Alteration of Extracellular Matrix Molecules and Perineuronal Nets in the Hippocampus of Pentylenetetrazol-Kindled Mice

The pathophysiological processes leading to epilepsy are poorly understood. Understanding the molecular and cellular mechanisms involved in the onset of epilepsy is crucial for drug development. Epileptogenicity is thought to be associated with changes in synaptic plasticity; however, whether extracellular matrix molecules—known regulators of synaptic plasticity—are altered during epileptogenesis is unknown. To test this, we used a pentylenetetrazole- (PTZ-) kindling model mouse to investigate changes to hippocampal parvalbumin- (PV-) positive neurons, extracellular matrix molecules, and perineuronal nets (PNNs) after the last kindled seizure. We found an increase in Wisteria floribunda agglutinin- (WFA-) and Cat-315-positive PNNs and a decrease in PV-positive neurons not surrounded by PNNs, in the hippocampus of PTZ-kindled mice compared to control mice. Furthermore, the expression of WFA- and Cat-315-positive molecules increased in the extracellular space of PTZ-kindled mice. In addition, consistent with previous studies, astrocytes were activated in PTZ-kindled mice. We propose that the increase in PNNs after kindling decreases neuroplasticity in the hippocampus and helps maintain the neural circuit for recurrent seizures. This study shows that possibility of changes in extracellular matrix molecules due to astrocyte activation is associated with epilepticus in PTZ-kindled mice.


Introduction
Epilepsy is one of the most common chronic neurological disorders with a prevalence of 0.4% to 1.0% [1]. To date, none of the drugs used to treat epilepsy are able to prevent its onset or reverse epileptogenesis [2]. Therefore, the current treatment of epilepsy merely prevents or suppresses seizures, rather than halting the process of developing epilepsy. The pathophysiological processes leading to epilepsy are poorly understood. A better understanding of the molecular and cellular mechanisms involved in the onset of epilepsy will help the development of therapeutic agents that alter or halt epileptogenesis.
Kindling is an experimental epilepsy model, where repeated electrical or chemical stimulation of certain fore-brain structures gradually causes stronger signals on electroencephalography and behavioral seizure activity [3,4]. Once established, kindling results in a permanent state of seizure susceptibility.
Although seizures can induce neuronal death, they also have nonfatal pathophysiological effects on neuronal structure and function [5]. For example, epileptogenesis is associated with changes to synaptic plasticity [6,7]. However, what causes and maintains these changes has not been clarified. The extracellular matrix (ECM)-which fills the extracellular domain of all organs and tissues and contains water and ions [8,9]-is thought to regulate synaptic plasticity; whether it is altered in the early stages of epilepsy is currently unknown.
ECM in the central nervous system (CNS) is formed from hyaluronic acid, tenescin-R, glycoproteins, and chondroitin sulfate proteoglycans (CSPGs) [8]. In the mature CNS, ECM molecules are distributed as neural granules or perineuronal nets (PNN) [8]. A PNN is a mesh-like special structure that covers the cell body, axonal origin, and proximal dendrites of specific neurons in the CNS [10]. In the cortex and hippocampus, PNNs mainly form around parvalbumin-(PV-) positive GABAergic neurons [10]. Although the exact function of PNNs is not known, they stabilize synaptic connections, regulate synaptic plasticity, participate in ECM linkage with the cytoskeleton, and promote neuronastrocyte interactions [11][12][13]. PNNs form late in development, and are thought to reduce synaptic plasticity leading to the termination of the critical period of the CNS. Indeed, it has been suggested that aggrecan, a component of PNNs, controls plasticity [14][15][16]. Once mature synapses are established and surrounded by aggrecan-containing PNNs, they are stable and poorly reorganized in adults [14,17].
Astrocytes-a group of specialized glial cells in the CNS that control ion concentrations, neurotransmitter homeostasis, metabolism, and synapse development and signaling [36]-are one of the major sources of both ECM components and modulators. During epileptic seizures, astrocytes respond rapidly by exhibiting a form and function change called reactive astrogliosis [37]. Whether reactive astrogliosis following seizures results in ECM remodeling-which plays a role in synaptic plasticity both during postnatal development and after injury-is currently unclear. Astrocytes control synaptic plasticity in glutamatergic synapses by releasing of D-serine [38]. Previous study has shown that the reduced Ca 2+ -dependent release of D-serine by astrocytes impairs initiation of synaptic plasticity [39]. Therefore, astrocyte can potentially play an important role in epileptogenesis.
The purpose of this study is to determine whether earlyonset seizure activity is associated with changes to ECM molecules and astrocytes in an experimental model of epilepsy. Increasing our understanding of ECM and astrocyte changes in vivo following an increase in seizure activity will help reveal the specific mechanistic role of astrocytes in epilepsy. The results obtained in this study may provide an opportunity to develop new therapeutic approaches to prevent seizures or their consequences.

