Inhibition of hyperactivity of the dorsal raphe 5‐HTergic neurons ameliorates hippocampal seizure

Abstract Aims Epilepsy, frequently comorbid with depression, easily develops drug resistance. Here, we investigated how dorsal raphe (DR) and its 5‐HTergic neurons are implicated in epilepsy. Methods In mouse hippocampal kindling model, using immunochemistry, calcium fiber photometry, and optogenetics, we investigated the causal role of DR 5‐HTergic neurons in seizure of temporal lobe epilepsy (TLE). Further, deep brain stimulation (DBS) of the DR with different frequencies was applied to test its effect on hippocampal seizure and depressive‐like behavior. Results Number of c‐fos+ neurons in the DR and calcium activities of DR 5‐HTergic neurons were both increased during kindling‐induced hippocampal seizures. Optogenetic inhibition, but not activation, of DR 5‐HTergic neurons conspicuously retarded seizure acquisition specially during the late period. For clinical translation, 1‐Hz‐specific, but not 20‐Hz or 100‐Hz, DBS of the DR retarded the acquisition of hippocampal seizure. This therapeutic effect may be mediated by the inhibition of DR 5‐HTergic neurons, as optogenetic activation of DR 5‐HTergic neurons reversed the anti‐seizure effects of 1‐Hz DR DBS. However, DBS treatment had no effect on depressive‐like behavior. Conclusion Inhibition of hyperactivity of DR 5‐HTergic neuron may present promising anti‐seizure effect and the DR may be a potential DBS target for the therapy of TLE.

The 5-HTergic neurons are mainly located in various subregions of the raphe nuclei that distributed near the midline of the brainstem. The dorsal raphe nucleus (DR) is the predominant source of 5-HTergic innervation of the forebrain, 18 implicating in diverse functions including anxiety, depression, and sleep-wake cycles. [19][20][21] But functions of the DR in epilepsy are limited investigated. It is reported that neuronal activities of different raphe subregions in different epilepsy models are heterogeneous. 22,23 Even in the DR, population firing and identified 5-HTergic neurons firing during seizure are diverse from each other. 24,25 What on earth the causal role of the DR and its 5-HTergic system in TLE is still unknown. Early nonspecific interventions studies found that electrolytic lesion of the midbrain raphe enhanced the epileptiform activity in hippocampus kindling model and DBS of the median raphe (MR) significantly inhibited convulsion in different types of seizure models. 8,26,27 However, a recent research indicated that activation of DR 5-HTergic neurons in DBA/1 mice reduced the respiratory arrest, but had limited effect on seizure itself. 28 These results indicate that the DR may have an important but heterogeneous role in TLE and more direct evidence of the function of DR 5-HTergic neurons in seizure should be provided. Here, by using calcium fiber photometry, optogenetics, and DBS, we investigate the role of DR and its 5-HTergic neurons in epilepsy.

| Animals
Pet-1 ETS gene is required for the differentiation of 5-HTergic neurons. 29 The Pet-cre mice (stock No. 012712) were genotyped according to the protocols provided by Jackson Laboratory and were bred to C57BL/6J mice. Male Pet-cre mice and C57BL/6J mice (at least 2 months old) were used. They were group housed with ad libitum access to standard pellet food and water in a 12/12 h light/ dark cycle (lights on 8 a.m.). Mice were singly housed after surgeries to facilitate recovery. All behavior experiments were performed between 9:00 a.m. and 7:00 p.m., which were approved by the Zhejiang University Animal Experimentation Committee and were in complete compliance with the National Institutes of Health Guide for the Care and Use of Laboratory Animal and followed the ARRIVE guidelines. 30

| Stereotaxic surgery
Stereotaxic surgeries were performed as our previous studies. 31,32 Briefly, under sodium pentobarbital anesthesia (50 mg/kg, i.p.), mice were head-fixed in a stereotaxic frame (512600, Stoelting) for injections and implantations with a heating pad. Virus (400 nl) was in- Co. LTD) was implanted in the DR and affixed with dental cement to deliver light. Two screws were placed over the cerebellum to serve as the reference and ground electrodes.
For DBS of the DR in the kindling model, kindling electrodes were implanted in the right ventral hippocampus for electrical stimulation and EEG recording, and the DR for the delivery of DBS. After behavioral tests, the place of the electrode implantation and viral expression in all mice were verified and only the mice with correct place were taken into data analysis. These criteria were pre-established.

| Hippocampal kindling model
After one-week surgery recovery, the CA3 of mice were stimulated with a constant-current stimulator (SEN-7203, SS-202J; Nihon Kohden) and the EEGs were recorded with a Neuroscan system (NuAmps, Neuroscan System). The stimulation intensity was started at 40 μA and was then increased in 20-μA steps every 1 min, until produced at least a 5-s after-discharge. This stimulation intensity was defined as after-discharge threshold (ADT) of each mouse, which indicates epileptogenic sensibility for that mouse and was used for grouping thereafter. Subsequently, all mice received 10 kindling stimulations daily (400 μA, 20 Hz, 2-s trains, 1-ms monophasic square-wave pulses). Seizure severity was classified into seizure stages 1-5 according to the Racine scale, among which stages 1-3 were considered as focal seizures (FS) and stages 4-5 as generalized seizures (GS). 33 The severity of seizure was scored by a trained observer who was unaware of the experimental groupings.

