = 11 slices; Fig. 4E–H) in the presence of 4-AP. There were still Ca2+ transients under THIP in 4/8 slices at 5 min and 6/8 at 10 min in the neocortex and in 4/11 slices at 5 min and 1/11 at 10 min in the CA1 region (6 mice). Control experiments with 4-AP alone showed no significant change in neuronal activity in the neocortex over time (RM one-way ANOVA, F(2.53–176.8) = 2.51, p = 0.07, n = 5 slices, 71 ROIs, 3 mice) or CA1 region (RM one-way ANOVA, F(2.29, 121) = 0.81, p = 0.46, n = 8 slices, 54 ROIs, 3 mice).
Spontaneous Ca2+ transients are reduced by increasing tonic inhibition in the neonatal neocortex and CA1 region. A,E, Maximal intensity projection of a neonatal neocortex, layer IV/V (A) and CA1 (E) region from acute brain slices of Thy1-GCaMP6s–expressing mice (P5–8) recorded in aCSF. Yellow circles: ROIs. Arrow: single ROI displayed in lower panels of B and F. B,F, Top panel: ΔF/F heatmap for the neonatal neocortex (B) and CA1 region (F) activity during each time-series recording. Rows: individual ROIs. Triangles: time between each time series. Bottom panel: ΔF/F trace of an individual ROI from B and F, respectively. C, Ca2+ peak transients by slice in the neocortex (Friedman test, p < 0.0001, Dunn test, n = 9 slices, 6 mice). D, Ca2+ peak transients by ROI in the neonatal neocortex (Friedman test, p < 0.0001, Dunn test, n = 95 ROIs, 6 mice). G, Ca2+ peak transients by slice in the CA1 region (Friedman test, p = 0.0003, Dunn test, n = 10 slices, 6 mice). H, Ca2+ peak transients by ROI in the CA1 region (Friedman test, p < 0.0001, Dunn test, n = 60 ROIs, 6 mice). White circles: matched individual Slices (C,G) or ROIs (D,H). Black lines: Median ± IQR.
Increasing tonic inhibition decreases 4-AP–induced Ca2+ transients in the neonatal neocortex and CA1 region. A,E, Maximal intensity projection of a neonatal neocortex, layer IV/V (A), and CA1 region (E) from acute brain slices of Thy1-GCaMP6s–expressing mice (P5–8) under 4-AP (10 µM). Yellow circles: ROIs. Arrow: single ROI displayed in lower panels of B and F. B,F, Top panel: ΔF/F heatmap of the neonatal neocortex activity (B) and CA1 region activity (F) during each time-series recording. Rows: ROIs. Triangles: time between each time series. Bottom panel: ΔF/F trace of an individual ROI from B and F, respectively. C, Ca2+ peak transients by slice in the neocortex (Friedman test, p = 0.0002, Dunn test, n = 8 slices, 7 animals). D, Ca2+ peak transients by ROI in 4-AP–treated slices of neonatal neocortex exposed to THIP (Friedman test, p < 0.0001, Dunn test, n = 120 ROIs, 7 mice). G, Ca2+ peak transients by slice in the CA1 region (Friedman test, p < 0.0001, Dunn test, n = 11 slices, 6 mice). H, Ca2+ peak transients by ROI in 4-AP–treated slices of neonatal CA1 region exposed to THIP (Friedman test, p < 0.0001, Dunn test, n = 89 ROIs, 6 mice). White circles: matched individual slices (C,G) or ROIs (D,H). Black lines: median ± IQR.
