Discovery of Cinnamylidene Derivative of Rhodanine with High Anthelmintic Activity against Rhabditis sp.

The treatment of parasitic infections requires the application of chemotherapy. In view of increasing resistance to currently in-use drugs, there is a constant need to search for new compounds with anthelmintic activity. A series of 16 cinnamylidene derivatives of rhodanine, including newly synthesized methoxy derivatives (1–11) and previously obtained chloro, nitro, and diethylamine derivatives (12–16), was investigated towards anthelmintic activity. Compounds (1–16) were evaluated against free-living nematodes of the genus Rhabditis sp. In the tested group of rhodanine derivatives, only compound 2 shows very high biological activity (LC50 = 0.93 µg/µL), which is higher than the reference drug albendazole (LC50 = 19.24 µg/µL). Crystal structures of two compounds, active 2 and inactive 4, were determined by the X-ray diffraction method to compare molecular geometry and search for differences responsible for observed biological activity/inactivity. Molecular modelling and selected physicochemical properties prediction were performed to assess the potential mechanism of action and applied in the search for an explanation as to why amongst all similar compounds only one is active. We can conclude that the tested compound 2 can be further investigated as a potential anthelmintic drug.


Introduction
Multidrug resistance (MDR) has become a factor seriously limiting the treatment of various diseases, including parasitic infections [1]. Parasitic infections are connected with the living environment and are widely distributed in tropical and subtropical areas. They are not transmitted directly from person to person. In the case of malaria, parasites are transmitted to people through the bites of infected mosquitoes. According to WHO's estimation, in 2020 there were about 241 million cases of malaria infections worldwide, which led to 627 thousand deaths [2]. Soil-transmitted helminth infections caused by different species of parasitic worms are also a big problem in the modern world. They are transmitted by eggs present in human faeces, which lead to contamination of the soil in areas where sanitation is poor. As reported by WHO, about 1.5 billion people are infected with soil-transmitted helminth infections [3]. More than half of those infected are children. Infections can cause anaemia, intestinal manifestations, general malaise and weakness, and impaired growth and physical development. Chemotherapy is the main direction to control control and fight against parasitic infections [4]. Thus, it is crucial to search for new anthelmintic drugs or other strategies allowing to overcome MDR.
Thus, this paper presents the synthesis of a series of new compounds (1-11) and structural analysis with X-ray crystallography for two selected compounds (2 and 4). Furthermore, the newly synthesized compounds (1-11) and these synthesized earlier (12)(13)(14)(15)(16) [8] were evaluated with regard to their anthelmintic activity. The bioactivity was observed only for one compound (2), which prompted us to search further for the explanation of the obtained results. We have analyzed the molecular geometries and intermolecular interactions of active compound 2 and inactive 4 in the crystals in order to search for structural features and geometrical parameters which could be responsible for the biological activity. Additionally, molecular docking studies have been performed to examine the putative mechanism of action and selected physicochemical properties have been estimated to find the plausible explanation as to why only one compound of the whole studied series exhibits the desired bioactivity.

Synthesis
The final compounds 1-11 were obtained within two-step synthesis routes (Schemes 1b and 2b) based on the procedures described previously for compounds 12-16 [8]. The synthetic pathway also required obtaining the rhodanine (Scheme 1a) and the carboxyalkylrhodanine acids (Scheme 2a). In the first step, the rhodanine and cinnamaldehyde derivatives were used to obtain the triethylammonium salts of final compounds 1, 8 and Thus, this paper presents the synthesis of a series of new compounds (1-11) and structural analysis with X-ray crystallography for two selected compounds (2 and 4). Furthermore, the newly synthesized compounds (1-11) and these synthesized earlier (12)(13)(14)(15)(16) [10] were evaluated with regard to their anthelmintic activity. The bioactivity was observed only for one compound (2), which prompted us to search further for the explanation of the obtained results. We have analyzed the molecular geometries and intermolecular interactions of active compound 2 and inactive 4 in the crystals in order to search for structural features and geometrical parameters which could be responsible for the biological activity. Additionally, molecular docking studies have been performed to examine the putative mechanism of action and selected physicochemical properties have been estimated to find the plausible explanation as to why only one compound of the whole studied series exhibits the desired bioactivity.

