Seizures in PPT1 Knock-In Mice Are Associated with Inflammatory Activation of Microglia

Infantile neuronal ceroid lipofuscinosis (INCL), the most severe form of neuronal ceroid lipofuscinoses, is caused by mutations in the lysosomal enzyme palmitoyl protein thioesterase 1 (PPT1). Typical symptoms of this disease include progressive psychomotor developmental retardation, visual failure, seizures, and premature death. Here, we investigated seizure activity and relevant pathological changes in PPT1 knock-in mice (PPT1 KI). The behavior studies in this study demonstrated that PPT1 KI mice had no significant seizure activity until 7 months of age, and local field potentials also displayed epileptiform activity at the same age. The expression levels of Iba-1 and CD68 demonstrated, by Western blot analysis, the inflammatory cytokine TNF-α content measured with enzyme-linked immunosorbent assay, and the number of microglia demonstrated by immunohistochemistry (IHC) were significantly increased at age of 7 months, all of which indicate microglia activation at an age of seizure onset. The increased expression of GFAP were seen at an earlier age of 4 months, and such an increase reached its peak at age of 6 months, indicating that astrocyte activation precedes microglia. The purinergic P2X7 receptor (P2X7R) is an ATP-sensitive ionic channel that is highly expressed in microglia and is fundamental to microglial activation, proliferation, cytokines release and epilepsy. We show that the ATP concentration in hippocampal tissue in PPT1 KI mice was increased using an enhanced ATP assay kit and demonstrated that the antagonist of P2X7R, A-438079, significantly reduced seizures in PPT1 KI mice. In contrast to glial cell activation and proliferation, a significant reduction in synaptic proteins GABAAR was seen in PPT1 KI mice. These results indicate that seizure in PPT1 KI mice may be associated with microglial activation involved in ATP-sensitive P2X7R signaling and impaired inhibitory neurotransmission.


Introduction
Neuronal ceroid lipofuscinosis (NCL) is a group of autosomal recessive neurodegenerative diseases characterized by the accumulation of ceroid lipofuscin deposition in the brain [1]. Common clinical symptoms of the disease include impairment of cognitive and motor functions, blindness, refractory seizures, and premature death [2][3][4][5]. Infantile neuronal ceroid lipofuscinosis (INCL) is the fastest developing and most severe type of NCL [6] caused by mutations in the CLN1 gene, which encodes palmityl protein thioesterase-1 (PPT1). As a depalmitoylase, PPT1 plays a critical role in regulating protein transport and protein-protein interactions. The mutation or deletion of PPT1 can induce a series of neurological pathological changes, including decreased synaptic plasticity, the activation of glial cells, and neural death [7][8][9].

Seizures in PPT1 KI Mice
To observe seizures in PPT1 KI mice, we recorded daily activities of PPT1 KI mice (5-7 months old) and WT mice (7 months old) for 24 h. PPT1 KI mice had no seizures at 5-6 months but exhibited typical seizures (over stage 4) at 7 months of age ( Figure 1A). The observation of seizure praxeology showed that seizure levels and duration varied largely in PPT1 KI mice. Seizure levels were mostly scaled as stage 2 (head nodding) and stage 3 (forelimb clonus), with few at stage 4 (rearing) and stage 5 (rearing and falling) ( Figure 1B). The longest seizure duration was up to 10 min 57 s, and the shortest was less than 1 min. These results demonstrate that PPT1 KI mice could suffer from seizures, as seen in PPT1 KO mice, and that these seizures occurred in late age in PPT1 KI mice. months but exhibited typical seizures (over stage 4) at 7 months of age ( Figure 1A). The observation of seizure praxeology showed that seizure levels and duration varied largely in PPT1 KI mice. Seizure levels were mostly scaled as stage 2 (head nodding) and stage 3 (forelimb clonus), with few at stage 4 (rearing) and stage 5 (rearing and falling) ( Figure  1B). The longest seizure duration was up to 10 min 57 s, and the shortest was less than 1 min. These results demonstrate that PPT1 KI mice could suffer from seizures, as seen in PPT1 KO mice, and that these seizures occurred in late age in PPT1 KI mice.