Material and Methods
2.1. Animals. Eleven-week-old male mice (C57BL/6N) were used for experiments. Mice were housed five to a cage under standard laboratory conditions. All procedures related to animal maintenance and experimentation were approved by the Committee for Animal Experiments at Kawasaki Medical School Advanced Research Center and conformed to the U.S. National Institutes of Health (NIH) Guide for the Care and Use of Laboratory Animals (NIH Publication No. 80-23, revised in 1996). We purchased the mice from Charles River Laboratories (Kanagawa, Japan). The mice were housed in a room with a 12-hour light/dark cycle (light on 8:00 a.m. and off 8:00 p.m.) with temperature maintained at 23-26°C. They were provided with nesting material, food, and water ad libitum.
2.2. Pentylenetetrazole Kindling Procedure. All mice were randomized into two groups (n = 15). Pentylenetetrazole (PTZ; Sigma-Aldrich Japan, Tokyo, Japan) was dissolved in saline to prepare concentrations of 20 mg/mL. A dose of 40 mg/kg was injected intraperitoneal for a total period of 15 days. Vehicle control mice were injected with saline. Seizure events during a 30 min period after each PTZ injection were observed. The resultant seizures were scored as follows [40]: stage 0 (no response), stage 1 (ear and facial twitching), stage 2 (myoclonic body jerks), stage 3 (forelimb clonus, rearing), stage 4 (clonic convulsions, turn onto the side), and stage 5 (generalized clonic convulsions, turn onto the back). On day 16, mice were sacrificed and the brains were removed.

Tissue Preparation.
We did following methods in Ueno et al. [15,41]. Mice were anesthetized with a lethal dose of sodium pentobarbital (120 mg/kg, i.p.) and transcardially perfused with 25 mL of phosphate-buffered saline (PBS) followed by 100 mL of 4% paraformaldehyde in PBS (pH 7.4). Brains were dissected and postfixed overnight at 4°C in the above fixative. The brains were then cryoprotected in 15% sucrose for 12 h followed by 30% sucrose for 20 h at 4°C. Next, the brains were frozen in an optimum cutting temperature compound (Tissue-Tek; Sakura Finetek, Tokyo, Japan) using a slurry of normal hexane in dry ice. Serial coronal sections with a thickness of 40 μm were obtained at -20°C using a cryostat (CM3050S; Leica Wetzlar, Germany). The sections were collected in ice-cold PBS containing 0.05% sodium azide.

Immunohistochemistry.
We treated the cryostat sections with 0.1% Triton X-100 with PBS at room temperature for 15 min. After three washes with PBS, we incubated the sections with 10% normal goat serum (ImmunoBioScience Corp., Mukilteo, WA) in PBS at room temperature for 1 h, we washed them three times with PBS, and we incubated them overnight at 4°C in PBS containing biotinylated WFA (B-1355, Vector Laboratories; 1 : 200) and/or the antibodies described in Antibodies and Lectins. After washing with PBS, we incubated the sections with Alexa Fluor 594conjugated streptavidin (S11227; Molecular Probes, Eugene, OR) and/or the corresponding secondary antibodies (described in Antibodies and Lectins) at room temperature for 2 h. We rinsed the labeled sections again with PBS, and we mounted them on glass slides with VECTASHIELD medium (H-1400; Vector Laboratories, Funakoshi Co., Tokyo, Japan). We stored the prepared slides at 4°C until we used them in the microscopy analysis. We used the following secondary antibodies for visualization: Alexa Fluor 488-conjugated goat anti-mouse IgG (ab150113; Abcam, Cambridge, MA; 1 : 1,000), Alexa Fluor 594-conjugated goat anti-guinea pig (A-11076; Thermo Fisher Scientific, Waltham, MA; 1 : 500), FITCconjugated anti-mouse IgM (sc-2082, Santa Cruz Biotechnology, Santa Cruz, CA, 1 : 1,000), Texas Red-conjugated goat anti-rabbit (TI-1000; Vector Laboratories, Funakoshi Co., Tokyo, Japan), and streptavidin-conjugated Alexa Fluor 594 (S11227, Thermo Fisher Scientific; 1 : 1,000).

