Thiazole-Bearing 4-Thiazolidinones as New Anticonvulsant Agents

: Here, we describe the synthesis and anticonvulsant activity of thiazole-bearing hybrids based on 2-imino-4-thiazolidinone and 2,4-dioxothiazolidine-5-carboxylic acid cores. The structure of target compounds was based on the following: ( i ) A combination of two thiazole cores; ( ii ) similarity to ralitolin’s structure; ( iii ) the compliance with structural requirements for the new anticonvulsants. Target compounds were synthesized via known approaches based on Knoenavegel reaction, alkylation reaction, and one-pot three-component reaction. Anticonvulsant properties of compounds were evaluated in two di ﬀ erent models—pentylenetetrazole-induced seizures and maximal electroshock seizure tests. Among the tested compounds 5 Z -(3-nitrobenzylidene)-2-(thiazol-2-ylimino)-thiazolidin-4-one Ib , 2-[2,4-dioxo-5-(thiazol-2-ylcarbamoylmethyl)-thiazolidin-3-yl]- N -(2-triﬂuoromethylphenyl)acetamide IId and (2,4-dioxo-5-(thiazol-2-ylcarbamoylmethylene)-thiazolidin-3-yl)acetic acid ethyl ester IIj showed excellent anticonvulsant activity in both models. The directions of compounds modiﬁcation based on SAR aspects were discussed. The results of the study provide a basis for further study of the anticonvulsant properties of selected thiazole-thiazolidinones.


Phenytoin
The present work is an extension of our ongoing efforts towards a search for new 4-thiazolidinone-based anticonvulsant agents [22,39]. Here, we addressed the design and synthesis of thiazole-thiazolidinones hybrids and evaluation of their anticonvulsant effects.

General Method for Amides of 2,4-Dioxothiazolidine-5-Carboxylic Acids Synthesis (II)
A mixture of 5 mmol of 2,4-dioxothiazolidine-5-carboxylic acid or 2,4-dioxothiazolidin-5ylidene)-acetic acid, 12 mL of thionyl chloride in 30 mL of dioxane was heated under reflux for 1 h, cooled, and treated with 100 mL of hexane. The precipitate was separated by filtration. To the solution of 5 mmol of the corresponding 2-aminoazole, 1 mL of triethylamine in 10 mL of anhydrous dioxane the solution of appropriate 5 mmol acid chloride in 10 mL of the same solvent was added. The mixture was heated for 20 min at 90 • C, cooled and diluted with 100 mL of water. The formed precipitate was filtered and recrystallized from an appropriate solvent.

Pharmacology
Adult random-bred albino-mice of either sex weighing 18-25 g, which were kept under the standard sanitary-hygienic conditions in the vivarium of the Central Research Laboratory of the National Pharmaceutical University (Ukraine), were used in the experiments. All animal procedures were approved by the institutional animal care and use committee and corresponding to the Law of Ukraine. During the experiment, the rules and principles adopted by the Helsinki Declaration on Humane Animal Welfare (2000), the Directive of the Council of the European Union on the protection of animals used for scientific purposes (2010) were observed. The animals were randomized into the next groups: Group 1-control group; Group 2-experimental group-mice treated with studied compounds or reference drugs.
The synthesized compounds were screened for their anticonvulsant activity by the pentylenetetrazole-induced (PTZ) seizures test and maximal electroshock seizure (MES) test [44,45]. Animals were administered the tested compounds once in the form of a thin suspencion stabilized with polysorbate-80 (Tween-80), at a dose of 100 mg/kg intragastrally.
In the PTZ test, in which seizures were induced by suppression of GABA-ergic inhibitory processes, sodium valproate (Depakin, Sanofi-Aventis, Gentilly, France) as a reference drug was administrated intragastrally (300 mg/kg) as an oral syrup. In the control group, the animals received purified water in an appropriate volume. Pentylenetetrazole (Corazol, Sigma-Aldrich, St. Louis, MO, USA) in the form of water solution (90 mg/kg) was injected subcutaneously 30 min after compound application. Subsequently, each mouse was placed in a separate cylindrical plastic container (5 L) and continuously monitored for 60 min.
MES test was used to evaluate the ability of compounds to prevent the generalization of seizure. Electroshock was applied via corneal electrodes. The ability of the compounds to prevent seizures was associated with the prevention of the spread of impulse through the nerve tissues. In the MES test, electric current-50 mA, a frequency 50 Hz during 0.2 s was used 30 min after the application of the tested compounds. Carbamazepine (Finlepsin, TEVA-Pharmaceuticals, Petah Tikva, Israel) was used as a reference drug and was administered intragastrically (40 mg/kg) in the form of a thin water suspension stabilized with a Tween-80 (polysorbate 80). The observation lasted 60 min. Latency of the convulsions; the number of clonic-tonic convulsions in one mouse; % of animals with clonic and tonic convulsions; the duration of the convulsive period (from the first to the last attack); and the lifetime of the animals before death (in mice with a lethal outcome) were calculated. The severity of seizures was evaluated according to a scale ranging from 1 to 6: 1-trembling; 2-circus movement; 3-clonic seizures; 4-clonic-tonic seizures with a lateral position; 5-tonic extension; 6-tonic extension leading to the animal's death. When seizures were not observed within 1 h, it was considered that the latent period was 60 min. The protection of animals from the development of clonic and tonic seizures and lethality were treated as the most significant indicators of anticonvulsant activity of the compounds [44].
Acute toxicity. The experiments were conducted on white male mice weighing 23-25 g. Compounds were dissolved in saline solution (0.9% NaCl) with l-2 drops of polysorbate 80 (Tween-80), after dissolution they were administered via intraperitoneal route. The LD 50 was evaluated for 4 or 5 different doses each on 6 animals and calculated by the Litchfield-Wilcoxon method [46,47].
Statistical analysis was carried out using a Statistica 10.0 by the methods of variation statistics. The average values and standard errors were calculated. The significance of the differences between groups was estimated according to the Student's criterion (t) in the case of normal distribution, and according to the nonparametric Mann-Whitney criterion (U) in the case of the absence of normal distribution. The results, which were determined in an alternative form (presence/absence of a certain feature), were evaluated using the Fisher's criterion (ϕ) [48].