Epileptiform Activity in Brain Slices of PPT1 KI Mice
To investigate the function of neural network in PPT1 KI mice, a low concentration of kainic acid (KA, 200 nM) was applied to the hippocampal slices, and γ oscillations in hippocampal CA3 from the WT mice were induced, as reported previously [40], but there was epileptiform activity from PPT1 KI mice at 7 months of age (Figure 2A,D). The typical examples of epileptiform activity in these slices are characterized from field potential recordings by the burst firing of bi-directional complex waves (0.1-0.4 Hz), composed of a large negative potential (0.5-2 mV, wave width 20-200 ms) followed by a large positive potential (0.2-2 mV, 50-300 ms) ( Figure 2D (d1) and (d2)). The γ oscillation in WT mice showed a stable increase over time, indicating a normal local neural network mediated by inhibitory interneurons (Figure 2B,C). In PPT1 KI mice, the epileptiform activity emerged 5 min after KA treatment, and the number of epileptiform activities reached stability at 10 min ( Figure 2E). These results suggest the impaired inhibition and neural network dysfunction in the hippocampus of 7-month-old PPT1 KI mice.

Epileptiform Activity in Brain Slices of PPT1 KI Mice
To investigate the function of neural network in PPT1 KI mice, a low concentration of kainic acid (KA, 200 nM) was applied to the hippocampal slices, and γ oscillations in hippocampal CA3 from the WT mice were induced, as reported previously [40], but there was epileptiform activity from PPT1 KI mice at 7 months of age (Figure 2A,D). The typical examples of epileptiform activity in these slices are characterized from field potential recordings by the burst firing of bi-directional complex waves (0.1-0.4 Hz), composed of a large negative potential (0.5-2 mV, wave width 20-200 ms) followed by a large positive potential (0.2-2 mV, 50-300 ms) ( Figure 2D (d1) and (d2)). The γ oscillation in WT mice showed a stable increase over time, indicating a normal local neural network mediated by inhibitory interneurons (Figure 2B,C). In PPT1 KI mice, the epileptiform activity emerged 5 min after KA treatment, and the number of epileptiform activities reached stability at 10 min ( Figure 2E). These results suggest the impaired inhibition and neural network dysfunction in the hippocampus of 7-month-old PPT1 KI mice.

Microglial Activation in Hippocampus Correlates with the Occurrence of Seizures
To investigate the expression level of Iba-1, a microglial marker, in PPT1 KI mice, we performed Western blot analysis. The results showed that the expression level of Iba-1 was significantly increased in the hippocampus of 7-month-old PPT1 KI mice compared

Microglial Activation in Hippocampus Correlates with the Occurrence of Seizures
To investigate the expression level of Iba-1, a microglial marker, in PPT1 KI mice, we performed Western blot analysis. The results showed that the expression level of Iba-1 was significantly increased in the hippocampus of 7-month-old PPT1 KI mice compared with age-matched WT mice (p = 0.04, permutation t test, n = 6) ( Figure 3A). post hoc LSD test, n = 21) ( Figure 3D). There were no significant changes in the expression level of CD68 in WT mice from age 1 to 7 months ( Figure 3E). These results indicate that the number of activated microglia was significantly increased in 7-month-old PPT1 KI mice. show that expressions of Iba-1 in the hippocampus of PPT1 KI mice and agematched WT mice from age 1 to 7 months (B, Iba-1 in PPT1 KI mice, 6 vs. 7 months old, p < 0.001, n = 35; C, Iba-1 in WT mice, n = 35); (D,E) Representative Western blots (top) and the bar graph (bottom) show expression of CD68 in the hippocampus of PPT1 KI mice and age-matched WT mice from age 1 to 7 months (D, CD68 in PPT1 KI mice, 1 vs. 4 months old, p = 0.049, 6 vs. 7 months old, p = 0.03, n = 21; E, CD68 in WT mice, n = 35); (F,G) Representative Western blots (top) and the bar graph (bottom) show expression of GABAARα1 in the hippocampus of PPT1 KI mice and agematched WT mice from age 1 to 7 months (F, GABAARα1 in PPT1 KI mice, 1 vs. 7 months old, p < 0.01, n = 35; G, GABAARα1 in WT mice, n = 35). * p < 0.05; ** p < 0.01; ***p < 0.001.

