Studies on Anticonvulsant Effects of Novel Histamine H3R Antagonists in Electrically and Chemically Induced Seizures in Rats

A newly developed series of non-imidazole histamine H3 receptor (H3R) antagonists (1–16) was evaluated in vivo for anticonvulsant effects in three different seizure models in Wistar rats. Among the novel H3R antagonists examined, H3R antagonist 4 shortened the duration of tonic hind limb extension (THLE) in a dose-dependent fashion in the maximal electroshock (MES)-induced seizure and offered full protection against pentylenetetrazole (PTZ)-induced generalized tonic-clonic seizure (GTCS), following acute systemic administration (2.5, 5, 10, and 15 mg/kg, i.p.). However, only H3R antagonist 13, without appreciable protective effects in MES- and PTZ-induced seizure, fully protected animals in the strychnine (STR)-induced GTCS following acute systemic pretreatment (10 mg/kg, i.p.). Moreover, the protective effect observed with H3R antagonist 4 in MES-induced seizure was completely abolished when animals were co-administered with the H3R agonist (R)-α-methylhistamine (RAMH, 10 mg/kg, i.p.). However, RAMH failed to abolish the full protection provided by the H3R antagonist 4 in PTZ-induced seizure and H3R antagonist 13 in STR-induced seizure. Furthermore, in vitro antiproliferative effects or possible metabolic interactions could not be observed for compound 4. Additionally, the predictive in silico, as well as in vitro, metabolic stability for the most promising H3R antagonist 4 was assessed. The obtained results show prospective effects of non-imidazole H3R antagonists as innovative antiepileptic drugs (AEDs) for potential single use against epilepsy.


Introduction
Epilepsy is considered as a brain disease characterized by repeated seizures due to different complex causes including idiopathic (ion channel or protein defects), symptomatic stroke, head injury Radioligand binding was performed as previously described using membranes from CHO-K1 cells stably transfected with the human H1 receptor (PerkinElmer., Waltham, MA, USA) [30,42]. Data were fitted to a one-site curve-fitting equation with Prism 6 (GraphPad Software, city, state, country) and Ki values were estimated from the Cheng-Prusoff equation [44].

Human Histamine H4 Receptor (hH4R) Binding Affinity for Selected Compounds 4, 7 and 13
H4R radioligand displacement assays were performed with Sf9 cell membrane preparations, expressing the hH4R, as described previously [28,42,43]. Assays ran in triplicates with four appropriate concentrations of the test compound. Data were analyzed by GraphPad Prism 7, using the "one-site competition" equation. 36 Radioligand binding was performed as previously described using membranes from CHO-K1 cells stably transfected with the human H1 receptor (PerkinElmer., Waltham, MA, USA) [30,42]. Data were fitted to a one-site curve-fitting equation with Prism 6 (GraphPad Software, city, state, country) and Ki values were estimated from the Cheng-Prusoff equation [44].

Human Histamine H4 Receptor (hH4R) Binding Affinity for Selected Compounds 4, 7 and 13
H4R radioligand displacement assays were performed with Sf9 cell membrane preparations, expressing the hH4R, as described previously [28,42,43]. Assays ran in triplicates with four appropriate concentrations of the test compound. Data were analyzed by GraphPad Prism 7, using the "one-site competition" equation. 40 Radioligand binding was performed as previously described using membranes from CHO-K1 cells stably transfected with the human H1 receptor (PerkinElmer., Waltham, MA, USA) [30,42]. Data were fitted to a one-site curve-fitting equation with Prism 6 (GraphPad Software, city, state, country) and Ki values were estimated from the Cheng-Prusoff equation [44].

Human Histamine H4 Receptor (hH4R) Binding Affinity for Selected Compounds 4, 7 and 13
H4R radioligand displacement assays were performed with Sf9 cell membrane preparations, expressing the hH4R, as described previously [28,42,43]. Assays ran in triplicates with four appropriate concentrations of the test compound. Data were analyzed by GraphPad Prism 7, using the "one-site competition" equation. 38 Radioligand binding was performed as previously described using membranes from CHO-K1 cells stably transfected with the human H1 receptor (PerkinElmer., Waltham, MA, USA) [30,42]. Data were fitted to a one-site curve-fitting equation with Prism 6 (GraphPad Software, city, state, country) and Ki values were estimated from the Cheng-Prusoff equation [44].

Human Histamine H4 Receptor (hH4R) Binding Affinity for Selected Compounds 4, 7 and 13
H4R radioligand displacement assays were performed with Sf9 cell membrane preparations, expressing the hH4R, as described previously [28,42,43]. Assays ran in triplicates with four appropriate concentrations of the test compound. Data were analyzed by GraphPad Prism 7, using the "one-site competition" equation. Radioligand binding was performed as previously described using membranes from CHO-K1 cells stably transfected with the human H1 receptor (PerkinElmer., Waltham, MA, USA) [30,42]. Data were fitted to a one-site curve-fitting equation with Prism 6 (GraphPad Software, city, state, country) and K i values were estimated from the Cheng-Prusoff equation [44].

