Decreased Clock Gene Bmal1 Mediates Epileptogenesis via PCDH19 in Temporal Lobe Epilepsy

Clock genes not only regulate the circadian rhythm of physiological activities but also participate in the pathogenesis of many diseases. Previous studies have found the abnormal expression of clock genes in epilepsy. However, as the core clock gene, the molecular mechanism of Bmal1 in the epileptogenesis and seizures of temporal lobe epilepsy (TLE) is still unclear. To dene the function of Bmal1, we rstly investigated the levels of Bmal1 and other clock proteins in the hippocampus in epilepsy. In the latency and chronic phases, the levels of Bmal1 were decreased compared with the control group. Knockout of Bmal1 in hippocampal dentate gyrus (DG) neurons of Bmal1 ox/ox mice by Synapsin 1 (Syn1) promoter AAV (adeno-associated virus) lowered the threshold of seizures induced by pilocarpine administration. High throughput sequencing analysis showed that PCDH19 (protocadherin 19), a gene associated with epilepsy, was regulated by Bmal1. And the expression of PCDH19 was also decreased in the hippocampus of epileptic mice. Furthermore, the levels of Bmal1 and PCDH19 were higher in the patients with no hippocampal sclerosis (no HS), compared to HS International League Against Epilepsy (ILAE) type I and III. Altogether, these data suggest that decreased expression of clock gene Bmal1 may participate in the epileptogenesis and seizures via PCDH19 in TLE. DG and CA1 of mice with TLE. Neuron-specic knockout of bmal1 in DG of Bmal1 ox/ox mice lowers the threshold of pilocarpine-induced seizures. With high-throughput sequencing and western blotting, the downstream gene PCDH19 regulated by Bmal1 was rstly identied and then detected in the hippocampus of epileptic mice. Furthermore, the expression of Bmal1 and PCDH19 were detected in HS type I, HS type III, and no HS. The levels of Bmal1 and PCDH19 protein in DG of HS type I and HS type III were decreased, compared with no HS group. These results suggest that these changes in Bmal1 and PCDH19 expression may be strongly associated with epileptogenesis and play an important role in the pathogenesis of TLE. with conditional deletion


Introduction
Temporal lobe epilepsy (TLE) is one of the common drug-resistant epilepsy. The common pathological change of TLE is hippocampal sclerosis (HS) which is characterized by severe neuronal loss and gliosis in one or more hippocampal regions [1]. Patients with TLE are often accompanied by cognitive impairment, memory loss, and mood impairments. Despite a lot of effort, the pathogenesis of TLE remains incompletely unclear [2][3][4]. Clinical observation nds that patients with TLE show a 24-hour nonuniform distribution of seizure occurrence, and the number of seizures had unimodal (afternoon) or bimodal (morning and noon) time peaks [5][6][7]. Spontaneous seizures in kainic acid/pilocarpine-induced and electrically stimulated models of TLE have been found to occur in a pattern of 24-hour non-uniform distribution [8,9]. These suggest that the epileptogenesis and seizures of TLE may be associated with the circadian rhythms.
Circadian rhythms are biological rhythms driven by a series of clock genes, such as Bmal1 (Brain and Muscle Arnt-Like Protein 1) and CLOCK (circadian locomotor output cycles kaput) [10]. As the core clock genes, Bmal1 and CLOCK in the cytoplasm heterodimerize and translocate to the nucleus, and then instigate transcription of target genes by interacting with E-box promoters. Among these target genes, some clock genes, such as Period (Per1/2/3) and Cryptochrome (Cry1/2), are involved in the positive and negative regulation of the circadian oscillation of Bmal1 and CLOCK. Clock genes and Clock-controlled genes (CCGs) underlie the rhythmic oscillations at a cellular and organismal level [11]. In the epileptic hippocampus, the levels of Bmal1, CLOCK, Cry and Per mRNA have been con rmed to be decreased.
Previous studies have shown that the expression levels of clock genes change after seizures, and can regulate downstream genes to directly affect epileptogenesis [12,13].
Several studies have investigated the role of Bmal1 in pathophysiology [14]. In the pilocarpine-induced epileptic rats, the level of Bmal1 mRNA is decreased in all comparison to the naive group [15]. The threshold of seizures induced by electrical stimulation in Bmal1 knockout (KO) mice is lower compared with the control group [16]. It should be pointed out that, in Ferraro's study, the ablation of Bmal1 is not speci c in the speci c brain nuclei and cellular types. Subsequent studies found that the ablation of Bmal1 in GLAST-positive astrocytes of the suprachiasmatic nucleus (SCN) alters circadian locomotor behavior and cognition in mice through GABA signaling [17]. Moreover, the Bmal1 deletion of astrocytes induces astrocyte activation and in ammatory gene expression via a cell-autonomous mechanism [18].
These studies suggest that abnormal expression of Bmal1 may be related to the epileptogenesis of TLE.
However, it remains unclear how the changes of Bmal1 expression in the hippocampus, especially in hippocampal neurons, affect the epileptogenesis and seizures of TLE.
Therefore, in this study, we investigated the expression and distribution of Bmal1 in the hippocampus of TLE and tested the effects of Bmal1 conditional KO (cKO) in hippocampal neurons of Bmal1 ox/ ox mice on epileptogenesis and seizures. And the downstream genes regulated by Bmal1 were identi ed through high-throughput sequencing. Moreover, Bmal1 and downstream gene (PCDH19) were detected in the hippocampal specimens of patients with TLE.