| In vivo fiber photometry
One week after surgery, fiber photometry was conducted during hippocampal kindling with the fiber photometry system (Nanjing Thinkertech) strictly according to our previous studies. 34,35 The GCaMP fluorescence was bandpass filtered and collected by a photomultiplier tube using a 488-nm diode laser that coupled into a 200μm optical fiber. An amplifier was used to convert the photomultiplier tube current output to voltage signals, which was further filtered through a low-pass filter (40 Hz cut-off). We analyzed data

| Light stimulation and DBS
473-nm (20 Hz, 10 ms/pulse, 5 mW) or 589-nm laser light (continuous, 5 mW) was delivered by the laser (IKECOOL Laser) through a 200μm optic fiber. The light stimulation was delivered immediately after kindling stimulation and stopped at the end of ADD manually.
To investigate the effect of DBS in the DR on the kindling model, mice were divided into four groups (sham, 1, 20, and 100 Hz). Mice in sham group were connected to the apparatus but no DBS was delivered while the other groups received DBS (monophasic squarewave pulse, 0.1 ms per pulse) immediately after kindling through a constant-current stimulator. The DBS threshold was 1/5 of the current intensity that induced abnormal behavior (300 μA for 1 Hz, 100 μA for 20 Hz, and 30 μA for 100 Hz).

| Assessment of depressive-like behavior
For all depressive-like behavioral tests, mice were transported in a holding cabinet to the testing room with dim light (~20 lux) where they were habituated at least 1 day before testing. 36 During the testing session, the behavior of each mouse was recorded by a video tracking system. Experiment and quantitative analyses were conducted by a trained experimenter who was unaware of the experimental groupings.
Sucrose preference test. Mice were housed individually and habituated with two identical bottles with 1% sucrose for 2 days, followed with 2-day identical bottles with water. Then, the experimental mice were presented with two bottles (one containing water and the other containing 1% sucrose) for 2 days, and the bottles' positions were switched after 1 day. The consumption of each fluid was measured, and sucrose preference was expressed as the percentage of (sucrose consumption)/(water consumption + sucrose consumption).
Tail suspension test. Each mouse was individually suspended 50 cm above the surface of a table using adhesive tape that was placed roughly 1 cm from the tip of the tail, which was videotaped from the side. Each mouse was tested only once for 6 min and the immobility time in the last 5 min was measured. Mice were considered immobile without initiated movements, including passive swaying.
Splash test. Each mouse was placed individually in their home cage and were sprayed with 10% sucrose solution on the back. The latency of grooming was recorded with a maximum of 5 min.

| Immunohistochemistry
Mice were sacrificed and perfused with PBS and 4% PFA. After

| Optogenetic inhibition of DR 5-HTergic neurons retards the acquisition of hippocampal seizure
To assess the causal role of DR 5-HTergic neurons in seizure, we then used optogenetics to selectively activate or inhibit DR 5-HTergic neurons in hippocampal kindling model.  Figure 2C). We discovered that photo-inhibition of DR 5-HTergic neurons significantly retarded the progress of seizure stage ( Figure 2D) and shortened ADD ( Figure 2E). This anticonvulsive effect mainly existed during the late period of the seizure acquisition, where seizures were completely inhibited during GSs but not FSs demonstrated by the significantly increased the number of stimulations in stage 0 ( Figure 2F,H). In contract, optogenetic activation of DR 5-HTergic neurons had no effect on seizure acquisition ( Figure 2D,E). Figure 2G demonstrates the representative ADDs and corresponding power spectrums of the 40th stimulation. Overall, Figure 2 demonstrates that selective inhibition of DR 5-HTergic neurons has an inhibition on hippocampal seizure.

| 1-Hz DBS of the DR retards the acquisition of hippocampal seizures through inhibition of DR 5-HTergic neurons
Next, we aimed to investigate whether DBS, an approach much suitable for clinical translation medicine, in the DR would have a therapeutic potential for hippocampal seizure. We implanted electrodes into the DR for the delivery of DBS and electrodes into the ventral hippocampus for EEG recording and kindling stimulation ( Figure 3A-C). We found that 1-Hz DR DBS significantly inhibited the progress of seizure stage ( Figure 3D) as well as shortened ADD ( Figure 3E).
Similar as that of optogenetic inhibiting 5-HTergic neuron, the antiseizure effect of DBS existed mainly during the late period of the kindling acquisition, where DBS significantly increased the number of stimulations in stage 0 during GSs ( Figure 3F,H). Whereas, DBS at the frequency of 20 or 100 Hz had no effect on the kindling acquisition ( Figure 3D-F). Representative ADDs and corresponding power spectrums were shown in Figure 3G. These data demonstrated that 1-Hz-specific DBS retarded the acquisition of hippocampal seizure.