Next, we further assessed if this high dose of THIP prevents synaptic transmission by measuring neuronal Ca2+ transients evoked by electrical stimulation (Fig. 5). We recorded Ca2+ transients in the neocortex and CA1 region during evoked neuronal firing through a monopolar stimulating electrode using different stimulus durations under baseline aCSF followed by THIP (paired recordings, Fig. 5A,B). THIP reduced the full imaged area intensity (area under the curve) at the three largest stimulus intensities for the neocortex and hippocampus (all p values <0.03, paired t tests, n = 5 slices; Fig. 5C,D). Additionally, THIP significantly reduced the percent neuronal firing probability across all stimulus durations in the neocortex and hippocampus and shifted the cumulative incidence curves to the right (p < 0.0001; log-rank Mantel–Cox test, n = 5 slices; Fig. 5E,F). Thus, our data show that enhancing tonic inhibition with THIP (50 µM) significantly reduces neuronal firing in the neonatal neocortex and the CA1 region without completely silencing the network.
Enhancing tonic inhibition reduces stimulus-evoked activity in the neonatal neocortex and CA1 region. A, Maximal intensity projection of neonatal neocortex layer IV/V at baseline and after two stimulus durations in aCSF (top) and under THIP 50 µM (bottom). Yellow circles: ROIs. Yellow box: magnification of individual ROI. B, Normalized ΔF/F trace from the t-series (panel A) using the whole slice intensity to account for both neuropil and ROI activity. Traces in aCSF (solid) and THIP 50 µM (dashed). C, Left: area under the curve for matched individual slices in aCSF (filled circles) and in THIP (white circles) in the neocortex and (D) CA1 region. Paired t test, n = 5 slices, 3 mice. Right: Boltzmann fit to mean AUC values in aCSF (filled circles) and THIP (white circles). Neocortex: aCSF, Max, 1,719 (1,372, 2,092); W50, 6.76 (4.61, 9.30). THIP, Max, 930 (773, 1,149); W50, 9.79 (8.09, 13.07); R2 = 0.75 and 0.78. For CA1: aCSF, Max, 1,507 (1,321, 1,699); W50, 4.32 (3.74, 5.26). THIP, Max, 623 (480, 875); W50, 4.17 (2.97, 8.75). R2 = 0.81 and 0.54. Shaded area, 95% CI. E, Left: percent of neurons (ROIs) firing in matched individual slices in aCSF (filled circles) and in THIP (white circles) in the neocortex and (F) CA1 region. Percent cells firing calculated as % of maximal cell response from all given stimulation durations. Right: cumulative incidence curves for the percent of neurons firing in aCSF (filled circles) and THIP (white circles) in the neocortex (log-rank Mantel–Cox test, p < 0.0001; hazard ratio aCSF/THIP = 3.06 (2.18, 4.28), n = 123 ROIs, 5 slices, 3 mice] and hippocampus (log-rank Mantel–Cox test, p < 0.0001; hazard ratio aCSF/THIP = 6.95 (3.92, 12.3), n = 49 ROIs, 5 slices, 3 mice]. Black thick lines in C and D: means, except for neocortex 5 ms AUC (C, left) and hippocampus 20 ms AUC (D, left), which are geometric means. Circles in E and F probability plots: mean ± 95% CI.