Synthesis
The final compounds 1-11 were obtained within two-step synthesis routes (Schemes 1b and 2b) based on the procedures described previously for compounds 12-16 [10]. The synthetic pathway also required obtaining the rhodanine (Scheme 1a) and the carboxyalkylrhodanine acids (Scheme 2a). In the first step, the rhodanine and cinnamaldehyde derivatives were used to obtain the triethylammonium salts of final compounds 1, 8 and 10 (Scheme 1b), which were reacted with hydrochloride acid to form the corresponding final products. In the case of other compounds (2-7, 9 and 11), the corresponding carboxyalkylrhodanine acid and cinnamaldehyde derivatives were reacted to access the intermediates, namely the triethylammonium salts. Next, hydrochloride acid was added in order to obtain the corresponding final products. 10 (Scheme 1b), which were reacted with hydrochloride acid to form the corresponding final products. In the case of other compounds (2-7, 9 and 11), the corresponding carboxyalkylrhodanine acid and cinnamaldehyde derivatives were reacted to access the intermediates, namely the triethylammonium salts. Next, hydrochloride acid was added in order to obtain the corresponding final products.   10 (Scheme 1b), which were reacted with hydrochloride acid to form the corresponding final products. In the case of other compounds (2-7, 9 and 11), the corresponding carboxyalkylrhodanine acid and cinnamaldehyde derivatives were reacted to access the intermediates, namely the triethylammonium salts. Next, hydrochloride acid was added in order to obtain the corresponding final products.   For two representative compounds, active 2 and inactive 4, X-ray crystallographic analysis was carried out. The projections of molecular geometries in the crystals of 2 and 4 are presented in Figure 2. The asymmetric unit of 2 consists of one molecule of the investigated compound and one solvent molecule of dimethyl sulfoxide (DMSO), while 4 crystallizes only with one molecule in the asymmetric unit.

Crystal Structures of 2 and 4
For two representative compounds, active 2 and inactive 4, X-ray crystallographic analysis was carried out. The projections of molecular geometries in the crystals of 2 and 4 are presented in Figure 2. The asymmetric unit of 2 consists of one molecule of the investigated compound and one solvent molecule of dimethyl sulfoxide (DMSO), while 4 crystallizes only with one molecule in the asymmetric unit. The investigated compounds possess two double bonds C5=C6 and C7=C8. As shown in Figure 2 for both presented compounds the same isomer (5Z, 7E) in crystals is observed. This isomer was also observed in other crystal structures determined earlier, containing rhodanine and aromatic rings linked with the same chain construct [8,29].
The 5-cinnamylidenerhodanine fragment is conjugated and almost planar. The values of single bonds (C6-C7 and C8-C9) in the linker are between the values of single and double bonds, namely 1.431(2) Å and 1.459(2) Å for 2, and 1.433(2) Å and 1.455(2) Å for 4. The angle between the planes containing rhodanine and aromatic rings is 11.71(8)° and 0.32(9)°, for 2 and 4, respectively. Similar values of bond lengths and interplanar angles are in agreement with these observed in the crystal structures of other derivatives containing this fragment [8,29].
The main intermolecular interactions motif is related to the O-H…O hydrogen bonds formation. One molecule of 2 interacts by carboxyl group with one molecule of DMSO ( Figure 3a), while in the case of 4, two molecules related by the inversion centre interact by carboxyl groups (Figure 3b). Furthermore, the crystal structures are stabilized by C-H…O and C-H…S contacts. The investigated compounds possess two double bonds C5=C6 and C7=C8. As shown in Figure 2 for both presented compounds the same isomer (5Z, 7E) in crystals is observed. This isomer was also observed in other crystal structures determined earlier, containing rhodanine and aromatic rings linked with the same chain construct [10,29].