Microscopy Imaging.
To quantify the density of PV-and WFA-positive PNNs and analysis of ECM fluorescence intensity, we used a confocal laser scanning microscope (LSM700; Carl Zeiss, Oberkochen, Germany) to obtain images of stained sections. Images (1024 × 1024 pixels) were saved as TIFF files using the ZEN software (Carl Zeiss). Briefly, we performed the analysis using a 10x objective lens and a pinhole setting that corresponded to a focal plane thickness of less than 1 μm. For observing ECM molecules and GAD67-and VGLUT1-positive synaptic terminals, samples were randomly selected and high-magnification images using a 100x objective lens were acquired. Prior to capture, the exposure time, gain, and offset were carefully set to ensure a strong signal but to avoid saturation. Identical capture conditions were used for all sections. Images from whole sections were acquired using a 10x objective lens of a fluorescence microscope (BZ-X; KEYENCE, Tokyo, Japan), and we merged them using the KEYENCE BZ-X Analyzer software (KEYENCE).

Quantification of Labeled PNNs and ECM Molecules.
Brain regions were determined in accordance with the mouse brain atlas of Paxinos and Franklin [42]. From each mouse, four serial coronal dorsal hippocampal sections (from -0.8 mm to -1.8 mm relative to bregma) were selected and processed for staining. All confocal images were acquired as TIFF files and analyzed with the NIH ImageJ software (NIH, Bethesda, MD; http://rsb.info.nih.gov/nih-image/). Stained PNNs (soma size above 60 μm 2 ) were manually tagged and counted within the area of interest. Background intensity was subtracted using unstained portions of each section. Labeled neuron density was calculated as cells/mm 2 . Quantifications were performed by a blinded independent observer. To quantify the fluorescence intensity of ECMpositive molecules, GAD67-and VGLUT1-positive synaptic terminals, WFA-and Cat-315-positive PNNs, we selected four sections from each mouse brain and stained them as described in Sections 2.4 and 2.5. The ellipse circumscribing the WFA-and Cat-315-positive PNNs was traced manually on 8-bit images of each section, and the gray levels for WFA or Cat-315 labeling were measured using the ImageJ software, which was assigned arbitrary units (a.u.). We manually outlined the parts excluding PNNs and measured the gray level with NIH ImageJ. Background intensity was subtracted using unstained portions of each section. We acquired all confocal images as TIFF files and analyzed them with NIH ImageJ. We coded the slides, and a blinded independent observer quantified them.
2.8. Data Analysis. Statistical analysis was conducted using the SPSS software (IBM Corp., Tokyo, Japan). To test for the area and PTZ kindling effects, statistical significance was determined by two-way analysis of variance followed by Bonferroni t-tests. The statistical significance threshold was set at p < 0:05. Data are expressed as the mean ± SEM of 6 animals per group. To examine whether neurodegeneration occurred following seizures, we first looked for changes to NeuNpositive cells in the hippocampus in fully kindled mice (Figure 2  To examine whether PNNs are associated with PVpositive neurons in the PTZ-kindled mouse hippocampus, we carried out a quantitative analysis (Figures 4(d)-4(g)).

Results
In the CA1 and CA3 areas, we found that the percentage of PV-positive neurons not surrounded by both WFA-and Cat-315-positive PNNs was significantly lower in PTZkindled mice (Figure 4  with Cat-315-positive molecules in the PTZ-kindled mouse hippocampus ( Figure 6(d)).
To examine whether PTZ kindling affects immune activation in the hippocampus, the morphology of iba-1positive microglia in the CA1 was examined ( Figure 6(b), 6(B ′ ), and 6(g)). We found no significant difference in the morphology of iba-1-positive microglia between control and PTZ-kindled mice.

