A Missense Mutation of the Gene Encoding Synaptic Vesicle Glycoprotein 2A (SV2A) Confers Seizure Susceptibility by Disrupting Amygdalar Synaptic GABA Release.

Synaptic vesicle glycoprotein 2A (SV2A) is specifically expressed in the membranes of synaptic vesicles and modulates action potential-dependent neurotransmitter release. To explore the role of SV2A in the pathogenesis of epileptic disorders, we recently generated a novel rat model (Sv2a(L174Q) rat) carrying a missense mutation of the Sv2a gene and showed that the Sv2a(L174Q) rats were hypersensitive to kindling development (Tokudome et al., 2016). Here, we further conducted behavioral and neurochemical studies to clarify the pathophysiological mechanisms underlying the seizure vulnerability in Sv2a(L174Q) rats. Sv2a(L174Q) rats were highly susceptible to pentylenetetrazole (PTZ)-induced seizures, yielding a significantly higher seizure scores and seizure incidence than the control animals. Brain mapping analysis of Fos expression, a biological marker of neural excitation, revealed that the seizure threshold level of PTZ region-specifically elevated Fos expression in the amygdala in Sv2a(L174Q) rats. In vivo microdialysis study showed that the Sv2a(L174Q) mutation preferentially reduced high K(+) (depolarization)-evoked GABA release, but not glutamate release, in the amygdala. In addition, specific control of GABA release by SV2A was supported by its predominant expression in GABAergic neurons, which were co-stained with antibodies against SV2A and glutamate decarboxylase 1. The present results suggest that dysfunction of SV2A by the missense mutation elevates seizure susceptibility in rats by preferentially disrupting synaptic GABA release in the amygdala, illustrating the crucial role of amygdalar SV2A-GABAergic system in epileptogenesis.

Introduction

Synaptic vesicle glycoprotein 2A (SV2A) is highly expressed in the brain including the cerebral cortex, limbic regions, and cerebellum, where it modulates action potential-dependent neurotransmitter release (Crowder et al., 1999; Janz et al., 1999; Xu and Bajjalieh, 2001; Custer et al., 2006; Chang and Südhof, 2009). Although the functional mechanisms of SV2A remain to be clarified, it is suggested that SV2A primes synaptic vesicles to fully respond to Ca2+ probably by interacting with the Ca2+ sensor protein synaptotagmin (Xu and Bajjalieh, 2001; Chang and Südhof, 2009; Nowack et al., 2010). In addition, previous studies suggest that SV2A plays an important role in the pathogenesis and treatment of epileptic disorders. This is because (1) SV2A-knockout mice exhibited severe seizures (Crowder et al., 1999; Janz et al., 1999), (2) SV2A serves as a specific binding site for certain antiepileptics (e.g., levetiracetam and its analogs; Lynch et al., 2004; Pollard, 2008; Kaminski et al., 2009; Correa-Basurto et al., 2015; Klitgaard et al., 2016) and (3) the expressional levels of SV2A are reported to be altered in various epileptic conditions both in animals (e.g., chemically- and electrically induced kindling) and humans (e.g., intractable temporal lobe epilepsy and focal cortical dysplasia; Matveeva et al., 2007; Feng et al., 2009; Ohno et al., 2009a, 2012b; Toering et al., 2009; van Vliet et al., 2009; Crèvecoeur et al., 2014; Serajee and Huq, 2015). Furthermore, a recent clinical study reported that a missense mutation (R383Q) in exon 5 of the SV2A gene resulted in intractable epilepsy, involuntary movements, microcephaly and developmental retardation (Serajee and Huq, 2015).

In order to explore the role of SV2A in modulating development of epileptic disorders (epileptogenesis), we recently generated a novel rat model (Sv2a rat) carrying a missense mutation (L174Q) in the Sv2a gene (Tokudome et al., 2016), using gene-driven ENU mutagenesis/MuT-POWER techniques (Mashimo et al., 2008). Sv2a rats were susceptible to PTZ seizures and to kindling development associated with repeated PTZ treatments or focal electrical stimulation of the amygdala. In addition, the Sv2a mutation significantly reduced depolarization-induced GABA release in the hippocampus. These findings suggest that SV2A plays a crucial role in the kindling epileptogenesis possibly by interacting GABAergic neurons. However, the detailed mechanisms underlying the regulation of seizure susceptibility by SV2A remain to be clarified.