| 1-Hz DBS of the DR has no effect on depressive-like behavior
Since the comorbid depression is common in epilepsy and closely interrelated with the activity of DR neurons, 36,38 we tested whether 1-Hz DBS of the DR would attenuate depressive-like behavior. We

| DISCUSS ION
Kindling-induced seizures are associated with activation of proinflammatory pathways 40

| Neuron activities of the DR in epilepsy
Activities of the DR were reported to be variable largely with increased, decreased, and not changed during seizure in different epilepsy models. [23][24][25] It should be noted that all of these researches were applied in anesthetized or decapitated animals.
In awake WAG/Rij rat model of absence epilepsy, spike and wave discharges propagate to the DR with a short delay with increased firing rate of the DR neurons. 25 However, in a same absence epilepsy model, level of 5-HT metabolite 5-hydroxyindoleacetic acid

| Optogenetic modulation of DR 5-HTergic neurons in epilepsy
In early studies, electrolytic lesion of the midbrain raphe and systematic lesion of the 5-HTergic system significantly increase the electrographic seizure activity while systematic enhancing the 5-HTergic system inhibit the seizure. 8 In addition, 5-HT1A/7 receptor or 5-HT2C receptor agonist significantly reduced epileptic discharges induced by maximal dentate activation in a partial complex model of TLE. 43,44 It should be noticed that these intervention studies modulate the whole brain 5-HTergic system without taking the heterogeneous function of the 5-HTergic sub-system into consideration. 37,45 Notably, conflicting reports about the role of SSRIs in epilepsy indicates both anti-seizure and pro-seizure results. 14,15,46,47 It indicates that precise investigation of the role of the 5-HTergic subsystems in epilepsy is necessary. Interestingly, in hippocampal kindling model, 2. The 5-HTergic subsystems may play different roles in different periods of epilepsy. In our study, the anti-seizure effect of inhibiting DR 5-HTergic neurons shows at the late period of the seizure acquisition. One explanation is that functions of DR 5-HTergic neurons may be more important in the late period of epilepsy, since calcium activities also demonstrated that degrees of participation of DR 5-HTergic neurons are higher during GS than during FS. Another explanation is the intrinsic characteristics of the central 5-HTergic system that SSRIs often take time to make an effect and acute intervention hardly play a role. 47 3. The 5-HTergic subsystems may play different roles in different models of epilepsy. In a DBA/1 mouse model of epilepsy, optogenetic activation of DR 5-HTergic neurons significantly suppresses respiratory arrest but has limited effect on acoustic stimulationinduced seizure intensity or pentylenetetrazole-induced GS. 28 However, the modulation of 5-HTergic system in epilepsy may be seizure type-dependent. In hippocampal kindling model of TLE, another type of FS with secondary GS, activities of DR 5-HTergic neurons increased aberrantly during hippocampal seizures.
Overall, DR 5-HTergic system plays a critical role in epilepsy, but more precise modulation of DR 5-HTergic neurons and circuits need to be studied in different models of epilepsy.

| DBS of the DR in epilepsy
DBS is developed into an optional treatment for medicationrefractory epilepsy and depression, although the optimal target and stimulation are still unclear. In this research, we discovered that DBS of the DR with the frequency of 1 Hz, but not 20 or 100 Hz, could mimic the anti-seizure effect of the inhibition of DR 5-HTergic neurons that significantly inhibit the seizure during the late period of the kindling acquisition. As low-frequency electrical stimulation of neurons may induce depotentiation or long-term depression, 48 it is likely that 1-Hz DR DBS is implicated in the depotentiation or inhibition of 5-HTergic neurons and thus produce anti-seizure effects. To confirm this hypothesis, activation of DR 5-HTergic neurons fully reverse the anti-seizure effect of 1-Hz DR DBS, which implied the participation of DR 5-HTergic neurons in the mechanism of 1-Hz DR DBS. In addition, modulation of the DR and its 5-HTergic systems is widely reported to have an effect on depressive-like behaviors. Acute or chronic activation of DR 5-HTergic neurons induce an antidepressive-like effect, 49,50 but inhibitions result in different consequences including increased floating behavior in forced swim test 51 or no effect. 52 In the present study, we found that 1-Hz DBS of the DR during kindling seizures has no effect on depressive-like behaviors in SPT, SP, and TST in epileptic mice. This can be due to the different timing of DBS modulation compared with previous study, in which DBS is applied during the behavior test, indicating that seizure-aligned stimulation of the DR does not affect the depressive-like behaviors of epileptic mice during inter-ictal period.
In conclusion, we found that DR 5-HTergic neurons are remarkably activated during seizure, and optogenetic inhibition of DR 5-HTergic neurons significantly retards the acquisition of hippocampal seizure. Further, 1-Hz DR DBS similarly inhibit ed hippocampal seizure which was blocked by the activation of DR 5-HTergic neurons. Together, inhibition of hyperactivity of DR 5-HTergic neuron may present promising anti-seizure effect and the DR is a potential target for the treatment of epilepsy.

ACK N OWLED G EM ENTS
This project was supported by grants from the National Natural Science Foundation of China (81630098, 82022071, 81871010, and 81821091).

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.