Anthelmintic Activity and Molecular Geometry
We have analyzed the geometries of active compound 2 and inactive 4 in the crystals and compared them with two other derivatives, namely 14 and 16, for which the crystal structures were determined earlier [10]. The compared compounds differ in the length of the linker between the rhodanine ring and carboxyl group and in the substituent at the aromatic ring. Figure 4 presents the overlap of all four compared molecules. The analyzed compounds show small differences in planarity of the 5-cinnamylidenerhodanine moiety (Figure 4a). The biologically active compound 2 shows the biggest deviation from the planarity of this fragment in comparison to the other three non-active compounds. Furthermore, the carbon atom of the methoxy substituent of 2 points to the opposite direction in comparison to 4. Mutual orientation of carboxyl group and rhodanine ring also shows differences (Figure 4b).

Anthelmintic Activity and Molecular Geometry
We have analyzed the geometries of active compound 2 and inactive 4 in the crystals and compared them with two other derivatives, namely 14 and 16, for which the crystal structures were determined earlier [8]. The compared compounds differ in the length of the linker between the rhodanine ring and carboxyl group and in the substituent at the aromatic ring. Figure 4 presents the overlap of all four compared molecules. The analyzed compounds show small differences in planarity of the 5-cinnamylidenerhodanine moiety (Figure 4a). The biologically active compound 2 shows the biggest deviation from the planarity of this fragment in comparison to the other three non-active compounds. Furthermore, the carbon atom of the methoxy substituent of 2 points to the opposite direction in comparison to 4. Mutual orientation of carboxyl group and rhodanine ring also shows differences (Figure 4b).

Docking Studies
For a deeper insight of the observed anthelmintic activity of 2 compared to the inactive 4 and two other inactive derivatives 14 and 16 with known crystal structures [10] the molecular docking study was performed. In analogy to albendazole (a chosen control probe in this study), we set the hypothesis that the mechanism of action of 2 is through the inhibition of microtubule polymerisation. Thus, the colchicine highly hydrophobic binding pocket postulated as the albendazole putative binding site [30], was selected for the molecular docking experiment.

Docking Studies
For a deeper insight of the observed anthelmintic activity of 2 compared to the inactive 4 and two other inactive derivatives 14 and 16 with known crystal structures [8] the molecular docking study was performed. In analogy to albendazole (a chosen control probe in this study), we set the hypothesis that the mechanism of action of 2 is through the inhibition of microtubule polymerisation. Thus, the colchicine highly hydrophobic binding pocket postulated as the albendazole putative binding site [30], was selected for the molecular docking experiment.
The re-docking experiment performed for the pre-prepared tubulin structure in complex with colchicine (PDB ID: 5NM5) [31] confirmed the good performance of the designed methodology, resulting in 0.77 RMSD calculated for corresponding heavy atoms of the docked colchicine and the active conformer observed in the crystal structure (PDB ID: 5NM5). The native ligand reached the best docking score (Table 2). Surprisingly, the second-best fit was obtained for inactive 14, followed by active 2. Both mentioned ligands adopt nearly identical poses in the tubulin β-2B hydrophobic pocket, allowing deeper penetration of the molecule in the binding site in comparison to the colchicine original location (Figure 7). The best-ranked rhodanine derivatives 2 and 14 form hydrophobic interactions with Cys241, Leu248, Ala250, Leu 255, and weak hydrogen bonds with Val238 The re-docking experiment performed for the pre-prepared tubulin structure in complex with colchicine (PDB ID: 5NM5) [31] confirmed the good performance of the designed methodology, resulting in 0.77 RMSD calculated for corresponding heavy atoms of the docked colchicine and the active conformer observed in the crystal structure (PDB ID: 5NM5). The native ligand reached the best docking score (Table 2). Surprisingly, the second-best fit was obtained for inactive 14, followed by active 2. Both mentioned ligands adopt nearly identical poses in the tubulin β-2B hydrophobic pocket, allowing deeper penetration of the molecule in the binding site in comparison to the colchicine original location (Figure 7). The best-ranked rhodanine derivatives 2 and 14 form hydrophobic interactions with Cys241, Leu248, Ala250, