Effects of Levetiracetam in Acute and Kindling Model of Epilepsy Through Adenosinergic Pathway in Mice: Possible Involvement of Adenosine A1 Receptor

Objective: The present study aimed to explore the possible levetiracetam mechanisms of action in the adenosine signaling systems using the PTZ-induced acute seizure and the PTZ-induced kindling model of epileptogenesis. Method: In acute model, male mice received caffeine (non-specic adenosine receptor antagonist), or dipropylcyclopentylxanthine (DPCPX) (specic A1 receptor antagonist) prior to levetiracetam. After 30 minutes, a convulsant dose of PTZ (100 mg/kg) was administered to determine whether caffeine or DPCPX have any antagonistic effects on anti-seizure activity of levetiracetam by analyzing the onset of rst myoclonic jerk (FMJ), generalized clonic seizures (GCS) and percent mortality. The chronic PTZ-induced kindling model was set to assess the gene expression changes in adenosine A1 receptor, inwardly-rectifying potassium channel (Kir3.2) and equilibrative nucleoside transporter-1 (ENT-1) through RT-qPCR. Data were analyzed using Origin statistical software version 8.5 and represented as Mean ± SEM. Results: In acute study, we found that caffeine (100 mg/kg) and DPCPX (25 mg/kg) reversed the anti-seizures effects of the levetiracetam signicantly by reversing the percent protection and shortening the delay in the FMJ and onset of GCS in animals. In kindling model of epileptogenesis, it was found that levetiracetam increased the gene expression of adenosine A1 receptor andKir3.2 in the brain. Furthermore, levetiracetam signicantly reduced the gene expression of ENT-1 in the brain that supposed to enhance the extracellular adenosine in the brain. Conclusion: Based on these results, it can be concluded that in addition to its action on SV2A vesicular protein, levetiracetam also prevent epileptogenesis by acting on the adenosine pathway in the CNS.


Introduction
Epilepsy is the most serious and debilitating neurological disorder characterized by recurrent intermittent seizures that affects more than 65 million population globally (Thurman, Beghi et al. 2011). It has been reported that rst seizure takes months to years after the occurrence of brain insults and this time period is known as seizure-free interval or latent period (Wong 2009). Various changes occurs during the latent period after the acute insult that later in life leads to the formation of epileptic circuit. These changes include the gliosis, neurodegeneration, mossy bers sprouting, dendritic remodeling, blood brain barrier damage, neuro-in ammatory mediators production and reorganization of synaptic connections (Pitkänen and Lukasiuk 2011).The insult-induced transformation of non-epileptic brain to epileptic, ultimately leading to epilepsy is known as epileptogenesis (Herman 2002). The time period between brain insult and onset of rst epileptic seizures may provide an opportunity during which an appropriate treatment stop or modify the epileptogenesis caused by brain insult (Löscher and Brandt 2010).
Classical antiepileptic drugs (AEDs) have various central and peripheral adverse effects that caused the non-compliance of the therapy by the patients. Novel drugs were developed with less adverse effects and broader range of effectiveness in epileptic patients. Among the novel AEDs, levetiracetam has been considered as the safest drug in term of adverse effects pro le (Briggs and French 2004). Levetiracetam has been used for epilepsy as monotherapy (Del Bianco et al., 2019, Tabrizi et al., 2019, adjuvant therapy and prophylactically for the prevention of post-traumatic seizures (Chen, Kuo et al. 2018). Recently, LEV has been used effectively for the management of neonatal seizures (Arican, Gencpinar et al. 2018) and neuropathic pain and hyperalgesia (Ardid, Lamberty et al. 2003). It has been found further that the antihyperalgesic effects of levetiracetam were due to its action on the adenosine A1 receptor, and pharmacological blockade of A1 receptor by caffeine or DPCPX at the periphery weakened the effects of this drug. However, the antiepileptic effects of levetiracetam through adenosine pathway was not reported previously. Hence,the current study focuses on evaluating the effects of levetiracetam on the adenosine A1 receptor in the CNS.
Adenosine is an endogenous anticonvulsant agent present in the brain that regulate hyper-excitation of the neurons by regulating directly on its receptors or indirectly by modulating the release of other neurotransmitters (Boison 2007). Adenosine acts on its receptor subtypes i.e. A1, A2a, A2b and A3 adenosine receptors. It has been reported that antiepileptic effect of adenosine is through its action on inhibitory A1 receptors (Gouder et al. 2003, Fredholm et al. 2005. Adenosine A1 coupled to G i/o GPCR that causes the inhibition of the adenylcyclase and cAMP and reduces the in ux of voltage-gated sodium channels at the presynaptic terminal. This further leads to the inhibition of the glutamate release from the nerve terminals (Kosmowska et al. 2020). Similarly, the postsynaptic hyperpolarization by A1 receptors occurred by activation of G-protein coupled inwardly-rectifying potassium channels that lower the hyperexcitability of neurons at postsynaptic terminals (Hill et al. 2020, Spanoghe et al. 2021).
Furthermore, studies showed that the Kir3.2 channels are involved in the action of A1R that causes hyperpolarization of the neurons and alleviate the seizure severity. Studies showed that Kir3.2 channels knock-out mice are more susceptible to seizures (Signorini et al. 1997).
Since, the effects of levetiracetam in the periphery for anti-hyperalgesic activity were evaluated through adenosine A1 receptor. We analyzed the antiepileptic effects of levetiracetam through modulation of adenosinergic pathway in acute and kindling model of epilepsy.