Human Histamine H4 Receptor (hH4R) Binding Affinity for Selected Compounds 4, 7 and 13
H4R radioligand displacement assays were performed with Sf9 cell membrane preparations, expressing the hH4R, as described previously [28,42,43]. Assays ran in triplicates with four appropriate concentrations of the test compound. Data were analyzed by GraphPad Prism 7, using the "one-site competition" equation.

Animals
Inbred male Wistar rats aged between six and eight weeks (body weight: 180-220 g, Central Animal Facility of the United Arab Emirates (UAE) University, Al Ain/Abu Dhabi, United Arab Emirates) were used. Animals were kept in an air-conditioned animal facility room with controlled temperature (24 ± 1 • C) and humidity (55 ± 15%) under a 12-h light/dark cycle. The animals were allowed free access to food and water. The experiments of the current study were carried out between 09:00 and 12:00, and all procedures were performed according to the guidelines of the European Communities Council Directive of 24 November 1986 (86/609/European Economic Community (EEC)) and were previously approved for epilepsy study by the College of Medicine and Health Sciences/United Arab Emirates University (Institutional Animal Ethics Committee, approval number; ERA_2017_5676).

Maximal Electroshock (MES)-Induced Seizure
As previously described [30,46], the seizures were induced in rats with a 50 Hz alternating current of 120 mA intensity. The current was applied through ear electrodes for 1 s. Protection against the spread of MES-induced seizure was defined as the abolition of the tonic hind limb extension (THLE) component of the seizure [25,29,47]. For the first screening of protective effects of test compounds 1-16 in the MES-induced seizure model, the animals were divided into eighteen groups of seven rats (n = 7) as follows; group (1) control group injected with vehicle (SAL), group (2) positive control group in which rats were injected with PHT at a dose of 10 mg/kg (this being the minimal dose of PHT that protected animals against the spread of MES-induced seizures without mortality), and groups (3)-(18) animals in the experimental groups were administered H3R antagonists 1-16 at a dose of 10 mg/kg. In another MES experiment in three separate groups of six rats (n = 6), H3R antagonist 4 with the most promising protection in MES model was administered at doses of 2.5, 5, or 15 mg/kg, i.p. 30-45 min before the MES challenge. In an additional abrogative experiment, the most effective dose of H3R antagonist 4 was designated for further analysis in which a separate group of six rats (n = 6) was co-injected with the selected dose of H3R antagonist 4 (30-45 min prior to MES test) with RAMH (10 mg/kg, i.p., 15-20 min before MES challenge) ( Figure 1, Table 2). antagonist 4 with the most promising protection in MES model was administered at doses of 2.5, 5, or 15 mg/kg, i.p. 30-45 min before the MES challenge. In an additional abrogative experiment, the most effective dose of H3R antagonist 4 was designated for further analysis in which a separate group of six rats (n = 6) was co-injected with the selected dose of H3R antagonist 4 (30-45 min prior to MES test) with RAMH (10 mg/kg, i.p., 15-20 min before MES challenge) ( Figure 1, Table 2).    In an additional abrogative experiment, the most effective dose of H3R antagonist 4 was designated for further analysis in which a separate group of six rats (n = 6) was co-injected with the selected dose of H3R antagonist 4 (30-45 min prior to MES test) with RAMH (10 mg/kg, i.p., 15-20 min before MES challenge) ( Figure 1, Table 2).   In an additional abrogative experiment, the most effective dose of H3R antagonist 4 was designated for further analysis in which a separate group of six rats (n = 6) was co-injected with the selected dose of H3R antagonist 4 (30-45 min prior to MES test) with RAMH (10 mg/kg, i.p., 15-20 min before MES challenge) ( Figure 1, Table 2).   In an additional abrogative experiment, the most effective dose of H3R antagonist 4 was designated for further analysis in which a separate group of six rats (n = 6) was co-injected with the selected dose of H3R antagonist 4 (30-45 min prior to MES test) with RAMH (10 mg/kg, i.p., 15-20 min before MES challenge) ( Figure 1, Table 2).   In an additional abrogative experiment, the most effective dose of H3R antagonist 4 was designated for further analysis in which a separate group of six rats (n = 6) was co-injected with the selected dose of H3R antagonist 4 (30-45 min prior to MES test) with RAMH (10 mg/kg, i.p., 15-20 min before MES challenge) ( Figure 1, Table 2).   In an additional abrogative experiment, the most effective dose of H3R antagonist 4 was designated for further analysis in which a separate group of six rats (n = 6) was co-injected with the selected dose of H3R antagonist 4 (30-45 min prior to MES test) with RAMH (10 mg/kg, i.p., 15-20 min before MES challenge) ( Figure 1, Table 2).