Astrocyte Is Activated at the Early Stage of PPT1 KI Mice
The expression of the astrocyte marker GFAP was determined by Western blot analysis. The level of GFAP was significantly elevated in PPT1 KI mice starting at 4 months of age and continued to increase with increased age until 6 months old (1 vs. 4 months, p = show expression of CD68 in the hippocampus of PPT1 KI mice and age-matched WT mice from age 1 to 7 months (D, CD68 in PPT1 KI mice, 1 vs. 4 months old, p = 0.049, 6 vs. 7 months old, p = 0.03, n = 21; E, CD68 in WT mice, n = 35); (F,G) Representative Western blots (top) and the bar graph (bottom) show expression of GABA A Rα1 in the hippocampus of PPT1 KI mice and age-matched WT mice from age 1 to 7 months (F, GABA A Rα1 in PPT1 KI mice, 1 vs. 7 months old, p < 0.01, n = 35; G, GABA A Rα1 in WT mice, n = 35). * p < 0.05; ** p < 0.01; ***p < 0.001.
To further demonstrate the correlation between microglial activation and seizures in PPT1 KI mice, the number of microglia in the hippocampus of 1-7-month-old PPT1 KI mice was measured. The number of hippocampal microglia did not change in PPT1 KI mice from age 1 to 6 months, but it was significantly increased at 7 months of age (p < 0.001, one-way ANOVA on ranks followed by post hoc LSD test, n = 35) ( Figure 3B). The expression level of Iba-1 in the hippocampus of WT mice had no change from age 1 to 7 months ( Figure 3C). We also analyzed expression levels of CD68, a marker of activated microglia and macrophages. The expression levels of CD68 in PPT1 KI mice showed a slow and small increase from age 1 to 6 months and a large increase at 7 months of age (1 vs. 4 months, p = 0.049, 6 vs. 7 months, p = 0.03, one-way ANOVA on ranks followed by post hoc LSD test, n = 21) ( Figure 3D). There were no significant changes in the expression level of CD68 in WT mice from age 1 to 7 months ( Figure 3E). These results indicate that the number of activated microglia was significantly increased in 7-month-old PPT1 KI mice.

Astrocyte Is Activated at the Early Stage of PPT1 KI Mice
The expression of the astrocyte marker GFAP was determined by Western blot analysis. The level of GFAP was significantly elevated in PPT1 KI mice starting at 4 months of age and continued to increase with increased age until 6 months old (1 vs. 4 months, p = 0.01; 6 vs. 7 months, p = 0.625, one-way ANOVA on ranks followed by post hoc LSD test, n = 21 in supplement data, Figure S3), indicating that increased astrocyte activation occurs at the early stage of PPT1 KI mice.

Age-Dependent Changes in Synaptic Proteins GluN2B and GABA A Rα1 in PPT1 KI Mice
The expression of excitatory synaptic protein, the GluN2B subunit of N-methyl daspartate receptors (NMDARs), and inhibitory synaptic protein GABA A Rα1 were further determined by Western blot analysis. GluN2B-containing NMDARs are expressed in hippocampal interneurons, and GABA A R mediates inhibitory synaptic transmission [41,42]. Our results showed the decreased expression of GluN2B in the hippocampus of PPT1KI mouse (GluN2B, 1 vs. 7 months, p < 0.01, one-way ANOVA on ranks followed by post hoc LSD test, n = 35 in supplement data, Figure S3) and a similar pattern of decreased expression of hippocampal GABA A Rα1 in PPT1KI mouse from age 1 to 7 months (GABA A Rα1, 1 vs. 7 months, p < 0.01, one-way ANOVA on ranks followed by post hoc LSD test, n = 35) ( Figure 3F), which indicates the impaired inhibitory synaptic transmission in the hippocampus of PPT1 KI mice with age. There was no age-related change in the expression of GABA A Rα1 in WT mice ( Figure 3G).