Enhancing tonic inhibition reduces SLEs in the neonatal neocortex and CA1 region in a dose-dependent manner in the 4-AP seizure model
Approximately 30–50% of neonatal seizures resist phenobarbital treatment (Riss et al., 2008), a drug that enhances synaptic GABAARs. As THIP enhanced tonic conductances and decreased firing in the neonatal neocortex and CA1 PCs, we next studied THIP's efficacy in reducing SLEs in these two brain regions using 4-AP (100 µM) to induce epileptiform activity (Fig. 6A,B). We recorded SLEs in acute brain slices at P5–8 and quantified them using a FFT (Glykys et al., 2009). As a comparison age group, we also recorded THIP's effect on adult brain slices, as GABA is mainly inhibitory in adults. After 30 min of stable SLEs, THIP decreased SLE power in a dose-dependent manner in the neonatal and adult neocortex and CA1 regions (Table 2). A low dose of THIP (1 µM) did not alter SLE in either brain region or age. THIP 10 µM significantly reduced SLE in both adult regions but was only effective in the neonatal neocortex and not in the CA1 region. THIP 50 µM significantly decreased SLE in both areas and ages (p = 0.009 for pup neocortex and CA1 region, paired t test, n = 7 and 8 slices; Fig. 6C, Table 2). The proportional reduction in SLE by the various THIP doses was statistically different in the neocortex (p < 0.0001, one-way ANOVA) and CA1 region (p = 0.008, Kruskal–Wallis test and p < 0.0001 one-way ANOVA) in both pups and adults, with 50 µM THIP showing the largest effect (Fig. 6D,E). The effect of THIP on SLE was not different between adults and pups (Fig. 6D,E). DZP (2 µM), a synaptic-GABAAR agonist binding to the α/γ interface (Sigel and Ernst, 2018), did not alter SLE in the neonatal neocortex (p = 0.92, paired t test, n = 8, 3 mice) or CA1 region (p = 0.20, paired t test, n = 6, 3 mice; Fig. 6D,E). In control conditions, there was no rundown of SLE in the neonatal neocortex under 4-AP alone for 60 min (30′ FFT power: 20.2 µV/Hz (15.8, 24.6); 60′: 19.1 µV/Hz (14.8, 23.3); p = 0.156, WSRT, n = 7, 3 mice).
Enhancing tonic inhibition reduces SLEs in the neonatal and adult neocortex and CA1 regions. A,B, Top: SLEs induced by 4-AP (100 µM) in the neocortex (layer IV/V) (A) and CA1 region (stratum pyramidale) (B) recorded with an extracellular field electrode in acute brain slices (P8). Middle: magnification of boxed segments from baseline and THIP perfusion, respectively. Bottom: FFT power area from top traces calculated every 30 s. C, FFT power area during baseline and after THIP perfusion (50 µM) in the neonatal neocortex (left) and CA1 region (right). Neocortex: n = 7 slices, 3 mice. CA1: n = 8 slices, 3 mice; paired t test. D, Ratio change in SLE power comparing various THIP concentrations (µM) in the pup neocortex (one-way ANOVA, F(2,23) = 23.97, p < 0.0001, Tukey post hoc test; left-to-right, n = 12, 7, 7) and the adult neocortex (one-way ANOVA, F(2,29) = 14.9, p < 0.0001, Tukey post hoc test; left-to-right, n = 13, 9, 10). DZP p = 0.713 (one-sample t test, n = 8). E, Same as D but in the pup CA1 region (Kruskal–Wallis test, p = 0.0008, Dunn's test; left-to-right, n = 7, 6, 8) and adult CA1 region (one-way ANOVA, F(2,23) = 20.8, p < 0.0001, Tukey post hoc test; left-to-right, n = 9, 9, 8). DZP p = 0.247 (one-sample t test on log transformation, n = 6). White circles: individual slices. F, Number of SLEs during baseline and THIP in pups. Note the lack of events during THIP (50 µM) perfusion. White circles: individual slices. Black lines: mean ± 95% CI, except DZP CA1 (E) and number of events in the neocortex (F), which are median ± IQR.
Effect of enhancing tonic inhibition on seizure-like activity in acute brain slices of neonatal and adult neocortex and CA1 region in 4-AP
We also compared epileptiform event length, inter-event interval, peak amplitude, and the total number of events (Fig. 6F) for each dose, age, and brain region (Table 3). The total number of epileptiform events was significantly reduced by 10 and 50 µM THIP in the neocortex of pups and adults. For CA1, only 50 µM THIP decreased the number of events in both pups and adults (Table 3). Changes in the other event characteristics were variable, becoming more pronounced with 50 µM. THIP (50 µM) completely silenced epileptiform activity in both neonatal brain regions (Fig. 6F, Table 3). The mean number of ictal events (event ≥10 s) was reduced to zero with 50 µM THIP in the neonatal neocortex [baseline, 1.83 (0.80, 2.87), n = 6; 6/7 slices with ictal events]. The neonatal CA1 region did not have ictal events except a single event in 1/8 slices. In adults, THIP (50 µM) reduced the number of ictal events in the neocortex [baseline, 8.13 (3.55, 12.7); THIP, 2.63 (0.26, 4.99); paired t test, p = 0.035, n = 8; 8/10 slices with ictal events] but not CA1 [baseline, 8.67 (1.17, 16.2); THIP, 2.17 (−1.25, 5.58)]; paired t test, p = 0.11, n = 6; 6/8 slices with ictal events and only 1 worsened with THIP]. Thus, our data demonstrate that enhancing tonic inhibition with THIP significantly reduces SLEs in the neonatal neocortex and CA1 region in a dose-dependent manner with similar effectiveness as in the adult brain.
