Exogenous microRNA 322-5p Reduced Neuronal Inammation via the TRAF6/IRF5/NF-kB Axis and Restored GAD1/GABA Expression in a Pilocarpine-induced Epileptic Rat Model

Objective: MicroRNAs (miRNAs) are small noncoding RNAs that control gene expression at the posttranscriptional level. Some dysregulated miRNAs have been shown to play essential roles in epileptic development. This study aimed to determine if microRNA-322-5p regulates seizure and seizure damage by targeting the NFκB-TLR4 associated inammatory signaling pathway. Methods: Pilocarpine-induced epileptic rat model was established. Immunohistochemical staining demonstrated the pathology of epilepsy. The expression of microRNA-322-5p, inammatory markers, NF-kB, TLR4, IL-1β and IL-6, and synaptic inhibitory molecules, GAD1 and GABA, were assessed by a quantitative polymerase chain reaction, and western blotting, respectively. Results: The expression of microRNA-322-5p was signicantly decreased in the SE (status epilepticus) rats compared with the normal counterparts. The reduction of miR-322-5p was accompanied by an increased level of pro ‐ inammatory cytokines such as IL-6 and TNF-α via increased NF-kB expression and reduced GAD1 and GABA expression. The exogenously increased miR-322-5p level by mimic molecules signicantly reduced the inammatory proles and increased GAD1 and GABA expressions in the S.E. rat brain compared to nontreated counterparts. Conclusions: Our ndings suggest that the restoration of miR-322-5p resulted in a signicantly reduced TRL4/IRF1/NF-kb associated inammatory circuit and increased GAD1 and GABA expression. These ndings suggest that miR-322-5p induction may be of therapeutic potential for neural damage as a result of repeated epileptic episodes.


Background
Status epilepticus (SE) is a common acute and critical illness in children. If it is not treated in time, it can cause irreversible brain damage due to high fever, circulatory failure, or neuronal excitotoxic damage. Long-term loss of learning and memory, mental retardation, mental retardation, and paralysis will remain.
In severe cases, it can endanger the lives of children, with high mortality and disability rates. According to the survey, the annual incidence of SE in all age groups is (10-40)/100,000 (1). SE is seen in 10%-20% of children with epilepsy, and the tonic-clonic status is the most common. The longer the duration of SE attack, the more serious the brain damage caused, and the fatality rate of SE is close to 20% (1). Therefore, active, rapid, and effective control of epileptic seizures is the key to saving patients' lives and improving their prognosis, especially in the eld of pediatrics.
Emerging evidence indicates that dysregulation of in ammatory circuits in the lesion sites represents a critical factor in the development of epilepsy. However, it remains underappreciated as to how the unbalanced regulation of in ammation contributes to the development of epilepsy and subsequent drug resistance (2). A strong association between in ammatory circuits and decreased activity of γ-Aminobutyric acid (GABA), a major inhibitory neurotransmitter, has been shown in epileptogenesis (3,4) Thus, GABA receptors and/or associated signaling have been the targets for therapeutic development. In addition, the level of GABA is regulated by a key enzyme, glutamate decarboxylase 1 (GAD1). Together, these signaling pathways and enzymes play a signi cant role in the development of epileptic activities (3). The etiology of epilepsy remains complex, the neuroin ammatory responses could arise from a variety of different sources, such as head trauma, infections of the central nervous system (CNS), CNS cancer and the aging process all can contribute to the development of epilepsy. The activation of Toll-like receptors (TLRs) via a nuclear protein released from dying cells such as high-mobility group box-1 (HMGB1) has been shown to contribute to chronic epilepsy in rat models as well as in human epileptic tissues. For instance, pilocarpine, a chemoconvulsant that induces epilepsy with chronic seizures, was mediated by the ampli cation of TRL4 associated signaling. However, the regulation of TRL4 signaling in the context of epilepsy has not been fully appreciated.
MicroRNAs (or miRs) represent a group of small noncoding RNAs that can speci cally bind to the 3' untranslated region (3'UTR) of mRNA targets and subsequently silence the expression by way of inhibiting the translation of the target proteins (5). Emerging evidence indicates that miRs also contribute to the development of epileptogenesis. For instance, a previous study reported speci c subsets of miRNAs that were upregulated and downregulated in a status epilepticus (S.E.) rat model (6). In our search of public databases, we identi ed miR-322-5p as one of the most decreased microRNA species in the S.E. rat brain. However, its role in epilepsy remains mostly unexplored. Connecting the importance of neuroin ammation, the role of GABA and its related signaling pathways in epilepsy, in this study, we aimed to demonstrate the relationship between miR-322-5p and its targeted signaling circuits. Finally, we investigated the feasibility of translating our research results to future clinical applications.

Materials And Methods
Pilocarpine induced epileptic rat model Sprague-Dawley rats (male, 20-day of age) were purchased from SLAC (Shanghai, People's Republic of China) and maintained in the following conditions: 12-hour light/dark cycle at a constant temperature of 23°C±2°C and with food and water ad libitum. All procedures were conducted according to the regulations of Animal experiment facility and approved by the Experimental Animal Ethical Committee of Zhejiang Provincial People's Hospital. Brie y, 48 rats were used to produce an epilepsy model through lithium chloride (LiCl) injection (3 mEq/kg, i.p.) followed by an injection of 1mg/kg methyl scopolamine bromide (18 hours later). 30 minutes after methyl scopolamine bromide injection, rats were injected with pilocarpine hydrochloride (100 mg/kg, i.p.) (7). Rats were then monitored to detect seizure activity for 2 h following pilocarpine administration. Seizures were determined and recorded in ve stages based on Racine's scale with slight modi cations. In short, stage 1: face and vibrissae twitching, ear rubbing on forepaws, chewing. Stage 2: nodding of the head, unilateral limb clonus. Stage 3: limb clonus, mild convulsions. Stage 4: rearing with bilateral forelimb clonus, tail hypertension, lockjaw, and whole-body convulsions. Finally, stage 5: rearing with body convulsions and collapsing with body rigidity. The onset of S.E. was de ned when stages 4 or 5 were reached, and the development of continuous motor activity(8). Chloral hydrate (400 mg/kg, i.p) and atropine (1mg/kg, i.p) was subsequently injected 30 minutes post onset of S.E. to attenuate seizure activity (9).

Animal grouping and intervention
After successful modeling of SE rats, the rats were divided into two experiments: Experiment 1, the expression of miR-322-5p, TRL4, TRAF6, RelA (NF-kb subunit) and IRF1 in SE rats at different times; SE rats divided into 4 subgroups of 1 week, 2 weeks, 3 weeks and 4 weeks (n=8 in each group). The control group consisted of eight normal rats. Experiment two, the effect of overexpression of miR-322-5p. The SE rats were divided into SE + miR-322-5p intervention group and SE + control(scrambled) (n=8 in each group). Overexpression of miR-322-5p, which targets TRAF6 and IRF5 mRNAs were designed and synthesized by Shanghai GeneChem Co., Ltd. (Shanghai, China). Please refer to Table 1 for the primer sequences used in this study. The microinjection of exogenous miR-322-5p was performed according to a previously established protocol. In short, rats were microinjected with 10 µl miR-322-5p (50 µmol/l) via a pre-implanted cannula, 30 minutes after SE modelling for 3 consecutive days. Following each injection, the needle was left in place for 5 min to allow complete diffusion of the injected material. The SE+control (scrambled) group received 10 µl scrambled RNA (50 µmol/l) and served as the negative control. The control group and the epilepsy group received the same volume of normal saline for microinjection. One week after treatment, all animals were put to death and specimens were kept.

Quantitative PCR reactions
Neuoin ammatory genes and miR-322-5p expression pro les were measured by real-time PCR reactions.
Total RNAs were extracted using a Trizol kit and the quality was checked by OD260/OD280 ratio (RNA samples with the 260/280 ratio >1.8 were used for experiments). cDNA was then reversed and transcribed and the qPCR reactions were performed using a LightCycler 96 (Hoffman-La Roche Ltd., Basel, Switzerland) following the preset procedure: 95°C initial denaturation for 5 minutes, 95°C denaturation for 30 seconds, 58°C annealing for 30 seconds, and 72°C extension 30 seconds for 30 cycles. The expressions of the genes of interest were determined by the −ΔΔCT method. All primer sequences can be found in Table 1.

Immunohistochemistry
Rats were rst anaesthetized by 10% chloral hydrate (i.p. injection) and perfused with cold 4% paraformaldehyde (Sigma, China) through the auricula sinistra for 30 min. Rats were then sacri ced by decapitation, and the hippocampal tissues were collected and xed in 4% paraformaldehyde for 2 h.
Subsequently, the tissues were para n-embedded, sliced (6 μm-thick), and stained. Nissl staining procedures were performed according to a previously established method (10).