In the present study, therefore, we further conducted behavioral and neurochemical studies to clarify the mechanisms (e.g., responsible brain regions and influences on synaptic amino acid release) underlying the seizure vulnerability in Sv2a rats. The present results show that the Sv2a mutation elevates excitability of the corticolimbic neural circuit, especially in the amygdala, by preferentially disrupting synaptic GABA release, illustrating the crucial role of amygdalar SV2A-GABAergic system in epileptogenesis.

Materials and Methods

Animals

Male Sv2a rats (Tokudome et al., 2016) were obtained from the National BioResource Project-Rat (F344-Sv2a NBRP-Rat No:0668). The Sv2a rat, carrying a single nucleotide mutation T521A, was first identified in a gene-driven ENU mutagenesis project in Kyoto University (Mashimo et al., 2008). Thereafter, Sv2a rats were backcrossed more than five generations on the F344/NSlc inbred background to eliminate mutations potentially induced by ENU mutagenesis elsewhere in the genome. Age-matched male F344 rats (Japan SLC, Shizuoka, Japan) were used as the control animal. The animals were kept in air-conditioned rooms under a 12-h light/dark cycle and allowed ad libitum access to food and water. All animal experiments were approved by the Animal Research Committees of Osaka University of Pharmaceutical Sciences and were conducted according to the Institutional Committees’ regulations on animal experimentation.

Evaluation of Seizure Susceptibility

To evaluate the seizure sensitivity, Sv2a or F344 rats were treated with an intraperitoneal dose of PTZ (30, 35, and 40 mg/kg for Sv2a rats; 35, 40, 45, and 50 mg/kg for F344 rats). PTZ-induced seizures were evaluated over 20 min after the drug treatment using a 6-point ranked scale as follows, 0: none response, 1: facial automatisms and twitching of the ears and whiskers, 2: convulsive waves propagating axially along the trunk, 3: myoclonic convulsions with a delay, 4: clonic convulsions, 5: repeated powerful clonic-tonic or lethal convulsions (Racine, 1972; Franke and Kittner, 2001; Kudryashov et al., 2007). The incidence of seizures was judged as positive when the animal showed a seizure score of 3 or more.

Analysis of Fos Protein Expression

To explore causal brain regions for PTZ seizures in Sv2a rats, expression of Fos protein, a biological marker of neural excitation, by the seizure threshold level of PTZ was analyzed. For this purpose, Sv2a and F344 rats were cumulatively injected with an increasing dose of PTZ (first dose: 10 mg/kg, second dose: 20 mg/kg, third dose: 30 mg/kg) with 30-min intervals. The incidence of seizures was monitored for 10 min immediately after each PTZ injection using a seizure scale described previously. Since the PTZ-induced seizures were observed only at 30 mg/kg, brain samples were obtained 2 h after the PTZ (30 mg/kg) injection under pentobarbital (80 mg/kg, i.p.) anesthesia.

After fixation with 4% formaldehyde solution, coronal sections (30 μm thickness) were cut from each brain using a Microslicer (DSK-3000, Dosaka, Kyoto Japan). The immunostaining of Fos protein was performed using a previously published method (Ohno et al., 2011, 2012a). Briefly, formalin-fixed sections were immunohistochemically stained with a goat c-Fos antiserum (Santa Cruz Biotechnology Inc., Santa Cruz, CA, USA) by the ABC method. Fos-immunoreactivity was visualized by the diaminobenzidine–nickel staining method and quantified by counting the number of Fos-positive neurons. Brain regions analyzed includes (1) cerebral cortices: mPFC, CgC, MC, SC, AIC, Pir, AuC, PRh-Ent; (2) basal ganglia and limbic regions: AcC, AcS, dlST, dmST, GP, LS, CA, DG, PMCo, BMA, BLA; (3) diencephalon: LHb, PT, AM, CM, VM, AH, PH, DMH (see Paxinos and Watson, 2007).