Glutamatergic and GABAergic Synaptic Terminals in the PTZ-Kindled Mouse Hippocampus.
To investigate whether synaptic terminals are altered in PTZ-kindled mice, we labeled excitatory (VGLUT1-positive) and inhibitory synaptic terminals (GAD67-positive) (Figures 7(a) and (b)). VGLUT1-and GAD67-positive synaptic terminals were found in the neuropil of the mouse hippocampus (Figure 7(b)). As expected, VGLUT1-positive synaptic terminals did not colocalize with GAD67-positive synaptic terminals (Figure 7(b)). We quantified VGLUT1-and GAD67positive signal intensities, excluding GAD67-positive neurons, in the hippocampus CA1 area and found that both VGLUT1 and GAD67 fluorescence intensities were higher in PTZ-kindled mice than in control mice (Figure 7

Discussion
The results of this study indicate that molecular components of the ECM, namely, WFA-and Cat-315-positive molecules, increase in the hippocampus of mice after kindling. We also found an increase in activated astrocytes in the hippocampus of PTZ-kindled mice, and we speculate that these activated astrocytes may be responsible for the increase in ECM molecule expression and secretion.
The present study showed an increase in WFA-and Cat-315-positive PNNs in the hippocampus after kindling acquisition. To our knowledge, no one has reported changes in the number of PNNs in PTZ-kindled mice. While previous studies have shown that aggrecan Cat-315-positive molecules increase from day 2 to 7 after status epilepticus in pilocarpine-induced epilepsy model rats [43] and phosphacan-positive PNNs decrease in the kainic acidinduced epilepsy model [44], neither of these studies looked at changes in Cat-315-positive PNNs. Although the results from these studies are not completely consistent with ours, all three studies demonstrate a change in PNNs using different animal models of epilepsy. Differences in the methods, animal species, and stimulation pathways used to induce epilepsy [45] may have contributed to the differences in results. It is widely thought that WFA-positive PNNs control plasticity. However, several reports have suggested the possibility that Cat-315-positive PNNs is involved in plasticity [14,41]. Aggrecan expression in PNNs signals the end of the critical period [17,46]. Therefore, this study suggested that the increase in PNNs after kindling decreases neuroplasticity in the hippocampus. Seizure activity is not randomly spread throughout the brain; it is generated and transmitted by specific anatomical pathways [47]. During development, many studies have shown that PNNs are formed in a stimulus-dependent manner that takes days [15,48,49]. Therefore, it makes sense that the number of PNNs increases to maintain the repetitive stimulation pathway caused by PTZ administration. Based on the role of PNNs in synaptic stability and their location around GABAergic interneurons, this structural change may contribute to the progression of epilepsy. The density of both WFAand Cat-315-positive PNNs was significantly higher only in the CA3 region. The clear reason is unknown. However, the CA3 region is suggested to play a critical role in generation of hippocampal epileptiform activity [50,51]. Moreover, it has been suggested that the CA3 area is susceptible to hyperexcitability [52].
In contrast, there was no change in the number of PVpositive interneurons after PTZ kindling, similar to results from other studies [44,53]. In the mouse hippocampus, around 70% of PV-positive neurons are surrounded by WFA-and Cat-315-positive PNNs, which is consistent with our results [54,55]. Our study is consistent with studies describing PNNs in the hippocampus in the past [54,56]. In this study, we show that the proportion of PV neurons covered with PNNs increases after kindling acquisition.
According to recent speculation, PNNs in the hippocampus do not turn over after maturation and can help maintain memories [57,58]. Because PNNs mainly surround PVpositive interneurons, the increase in PNNs can affect the GABAergic circuit [59]. In particular, since PNNs are thought to inhibit synaptic plasticity [14,17], the decrease in PV-positive neurons not surrounded by PNNs observed in this study indicates a decrease in plasticity in the hippocampus of mice after kindling acquisition. It has long been known that changes to synaptic plasticity are associated with epileptogenesis [6,7]. For instance, neural circuits that maintain high excitability are caused by changes in synaptic plasticity due to repetitive stimulation in patients with epilepsy and experimental animal models of epilepsy [60][61][62]. However, the underlying cause of this change in synaptic plasticity is unknown. We speculate that changes to the ECM, and in particular PNNs, may reduce synaptic plasticity in epilepsy.