Iba-1 Immunoreactivity and Morphological Changes of Microglia in Hippocampus of PPT1 KI Mice
To further determine the relationship between seizures and microglial activation in PPT1 KI mice, we performed in situ IHC in the hippocampus of 7-month-old PPT1 KI mice and age-matched WT mice ( Figure 4A). The microglia of the WT mice remained in a resting state with slender branches, whereas the microglia had larger somata with shorter, thicker and dense protrusions in the hippocampus of 7-month-old PPT1 KI mice, indicating that these microglia are highly activated. Our results showed that the number of microglia in the hippocampal CA1 and CA3 of PPT1 KI mice was significantly increased compared to those of age-matched WT mice (CA1: p = 0.003, CA3: p = 0.01, permutation t test, n = 12) ( Figure 4B,C). Our data further showed that microglial activation was significantly increased in PPT1 mice with seizures at 7 months of age, which suggests that activated microglia likely contributed to seizures in the PPT1 KI mice.

Age-Related Neuronal Loss in PPT1 KI Mice
To test whether there is neuronal loss in PPT1 KI mice, we used DAB staining IHC in the hippocampus of PPT1 KI mice at ages 3, 5 and 7 months. Our data demonstrated that there was no significant neuronal loss in 3-and 5-month-old PPT1 KI mice, but there was 46% neuronal loss in 7-month-old PPT1 KI mice compared to 5-month-old PPT1 KI mice (5 vs. 7 months, p < 0.001, one-way ANOVA on ranks followed by post hoc LSD test, n = 12) and 7-month-old WT mice (p < 0.001, Student's t test, n = 12) ( Figure 4D,E). Therefore, these results indicate that PPT1 KI mice with seizure are associated with dramatic neuronal death and microglial activation.

Cytokines IL-1β and TNF-α Changes in PPT1 KI Mice
We measured the contents of two common cytokines IL-1β and TNF-α by ELISA. Our results showed that the concentrations of IL-1β slightly increased in PPT1 KI mice

Cytokines IL-1β and TNF-α Changes in PPT1 KI Mice
We measured the contents of two common cytokines IL-1β and TNF-α by ELISA. Our results showed that the concentrations of IL-1β slightly increased in PPT1 KI mice compared with WT, but there was no statistically significant difference (p = 0.1, permutation t test, n = 12) ( Figure 5A). The expression levels of TNF-α in PPT1 KI mice were significantly increased compared to their WT littermates (WT vs. PPT1 KI, p = 0.02, permutation t test, n = 10) ( Figure 5B).

Inhibition of ATP-Sensitive P2X7R Repressed Seizure in PPT1 KI Mice
We measured the concentration of ATP in the hippocampus of 7-month-old PPT mice (seizure state and resting state) and age-matched WT mice. The ATP concentra of PPT1 KI mice was significantly higher in the seizure state than the resting state 0.043, permutation t test, n = 6) ( Figure 6A). The ATP concentration in the resting sta PPT1 KI mice was higher than that of WT mice (p = 0.048, permutation t test, n = 6) ( Fig  6A).
Intraperitoneal injection of A 438079 (30 mg/kg), an ATP-sensitive P2X7 antago into 7-month-old PPT1 KI mice significantly reduced the duration and number of seiz (seizure duration, p = 0.01, number of seizures, p = 0.01, Permutation t test, n = 8). T was no significant change in the saline (30 mg/kg) treatment group ( Figure 6B,C). C cally, A 438079 treatment significantly reduced the number of microglia in hippocam CA1 and CA3 of PPT1 KI mice (CA1, p = 0.004, CA3, p = 0.02, permutation t test, n = ( Figure 6D-F). These data indicate that the P2X7R inhibitor alleviated seizures, sugges the possible microglia involvement in the seizures of PPT1 KI mice via P2X7R activat