Methodology Chemicals
Pentylenetetrazole,1,3-dipropyl-8-cyclopentylxanthine (DPCPX), caffeine and Diazepam were purchased from Sigma Aldrich (USA). Levetiracetam(obtained from Helix Pharmaceutical, Pakistan). All drugs were diluted in distilled water except DPCPX, which was dissolved in 5% DMSO. The test compounds were given intraperitoneally with dosing volume of about 10ml/kg of the body weight.

Animals
Male NMRI mice weighing between 18-22 g were used in this study. Animals were provided with food and water ad libitum. Experiments were performed in the day time and were conducted according to the institutional ethical guidelines and after getting approval from the Institutional Animal Care and Use Committee (IACUC) (license number 2018-006).
Acute PTZ-induced seizure model Animals (n = 8) were brought to the experimental room one hour prior to the start of the experiment. They were grouped as shown in table 1. Onset of rst myoclonic jerk (FMJ), generalized clonic seizures (GCS) and percent mortality were recorded after the administration of PTZ 110 mg/kg.
Animals (n = 8) were grouped as shown in the table 2. The seizure scoring was recorded over a period of 30 minutes after the administration of PTZ45mg/kg on each alternate day. Scores were noted according to the Racine's scale (Racine 1972) i.e. score 1 (mouth and facial twitching), score 2 (slight body jerks), score 3 (tonic seizures with tail erection), score 4 (tonic clonic seizure with posture on one side of the body) and score 5 (tonic clonic seizure with inability of the animal to maintain posture). Animals were considered kindled when 3 consecutive occasion of score 4/5 appeared in the animals of PTZ group. A total of 13 injections were required to kindle all the animals in PTZ-treated group. Twenty four hours after the completion of kindling, animals were anesthetized by administration of ketamine/xylazine cocktail. Cardiac perfusion was performed with 1x PBS. Animals were decapitated by guillotine apparatus. Brain was removed carefully; hippocampus and cortex were isolated and processed for the extraction of total RNA.