Chemically-Induced Seizures
In this study and according to previously used protocols, two chemical agents have been used to induce seizures, namely pentylenetetrazole (PTZ, 60 mg/kg, i.p.) and strychnine (STR, 3.5 mg/kg, i.p.) [27,47]. PTZ 60 mg/kg or STR 3.5 mg/kg were administered i.p. to all groups of seven rats (n = 7), that is, animals pretreated with vehicle or test compounds 1-16. Vehicle, VPA (300 mg/kg, i.p.), [26,[47][48][49] or test compounds 1-16 (10 mg/kg, i.p.) were systemically administered 30-45 min before PTZ (60 mg/kg, i.p.) or STR (3.5 mg/kg, i.p.) injection, and animals were directly observed for any signs of convulsion for a duration of 30 min. The observed scores as well as percent protection against generalized tonic-clonic seizure (GTCS) were graded and scaled according to the Racine scale (stage 0, no change in behavior; stage 1, stereotype mouthing, eye blinking, and/or mild facial clonus; stage 2, head nodding and/or severe facial clonus; stage 3, myoclonic jerk in forelimbs; stage 4, clonic convulsions in the forelimbs with rearing; stage 5, generalized colonic convulsions associated with loss of balance [50]) to evaluate the protection provided by the respective test compound against seizures. The animals were divided into eighteen groups of seven rats (n = 7) and treated as follows.
(1) Control group injected with vehicle + PTZ (SAL group), (2) positive control group in which rats were injected with VPA (300 mg/kg) + PTZ (VPA group), and animals in the test groups (3)- (18) were systemically administered with test compounds 1-16 (10 mg/kg), respectively. An additional PTZ experiment was conducted using three different groups of seven rats (n = 7), the H3R ligand with the most promising protection in PTZ model was administered at doses of 2.5, 5, or 15 mg/kg, i.p. 30-45 min before the PTZ challenge. In an additional abrogative PTZ experiment, the most effective dose of H3R antagonist 4 was selected for further analysis, namely in a separate group of seven rats (n = 7), the selected dose of 4 was co-injected (30-45 min prior to PTZ test) together with RAMH (10 mg/kg, i.p., 15-20 min before PTZ challenge) (Figure 2, Table 2). The same experimental procedure was followed in the STR-induced seizure model applying the reference drug VPA, which was used at a dose of 300 mg/kg, i.p. (Figure 3, Table 2). 0, no change in behavior; stage 1, stereotype mouthing, eye blinking, and/or mild facial clonus; stage 2, head nodding and/or severe facial clonus; stage 3, myoclonic jerk in forelimbs; stage 4, clonic convulsions in the forelimbs with rearing; stage 5, generalized colonic convulsions associated with loss of balance [50]) to evaluate the protection provided by the respective test compound against seizures. The animals were divided into eighteen groups of seven rats (n = 7) and treated as follows.
(1) Control group injected with vehicle + PTZ (SAL group), (2) positive control group in which rats were injected with VPA (300 mg/kg) + PTZ (VPA group), and animals in the test groups (3)- (18) were systemically administered with test compounds 1-16 (10 mg/kg), respectively. An additional PTZ experiment was conducted using three different groups of seven rats (n = 7), the H3R ligand with the most promising protection in PTZ model was administered at doses of 2.5, 5, or 15 mg/kg, i.p. 30-45 min before the PTZ challenge. In an additional abrogative PTZ experiment, the most effective dose of H3R antagonist 4 was selected for further analysis, namely in a separate group of seven rats (n = 7), the selected dose of 4 was co-injected (30-45 min prior to PTZ test) together with RAMH (10 mg/kg, i.p., 15-20 min before PTZ challenge) (Figure 2, Table 2). The same experimental procedure was followed in the STR-induced seizure model applying the reference drug VPA, which was used at a dose of 300 mg/kg, i.p. (Figure 3, Table 2).

Statistical Analysis
For statistical comparisons, the software package SPSS 25.0 (IBM Middle East, Dubai, UAE) was used. All data are expressed as the means ± standard error of mean (SEM). K i values at the H3R are given as means with the 95% confidence interval. Following normal distribution assessment, the anticonvulsant effects observed for H3R antagonists 1-16 in MES-, PTZ-, and STR-induced convulsion models were analyzed using one-way analysis of variance, followed by the Bonferroni post hoc test for multiple comparisons. The criterion for statistical significance was set at p < 0.05.

Statistical Analysis
For statistical comparisons, the software package SPSS 25.0 (IBM Middle East, Dubai, UAE) was used. All data are expressed as the means ± standard error of mean (SEM). Ki values at the H3R are given as means with the 95% confidence interval. Following normal distribution assessment, the anticonvulsant effects observed for H3R antagonists 1-16 in MES-, PTZ-, and STR-induced convulsion models were analyzed using one-way analysis of variance, followed by the Bonferroni post hoc test for multiple comparisons. The criterion for statistical significance was set at p < 0.05.