Patient selection
The patients included in the study were obtained from the les of the Department of Neurosurgery, The First A liated Hospital of Xi'an Jiaotong University. We examined 16 specimens obtained from patients undergoing surgery for medically intractable TLE. All procedures were performed with the informed consent of the patients or legal next-of-kin and were approved by the Committee on Human Research at

Animals
Adult C57/BL6J mice weighing 20-25 g were purchased from the experimental animal center of Medical College in Xi'an Jiaotong University. Bmal1 ox/ ox mice were purchased from The Jackson Laboratory (Stock No: 007668). The animals were housed under controlled humidity (55 ± 5%) and temperature (20 ± 2 °C) with a normal 12 h light/12 h dark cycle. Water and food were available ad libitum. All animal procedures were carried out in line with the National Institutes of Health Guide for the Care and Use of Laboratory Animals and were rati ed by the Institutional Animals Care and Use Committee. All procedures performed in studies involving animals were approved by the Ethics Committee of Xi'an Jiaotong University (Ethics and Science # G-83) in full accordance with the ethical guidelines of the National Institutes of Health for the care and use of laboratory animals. All efforts were made to minimize suffering and the number of rats used for the experiments.

Study design and experimental endpoints
To investigate the levels of proteins after seizures, 3 of 20 mice were randomly taken out and given saline vehicle as a control group, and the rest were administrated by pilocarpine to induce seizures. According to the Racine categories, 14 mice had seizures that reached Racine category 4 and were included in the epilepsy model group. Subsequently, 3 mice were randomly taken out at different time points (1d, 3d, 14d, 60d). Each mouse brain was divided into two parts along the longitudinal ssure. One part was used for immunoblotting experiments. The other part was xed with 4% paraformaldehyde solution in PBS and used for immuno uorescence experiments. For evaluating the effects of Bmal1 conditional KO (cKO) on the susceptibility to seizures, 26 Bmal1 ox/ ox mice were randomly divided into 2 groups and injected Syn1-mCherry and Syn1-Cre AAVs respectively. Four weeks after AAVs injection, 3 mice were randomly selected from each group of animals to test the e ciency of Bmal1 cKO and transcriptome sequencing.
Behavioral tests of 20 mice from the two groups are evaluated according to Racine categories. After behavioral tests, mice were sacri ced and collected samples for immunoblotting and immuno uorescence experiments.