RNA Isolation
Isolation of RNA from hippocampus and cortex was done through trizol reagent. Approximately 30-40 mg of tissue was placed in 2 ml eppendorf tube, and 1 ml of trizol reagent was added. The collected tissue was homogenized by using homogenizer. For the isolation of RNA, 0.2ml of chloroform was added in each tube containing homogenized tissue in trizol. The content was incubated for 2-3 minutes at room temperature, and centrifuged for 12000g for duration of 15 minutes. After centrifugation, the upper aqueous layer containing RNA was isolated in the new RNase free tube, and 0.5 ml isopropanol was added for precipitation of RNA. Incubate for 10 minutes, and then centrifuged for 10 minutes at 12000 x g. A pellet of RNA will be settled down at the base of the tube. The supernatant was discarded carefully leaving the RNA pellet inside the tube. For the puri cation of RNA pellet, 1 ml of 75% ethanol was added to the tube and after brief vertex, centrifuged the tubes at 7500g for 5minutes. Remove the supernatant and air dry the RNA pellet inside the tube. Pellet was dissolved in 20-40µl nuclease free water.
Quanti cation of RNA The total RNA in each sample was quanti ed with Nanodrop. First the reading was adjusted to zero by nuclease free water. Purity of the isolated RNA was investigated through the value of A260:A280 in nanodrop. If the value of this ratio falls within the range of 2±0.25 then the samples were found to be free from contamination. Samples were stored at -20 0 C until used for the synthesis of cDNA. cDNA synthesis Following RNA isolation from brain tissue samples cDNA were synthesized from isolated RNA samples using cDNA synthesis kit (Invitrogen, cDNA Synthesis kit 1622). To an RNase free tube 1 µg of total RNA1, µl 10X reaction buffer with MgCl 2 , and 1µg DNase-1 were mixed and the nal volume was made up with nuclease free water to 10 µl. The mixture was incubated at 37 o C on water bath for a duration of 30 minutes. After incubation 1µl 50mM EDTA was added to the tube and incubated at 65 o C for 5 minutes.
To this mixture I µl of Oligo dT 18 was added. The volume of this mixture became 12 µl. To this mixture the following reagents were added in the indicated order, 4 µl of 5X reaction buffer, 1µl Ribolock RNase inhibitor, 2 µl dNTP mixture and 1 µl reverse transcriptase were added. The nal volume of the mixture in the tube became 20µl. Incubate for 60 minute at 42 o C followed by 70 o C for a duration of 5 minutes.

RT-qPCR reaction
RT-qPCR reactions were prepared in low pro le PCR tubes. For ampli cation of the cDNA SYBR TM Green PCR Master Mix (ThemoFisher TM , UK) was used according to the protocol of the manufacturer. 1µl cDNA, 0.5 µl forward and 0.5 µl reverse primer, 3 µl DFW and 5 µl SYBR green mixture were added the tube in triplicate. Reactions were placed in the RT-qPCR machine. The protocol was set at 95 o C (hotstart) for 10 minutes followed by 40 cycles of 95 o C for 15s(denaturation), 60 o C for 15s(annealing), and 72 o C for 30s(extension) and melting curve analysis between65C to 95C (With 5s increment). After the completion of the PCR, data was exported to excel le and analyzed in Origin 8.5 statistical software.

Expression of Gene
Quanti cation of genes were done by using ΔΔCt method. Initially, the ΔCt values of the samples were calculated by subtracting the Cq value of GAPDH of the respective samples from Cq values of the samples. ΔCt value of all other groups were subtracted from the ΔCt value of the control group to obtain ΔΔCt values.

Results
Caffeine and DPCPX attenuated the anti-seizure activity of levetiracetam in acute PTZ-induced seizure model In the acute PTZ-induced seizure model, it was observed that the administration of caffeine (100 mg/kg) or DPCPX (25 mg/kg) signi cantly reduces the anti-seizure activity of LEV (200mg/kg). The onset of rst myoclonic jerk (FMJ) in LEV was observed at 146.66 ± 6.76 s which was signi cantly (P < 0.001) delayed response compared to the PTZ control group in which the FMJ occurred at 48.33± 3.75s (Fig. 1). It was interesting to nd that administration of caffeine (100mg/kg) or DPCPX (25mg/kg) prior to the LEV (200mg/kg) treatment, reverses the anti-seizure effects of LEV by reducing the delay of the FMJ to 76±14.52s and 100±16.32s respectively, as compared to the LEV 200mg/kg (146.66±6.76s, P < 0.05, Fig.1).
Furthermore, LEV (200mg/kg) followed by PTZ 100mg/kg did not produced any mortality and confers 100% protection as compared to the PTZ control group. It was also observed that caffeine (100mg/kg) and DPCPX (25 mg/kg) completely abolished the percent protection caused by LEV as shown in table-4.
Similarly, table-4also shows that administration of PTZ leads to the onset of generalized clonic seizure (GCS) at 66.66 ± 6.009s. On the other hand, LEV prevented the onset of GCS remarkably in mice administered with acute convulsive dose of PTZ. Caffeine (100mg/kg) and DPCPX (25mg/kg) reversed the effects of LEV by producing the GCS in115 ± 16.52s and 142.33 ± 12.170s respectively, as shown in table-4.