Results and Discussion
The design of molecule structure of target compounds was based on: Our former findings [22,39]; the compliance with structural requirements for potential anticonvulsants (see above); combination of thiazole and thiazolidinone cores in one molecule [49,50]; structural similarity with ralitolin ( Figure 2 MES test was used to evaluate the ability of compounds to prevent the generalization of seizure. Electroshock was applied via corneal electrodes. The ability of the compounds to prevent seizures was associated with the prevention of the spread of impulse through the nerve tissues. In the MES test, electric current-50 mA, a frequency 50 Hz during 0.2 s was used 30 min after the application of the tested compounds. Carbamazepine (Finlepsin, TEVA-Pharmaceuticals, Petah Tikva, Israel) was used as a reference drug and was administered intragastrically (40 mg/kg) in the form of a thin water suspension stabilized with a Tween-80 (polysorbate 80). The observation lasted 60 min. Latency of the convulsions; the number of clonic-tonic convulsions in one mouse; % of animals with clonic and tonic convulsions; the duration of the convulsive period (from the first to the last attack); and the lifetime of the animals before death (in mice with a lethal outcome) were calculated. The severity of seizures was evaluated according to a scale ranging from 1 to 6: 1-trembling; 2-circus movement; 3-clonic seizures; 4-clonic-tonic seizures with a lateral position; 5-tonic extension; 6-tonic extension leading to the animal's death. When seizures were not observed within 1 h, it was considered that the latent period was 60 min. The protection of animals from the development of clonic and tonic seizures and lethality were treated as the most significant indicators of anticonvulsant activity of the compounds [44].
Acute toxicity. The experiments were conducted on white male mice weighing 23-25 g. Compounds were dissolved in saline solution (0.9% NaCl) with l-2 drops of polysorbate 80 (Tween-80), after dissolution they were administered via intraperitoneal route. The LD50 was evaluated for 4 or 5 different doses each on 6 animals and calculated by the Litchfield-Wilcoxon method [46,47].
Statistical analysis was carried out using a Statistica 10.0 by the methods of variation statistics. The average values and standard errors were calculated. The significance of the differences between groups was estimated according to the Student's criterion (t) in the case of normal distribution, and according to the nonparametric Mann-Whitney criterion (U) in the case of the absence of normal distribution. The results, which were determined in an alternative form (presence/absence of a certain feature), were evaluated using the Fisher's criterion (φ) [48].
The evaluation of anticonvulsant activity of the tested compounds was undertaken within the known anticonvulsant drug development program protocol [55]. The procedure involved PTZ-induced seizures and MES-test. For anticonvulsant activity study, N-unsubstituted 2-imino-thiazolidinones were used in the form of N-potassium salts.
At the first stage, a PTZ-induced seizures model was used. The mechanism of attack development was based on the suppressive effect of PTZ on GABA-receptors, resulting in a reduction in the inhibitory effect of GABA in the central nervous system [45]. Sodium valproate was used as a reference drug ( Table 1). The anticonvulsant activity of the latter was associated with blockade of sodium channels, increased activity of glutamate decarboxylase, and suppression of GABA transaminase, which led to an increase in the level of GABA in the central nervous system [56]. At the experimental conditions, sodium valproate (300 mg/kg) showed good anticonvulsant effect: A significant decrease of the latent period of seizure development (by 6 times comparing to control); decrease (2.4 times) of the amount of clonic-tonic seizures per animal; decrease of the percentage of mice with clonic and tonic seizures (50 vs. 100% and 33.33 vs. 91.67%, p < 0.01); decreasing of the severity of the seizures by 2.8 times; decrease of the duration of the seizure period by 1.9 times; and decrease of the mortality. Among the tested compounds, three of the most active compounds Ib, IId, IIj were identified.
Compound IIj significantly reduced (by 3.3 times) the duration of the convulsive period (2.93 ± 1.99 min vs. 9.68 ± 1.96 min, p < 0.05) and the level of mortality by 2.5 times (33.33% vs. 83.33%, p < 0.05). The latent period, the proportion of animals with clonic seizures and severity of convulsion were decreased, but the compound's effect was inferior to the sodium valproate.
Seven compounds: Ie, IIb, IIf, IIe, IIg, IIh, and IIi did not show either anticonvulsant or pro-convulsive activity. For the two compounds Ia and Ic, the minor anticonvulsant properties were established since they significantly affected only one or two of the studied parameters. There was a significant decrease in the control of the number of convulsion per animal and the decrease in the duration of the convulsive period under compound Ia action. Compound Ic contributed to the lengthening of the latent period and the decrease in the amount of clonic-tonic convulsions per mouse.
Compounds Id and IIc showed pro-convulsant properties. Compound IIc, which showed moderate pro-convulsant properties, was characterized by a significant reduction in the life of animals before death and, accordingly, the duration of the convulsive period. In all animals, under compound IIc action the most severe tonic convulsions were observed. Compound Id induced the development of tonic seizures in 100% of cases, indicating its pro-convulsant properties. At the second stage, the most active compounds Ib, IId, and IIj were selected to evaluate their anticonvulsant action in the MES model. The results of the experiment are shown in the Table 2.Carbamazepine was used as a reference drug, due to its ability to block sodium channels that play a leading role in the development of primary-generalized convulsions [57].There was a significant reduction of the severity of the convulsion by 1.2 times (4.78 ± 0.28 vs. 5.53 ± 0.13 points, p < 0.05), the duration of the convulsive period by 2.4 times (0.10 ± 0.03 vs. 0.24 ± 0.03 min, p < 0.01) and the duration of the recovery period by 1.9 times (0.41 ± 0.14 vs. 0.80 ± 0.16 min, p < 0.01) under carbamazepine treatment.  The most active of the synthesized compounds were evaluated for their approximate LD 50 (intraperitoneal administration) in mice [46,47]. Tested compounds were relatively non-toxic and well tolerated by the experimental animals as demonstrated by their LD 50 (Ib-263 ± 27.1 mg/kg, IId-610 ± 40.5 mg/kg, IIj-320 ± 31.5 mg/kg).
The analysis of the structure-activity relationships in the group of 2-imino-4-thiazolidinones (I) reveled the following patterns: The presence of nitrobenzylidene fragment at C5 position of main core are preferred for anticonvulsant activity realization. The replacement of the nitro-group by fluorine atom (compound Ia) or dimethylamino group (compound Ic) led to reduction of the anticonvulsant effect. The introduction of the substituent in the N3 position of the thiazolidinone core led to only imino-form of compounds and the impossibility of amino-iminotautomeric transformation. Such modification of 2-iminothiazolyl-4-thiazolidinones led to a decrease of anticonvulsant activity (compound Ie). The replacement of the arylidene fragment by isatine moiety that can be treated as annulation of benzylidene fragment (compound Id), and it was not the optimal direction for compound structure modification. Compound Id possesses even pro-convulsant activity. For such group of compounds I with different substituent in the C5 benzylidene fragment the strong antimicrobial activity [43,58] and inhibition of SHP-2 (non receptor protein tyrosine phosphatase that mediates cell signaling by growth factors and cytokines acting via the RAS/MAP kinase pathway [59]) also was detected. These alone with a new trend in the creation of new anticonvulsants [60] can be treated as a benefit, especially within a polypharmacological approach [10,11], when the compound possessed several types of activity and can be used as a baseline for further optimization.
For the 2,4-thiazolidinone-5-carboxylic acid derivatives (II), the next aspects of structure-activity relationships were funded: The complication of C5 moiety (compounds IIh and IIi) led to activity decreasing; the presence of N3 substituent was essential (crucial decreasing of activity of IIf compare to IIj). The nature of the substituents in the N3 fragment also plays a significant role: The absence of C=O group (compound IIb), as well as the replacement of CF 3 (IId) group by OMe group (IIc) provided the activity decreasing. Saturated analogs (with C5-single bond) of compound IIj also possessed the strong anticonvulsant activity [61] that confirmed the potency of such direction for the new anticonvulsants design.
Hit-compounds Ib, IId, and IIj showed pronounced anticonvulsant properties in both the PTZ-model and the MES-test. It can be assumed that these compounds have a mixed mechanism of action aimed at increasing the inhibitory processes in the central nervous system by increasing GABA-ergic activity, and reducing the processes of excitation in the CNS by blocking the sodium channels. The results of the study provide a basis for further study of the anticonvulsant properties of thiazole-thiazolidinones.