Effect of enhancing tonic inhibition on seizure-like event characteristics in the neocortex and CA1 region of adult and neonatal acute brain slices in 4-AP
THIP reduces seizure-like activity in the neonatal neocortex and CA1 region in the low-Mg2+ model of pharmacoresistant seizures
We next examined whether enhancing tonic inhibition reduces SLEs in the low-Mg2+ model of pharmacoresistant seizures to verify our 4-AP model findings (Walther et al., 1986; Zhang et al., 1994). Prolonged (1 h) exposure to this medium renders phenobarbital ineffective at reducing SLE (Glykys et al., 2009; Dzhala et al., 2010). Acute brain slices were incubated for 1 h in a low-Mg2+ medium before recording. Based on our findings above, we focused on neonates and 50 µM THIP. We recorded SLEs during baseline (30 min), THIP (30 min), and washout (15–30 min) in the neocortex and CA1 region (Fig. 7A,B). THIP (50 µM) significantly reduced but did not abolish SLEs in the neonatal neocortex and CA1 region (p < 0.0001, Friedman test, n = 8, and p = 0.0083, One-way ANOVA, n = 7, respectively; Fig. 7C). During the washout period, the FFT power increased and was not statistically different from the baseline. There was no statistical difference in THIP efficacy between the neocortex and CA1 region (p = 0.234, unpaired t test; Fig. 7D).
Enhancing tonic inhibition reduces SLEs in the neonatal neocortex and CA1 region in the low-Mg2+ model of pharmacoresistant seizures. A,B, Top: SLEs induced by low-Mg2+ aCSF in the neocortex, layer IV/V (A), and CA1 region (stratum pyramidale) (B) recorded with an extracellular field electrode in acute brain slices (P8). Middle: magnification of boxed segments from baseline, THIP perfusion, and washout (W/O). Bottom: FFT power area from top traces calculated every 30 s. C, FFT power area during baseline, THIP, and washout conditions in the neonatal neocortex (left) and CA1 region (right). Neocortex: Friedman test, p < 0.0001, Dunn test, n = 8. CA1, RM one-way ANOVA, F(2,12) = 7.33, p = 0.0083, Tukey post hoc test, n = 7. D, Ratio change in SLE with THIP (unpaired t test, p = 0.234). E, Number of SLEs during baseline, THIP perfusion, and W/O in the pup neocortex (RM one-way ANOVA, F(2,14) = 24.4, p < 0.0001, Tukey’s test, n = 8, 3 animals) and CA1 region (RM one-way ANOVA, F(2,12) = 10.8, p = 0.0021, Tukey’s test, n = 7, 3 animals). Lines link paired individual slices (white circles). White circles: individual slices. Black lines: mean ± 95% CI, except for the neocortex power area (C), which is median ± IQR.