Pilocarpine treatment and seizure assessment
Mice were administered intraperitoneally with methylscopolamine (1 mg/kg body weight) 30 min before injection of pilocarpine (300 mg/kg body weight) in 0.2 ml sterile saline vehicle (0.9% NaCl). Control mice received an equivalent volume of saline vehicle. The severity of seizure behavior was observed for 1 h and assessed using the following standard. Categories 1-2, one or more of the symptoms including facial automatisms, tail stiffening, and wet-dog shakes; Categories 1-2 were considered as a group to avoid subjectivity in assessing the seizures. Category 3 clonic unilateral forelimb myoclonus in addition to the symptoms above; Category 4 bilateral forelimb myoclonus and rearing; Category 5 generalized clonic-tonic convulsions and loss of postural control. The seizure categories were separately evaluated by two observers. A mouse experiencing continuous category 3-5 seizure events (30-90 sec) was considered to undergone status epilepticus (SE). Diazepam (10 mg/kg, i.p.) was administered to terminate SE 1 h after the onset. Mice were monitored for 2 h/day, 7 days/week for the occurrence of spontaneous seizures (category [3][4][5]. Only after the occurrence of a seizure was a mouse identi ed as epilepsy mice (i.e. the observer was not aware of a priori of the treatment for any mouse). Methylscopolamine (S1978) and pilocarpine (S4231) were purchased from Selleck. Diazepam was purchased from Shanghai Shyndec Pharmaceutical Co., Ltd.

Total protein preparation and Immunoblotting
All mice were sacri ced under deep anesthesia with iso uorane (VETEASY, RWD Life Science). Then open the skull and carefully remove the brain tissue. Put the brain tissue in pre-cooled PBS, and separate the hippocampus from brain tissue with ne tweezers. Hippocampus tissues and cells were homogenized in radioimmunoprecipitation assay (RIPA) buffer (25 mM Tris·HCl at pH 7.6, 150 mM NaCl, 0.1% SDS, 1% NP-40, 1% sodium deoxycholate, and protease inhibitor (ThermoFisher, 89901)). The homogenates were then centrifuged at 12,000 rpm for 20 min at 4 °C. The protein concentration of the tissue lysate was measured with the bicinchoninic acid (BCA) assay (Thermo Fisher Scienti c, 23227) after the manufacturer's protocol. Protein lysates were mixed with one-third volume of 4 × loading buffer. The supernatants were boiled and separated by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE). Proteins were electrophoresed and transferred to PVDF lter membranes (Merck Millipore, ISEQ00010). The membranes were blocked by incubation for 1 h with TBST (0.1% Tween 20) containing 5% non-fat milk powder (w/v), then incubated with primary antibodies. Corresponding HRP-conjugated secondary antibodies were subsequently incubated for 2 h at room temperature.

High throughput sequencing (RNA sequencing analysis)
Mice were anesthetized with iso uorane (VETEASY, RWD Life Science) and perfused transcardially with phosphate-buffered saline (PBS). Put the brain tissue in pre-cooled PBS, and separate the hippocampus from brain tissue with ne tweezers. Total RNA of the hippocampus was extracted by RNeasy Mini Kit   (QIAGEN, 74106). RNA samples were then further puri ed with magnetic oligo(dT) beads after denaturation. Puri ed mRNA samples were reverse transcribed into the rst-strand cDNA, and a second cDNA was further synthesized. Fragmented DNA samples were blunt-ended and adenylated at the 3' ends. Adaptors were ligated to construct a library. DNA was quanti ed by Qubit (Invitrogen). After cBot cluster generation, DNA samples were then sequenced by an Illumina Hiseq2500 SBS from Genergy Biotechnology Co., Ltd. (Shanghai, China). Raw data were converted into Fastq format. The number of transcripts in each sample was calculated based on the number of fragments per kilobase of transcript per million fragments mapped (FPKM); Cuffnorm software was used to calculate the FPKM value for each sample, and the values were log2 transformed. DESeq2 software was used to calculate the differential gene expression between different samples. FDR (adjusted P-value) ≤ 0.05 was used to screen up-regulated or down-regulated RNAs. For the KEGG pathway analysis, the entire set of genes was used as the background list, the differential genes were used as the candidate list, and P was calculated. Signi cant genes were categorized based on gene functions. Data analysis at Genergy Biotechnology Co., Ltd. (Shanghai, China).