In Vivo Microdialysis Study

Sv2a or F344 rats were anesthetized with pentobarbital (40 mg/kg, i.p.) and fixed in a stereotaxic instrument (Narishige, SR-6, Tokyo, Japan). A guide cannula (1 mm diameter) was inserted into a position 2 mm above the amygdala (P: 2.8 mm, L: 4.8 mm, H: -6.1 mm; Paxinos and Watson, 2007) and fixed to the skull using dental cement. After a recovery period of about 1 week, animals with a chronically implanted guide cannula were subjected to microdialysis experiments. Briefly, a dialysis probe (Eicom, A-I-10-02, Kyoto, Japan) was inserted into the amygdala through a guide cannula and aCSF containing (in mM): NaCl 140, KCl 2.4, MgCl2 1.0, CaCl2 1.2, NaHCO3 5.0, was perfused at a flow rate of 1 μL/min using a microperfusion pump (Eicom, ESP-32, Kyoto, Japan). The dialysate samples were collected into a microtube every 10 min (10 μL/sample). To evaluate the depolarization-evoked synaptic release, high concentration (100 mM) K+-containing aCSF was perfused for 60 min through the dialysis probe.

The dialysate samples were analyzed for GABA and glutamate levels using a HPLC-ECD system. GABA and glutamate were derivatized with o-phthalaldehyde before the HPLC injection and separated on a cation exchange column (Eicom, 3.0φ × 150 mm; Eicompak SC-5ODS, Kyoto, Japan). The mobile phase consisted of 0.1 M phosphate buffer, 5 mg/L EDTA 2Na, pH6.0, with 27% methanol pumped at a flow rate of 500 μL/min. All data were analyzed by using eDAQ Power Chrom (eDAQ Pty Ltd, Denistone East, NSW, Australia). Extracellular GABA and glutamate levels were expressed as a percentage of the basal control level, which was the mean of the three points before the high K+ application, in each animal. The AUC of the high K+-evoked GABA or glutamate release was also estimated by the trapezoidal approximation method.

Immunofluorescence Double Staining

Sv2a or F344 rats were decapitated under pentobarbital (80 mg/kg, i.p.) anesthesia and brains were removed from the skull. After fixation with 4% paraformaldehyde solution for 24 h, the brain was dehydrated and embedded in paraffin. Formalin-fixed and paraffin-embedded amygdaloid tissues were cut into 4 μm thick sections and the sections were subjected to immunofluorescence double staining with anti-SV2A and anti-Gad1 (Ohno et al., 2012b). Sections were incubated with a goat anti-rat SV2A (dilution 1:500, Santa Cruz Biotechnology, Dallas, TX, USA) and mouse anti-human Gad1 (dilution 1:1000, Santa Cruz Biotechnology, Dallas, TX, USA) for 42 h at 4°C, and then with an FITC (green fluorescence)-conjugated rabbit anti-goat IgG secondary antibody (dilution 1:500, Sigma–Aldrich, St. Louis, MO, USA) and TRITC (red fluorescence)-conjugated rabbit anti-mouse IgG secondary antibody (dilution 1:500, Sigma–Aldrich, St. Louis, MO, USA) to probe SV2A and Gad1, respectively. Immunofluorescence images were obtained with a confocal laser scanning microscope (Carl Zeiss Japan, LSM 700 ZEN, Tokyo, Japan). To quantify SV2A and Gad1 expression, digital images of the amygdala were stored and the integrated optical density was measured by computer analysis with ImageJ software (ver. 1.42, NIH).

Statistical Analysis

Statistical significance of differences between two groups was performed by Mann–Whitney’s U-test (behavioral scores) or Student’s t-test (Fos and SV2A expression). Comparisons of seizure incidence rate were done by X2 test. Differences in GABA and glutamate release (in vivo microdialysis) were analyzed by two-way ANOVA followed by Tukey’s post hoc test. A P-value of less than 0.05 was considered statistically significant.