The total number of epileptiform events decreased in the neocortex and CA1 region with THIP (50 µM) (Fig. 7E; Table 4). Event length decreased, and inter-event interval increased in the neonatal neocortex and CA1 region during THIP perfusion (Table 4). The number of ictal events per slice (event ≥10 s) was reduced to zero with THIP (50 µM) in the neocortex and increased upon washout [baseline, 0.63 (0.003, 1.25); THIP 0; washout, 0.25 (−0.14, 0.64); n = 5; 5/8 slices with ictal events]. The CA1 region behaved similarly without the events being completely abolished [baseline, 5.57 (2.80, 8.34); THIP, 0.14 (−0.21, 0.49); washout, 0.29 (−0.17, 0.74); RM one-way ANOVA, F(1.03, 6.16) = 25.7, p = 0.002, Tukey post hoc test p = 0.005 between baseline and THIP and p = 0.006 between baseline and washout; n = 7]. Thus, our data demonstrate that enhancing tonic inhibition also significantly reduces SLEs in the neonatal neocortex and CA1 region in the Low-Mg2 model of pharmacoresistant seizures.
Effect of enhancing tonic inhibition on seizure-like event characteristics in the neocortex and CA1 region of neonatal acute brain slices in low-Mg2+
Discussion
Our results can be summarized as follows: (1) tonic conductances are present in the neonatal CA1 region and neocortex, and THIP enhances these conductances. (2) Enhanced tonic inhibition reduces spontaneous and evoked neuronal Ca2+ transients. (3) Enhancing tonic inhibition reduces SLEs in the adult and neonatal neocortex and CA1 regions in two different in vitro seizure models.
We found that neocortical and CA1 pyramidal neurons have inhibitory tonic conductances during early brain development (P5–8), which was enhanced by THIP at 10 and 50 µM. Tonic currents have been measured previously in the neonatal neocortex, and δ- and α5-containing extrasynaptic GABAARs are expressed early on during development (Kilb et al., 2013). Our recordings show that CA1 and neocortical layer V PCs tonic conductances increased with THIP at 10 and 50 µM, the last one being close to THIP's EC50 [44 µM in slices (Drasbek and Jensen, 2006) and 51.8 µM in oocytes (Stórustovu and Ebert, 2006)]. Our results are at odds with the study of Sebe et al., 2010, which found that THIP, at 1 µM, increased the tonic currents in only 1 out of 11 neonatal neocortical layer V neurons while decreasing it in the rest of the recorded cells. The reason for this discrepancy is unclear, as THIP is an agonist that is expected to increase tonic currents, and multiple prior studies have found that THIP enhances tonic inhibitory currents (Karobath and Lippitsch, 1979; Drasbek and Jensen, 2006; Meera, 2011). It is uncertain if THIP 50 µM also enhances α5-mediated tonic currents in the newborn brain. In oocytes, the EC50 of THIP on α5-containing GABAAR varies between 40 and 129 µM depending on the β-subunit composition and above 143 µM for α1- and α3-containing GABAAR (Ebert et al., 1994). While off-target effects are likely at these higher doses of THIP (Meera, 2011), most importantly, our results demonstrate that CA1 and neocortical layer V PCs have a tonic conductance during early development that is enhanced by THIP regardless of the exact subunit composition of its target receptors. Future studies will address the role of α5 versus δ-containing GABAAR mediating inhibitory tonic conductances in the neonatal period.
The modulation of tonic inhibition and how it affects a neuronal network is complex (Farrant and Nusser, 2005; Glykys and Mody, 2007b). GABAAR subunit composition varies between brain regions and neuron subtypes (Sperk et al., 1997), and the δ- and α5-containing extrasynaptic GABAARs are present in principal cells and interneurons (Sun et al., 2004; Kilb et al., 2013). Also, within one cell type, more than one receptor may contribute to tonic conductances, depending on the level of extracellular GABA (Scimemi et al., 2005; Glykys et al., 2008). Most of the interneurons in the hippocampus are PV+ and δ-subunit containing, responding to THIP (Lee and Maguire, 2013; Yu et al., 2013), forming partnerships with α1-containing GABAARs (Glykys et al., 2007). On the other hand, a few number of somatostatin interneurons express δ-subunits (Milenkovic et al., 2013), while CCK+ interneurons do not express α1-GABAARs thus most likely not expressing δ-containing GABAARs (Ferando and Mody, 2014). Therefore, changes in tonic current can result in various network alterations, including gamma oscillations, which are hard to predict (Ferando and Mody, 2015). Reducing the tonic inhibitory current in hippocampal interneurons enhanced their excitability and the inhibitory drive to PCs (Semyanov et al., 2003). Yet, when we used large THIP doses to increase tonic inhibition, the network's net effect was dampened, decreasing seizure-like activity. Thus, our data suggest that large increases in tonic conductance may be needed to reduce clinical seizures during status epilepticus.