Antiproliferative Activity
Human embryonic kidney (HEK)-293 cell line (ATCC CRL-1573) was kindly donated by Prof. Dr. Christa Müller (Pharmaceutical Institute, Pharmaceutical Chemistry I, University of Bonn). Hepatoma HepG2 (ATCC HB-8065) cell line was kindly donated by the Department of Pharmacological Screening, Jagiellonian University Medical College. The cell cultures' growth conditions in presence of H3R antagonist 4 were applied as described before [46,51]. The cells' viability was assessed after 72 h of incubation with H3R antagonist and the following references: doxorubicin (Sigma-Aldrich) and carbonyl cyanide 3-chlorophenylhydrazone (CCCP) (Sigma-Aldrich). CellTiter 96 ® AQueous non-radioactive cell proliferation assay (MTS) was purchased from Promega ® and added to each well in the volume of 20 µ L. The cells were then incubated for 2-5 h. The microplate reader EnSpire (PerkinElmer Ltd., Waltham, MA, USA) was used to measure the absorbance at 490 nm.

Antiproliferative Activity
Human embryonic kidney (HEK)-293 cell line (ATCC CRL-1573) was kindly donated by Prof. Dr. Christa Müller (Pharmaceutical Institute, Pharmaceutical Chemistry I, University of Bonn). Hepatoma HepG2 (ATCC HB-8065) cell line was kindly donated by the Department of Pharmacological Screening, Jagiellonian University Medical College. The cell cultures' growth conditions in presence of H3R antagonist 4 were applied as described before [46,51]. The cells' viability was assessed after 72 h of incubation with H3R antagonist and the following references: doxorubicin (Sigma-Aldrich) and carbonyl cyanide 3-chlorophenylhydrazone (CCCP) (Sigma-Aldrich). CellTiter 96 ® AQueous non-radioactive cell proliferation assay (MTS) was purchased from Promega ® and added to each well in the volume of 20 µL. The cells were then incubated for 2-5 h. The microplate reader EnSpire (PerkinElmer Ltd., Waltham, MA, USA) was used to measure the absorbance at 490 nm.

Prediction of In Silico Metabolism
The in silico prediction for sites of metabolism of H3R antagonist 4 was performed by MetaSite 5.1.1 provided by Molecular Discovery Ltd. (Borehamwood, Hertfordshire, UK). The most probable sites of metabolism were predicted during this study by the liver computational model [52] (Figure 5).

Metabolic Stability
Rat liver microsomes (RLMs) purchased form Sigma-Aldrich (St. Louis, MO, USA) were used for in vitro determination of compound 4's metabolic stability. The NADPH regeneration system was purchased from Promega (Madison, WI, USA). All experiments were performed as described before [26,45,46]. The reaction mixture was preincubated at 37 • C for 5 min, and then the reaction was started by adding 50 µL of NADPH regeneration system. The reaction was ended after 120 min by the addition of cold methanol (200 µL). The mixture was then centrifuged at 14,000 rpm for 15 min and the ultra-performance liquid chromatography-mass spectrometry (UPLC/MS) analysis of the supernatant was conducted. Ion fragment analyses were also performed for metabolic pathways' determination.

Metabolic Interactions
The metabolic interactions of H3R antagonist 4 were determined by the luminescent P450-Glo™ 3A4 and 2D6 assays purchased from Promega ® . All enzymatic reactions were performed according to the manufacturer protocols and as previously described [26]. The references ketoconazole (KE, CYP3A4 inhibitor) and quinidine (QD, CYP2D6 inhibitor) were obtained from Sigma-Aldrich. The luminescence was measured with a microplate reader EnSpire (PerkinElmer). The final concentration of H3R antagonist 4 was in the range of 0.1-25 µM, whereas reference inhibitory effects ranged from 0.01 µM to 10 µM. The IC 50 values' calculations were performed by GraphPad Prism™ software (version 5.01, San Diego, CA, USA).

In Vivo Seizure Models
With regard to the MES test, the current one used in the study produced seizures in 100% of animals without mortality. Likewise, the dose of PTZ and STR used in the present study formed seizures (score 4-5) in 100% of animals without mortality.