Inhibition of ATP-Sensitive P2X7R Repressed Seizure in PPT1 KI Mice
We measured the concentration of ATP in the hippocampus of 7-month-old PPT1 KI mice (seizure state and resting state) and age-matched WT mice. The ATP concentration of PPT1 KI mice was significantly higher in the seizure state than the resting state (p = 0.043, permutation t test, n = 6) ( Figure 6A). The ATP concentration in the resting state of PPT1 KI mice was higher than that of WT mice (p = 0.048, permutation t test, n = 6) ( Figure 6A).
Intraperitoneal injection of A 438079 (30 mg/kg), an ATP-sensitive P2X7 antagonist, into 7-month-old PPT1 KI mice significantly reduced the duration and number of seizures (seizure duration, p = 0.01, number of seizures, p = 0.01, Permutation t test, n = 8). There was no significant change in the saline (30 mg/kg) treatment group ( Figure 6B,C). Critically, A 438079 treatment significantly reduced the number of microglia in hippocampal CA1 and CA3 of PPT1 KI mice (CA1, p = 0.004, CA3, p = 0.02, permutation t test, n = 12) ( Figure 6D-F). These data indicate that the P2X7R inhibitor alleviated seizures, suggesting the possible microglia involvement in the seizures of PPT1 KI mice via P2X7R activation.

Discussion
In this study, behavioral experiments showed that PPT1 KI mice had no seizures until 7 months of age. The behavioral finding was confirmed by ex vivo LFP recording in PPT1 KI mouse. Western blot and immunoreactivity experiments demonstrated microglia acti-