Despite a depolarizing EGABA in neonates, we demonstrate that significantly augmenting neonatal tonic inhibition reduces neuronal activity in the form of spontaneous and evoked Ca2+ transients and shifts the input–output curves to the right. Previously, enhancing tonic inhibition was shown to decrease neuronal firing in adult mice by shifting their input–output curves, similar to our findings in neonates (Mitchell and Silver, 2003). This is not a surprise as tonic conductances mediate ∼75% of the total amount of inhibition that neurons receive, and these conductances exert a lasting influence over resting membrane potential compared to synaptic GABAARs (Mody and Pearce, 2004; Cherubini, 2012). We demonstrated that THIP 50 µM raises the rheobase, hyperpolarizes the membrane potential, and decreases the input resistance, making these neurons less excitable resulting in shunting inhibition (Staley and Mody, 1992; Song et al., 2011). This is not unexpected, as altering tonic currents can affect the membrane potential (Cope et al., 2005). Interestingly, the frequency of spontaneous excitatory synaptic currents was not different when THIP was present compared to baseline, yet their frequency was low at baseline, and we speculate that some of them may not be action potential mediated. Significantly, THIP 1 µM did not worsen SLEs in neonatal slices, suggesting that the THIP's net effect is not pro-convulsive at this low dose, even when GABA has a more depolarizing action at this age.
We also showed that a high dose of THIP significantly enhanced tonic inhibition yet did not entirely silence neuronal firing. The large tonic conductances under THIP were recorded using a CsCl internal solution, which enhances Cl−-mediated currents. However, when using a more physiological K-gluconate internal solution, 50 µM THIP shifted the input–output curve to the right without affecting spontaneous excitatory synaptic transmission. Also, using Ca2+ imaging, we showed that while the input–output curves shifted to the right under THIP, neuronal activity was still present and that spontaneous firing still occurred at baseline. Lastly, during seizures induced in Low-Mg2+, THIP (50 µM) significantly decreased SLE but did not abolish it altogether. Thus, enhancing tonic inhibition to a high degree does not silence neuronal networks but can reduce seizure-like activity.
Seizure-like activity in acute brain slices decreased with THIP in a dose-dependent manner. A low THIP dose (1 µM) did not reduce SLEs in neonates or adults in the neocortex or CA1 region. This is consistent with prior reports showing the ineffectiveness of THIP and other δ-GABAAR agonists at low doses (Hansen et al., 2004; Sharopov et al., 2019). THIP increased its efficacy in decreasing SLEs at higher doses. Interestingly, we observed the same effect at 50 µM in neonates and adults, suggesting that enhancing tonic inhibition may be a viable therapeutic option at both ages. A recent phase 2 clinical trial in patients 12 years and older evaluated different doses of ganaxolone, a synthetic neuroactive steroid that acts on GABAA extrasynaptic receptors, and showed promising results for treating refractory status epilepticus (Vaitkevicius et al., 2022). Therefore, enhancing tonic inhibition may also be a valuable approach in treating refractory neonatal seizures.