Anticonvulsant Screening of H3R Antagonists 1-16 in MES-Induced Seizure
The preliminary screening for anticonvulsant activities of acute systemic pretreatment with H3R ligands 1-16 on MES-induced seizures in rats was carried out and the observed results were compared with the protective effect of the reference antiepileptic drug PHT in MES-induced seizure in rats (Tables 2 and 3 Figure 1). Moreover, the protection observed with H3R antagonist 4 at a dose of 10 mg/kg, i.p. was significantly higher than that found for  H3R antagonists 3, 5, 6, 7 (Table S1, Supplementary Material). On the other hand, the results show that animals pretreated with 2.5, 5, and 15 mg/kg of H3R antagonist 4 were protected to a significantly lesser extent against seizures when compared with the H3R antagonist 4 (10 mg)-treated group with [F (1,11) Tables 2 and 3). However, pretreatment with 2.5 or 5 mg/kg of H3R antagonist 4 provided significantly lower protection when compared with that provided with 10 or 15 mg/kg of the same compound (all p < 0.05) (Figure 2, Tables 2 and 3). Furthermore, Figure 2 shows the reversal of H3R antagonist 4-provided protection when co-injected with 10 mg/kg of histamine H3R agonist RAMH. The observed results showed that RAMH failed to reverse the H3R antagonist 4-provided protection in PTZ-induced seizure model with [F (1,12) = 2.40; p = 0.147] for 6-treated group versus 4 + RAMH-treated group ( Figure 2). Importantly, RAMH alone did not affect seizure score when compared with effects of the saline-treated group with p = 0.232 ( Figure 2, Tables 2 and 3 ). Table 3. Effects of H3R antagonists 4 and 13 on THLE duration, average score, and percentage of provided protection against generalized tonic-clonic seizure (GTCS) in MES-, PTZ-, and STR-induced convulsion models.  In STR-induced seizure in rats, H3R antagonist 4 with the most promising effect in MES-and PTZ-induced seizures failed to provide any appreciable protection in STR-induced seizure when compared with the saline-treated group after 30 min of observation time with [F (1,12) = 4.50; p = 0.055] (Tables 2 and 3). However, H3R antagonist 13 without protection in MES-induced seizure and with weak protection in PTZ-induced seizure exhibited a moderate protective effect in STR-induced seizure when compared with the saline-treated group with [F (1,12) = 63.00; p < 0.001] (Figure 3, Table 3). Notably, the reference antiepileptic drug VPA (VPA 300 mg/kg, i.p.) showed significant protection when compared with saline-treated group after 30 min observation time with [F (1,12) = 236.308; p < 0.001] (Figure 3, Tables 2 and 3). Moreover, the results showed that acute systemic pretreatment with H3R antagonist 13 at a lower dose (2.5 mg/kg, i.p.) failed to exhibit protection against STR-induced seizures when compared with the saline-treated group with [F (1,12) = 1.00; p = 0.337] (Figure 3, Tables 2 and 3). Furthermore, no significant differences in the protection provided by H3R antagonist 13 (10 mg/kg, i.p.) were observed when 5 or 15 mg/kg of the same compound was administered with [F (1,12) Table 3).

Antiproliferative Assay
The effect of H3R antagonist 4 (0.1-100 µM) on proliferation of HEK-293 and hepatoma HepG2 cell lines was assessed and compared with the reference cytostatic drug doxorubicin (DX) and hepatotoxin carbonyl cyanide 3-chlorophenylhydrazone (CCCP) (Figure 4A,B). The results showed that H3R antagonist 4 significantly decreased proliferation of HepG2 cells only at the highest concentration used (100 µM), however, decline in proliferation of HEK-293 cells was achieved with two doses of the same compound, namely 10 and 100 µM (p < 0.001) ( Figure 4A,B).

In Silico Metabolic Stability
The computational procedure MetaSite.5.1.1 provided by Molecular Discovery Ltd. indicated the sixth position of naphthalene moiety as the most probable site of H3R antagonist 4 metabolism (blue circle marked; Figure 5). Moreover, the azepane moiety was also shown to be susceptible for metabolic biotransformations (the darker red color of the marked functional group indicates its higher probability of being involved in the metabolism pathway; Figure 5). The predicted in silico most probable metabolic routes included hydroxylation at the sixth position of naphthalene or azepane group followed by the compound's oxidative degradation of the aliphatic alkyl chain (data not shown).

In Silico Metabolic Stability
The computational procedure MetaSite.5.1.1 provided by Molecular Discovery Ltd. indicated the sixth position of naphthalene moiety as the most probable site of H3R antagonist 4 metabolism (blue circle marked; Figure 5). Moreover, the azepane moiety was also shown to be susceptible for metabolic biotransformations (the darker red color of the marked functional group indicates its higher probability of being involved in the metabolism pathway; Figure 5). The predicted in silico most probable metabolic routes included hydroxylation at the sixth position of naphthalene or azepane group followed by the compound's oxidative degradation of the aliphatic alkyl chain (data not shown).

In Vitro Metabolic Stability
The UPLC analysis of the reaction mixture of H3R antagonist 4 incubated for 120 min with rat liver microsomes (RLMs) showed that ~18% of this compound was converted into four metabolites, namely M-I-M-IV ( Figure 6A). The MS determination of the molecular mass of the main metabolite M-I (+32 units) and the comparison of the ion fragments analyses of H3R antagonist 4 and metabolite M-I suggests the double-hydroxylation at the 6-n-hexylazepane moiety as the main metabolic pathway ( Figure 6B). Moreover, the hydroxylation reactions at the naphthalene moiety and at the aliphatic alkyl chain were also identified (metabolites M-II-M-IV, Figure 6C).