Discussion
In this study, behavioral experiments showed that PPT1 KI mice had no seizures until 7 months of age. The behavioral finding was confirmed by ex vivo LFP recording in PPT1 KI mouse. Western blot and immunoreactivity experiments demonstrated microglia activation and increased ATP concentration in PPT1 KI mice with seizure. Furthermore, inhibition of P2X7R, an inotropic ATP receptor mainly expressed in microglia, reduced the number and total duration of seizures in PPT1 KI mice.
Seizures and epileptiform activities observed in 7-month-old PPT1 KI mice both in vivo and ex vivo indicate the impaired hippocampal neural networks in 7-month-old PPT1 KI mice. A foundational mechanism of seizures is imbalance between excitatory and inhibitory neurotransmission. Previous studies on the pathogenesis of seizures in PPT1 KO mice were attributed to the loss of GABAergic interneurons [7]. The PPT1 KI mice had motor deficits during 3-5 months of age [12], but seizure onset was at 7 months of age. The reason for this late-onset seizure is probably related to the sufficient expression level of the GABA A Rα1 maintained in PPT1 KI mice until 5 months of age. While the PPT1 KI mice exhibit behavioral seizures at 7 months of age, it is possible that the mice experience electrographic seizures that manifested as behavioral seizures only later [43].
Some studies have suggested that astrocytes are involved in epilepsy [44]. The astrocytes of PPT1 KO mice showed functional and morphological abnormalities, and their survival rate was lower than that of WT mice [45]. Our results showed that the number of astrocytes significantly increased even at an early age (3 months) in PPT1 KI mice, which is in line with previous reports [46,47]. Although it is not consistent with the timing of seizure onset in these mice, the early and dramatic increase in GFAP expression in PPT1 KI mice in this study and other reports [46] suggest that astrocytes play a role in neuroinflammation and microglial activation in PPT1 KI mice in response to neuronal damage and death in these mice.
Dramatic neuronal death was observed in the hippocampus in PPT1 KI mice by our immunoreactivity data. The mechanisms for neuronal death in PPT1 KI mice have been studied, and the abnormalities in NMDARs in PPT1 KI mice appear to be involved [42]. The increased expression of GluN2B in neurons of PPT1 KO mice may render neurons vulnerable to excitotoxicity [48]. Our data showed that the expression levels of GluN2B were not significantly altered in PPT1 KI mice until age 6-7 months, the reason for the downregulation of GluN2B at such a late age is currently unknown but may be related to the reduced neuronal numbers in these mice. Nevertheless, the increased inflammatory cytokines such as TNF-α measured in this study are known to mediate neuronal death [25,26]. The reduced expression of GABA A Rα1 becomes evident at the late stage of PPT1 KI mice, suggesting that the reduced inhibition mediated by GABA A Rα1 may also contribute to increased excitotoxicity, seizure, and neuronal loss in these mice.
Although there is existing neuronal loss in PPT1 KI mice, this study demonstrated that the number of activated microglia was significantly increased in PPT1 KI mice with seizures, which is in agreement with previously published reports [12,49]. The morphological changes in microglia of PPT1 KI mice were consistent with the microglial activation observed in the hippocampus of WT mice with seizures induced by KA [50]. In humans, the expression of microglial markers in the hippocampus is significantly higher in a human brain with seizures than in normal individuals [51]. The duration and frequency of seizures are correlated with microglial activation [49].
Critically, unlike in astrocyte activation, which precedes seizure, microglial activation is concurrent with seizure onset in PPT1 KI mice, emphasizing a role of microglial activation in seizure onset. The binding of P2X7R located on microglial membrane to ATP, released from astrocytes and/or neurons, is an important mechanism to activate microglia and release microglial cytokines (IL-1β and TNF-α) [36]. The inflammatory cytokine IL-1β expression was increased in the brains of epileptic mice [52,53] and in stimulated PPT1 KO microglia [45]. These cytokines act on neurons and astrocyte to release ATP and glutamate, and consequently cause neuronal hyperexcitation, excitotoxicity and seizures [54,55] (Figure 7).