The extracellular space (ECS) is an important mediator of neuronal plasticity [63]. In this study, the expression of WFA-and Cat-315-positive molecules present in the ECS increased in the PTZ-kindled model. It is unclear whether ECM molecules present outside PNNs have the same functions. ECM molecules in the developing CNS control nerve migration, axonal outgrowth, synapse formation, and synaptic maturation [64][65][66][67]. While PNNs are abundantly expressed in the spinal cord, it is not the PNNs but the ECM molecules that suppress axonal extension during regeneration after spinal cord injury [68,69]. The structure of the net itself is not important since a random collection of ECM molecules has the same effect. Therefore, we propose that the increase in ECM molecule expression in the ECS observed in PTZ-kindled mice decreases synaptic plasticity in the hippocampus.
In mice lacking tenascin-R, which is an extracellular matrix glycoprotein, the progression of kindling due to electrical stimulation is delayed [70]. Furthermore, enzymatic removal of hyaluronan by hyaluronidase treatment reduced kainate-induced hippocampal mossy fiber sprouting, one of the salient features associated with temporal lobe epilepsy [71]. Seizure upregulates the expression of multiple ECM molecules, including tenascin-C, tenascin-R, neuronal pentraxin 2, and hyaluronan [72]. Elevated levels of CSPG, the main component of the brain extracellular matrix, have been seen in patients with temporal lobe epilepsy [73,74].
Similar to these reports, the results of this study suggest that the ECM is involved in the onset of epilepsy by helping to form and maintain hyperexcitable networks during kindling.
In the present study, we found that astrocytes, but not microglia, were activated in PTZ-kindled mice. Activated astrocytes are observed in the hippocampus of animal models of epilepsy [75,76] and in the human hippocampus with temporal lobe epilepsy [77]. Reactive astrocytes show increased expression of GFAP and are present in large numbers due to reactive astrocytosis [78,79]. Astrocytes can increase the volume of the ECS by producing ECM molecules such as CSPG [80,81]. In addition, astrocytes play a role in synaptic interactions and regulate synaptic strength [82]. Whether astrocyte activation leads to the increase in ECM molecules following PTZ kindling is unknown. It is reasonable to assume that both astrocyte activation and increased ECM molecules may be directly related to the imbalance in excitation and inhibition, as well as changes in synaptic plasticity during epileptogenesis. It has been suggested that the development of drugs that target ECM synthesis and degradation will lead to effective antiepileptic treatments [72]. In addition, our study suggests that therapeutic agents targeting astrocytes may also be effective. Further studies are needed to determine whether inhibiting astrocyte activation suppresses ECM molecule secretion and epilepsy formation. Seizures are the result of an imbalance of excitation and inhibition caused by excessive excitability and/or imperfect GABAergic circuits. We showed that VGLUT1-and GAD67-positive synaptic terminals increased in PTZkindled mice. An increase in VGLUT1 in the hippocampus has also been shown in other animal models of epilepsy, which is consistent with the results of this study [83]. One of the glutamate receptors, the NMDA receptor, is the main focus of research on the molecular mechanisms underlying epileptogenesis. Much evidence indicates that NMDA receptors are involved in the pathogenesis of epilepsy firing [84][85][86] and in seizure-induced selective excitotoxic cell death in the hippocampus [87,88].
In this study, we show that the expression of PV protein in PV-positive neurons increased in the hippocampus after PTZ kindling. PV, which is a calcium-binding protein, regulates intracellular calcium dynamics in PV-expressing neurons [89,90]. In PV-positive neurons of the somatosensory cortex, the expression level of PV protein depends on the input stimulus [91]. PTZ, a noncompetitive GABA A receptor antagonist, induces status epilepticus, a state of increased neural excitation in the hippocampus [92]. Thus, it is possible that the high excitability caused by status epilepticus in PTZ-kindled mice leads to the increase in PV protein expression in the hippocampus. In fact, we found that VGLUT1-positive excitatory synaptic terminals increased in the hippocampus of PTZ-kindled mice. Furthermore, seizures dramatically increase mRNA levels and protein expression of the brain-derived neurotrophic factor (BDNF) both in epilepsy animal models and humans with epilepsy [93][94][95]. BDNF is known to increase PV protein expression [96,97], and we speculate that increased BDNF expression increased PV protein expression in this study. Further studies are needed to determine how BDNF expression changes in PTZ-kindled mice.

Conclusions
We found that astrocyte activation, increased expression of ECM molecules in the ECS region, and increased PNN formation occurred simultaneously in PTZ-kindled epilepsy model mice. We propose that changes in synaptic plasticity due to astrocyte activation and increased ECM molecules may contribute to epileptogenicity.

Data Availability
The data used to support the findings of this study are included within the article.