Statistical analysis
Differences in protein expression levels at different time points and the uorescence intensity among three groups were analyzed by using one way ANOVA. Differences in protein expression levels and the uorescence intensity between two groups were analyzed by using Student's t-test. Data of mortality rate were carried out using Fisher's exact test. The data are presented as mean ± standard error of the mean (SEM). Statistical signi cance was set at p < 0.05. The detailed statistical tests used for each analysis are stated in the gure legends. All the statistical analyses were performed with RStudio software (version 1.3.1093; https://rstudio.com/products/ rstudio/).

Dynamic changes in Clock genes in the hippocampus after seizures
Previous studies have reported the changes in mRNA expression levels of some Clock genes [15], but the protein expression changes of these genes are still unclear. To investigate the role of circadian rhythm proteins during epileptogenesis, we examined the dynamic changes in protein levels of these proteins in mice hippocampus at the different time points after pilocarpine-induced status epilepticus (SE). The acute phase was de ned as 1 to 3 days post SE [19]. The latent phase was de ned as a seizure-free period that can last weeks [20,21]. The chronic phase was de ned as mice exhibit spontaneous, recurrent seizures. Compared with the control group, Bmal1 expression was decreased at the latent phase (14 days -post-SE) and chronic phase (60 days -post-SE) ( Fig. 1A and B). Clock expression was decreased at the acute (1 day -post-SE) and chronic phase, although one way ANOVA analysis is not statistically signi cant (Fig. 1C). There was no statistically signi cant change in the expression levels of Per2 and Cry1 protein (Fig. 1D -E). Because the bulk sample of hippocampus for immunoblotting detection cannot clarify which subregions of the hippocampus have decreased Bmal1 expression. The abnormal distribution of Bmal1 in the hippocampus was labeled by immuno uorescence. Compared to control, the uorescence intensity of Bmal1 staining was reduced in the CA1 and dentate gyrus (DG) at the chronic phase ( Fig. 2A -D).

Conditional knockout of Bmal1 in DG neurons increased the susceptibility to seizures
To clarify the effect of decreased Bmal1 on the epileptogenesis, AAV2/9-hSyn-Cre-mCherry or control virus were injected into the bilateral hippocampal DG area of Bmal1 ox/ ox mice (Fig. 3A). Neuron-speci c knockout of Bmal1 (Bmal1 cKO) in DG signi cantly shortened the latency for seizures (Fig. 3B). Latency refers to the time from the pilocarpine administration (i.p.) to seizures of Racine category 4 [22]. Although there was no statistical signi cance, Bmal1 cKO increased the mortality rate resulted from seizures (Fig. 3C). The e ciency of Bmal1 cKO in Bmal1 ox/ ox mice was veri ed by immunoblotting. Compared with the control virus group, Bmal1 protein expression was signi cantly decreased in the cKO group (Fig. 3D -E), while Cre protein expression was signi cantly up-regulated in the cKO group (Fig. 3D -F).

Protocadherin 19 (PCDH19) as a potential candidate gene regulated by Bmal1
To further clarify the mechanism by which Bmal1 cKO increased the susceptibility to seizures induced by pilocarpine administration, the hippocampal tissues of Bmal1 cKO and control tissues were subjected to high-throughput sequencing. FDR (adjusted P-value) ≤ 0.05 was used to screen up-regulated or downregulated mRNAs between the different samples. The 25 up-regulated and 19 down-regulated genes are displayed through the volcano plot and heatmap plot (Fig. 4A -C). Among these genes, PCDH19 is a gene that has been reported to be closely related to epilepsy [23]. Mutations in this gene on human chromosome X are associated with sporadic infantile epileptic encephalopathy and a female-restricted form of epilepsy [24]. PCDH19 was detected on the brain tissue slices of Bmal1 cKO (Fig. 4D). The uorescence intensity of PCDH19 was signi cantly in the cKO group, compared with the control group (Fig. 4E). The level of PCDH19 in cKO was also down-regulated in the cKO group, compared with the control group (Fig. 4F -G).