Western blotting experiments
Total cellular proteins from the brain homogenates were extracted using RIPA buffer (Santa Cruz Biotechnology, CA, USA) supplemented with a cocktail of protease inhibitors (Roche, Basel, Switzerland).

Results
Status epilepticus (S.E.) rat brain showed an elevated TRL4/NF-kB expression pro le and decreased GABA expression.
A pilocarpine-induced epileptic rat model was established to examine the role of TRL4/NF-kB circuit in epileptogenic development. The representative hippocampal section from the S.E. rat brain showed disorganized neurons compared to their control counterparts ( Figure 1A). The amount of GABA was signi cantly reduced in the S.E. rat homogenates for the validation of the model (insert, Figure 1B). We then examined the components of the Toll-like receptor 4 (TRL4) in ammatory circuit and found that there was a time-dependent increase in the expression of TRL4, TRAF6, RelA (NF-kb subunit) and IRF1 ( Figure 1C), suggesting the chronic in ammation had occurred in the S.E. brains.
A negative association was observed between TRL4/TRAF6 circuit and the level of miR-322-5p in S.E. rat brain tissue Subsequently, we searched a public database of microRNA pro lings in S.E. rats (11)and found that one of the most affected microRNA speices was miR-322-5p (Figure 2A). Based on three target gene prediction tools (miRmap, TargetScan, and PITA), we identi ed that one of the key TLR4/NF-kB signaling components, TRAF6, was the target of miR-322-5p where miR-322-5p binds to the 3'UTR of TRAF6 (upper panel, Figure 2B). In addition, analysis from a database (12) (GSE16969) indicates a relative strong negative relationship between the expression of miR-322-5p and the incidence of S.E. (miR-322-5p was found lower in the patients with intractable epilepsy, lower panel, Figure 2B). Our in vivo data supported this notion where the level of miR-322-5p sharply reduced upon the S.E. induction by pilocarpine (one week after induction) and continued to decrease over the next 3 weeks until reached the lowest level (approximately 50% compared to that in the control counterparts) ( Figure 2C). Subsequently, our qPCR analyses of the S.E. rat brain tissues also demonstrated a signi cantly increased mRNA level of TRL4, TRAF6, IRF5, IL-1β, and RelA ( Figure 2D).
Exogenous miR-322-5p treatment signi cantly reduced the neuroin ammation After establishing that the reduced level of miR-322-5p in the S.E. rats was associated with the increased TRAF6/NF-kB in ammatory circuit, we aimed to determine the therapeutic potential of miR-322-5p. The brain tissues from rats that were injected with exogenous miR-322-5p immediately post S.E. induction clearly showed a higher number of viable neurons in the CA3 region compared to the S.E. rats without treatment ( Figure 4A). In addition, rats injected with miR-322-5p showed a signi cantly lower level of neuroin ammation, as re ected by the reduced expression of TRAF6, RelA, IL-6, and TRL4 than the S.E. rats without treatment ( Figure 4B). The level of miR-322-5p was observed to be elevated in the miR-322-5p injected rat brain compared to their control counterparts ( Figure 4C). Finally, the western blots of the brain homogenates demonstrated that a lower expression of neuroin ammatory markers including TRAF6, TLR4, RelA, and increased expression of GAD1 and GABA in the rat brain received Discussion Accumulating evidence indicates a close correlation between chronic in ammation and epileptogenesis in patients. However, the signaling network underlying responsible for the persistence of neuroin ammation remains to be explored. The classical TLR4/NF-kB signaling pathway is established to be responsible for in ammatory responses in different diseases including epilepsy (13). It has been postulated that prolonged use of anti-epileptic drugs may contribute towards acute/chronic in ammation, oxidative stress, leading to recurrent epilepsy, and resistance to drugs (14). In this study, we identi ed elevated TLR4/IRF5/NF-kB signaling in the brain of S.E. rats. Previous studies demonstrated that the in ammatory mediators, such as tumor necrosis factor (TNF)-α, IL-1β, and IL-6 may contribute towards epileptic development and progression. More importantly, elevated expression of TLR4 was identi ed in surgical resection specimens of drug-resistant temporal lobe epilepsy (TLE), focal cortical dysplasia (FCD), and tuberous sclerosis complex (TSC) patients (15). The inhibition of HMGB1/TLR4 was shown to reduce the incidence of seizures (16). Moreover, TRAF6, a downstream effector of TRL4 signaling, has been shown to play an essential role in Th1 type in ammation and involved in inhibiting autophagy and promoting oxidative stress in hemorrhaged rat brain (17). More importantly, the inhibition of TLR-4/NF-κB signaling and IL-1β in the hippocampus was shown to attenuate the severity of S.E. in rats(18). These ndings signi ed that TLR4/NF-kB signaling could play an important role in the development of epilepsy.
Here, we demonstrated another potential regulatory mechanism by which TRL4 signaling was ampli ed in the S.E. rat brains. miR-322-5p was one of the most reduced microRNA species after the S.E. establishment in our study as well as in another study (11). Our results showed that miR-322-5p downregulated TRAF6 and IRF5, both activating components of the NF-kB signaling pathway. A previous study indirectly supported our notion where an increased level of miR-322-5p elevated EZH2 and activated Akt/GSK3β pathway, thereby protecting myocardial cells from ischemic reperfusion injury (19).
On the contrary, in the S.E. rat brain, miR-322-5p was signi cantly reduced and associated with a higher level of neruoin ammation.
We then veri ed these observations in GABAergic M213-2O cells where LPS treatment signi cantly induced the TLR4/IRF5/NF-kB in ammatory circuit. Previous reports also supported that peripheral in ammation contributed to the development of seizures and TLR4 signaling appeared to be instrumental (20). In ammatory cytokines such as IL-6 and IL-1β signi cantly increased upon TLR4 activation and consistently reported in epileptic patients, indicating a chronic in ammatory state (21). In addition, increased IL-6 was shown to mediate the degradation of GABAergic neurons in the mouse brain (22). Our ndings were in agreement with these reports where increased IL-6 and IL-1βlevel was associated with decreased GABA and GAD1 expression. More importantly, we provided evidence that exogenous miR-322-5p (using mimic molecules) resulted in the restoration of GAD1 and GABA in LPStreated M213-2O cells, accompanied with a sign ciantly reduced expression of the neuroin ammatory markers, NF-kB, TRL4 and IRF.
Finally, we examined the feasibility of exogenous miR-322-5p as a Preventive measure. Rats injected with exogenous miR-322-5p immediate after S.E. induction showed a signi cantly lower expression of TLR4, IRF5, and NF-kB, while increasing GAD1 and GABA, compared to their counterparts which received scramble. This nding suggested the feasibility of exogenous miR-322-5p as an inhibitor of neuroin ammation. Emerging evidence demonstrates that stereotactic injection or antagomirs or mimic molecules into the hippocampal regions not only is feasible but also reduces the incidence of seizures in rodent models (23). Thus, our ndings provided support for injecting exogenous miR-322-5p as a Preventive measure for reducing chronic neuroin ammatory circuits, namely, the TRL4/TRAF6/NF-kB pathway.