In Vitro Metabolic Stability
The UPLC analysis of the reaction mixture of H3R antagonist 4 incubated for 120 min with rat liver microsomes (RLMs) showed that~18% of this compound was converted into four metabolites, namely M-I-M-IV ( Figure 6A). The MS determination of the molecular mass of the main metabolite M-I (+32 units) and the comparison of the ion fragments analyses of H3R antagonist 4 and metabolite M-I suggests the double-hydroxylation at the 6-n-hexylazepane moiety as the main metabolic pathway ( Figure 6B). Moreover, the hydroxylation reactions at the naphthalene moiety and at the aliphatic alkyl chain were also identified (metabolites M-II-M-IV, Figure 6C).

Metabolic Interactions
The results observed in luminescent CYP3A4 P450-Glo™ assay showed no inhibitory effects of H3R antagonist 4 on CYP3A4 activity at examined concentrations of 0.1-25 µM ( Figure 7A). Moreover, the observed results in luminescent CYP2D6 P450-Glo™ assay revealed that the inhibitory effect of H3R antagonist 6 was detected with an IC 50 value of 0.67 µM ( Figure 7B).

Metabolic Interactions
The results observed in luminescent CYP3A4 P450-Glo™ assay showed no inhibitory effects of H3R antagonist 4 on CYP3A4 activity at examined concentrations of 0.1-25 µM ( Figure 7A). Moreover, the observed results in luminescent CYP2D6 P450-Glo™ assay revealed that the inhibitory effect of H3R antagonist 6 was detected with an IC50 value of 0.67 µM ( Figure 7B).

Histamine H1 Receptor Affinity
As the two histamine receptors H1R and H3R play a pivotal role in the anticonvulsant activity of the central histamine [41], selected H3R antagonists (6, 9, and 15) were examined in a binding assay at the human histamine H1R and showed weak affinities (915 ≤ Ki ≤ 1338 nM), being 10-fold lower than at the H3R.

Histamine H4 Receptor Affinity
As H3R represents the highest degree of homology with H4R [19], the potential interaction with this receptor subtype for selected H3R antagonists (4, 7, and 13) was evaluated. The observed results showed that none of the tested H3R antagonists displayed affinity for hH4R (Ki > 10,000 nM).

In Vivo Anticonvulsant Activity
The results demonstrated that H3R antagonist 4 exhibited the most promising protection against MES-induced seizures when animals were administered with 10 mg/kg i.p., and as compared with the saline-treated group of animals (Tables 2 and 3). However, lower doses (2.5 and 5 mg/kg, i.p.) and a higher dose (15 mg/kg, i.p.) failed to increase the (10 mg) H3R antagonist 4-provided protection ( Figure 1, Table 3). Accordingly, the observed results show a dose-response relationship of the protection provided and the presence of a ceiling effect for H3R antagonist 4 in the MES-induced seizure model achieved with a dose of 10 mg/kg, i.p. (Figure 1, Table 3). Notably, the protective effect

In Vitro Histamine H3 Receptor Affinity of Test Compounds 1-16
All novel H3R antagonists were evaluated in the form of hydrogen oxalate salts in radioligand displacement assays. Tested H3R antagonists 1-16 displayed affinities at the human H3R in a nanomolar concentration range (K i : 36-137 nM, Table 1). Compounds 13-15 displayed K i values similar to that of pitolisant (12 nM, Table 1), while the 4-fluorophenyl derivatives were the least potent ones (11-13; K i : 83.6-137.2 nM).

Histamine H1 Receptor Affinity
As the two histamine receptors H1R and H3R play a pivotal role in the anticonvulsant activity of the central histamine [41], selected H3R antagonists (6, 9, and 15) were examined in a binding assay at the human histamine H1R and showed weak affinities (915 ≤ K i ≤ 1338 nM), being 10-fold lower than at the H3R.

Histamine H4 Receptor Affinity
As H3R represents the highest degree of homology with H4R [19], the potential interaction with this receptor subtype for selected H3R antagonists (4, 7, and 13) was evaluated. The observed results showed that none of the tested H3R antagonists displayed affinity for hH4R (K i > 10,000 nM).