In the low-Mg2+ model of neonatal pharmacoresistant seizures, we similarly observed that enhancing tonic inhibition reduced SLEs in the neocortex and CA1 region. The number of epileptiform events was not completely abolished, as was observed in the 4-AP model. Still, the total event number was significantly reduced with THIP perfusion. Reasons for this difference could include the severity of seizures in the low-Mg2+ model, which represents more of a pharmacoresistance model displaying late recurrent discharges resistant to anticonvulsants not seen in the 4-AP model (Perreault and Avoli, 1991; Zhang et al., 1995). These two models also have differences in their mechanisms, as 4-AP functions as a K+-channel blocker inducing intrinsic depolarization in neurons compared to removing the Mg2+ block of NMDA receptors in the low-Mg model. The latter also has a relative energy failure that may contribute to the transformation to late recurrent events (Schuchmann et al., 1999). Notwithstanding, THIP decreased neonatal seizure-like activity in both models.
We previously showed that phenobarbital, the first-line anticonvulsant for neonatal seizures and a drug that acts on synaptic receptors, was ineffective in both of these models (Glykys et al., 2009; Dzhala et al., 2010). Consistent with these prior results, we demonstrated that DZP did not lessen neonatal seizures in the neocortex or CA1 region. The lack of an effect of phenobarbital and DZP on neonatal seizures is related, among multiple reasons, to high neuronal Cl− during early development (Ben-Ari, 2002) and an activity-dependent accumulation of Cl− (Dzhala et al., 2010). Thus, intraneuronal Cl− accumulation during seizures makes EGABA even more depolarizing (thus more excitatory), contributing to decreased efficacy of synaptic GABAergic anticonvulsants (Burman et al., 2022). Also, during prolonged seizures in cultured neurons, synaptic GABAARs are rapidly internalized, further reducing the effectiveness of GABAergic anticonvulsants like phenobarbital (Goodkin, 2005; Goodkin et al., 2007). In contrast, the cell-surface expression of extrasynaptic GABAARs and the corresponding tonic current density are maintained at these acute time points (Yu et al., 2013).
Research on the changes in δ- and α5-containing GABAARs during status epilepticus in neonates is an area that needs further studies. Acute and subacute changes of some GABAAR subunits in the hippocampus occur after kainic acid in P9 rats, but not all subunits have been studied (Laurén et al., 2005). This group observed lower levels of α4 subunit in CA3c and dentate gyrus, no change in α5, and low levels of γ2 in all hippocampal regions 6 h after kainic injection. Unfortunately, the δ-subunit was not evaluated. Our electrophysiological data suggests that tonic conductances can be enhanced during the most acute period of seizures to decrease seizure-like activity, suggesting that even if some extrasynaptic subunits are internalized, there are enough to be activated. More research is needed to understand GABAAR subunit changes during status epilepticus in neonates. Yet, our functional studies suggest that enhancing tonic inhibition may be better for treating refractory neonatal seizures.
A limitation of our study is that we did not perform in vivo studies to determine if enhancing tonic inhibition would suppress seizure activity. THIP, at low doses, did not decrease seizures in the adult pentylenetetrazole kindling model (Simonsen et al., 2017). Also, most recently, THIP showed no difference from placebo regarding the Clinical Global Impression severity score to treat Angelman syndrome (Keary et al., 2023). However, the doses previously used have not been selected to treat status epilepticus. Future in vivo studies will address whether high doses of THIP or other tonic inhibitory conductance-enhancing drugs can decrease neonatal seizures. While THIP has a “sedative” side effect, patients with refractory neonatal seizures are intubated, making this less of a concern.
In conclusion, we demonstrate that neonatal PCs in the neocortex and CA1 region have tonic conductances that THIP augments. This significant increase in tonic inhibition makes these neurons less excitable and decreases seizure-like activity in two different in vitro seizure models. Thus, enhancing tonic inhibition could be a viable treatment for neonatal seizures, especially in pharmacoresistant cases.
Footnotes
J.G. was funded by NIH/NINDS R01NS115800 and the Iowa Neuroscience Institute. This research was partly supported by the University of Iowa Hawkeye Intellectual and Developmental Disabilities Research Center (HAWK-IDDRC) P50 HD10355.
The authors declare no competing financial interests.
- Correspondence should be addressed to Joseph Glykys at joseph-glykys{at}uiowa.edu.