Results

Seizure Susceptibility of Sv2a Rats

Sv2a rats have a missense mutation T521A in the Sv2a gene, which results in the substitution of Leu 174 to Gln (Figure ). The mutation site is located in the first transmembrane region of SV2A that was reportedly essential for the normal function and structure of SV2A (Chang and Südhof, 2009). In order to confirm seizure susceptibility of Sv2a rats, we evaluated the responses of Sv2a and F344 rats to PTZ injections (30–50 mg/kg, i.p.). While gross behaviors of Sv2a rats were normal, these animals showed high susceptibility to PTZ seizures, yielding significantly higher seizure scores at 35 mg/kg [U(17) = 20.5, P < 0.05] and 40 mg/kg [U(17) = 12.0, P < 0.01], and a higher seizure incidence at 40 mg/kg (X2 = 6.34, P < 0.05) than F344 rats (Figure ).

Seizure susceptibility of (A) Structure of SV2A showing the mutation site L174Q. SV2A possesses a 12-transmembrane structure which contains a putative binding site for synaptotagmin in the N terminal region and glycosylation sites in a long intravesicular loop between transmembrane regions 7 and 8. (B) Susceptibility of Sv2a rats to PTZ-induced seizures. Each point represents the mean ± SEM of 8–11 animals. ∗P < 0.05, ∗∗P < 0.01 significantly different from F344 rats.

Fos Expression Analysis

To explore the brain sites most sensitive to the seizure threshold level of PTZ, Sv2a and F344 rats were cumulatively injected with an increasing dose of PTZ (10, 20, or 30 mg/kg, i.p.) with 30-min intervals. Under these conditions, the first two does did not evoke any seizure either in Sv2a or F344 rats. However, subsequent 30 mg/kg PTZ induced clonic or tonic-clonic seizures in five out of seven Sv2a rats while none of seven F344 rats tested exhibited seizures.

We next compared the expression of Fos protein, a biological marker of neural excitation, in various regions of the brain between the Sv2a (PTZ seizure-positive, N = 5) and F344 (PTZ-negative, N = 7) rats (Figure ). Sv2a rats showed considerably higher Fos expression than F344 rats in most of the cerebral cortex, reflecting the generalized seizure property of PTZ seizures (Figure ). PTZ-induced Fos expression in Sv2a rats were statistically significant in the SC and Pir. On the other hand, the seizure threshold level of PTZ region-specifically elevated Fos expression in the amygdala among 21 subcortical regions examined including basal ganglia, limbic regions, and diencephalon (Figures ).

Topographical Fos expression analysis following PTZ-induced seizures. Schematic illustrations of the brain sections selected for quantitative analysis of Fos protein expression. Filled boxes in each section indicate the sample areas analyzed. Red boxes represent the regions where Fos expression was significantly elevated by PTZ treatments.

Regional changes in Fos expression levels by the seizure threshold level of PTZ. Graphs show Fos expression in the cerebral cortical regions (A), basal ganglia and limbic regions (B), and diencephalon (C). Typical photos of Fos expression in the amygdala are also shown in the right bottom. Each column represents the mean ± SEM of five Sv2a (PTZ seizure-positive) and seven F344 (PTZ seizure-negative) rats. ∗P < 0.05, ∗∗P < 0.01 significantly different from F344 rats. Scale bar: 100 μm.

GABA and Glutamate Release in the Amygdala

Since Sv2a rats exhibited a region-specific excitation of the amygdala by PTZ, we conducted in vivo microdialysis studies to evaluate synaptic release of GABA and glutamate in the amygdala. As shown in Figure , high K+ (depolarization) stimuli evoked GABA release both in Sv2a and F344 rats. However, the depolarization-evoked GABA release was largely diminished by the Sv2a mutation [F(1,218) = 56.72, P < 0.001] (Figure ). A comparison of AUC also revealed a significant reduction in depolarization-evoked GABA release in Sv2a rats. On the other hand, in contrast to GABA release, high K+-evoked glutamate release was not significantly affected by the Sv2a mutation (Figure ).