Conclusions
In the graphical summary of Figure 5, the present study provided preclinical in vitro and in vivo models demonstrating the essential role of TRL4/TRAF6/NF-kB as a neuroin ammatory signaling pathway associated with pilocarpine-induced epilepsy; miR-322-5p functioned as a negative regulator of this in ammatory circuit and could serve as a potential therapeutic tool for suppressing chronic in ammation and preventing the decreased level of GAD1 and GABA, thereby reducing the incidence of epileptic episodes. Further investigation is warranted for the safety of exogenous miR injection.      injected rats. A signi cantly lower levels of TRL4, TRAF6, IRF5, IL-6, and NF-kB were observed in the miR-322-5p group as compared with the scramble group. (C) Similarly, the miR-322-5p injected rat brain contained signi cantly higher miR-322-5p than the ones from the control group. (D) Western blot of the brain homogenate. The neuroin ammatory markers, TLR4, RelA (NF-kB), and TRAF6 were decreased compared to the scrambling control tissues.

Figure 5
A pilocarpine-induced rat epileptic model was created to demonstrate the important role of the neuroin ammatory circuit, TRL4/TRAF6/NF-kB in the epileptogenic process. In the status epilepticus (S.E) rat brain, miR-322-5p was found signi cantly lower than the normal counterparts and TRAF6 and IRF5 were found to be miR-322-5p's targets. An increased level of miR-322-5p was shown to signi cantly reduce the neuroin ammatory markers and restore the level of GAD1 and GABA.