Figure 7.
ATP-driven P2X7 activation in microglia likely contributes to seizures at a late age in PPT1 KI mice. Increased ATP concentration in the hippocampus of PPT1-deficient mice was observed in our study, which likely activate P2X7R on the surface of microglia, leading to inflammatory cytokines release. Inflammatory cytokines cause ATP and glutamate release from activated astrocytes, neuroinflammation, neuronal hyperexcitability and epileptiform discharges [34,54]. Neural damage and ATP release by activated microglia form a positive feedback loop in PPT1 KI mice.
The involvement of inflammatory cytokine and microglia in seizures was also demonstrated by the evidence that the inhibition of cytokine production attenuates neuroinflammation and seizures [56,57]. Application of PLX3397, a potent inhibitor of colony stimulating factor-1 receptor expressed on the cell surface of microglia depletes microglial activation, by which neuronal injury and seizure activity were reduced and clinical outcome was improved in PPT1 KO mice [54], which both demonstrate a detrimental impact on the microglia in the CNS of CLN1 mice.
In our study, the increased expression of cytokines TNF-α and an increasing trend of IL-1β expression may support the increased inflammation and neuronal injury being associated with microglial/astrocyte activation in PPT1 KI mice.
Our data further showed that the ATP concentration significantly increased in the hippocampal tissue of PPT1 KI mice in the seizure state, suggesting that astrocytes and neurons in these mice are under stress and release a large amount of ATP [58,59]. The increased ATP likely acts on microglia and neurons in PPT1 KI mice. Therefore, the interference in the role of ATP may be effective for the treatment of INCL. In this study, administration of A 438079, a P2X7R antagonist to PPT1 KI mice, significantly reduced the number and total duration of seizures. This result was in agreement with a previous report that P2X7R inhibitors reduce microgliosis and seizures in kainic acid-induced epileptic mice [33,60,61]. . ATP-driven P2X7 activation in microglia likely contributes to seizures at a late age in PPT1 KI mice. Increased ATP concentration in the hippocampus of PPT1-deficient mice was observed in our study, which likely activate P2X7R on the surface of microglia, leading to inflammatory cytokines release. Inflammatory cytokines cause ATP and glutamate release from activated astrocytes, neuroinflammation, neuronal hyperexcitability and epileptiform discharges [34,54]. Neural damage and ATP release by activated microglia form a positive feedback loop in PPT1 KI mice.
The involvement of inflammatory cytokine and microglia in seizures was also demonstrated by the evidence that the inhibition of cytokine production attenuates neuroinflammation and seizures [56,57]. Application of PLX3397, a potent inhibitor of colony stimulating factor-1 receptor expressed on the cell surface of microglia depletes microglial activation, by which neuronal injury and seizure activity were reduced and clinical outcome was improved in PPT1 KO mice [54], which both demonstrate a detrimental impact on the microglia in the CNS of CLN1 mice.
In our study, the increased expression of cytokines TNF-α and an increasing trend of IL-1β expression may support the increased inflammation and neuronal injury being associated with microglial/astrocyte activation in PPT1 KI mice.
Our data further showed that the ATP concentration significantly increased in the hippocampal tissue of PPT1 KI mice in the seizure state, suggesting that astrocytes and neurons in these mice are under stress and release a large amount of ATP [58,59]. The increased ATP likely acts on microglia and neurons in PPT1 KI mice. Therefore, the interference in the role of ATP may be effective for the treatment of INCL. In this study, administration of A 438079, a P2X7R antagonist to PPT1 KI mice, significantly reduced the number and total duration of seizures. This result was in agreement with a previous report that P2X7R inhibitors reduce microgliosis and seizures in kainic acid-induced epileptic mice [33,60,61].
In summary, we provide evidence to demonstrate that PPT1 KI mice have seizures, which is associated with inflammatory activation of microglia involved in ATP-P2X7R signaling.

Animals
Homozygous mutant mice of PPT1 c.451C > T/c.451C > T, termed PPT1 KI, were generously gifted by our collaborator, Dr. Anil B. Mukherjee, Eunice Kennedy Shriver National Institute of Child Health and Human Development, National Institutes of Health (Bethesda, MD, USA) (Supplemental Figure S1). Mice were housed at Xinxiang Medical University Laboratory Animal Center and bred in a pathogen-free facility with 12 h light/12 h dark cycles with access to water and food ad libitum. There were 74 PPT1 KI mice and 21 age-matched wild-type (WT) littermates, C57BL6, used in this study. The smallest possible number of mice was used in the experiments according to sample size test (MSST v. 6.0.6). All animal procedures were performed according to guidelines approved by the committee on animal care at Xinxiang Medical University.

Behavioral Scaling in Seizures
PPT1 KI mice aged from 5-7 months old and 7-month-old WT mice were placed into the observation cage with water and food at constant room temperature (25 ± 2 • C). Mice were videotaped by the Plexon CinePLex Studio Application Version 3.7 (Plexon, Dallas, TX, USA) for 24 h in 12 h/12 h day/night cycle (lights on at 8 a.m.). The spontaneous seizure behavior repertoire of mice was classified into five stages according to the Racine scale [62], including: (1) mouth and facial movements, (2) head nodding, (3) forelimb clonus, (4) rearing, and (5) rearing and falling. We recorded the duration and number of seizures within 24 h.