Effects of levetiracetam in PTZ-induced kindling model
It was observed that the repeated administration of subconvulsive dose of PTZ (45 mg/kg, i.p.) on alternate day gradually increased the seizure scores to 4-5 (clonic-tonic seizures) in mice (Fig. 2). Diazepam 7.5mg/kg did not produced seizures in any animal of the group. Furthermore, LEV (200mg/kg, i.p.) signi cantly (P<0.001, Fig. 2) reduced the seizure score in PTZ-induced kindled animals. Interestingly, none of the animals achieved a score above 3, thus showed a signi cant antiepileptogenic activity.

Effect of levetiracetam on the gene expression of adenosine A1R in kindling model
Adenosine A1R has been one of the initial pathways involved in the regulation of the neuronal hyperexcitation. It was found that LEV 200mg/kg increased the fold change mRNA expression in adenosine A1R in the RT-qPCR by 1.76±0.144 and 1.45±0.124 times in the hippocampus and cortex respectively, as compared to the PTZ-treated group (n = 8, Fig. 1, P < 0.05) in kindling model of epilepsy. Furthermore, it was also observed that LEV 200mg/kg only group decreased the fold change of A1R in mice hippocampal and cortical region. Results also revealed insigni cant upregulation of A1R expression in the hippocampus and cortex in diazepam-treated group in mice (Fig 1, P<0.05). Data obtained in the RT-qPCR inferred that levetiracetam showed its antiepileptic effects in PTZ-Induced kindling model by acting on the adenosine A1 receptor by up-regulating the A1 receptor mRNA expression in the hippocampal and cortical regions.
Gene expression of Kir3.2 in the brain of PTZ-kindled mice receiving treatment of levetiracetam Adenosine A1 receptor mediated potassium inwardly-rectifying channels (Kir3.2) play a key role in regulation of neuronal over-excitation. Figure 4 shown that LEV (200 mg/kg, i.p.) prevented the PTZinduced seizures in kindling model by increasing the mRNA expression of the Kir3.2in hippocampus (fold change: 2.089 ± 0.214) and cortex (fold change: 1.57 ± 0.058), as compared to the PTZ-treated group (n = 8, Fig.4, **P < 0.01). Furthermore, diazepam 7.5mg/kg has shown little effect on the mRNA expression of the Kir3. When compared with PTZ control group, diazepam treatment was found to decrease the ENT1 expression however this reduction was non-signi cant both in cortex and hippocampus.