Synaptic GABA and glutamate release in the amygdala. (A) GABA release. (B) Glutamate release. Each point represents the mean ± SEM of 10 or 5 animals. Depolarization stimulation was given by applying high concentration (100 mM) K+-containing aCSF (High K+) for 60 min through the dialysis probe. AUC (the area under the curve) comparisons of GABA and glutamate release are shown in the right. ∗P < 0.05, ∗∗P < 0.01 significantly different from F344 rats. Each point represents the mean ± SEM of 10 or 5 animals.

SV2A and Gad1 Double Staining in the Amygdala

We further conducted immunofluorescence double staining of SV2A with Gad1 (also known as GAD67), a marker protein of GABAergic neurons, in the amygdala. As shown in Figure , SV2A (green) was mostly co-stained with Gad1 (red) both in Sv2a and F344 rats, illustrating a specific expression of SV2A in the amygdalar GABAergic neurons. In addition, there were no significant differences in expressional levels of SV2A and Gad1 between Sv2a and F344 rats (Figure ).

SV2A and Gad1 double staining in the amygdala. (A) Photos showing representative double staining of Gad1 (left: red), SV2A (center: green) in the amygdala of Sv2a and F344 rats. Merged photos (right) revealed a predominant co-expression of SV2A with Gad1 (GABAergic neurons) in the amygdala. (B) Expression levels of SV2A and Gad1 in Sv2a and F344 rats. ROD, relative optical density. Each column represents the mean ± SEM of four animals. Scale bar: 50 μm.

Discussion

The present study confirmed that Sv2a rats carrying a missense mutation, L174Q, in the Sv2a gene were highly sensitive to PTZ-induced seizures, supporting the notion that SV2A plays the crucial role in controlling seizure susceptibility. Although a complete deletion of SV2A is known to cause premature death with severe seizure incidence (Crowder et al., 1999; Janz et al., 1999), the behavioral phenotype of Sv2a rats mimicked those reported in heterozygous SV2A-deficient mice (Kaminski et al., 2009), implying that the Sv2a mutation causes a partial loss of the SV2A function.

Fos protein is the immediate early gene product and is widely used as a cellular marker of neural excitation. Specifically, brain mapping analysis of Fos expression is useful to identify brain regions related to disease conditions (e.g., epilepsy, emotional disorders and cognitive impairments) or responses to various pathophysiological and pharmacological stimuli (e.g., pain, body temperature, stress, and drug treatments; Morgan and Curran, 1989; Herrera and Robertson, 1996; Ohno et al., 2009b, 2011, 2012a, 2015; Mukai et al., 2013; Fumoto et al., 2014; Iha et al., 2016). In the present study, we treated animals with the seizure threshold dose (30 mg/kg, i.p.) of PTZ which first evoked seizures only in Sv2a rats. Under these conditions, the PTZ treatment region-specifically elevated Fos expression in the cerebral cortex (e.g., SM and Pir), amygdala. Since PTZ evoked generalized clonic or tonic-clonic seizures associated with a wide-spread excitation of the cerebral cortex (Figure ), we could not specifically identify the causative regions in the cortical regions. Nonetheless, the present results clearly illustrates that the amygdala is most sensitive to and potential seizure initiation site for PTZ seizures in Sv2a rats. In addition, in vivo microdialysis demonstrated that the Sv2a mutation preferentially impaired depolarization-evoked synaptic release of GABA in the amygdala without affecting glutamate release. These findings provide important information for our understanding of the SV2A function in modulating seizure susceptibility. Although the hippocampus was less sensitive to PTZ seizures than the amygdala, our results do not deny the potential role of hippocampus in seizure vulnerability of Sv2a rats since it is known that PTZ evokes seizures by activating limbic regions including the hippocampus (Humpel et al., 1993; Löscher and Ebert, 1996; Szyndler et al., 2009) and that, indeed, a sufficient dose (70 mg/kg, i.p.) of PTZ increases Fos expression in the hippocampus (Bastlund et al., 2005). In addition, we previously demonstrated that the Sv2a mutation also disrupted depolarization-evoked synaptic release of GABA in the hippocampus. Furthermore, the piriform cortex also seems to be partly involved in the seizure vulnerability in Sv2a rats since the threshold level of PTZ significantly elevated Fos expression in the posterior part of this structure.