In Vivo Anticonvulsant Activity
The results demonstrated that H3R antagonist 4 exhibited the most promising protection against MES-induced seizures when animals were administered with 10 mg/kg i.p., and as compared with the saline-treated group of animals (Tables 2 and 3). However, lower doses (2.5 and 5 mg/kg, i.p.) and a higher dose (15 mg/kg, i.p.) failed to increase the (10 mg) H3R antagonist 4-provided protection ( Figure 1, Table 3). Accordingly, the observed results show a dose-response relationship of the protection provided and the presence of a ceiling effect for H3R antagonist 4 in the MES-induced seizure model achieved with a dose of 10 mg/kg, i.p. (Figure 1, Table 3). Notably, the protective effect of H3R antagonist 4 (10 mg/kg, i.p.) was similar to that observed in the group treated with the reference antiepileptic drug PHT (10 mg/kg, i.p.), and was significantly higher than that observed for H3R ligands 3, 5, 6, 7, 8, and 14 ( Figure 1, Table 1). The latter observations are in agreement with recent preclinical outcomes that showed a dose-dependent anticonvulsant effect of several H3R antagonists tested in several animal models of seizures. Also, the present results are in line with earlier observations for the H3R antagonist PIT in a photosensitivity seizure model in adult patients, and in agreement with earlier observations for several H3R antagonists tested for their anticonvulsant potential in different animal models of seizures [25][26][27][28][29][30]33,40,54]. An additional test in the present study showed that the protection observed for H3R antagonist 4 was abolished when animals were co-administered with the CNS penetrant histamine H3R agonist RAMH (10 mg/kg i.p.) (Figure 1, Table 3), proposing that the provided protective effect of H3R antagonist 4 in the MES-induced seizure model involves, at least to some extent, H3R blockade provided by H3R antagonist 4. Notably, these observations are consistent with the earlier preclinical results for numerous imidazole-and non-imidazole-based H3R antagonists [14,[25][26][27][28][29][30]40,54].
H3Rs are auto-receptors presynaptically positioned on histaminergic neurons with an inhibitory effect on the biosynthesis and release of histamine [21]. Therefore, blocking H3Rs by selective H3R antagonists, such as H3R antagonist 4, would escalate neuronal release of brain histamine, providing the protection in the MES-induced seizure in rats. The latter proposed mechanism underlying the anticonvulsant effect of H3R antagonist 4 is also in line with previous preclinical observations in animal seizure models in which high doses of several centrally acting H1R antagonists used as anti-allergic drugs promoted the development of convulsions of tested animals, indicating the involvement of H1R antagonism, and consequently, brain histaminergic neurotransmission in the seizure promotion. Noticeably, similar protective effects of imidazole-based and non-imidazole-based H3R antagonists were earlier described to be abolished either by H3R agonists or by centrally acting H1R antagonists, suggesting an interaction of the H3R antagonism-released histamine with postsynaptically located H1Rs on neurons [14,[25][26][27][28][29][30]53,54].
In the PTZ-induced seizure mode, acute systemic administration of VPA (100 mg/kg, i.p.) as well as H3R antagonists 4, 7, and 11 (10 mg/kg, i.p.) showed full protection ( Figure 2, Table 3). Furthermore, the anticonvulsant effects observed for H3R antagonist 4 at lower doses (2.5 and 5 mg/kg, i.p.) or a higher dose (15 mg/kg, i.p.) showed a dose-dependent protection in the PTZ-induced seizure model ( Figure 2, Table 3). However, the protective effect of H3R antagonist 4 (10 mg/kg, i.p.) was not reversed when rats were pretreated with RAMH (10 mg/kg, i.p.) before PTZ challenge, suggesting that the protection observed for H3R antagonist 4 in PTZ-induced seizure is not facilitated through modulation of central histaminergic neurotransmission (Figure 2, Table 3). The latter observation may be explained with the differences in the triggers or the mechanisms and types of seizures each model represents (MES is considered as a model of generalized tonic-clonic seizures, whereas PTZ (60 mg/kg, i.p.) induces generalized myoclonic and/or tonic-clonic seizures) [11,[53][54][55]. The failure of RAMH to reverse the protections provided by H3R antagonist 4 might be explained with its capability to reduce the suppression of glutamatergic and GABAergic synaptic transmission through blockade of H3 heteroreceptor function in CNS, necessitating additional future investigations of whether H3R antagonist 4 modulated the release of inhibitory neurotransmitters, for example, GABA [56].
In the STR-induced seizure model, the results showed that acute systemic administration of H3R antagonists 4, 7, and 11 (10 mg/kg, i.p.) failed to exhibit appreciable protection during 30 min of the time observation, whereas H3R antagonist 13 with moderate protection in PTZ and without any considerable protection in MES demonstrated reasonable protection following 30 min during the time observation (Figure 3, Table 3). Moreover, the lower doses (2.5 and 5 mg/kg, i.p.) as well as the higher dose (15 mg/kg, i.p.) of H3R antagonist 13 failed to exhibit a dose-dependent protection against STR-induced seizure in rats ( Figure 3, Table 3). In resemblance to the observations in PTZ-induced seizure, an additional experiment showed that the H3R antagonist 13-provided moderate protection in STR-induced seizure was not reversed when animals were co-injected with 10 mg/kg i.p. of the CNS penetrant histamine H3R agonist RAMH 30-45 min before STR challenge ( Figure 3, Table 3). These findings in STR-induced seizure further comprehend the present results for the protective effects of the H3R antagonist 4 in PTZ-induced seizure model, as both models are considered as chemically-induced seizure models. The latter observations for H3R antagonists 1-16 in the STR-induced seizure model show that the moderate protection provided with H3R antagonist 13 in STR model is also not facilitated through central histaminergic neurotransmission. Notably, STR is an established competitive antagonist of the inhibitory amino acid glycine. Therefore, the inability of H3R antagonist 13 to afford a dose-dependent protection against STR-induced seizure model advocates little or no modulation effect of H3R antagonist 13 on the glycine receptors, because the mechanisms underlying STR-induced seizures are supposed to be attributed to its blocking activity on glycine receptors in the brain as well as in the spinal cord [56]. Notably, most marketed AEDs were not effective in all conducted convulsion models during preclinical drug development. Accordingly, carbamazepine, oxcarbazepine and PHT were found to be, and are still, highly effective in MES-induced model in rodents, however, they failed to protect against convulsions in rodents induced by PTZ, STR, or picrotoxin [57,58]. On the contrary, ethosuximide and tiagabine, which show high protection in chemically-induced convulsion models in rodents, lack protection in the MES-induced model when used at nontoxic doses [58]. Nonetheless, the diversity in preclinical activities detected for numerous AEDs was translated into the clinical utility of PHT, carbamazepine, and oxcarbazepine, but not ethosuximide or tiagabine, in patients diagnosed with generalized tonic-clonic convulsions.