Decreased expression of PCDH19 in the hippocampus of epileptic mice
Although PCDH19 mutations cause epilepsy, the expression of PCDH19 in acquired epilepsy has not been reported. Firstly, the distribution of PCDH19 in the hippocampus of TLE mice was labeled by immuno uorescence. Compared to control, the uorescence intensity of PCDH19 staining was faint in the CA1 and DG at the chronic phase ( Fig. 5A -D). The levels of PCDH19 protein were also decreased at the latent phase and the chronic phase ( Fig. 5E -F).

Abnormal expression of Bmal1 and PCDH19 in hippocampal sclerosis of patients with TLE
Hippocampal sclerosis (HS) is the common histopathology in patients with drug-resistant TLE [1]. The three types of hippocampal sclerosis are classi ed as HS International League Against Epilepsy (ILAE) type I (severe neuronal cell loss in CA1 and CA4, 50-60% granule cell loss and granule cell dispersion (GCD)), HS ILAE type II (CA1 predominant neuronal cell loss and GCD, but usually lack severe granule cell loss) and HS ILAE type III (CA4 predominant neuronal cell loss and 35% granule cell loss) [1]. In the present study, Bmal1 and PCDH19 were detected in DG of hippocampal sclerosis tissues (HS type I and III) and the tissues without hippocampal sclerosis (no HS) by immuno uorescence. In DG, the intensity of Bmal1/NeuN and PCDH19/NeuN in the HS type I group were signi cantly reduced compared with no HS group (Fig. 6A -D). The intensity of Bmal1/NeuN and PCDH19/NeuN in the HS type III group was reduced compared with no HS group, although there was no statistical signi cance. Furthermore, the level of Bmal1 and PCDH19 in HS type I and HS type III were decreased, compared with no HS group (Fig. 6E -F).