Discussion
Conventional AEDs have various adverse effects and therefore efforts are made to develop novel drugs with less side effects and broader range of effectiveness in epileptic patients. Among the novel AEDs, levetiracetam has been considered as the safest drug in term of adverse effects pro le (Briggs and French 2004). It is also known that levetiracetam partially protect seizures in animals in acute seizure models and demonstrated a signi cant suppression of seizures in kindling model of epilepsy (Klitgaard 2001). Previously, studies showed that levetiracetam acts peripherally on A1 receptor to produce its antihyperalgesic effects and this effects was antagonized by the adenosine receptor antagonists caffeine (non-selective A1 and A2 receptor antagonist) and DPCPX (selective A1R antagonist) (Micov, Tomić et al. 2010). Effects of levetiracetam on adenosinergic pathway in the CNS were not investigated completely. Therefore, in this study we evaluated the antiepileptic action of levetiracetam through adenosinergic pathway speci cally through its action on A1 receptors.
Adenosine A1R suppressed the hyperexcitability of the neuronal network through two mechanisms i.e. by inhibiting the release of the neurotransmitters from the presynaptic terminals and by hyperpolarizing the postsynaptic terminal. The hyperpolarizing effects of A1R was because of the activation of GPCR-linked potassium inwardly rectifying channels (Kir3.2). These channels are also activated by other receptors such as GABA-B receptors. It was observed that the effects of GABA mediated inhibition through Kir3.2 are not that pronounced than the ones due to the activation of A1R. Furthermore, for the optimal effects of Kir3.2 channels presence of adenosine were required (Hill, Hickman et al. 2020).
The current study was divided into two main phases' i.e. acute phase and chronic kindling phase. In acute study, it was found that antiepileptic activity of levetiracetam was antagonized by the caffeine and DPCPX which inferred that levetiracetam may partially acts on the adenosine receptor pathway to alleviate seizures. To further con rm the results of acute study, the changes in the A1R and its associated GIRK2 channels and ENT1 mRNA expression were investigated in the kindling model of epileptogenesis.
It was found that levetiracetam increased the expression of A1R and its associated Kir3.2 mRNA expression in kindling and parallel decrease in the mRNA expression of ENT1 in the hippocampus and cortex.
Adenosine A1 receptor density in TLE in human subjects were reported to be increased (Angelatou, Pagonopoulou et al. 1993) or decreased (Glass, Faull et al. 1996) in temporal neocortical region of the brain. However, it has been reported in animal models that A1 receptor density decreased in hippocampal pyramidal neurons in rats (Ochiishi et al. 1999, Ekonomou et al. 2000, Rebola et al. 2003. It has been found that A1R knockout mice leads to the development of status epilepticus after the cerebral injury (Kochanek, Vagni et al. 2006).Here we have found that adenosine A1 receptor density decreased in cortical and hippocampal region of the brain isolated from the control PTZ group. However, treatment of levetiracetam caused a gradual increase in the gene expression of adenosine A1 receptor in the brain of the mice that results in the reduction of the seizure due to their inhibitory tone on the neurons caused by activation of adenosine A1 receptors. Previously it has been shown that knock-out of Kir3.2 channels increased seizure susceptibility in animals (Signorini et al. 1997).It was known that GABA-B receptormediated hyperpolarization may also be due to the activation of Kir3.2 channels. There are other studies giving evidence that levetiracetam does not acts on the GABA receptors (De Smedt et al. 2007) thus indicating that the activity of levetiracetam mediated action on the Kir3.2 is due activation of A1 receptor in the brain.
It has also been revealed that the expression of ENT1 increased in patients suffering from epilepsy and also in animal models of epilepsy and pharmacological inhibition of ENT1 transporters in epileptic animal models decreased the severity of seizures (Ho et al. 2020). Data of the present study revealed that the increased expression of the ENT1 mRNA seen in the kindling model of epilepsy is ameliorated by levetiracetam. It may be hypothesized that increase in ENT1 expression could be due to the seizure induced astragliosis and levetiracetam decreased the gene expression of ENT1 by preventing astragliosis in animal model (Devinsky et al. 2013). Therefore, it can be suggested that levetiracetam prevent the hallmark feature of epilepsy i.e. astragliosis.

Conclusion
Based on the present experimental ndings, it can be concluded that levetiracetam increases the adenosine A1 receptor and Kir3.2 channel mRNA expression both in hippocampus and cortical regions of the brain and therefore it reduces the seizure scores. Adenosine A1 receptor activation has been known for preventing the hyperpolarization of neurons through Kir3.2 channels located at postsynaptic terminals while at presynaptic terminal adenosine A1 receptor decreased the release of glutamatergic     Seizure scores in PTZ-induced kindling model of epilepsy. Levetiracetam signi cantly reduced the mean seizure score in the PTZ-induced kindling. Data is represented as Mean ± SEM and analyzed through Oneway ANOVA using Origin statistical software. ***P< 0.001 is considered signi cant vs. PTZ-treated group.