ADME-Tox Properties
The most promising H3R antagonist 4 was selected to evaluate its ADME-Tox properties, which included the determination of the in vitro safety profile by applying eukaryotic cell lines, the in silico and in vitro determination of the metabolic stability and main metabolic routes, and the in vitro assessment of potential drug-drug interactions by bioluminescent enzymatic assays.
In order to determine the potential toxicity, the standard colorimetrical MTS test was used. This test allowed one to follow up the influence of H3R antagonist 4 on the human embryonic kidney (HEK-293) and hepatoma HepG2 cell lines' proliferation. The antiproliferative activity of H3R antagonist 4 differed between both used cell lines. HEK-293 cells were more susceptible for antiproliferative activity as revealed in the statistically significant decrease of HEK-293 cell line viability (p < 0.001), which was observed at concentrations of 10 and 100 µM, whereas in the HepG2 cell line, the antiproliferative effect was achieved only at a concentration of 100 µM ( Figure 4A,B). However, the obtained results for HEK-293 indicated a satisfying safety profile of H3R antagonist 4 in comparison with the reference DX, which significantly decreased HEK-293 viability at a much lower concentration (1 µM) ( Figure 4A). Moreover, no significant hepatotoxic effect was observed for H3R antagonist 4, as the reference drug CCCP decreased HEpG2 viability at 10 µM, whereas H3R antagonist 4 only did so at 100 µM ( Figure 4B).
In further in vitro metabolic stability studies, the results observed for H3R antagonist 4 applying RLMs revealed good metabolic stability of H3R antagonist 4, as only~18% of the substrate was metabolized and converted into four different metabolites after 120 min of incubation at 37 • C ( Figure 6). Moreover, the in silico data and the MS fragmentation analyses showed one hydroxylation on the azepane moiety and another hydroxylation on the hexyl chain, followed by oxidative degradation of the alkyl chain as the main metabolic routes of H3R antagonist 4 ( Figure 5).
Furthermore, H3R antagonist 4 did not affect CYP3A4, whereas CYP2D6 was inhibited in a more pronounced manner. However, this effect was observed with a 67-fold higher IC 50 (IC 50 = 0.67 µM) as compared with the reference drug QD with a calculated IC 50 value of 0.01 µM (Figure 7).

Conclusions
The tested series of H3R antagonists showed in vitro affinity at the hH3R in the nanomolar range. The most promising H3R antagonist 4 having the 1-(6-(naphthalen-2-yloxy)hexyl)azepane pharmacophore exhibited an affinity for hH3R (K i = 69.3 nM). The in vivo anticonvulsant results revealed that H3R antagonist 4 exhibited most promising protection following acute systemic administration in MES-and PTZ-induced seizure models in rats. Moreover, the protection observed for H3R antagonist 4 in the MES-induced seizure model was fully reversed when rats were pretreated with the CNS-penetrant H3R agonist RAMH. However, RAMH failed to abrogate the protective effects observed for H3R antagonist in PTZ-or STR-induced seizure models, indicating that histaminergic pathways appear to be involved in the provided anticonvulsant efficacy of H3R antagonist 4 in only the MES-induced seizure model, but additional pharmacological properties of the compounds or their metabolites cannot be fully excluded. Moreover, ADME-Tox parameters' screening revealed satisfying low cytotoxicity, good metabolic stability, as well as no inhibition of CYP3A4 activity and moderate inhibition of CYP2D6 activity. Therefore, the overall experimental observations provide promising potential for the novel H3R antagonist 4 to be used as a potential template for further drug design and synthesis in the search for potent and active in vivo H3R antagonists, for example, as AED drugs with a high safety profile. Nonetheless, a battery of additional seizure models with different species is still required to further corroborate the current results observed for H3R antagonist 4, and to strengthen the translational value of its potential applicability in the therapeutic management of epilepsy.

Conflicts of Interest:
The authors declare no conflict of interest.