Discussion
In the present study, we found that Bmal1 protein was reduced in the hippocampal DG and CA1 of mice with TLE. Neuron-speci c knockout of bmal1 in DG of Bmal1 ox/ ox mice lowers the threshold of pilocarpine-induced seizures. With high-throughput sequencing and western blotting, the downstream gene PCDH19 regulated by Bmal1 was rstly identi ed and then detected in the hippocampus of epileptic mice. Furthermore, the expression of Bmal1 and PCDH19 were detected in HS type I, HS type III, and no HS. The levels of Bmal1 and PCDH19 protein in DG of HS type I and HS type III were decreased, compared with no HS group. These results suggest that these changes in Bmal1 and PCDH19 expression may be strongly associated with epileptogenesis and play an important role in the pathogenesis of TLE.
Circadian rhythm genes, as the important transcription factors, not only control rhythmic physiological activities such as sleep and hormone secretion, but also Clock-controlled genes (CCGs) are involved in neurodegenerative diseases, such as Alzheimer's disease and Parkinson's disease [25,26]. In previous studies, the levels of Bmal1, CLOCK, Cry and Per mRNA have been con rmed to be decreased after druginduced and electrically stimulated seizures in animal models [15]. The threshold of seizures induced by electrical stimulation in Bmal1 knockout (KO) mice is lower compared with the control group [16]. In the study, the knockout of clock genes is systemic, so it is unclear which organs or tissues of clock genes are knocked-out directly affect seizures. Here, the role of Bmal1 in the epileptogenesis and seizures for TLE was examined in Bmal1 ox/ ox mice with conditional deletion of Bmal1 gene in hippocampal neurons.
Our results show that Bmal1 was decreased in DG of mice and patients, and neuron-speci c knockout of Bmal1in DG of Bmal1 ox/ ox mice signi cantly shortened the latency for seizures. This suggests that bmal may be involved in the epileptogenesis of TLE. In our study, the changes in Bmal1 expression mainly occurred in the CA1 and DG neurons. Therefore, we did not perform knockout and functional veri cation of Bmal1 in hippocampal astrocytes. The changed functions of astrocytes caused by Bmal1 de ciency may be involved in the pathogenesis of TLE. Astrocyte-speci c Bmal1 deletion induces astrocyte activation and in ammatory gene expression in vitro and in vivo and alters circadian locomotor behavior and cognition through GABA signaling in mice [17,18]. Astrocyte activation and gliosis are one of the common pathological symptoms of TLE [27].
At present, the molecular mechanism of Bmal1 involved in epileptogenesis has not been reported. In the present study, by using high-throughput sequencing, 25 up-regulated or 19 down-regulated mRNAs in Bmal1 cKO mice were screened (FDR ≤ 0.05). As one of the candidate genes, PCDH19 is a cell adhesion molecule belonging to the cadherin family. Its prominent expression is in the nervous system especially in limbic areas and cortex [24]. PCDH19 mutations result in an epileptic syndrome known as EIEE9 (OMIM # 300088). A mechanism of cellular interference has been suggested, wherein the coexistence of neurons expressing wild-type (WT) or mutant PCDH19 disrupts cell-cell interactions [28]. PCDH19 downregulation has been proved to bind and regulate GABA A Rs kinetics, and increase the frequency of action potential ring [23]. PCDH19 downregulation in rat hippocampal neurons also affects the dendrite morphology [29]. Interestingly, Clock ox/ ox mice with conditional deletion of the Clock gene in excitatory neurons also show speci c spine defects and increased excitability [12]. Considering that Bmal1 and Clock are involved in transcription in the form of Bmal1:Clock complex, these indicate that the abnormal expression of Bmal1 and Clock in neurons may cause similar phenotypes and affect the epileptogenesis.
In epilepsy, DG cells formed excessive de novo excitatory connections and recurrent excitatory loops, leading to the ampli cation and propagation of excessive recurrent excitatory signals [30]. The granule cells' aggregate excitability has the potential to provide a therapeutic target [31]. In the present study, the expression of Bmal1 was signi cantly reduced in DG of patients and mice with TLE and lowered the seizure threshold via PCDH19. Therefore, DG was chosen as the target for Bmal cKO with AAV.
Clinical and animal experiments nd that patients and animals with TLE show a 24-hour non-uniform distribution of seizure occurrence. These suggest that the epileptogenesis and seizures of TLE may be associated with the circadian rhythms. The two hypotheses have been proposed in seizures of TLE: (1) Rhythmic activity of molecules causes an increase in excitability periodically exceed the seizure threshold, displaying the behavioral seizures.
(2) Oscillation of neuronal excitability in the suprachiasmatic nucleus (SCN) modulates the rhythmic excitability in the hippocampus via neural projections [32]. Previous studies have found that Bmal1 expression level in hippocampus still presents the circadian rhythmic oscillation in epilepsy [15]. Although the connection of nerve bers between the suprachiasmatic nucleus and the dentate gyrus is not clear, the circadian rhythmic activity of DG has been reported in TLE [33,34]. These suggest that the rhythmic activity of Bmal1 may cause increased excitability via PCDH19, and then periodically exceed the seizure threshold.
However, one of the disadvantages of this study is that it cannot arti cially reduce the expression of Bmal1 while maintaining its periodic expression oscillation characteristics in Bmal1 ox/ ox mice with conditional deletion of the Bmal1 gene. We did not study the effect of Bmal1 KO in SCN on the epileptogenesis and seizures of TLE. Because we have not yet determined the expression changes of circadian rhythm molecules in SCN of TLE. The role of Bmal1 in the SCN of TLE will continue to be explored in the future.
In conclusion, we have disclosed a new biological function of Bmal1 in the epileptogenesis and seizures of TLE and found a downstream gene regulated by Bmal1. Our research ndings may help in developing chronotherapy for mTLE, based on the chronobiology of spontaneous seizures. More detailed understandings of the role of Bmal1 and other Clock genes in the brain are required and may give novel insights into the mechanism underlying epileptogenesis and seizures of TLE.

Declarations
Ethics approval and consent to participate This study was performed according to the Helsinki Declaration and approved by the Ethics Committee of Xi'an Jiaotong University (Ethics and Science # G-83) in full accordance with the ethical guidelines of the National Institutes of Health for the care and use of laboratory animals. We confrm that we have read the Journal's position on issues involved in ethical publication and afrm that this work is consistent with those guidelines.

Consent for publication
Not applicable.

Availability of data and materials
The datasets used and analyzed in this study are available from the corresponding authors on reasonable request.