Slice Preparation and Extracellular Field Recordings
KI and WT mice were anesthetized by intraperitoneal injection of pentobarbital (40 mg/kg). The brains were removed and stored in a cold slicing solution (225 mM sucrose, 3 mM KCl, 6 mM MgSO 4 ; 1.2 mM NaH 2 PO 4 ; 24 mM NaHCO 3 , 10 mM glucose, and 0.5 mM CaCl 2 ). For local field potential (LFP) recordings, 350 µm horizontal sections were made using a Leica VT1000S vibratome (Leica Microsystems, Milton Keynes, England, UK). The hippocampal slices were transferred to an interface recording chamber and incubated for 1 h. Slices were maintained at a temperature of 32 • C and at the interface between the artificial cerebrospinal fluid (ACSF) and warm humidified carbon gas (95% O 2 -5% CO 2 ). The ACSF contained (in mM): 126 NaCl; 3 KCl; 2 MgSO 4 ; 1.25 NaH 2 PO 4 ; 24 NaHCO 3 ; 10 glucose, and 2 CaCl 2 . Extracellular field recordings were performed using the interface-recording chambers. Recordings were obtained using normal ACSF-filled 1-3 MΩ borosilicate glass microelectrodes (Sutter Instrument, Novato, CA, USA). The γ oscillation from hippocampal CA3 was induced through the perfusion of ACSF containing kainic acid (200 nM). Data were band-pass filtered online between 0.5 Hz and 2 kHz using the Axoprobe amplifier and a Neurolog system NL106 AC/DC amplifier (Digitimer Ltd. Cambridge, England, UK). The data were digitized at a sample rate of 5-10 kHz using a CED 1401 plus ADC board (Digitimer Ltd. Cambridge, England, UK). Electrical interference from the main supply was eliminated from extracellular recordings online using 50 Hz noise eliminators (HumBug; Digitimer Ltd. Sequim, WA, USA).
Data were analyzed offline using Spike 2 software (CED, Cambridge, UK). Power spectra were generated to provide a quantitative measure of the frequency components in a stretch of recording, where power, a quantitative measure of the oscillation strength, was plotted against the respective frequency. Power spectra were constructed for 60 s epochs of extracellular field recordings using a fast Fourier transform algorithm provided by Spike 2. The parameters used for measuring the oscillatory activity in the slice were the peak frequency (Hz) and area power (µV2). In this study, the area power was equivalent to the computed area under the power spectrum between the frequencies of 20 and 60 Hz.

Enzyme-Linked Immunosorbent Assay (ELISA)
The protein concentrations of IL-1β and TNF-α were quantified using an ELISA kit (4A Biotech Co., Ltd., Beijing, China) according to the manufacturer's instruction for 7-month-old PPT1 KI mice and age-matched WT mice. The optical density was detected at a wavelength of 450 nm, and the concentration of the target protein was calculated according to the standard curve and normalized against the protein of the samples. Results were expressed as pg/mg protein. Sensitivity: the lowest detectable IL-1β and TNF-α dose is less than 15 pg/mL. Both intra-and interassay CV% are less than 10%.

ATP Assays
The unilateral hippocampus was harvested from PPT1 KI mice with seizures and without seizures and age-matched WT control mice. The ATP concentrations of these hippocampi were assessed using an enhanced ATP assay kit (Beyotime, Shanghai, China) according to the manufacturer's protocol. ATP consumption was calculated using the ATP standard curve.

Drug Administration
The specific P2X7R inhibitor A 438079 (HY-15488, MedChemExpress, Shanghai, China) was dissolved in normal saline and delivered via an intraperitoneal injection to PPT1 KI mice with seizures (30 mg/kg, twice a day for two days). Seizure behaviors were recorded by monitoring for 24 h after drug application.

IHC and Confocal Imaging
Mouse brains were fixed in 4.0% paraformaldehyde at room temperature after perfusion with cold 1 × phosphate-buffered saline (PBS). The brains were sectioned coronally. Before staining, the slices were incubated in a citrate buffer (pH 6) 25 min at 95 • C for antigen retrieval and then permeabilized with 0.1% Triton X-100 in PBS. After blocking with 10% normal goat serum, the slices were incubated with primary antibodies overnight