Preferential modulation of synaptic GABA release by SV2A was further supported by the SV2A expression pattern in amygdala GABAergic neurons. In that, most of SV2A-immunoreactivity was expressed in the amygdala neurons and dendrites, which were co-stained with antibodies against SV2A and Gad1. These findings are consistent with our previous findings that SV2A was predominantly expressed in GABAergic neurons, but only rarely in glutamatergic neurons in the mouse hippocampus (Ohno et al., 2012b). Therefore, SV2A seems to specifically regulate synaptic GABA release both in the amygdala and hippocampus. Since the amygdala and hippocampus are potential causative sites for various epileptic disorders (Löscher and Ebert, 1996; Morimoto et al., 2004; Avoli and de Curtis, 2011), the SV2A-GABAergic system in these structures is likely to be involved in pathogenesis of SV2A-related epileptic disorders.

The detailed mechanisms underlying SV2A dysfunction by the Sv2a mutation are currently unknown. Interestingly, a previous study showed that the neighboring missense mutations (D179A and E182A) of the two charged polar amino acids to the non-polar alanine abolished the normal function of SV2A, possibly by disrupting the protein folding and/or trafficking into synaptic membranes (Chang and Südhof, 2009). Thus, the changes in polarization of the first transmembrane region by the substitution of hydrophobic non-polar Leu 174 to polar Gln may impair the integrity of SV2A function. In addition, we previously showed that the Sv2a mutation specifically reduced the expression level Syt1, the Ca2+ sensor protein modulating synaptic release, among exocytosis regulatory proteins examined (Tokudome et al., 2016). Since SV2A is suggested to prime synaptic vesicles by interacting with the Syt1 (Xu and Bajjalieh, 2001; Chang and Südhof, 2009; Nowack et al., 2010), the disruption of synaptic GABA release by the Sv2a mutation may result from reduced expression of Syt1. Further studies are required to delineate molecular mechanisms of SV2A dysfunction by the Sv2a mutation.

The present study supports the clinical view that dysfunction of SV2A is involved in the pathogenesis of epilepsy, including intractable temporal lobe epilepsy and focal cortical dysplasia epilepsy (Feng et al., 2009; Toering et al., 2009; van Vliet et al., 2009; Crèvecoeur et al., 2014). Indeed, a recent study showed that a missense mutation R383Q in the SV2A gene caused intractable epilepsy and involuntary movements, which were accompanied by developmental retardation (Serajee and Huq, 2015). Thus, the Sv2a rat may be useful for exploring the epileptogenic mechanisms of SV2A-related epileptic disorders. Furthermore, since SV2A is known as a specific binding site for certain antiepileptics (e.g., levetiracetam, brivaracetam, and seletracetam; Lynch et al., 2004; Pollard, 2008; Kaminski et al., 2009; Correa-Basurto et al., 2015; Klitgaard et al., 2016), the Sv2a rat may also be useful as a novel animal model for analyzing the action mechanisms of the levetiracetam-analogs.

Conclusion

We confirmed high susceptibility of Sv2a rats to PTZ seizures. Treatment of Sv2a rats with PTZ at threshold level specifically elevated Fos expression in the amygdala, suggesting that the amygdala is the potential site responsible for seizure vulnerability in Sv2a rats. In addition, the Sv2a mutation preferentially reduced depolarization-evoked GABA, but not glutamate, release in the amygdala. The preferential disruption of GABA release due to the Sv2a mutation was supported by the specific expression of SV2A in GABAergic neurons. The present study suggests that dysfunction of SV2A by the missense mutation elevates seizure susceptibility by disrupting amygdalar synaptic GABA release, illustrating the crucial role of the SV2A-GABAergic system in epileptogenesis.

Author Contributions

YO designed research. KT, TO, SS, RT, NK, TM, and YO performed pharmacological and neurochemical research. KT, TO, SS, RT, NK, TM, and YO analyzed data. KT, TM, TS, MS, and YO wrote the paper.

Conflict of Interest Statement

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

The reviewer TK and handling Editor declared their shared affiliation, and the handling Editor states that the process nevertheless met the standards of a fair and objective review.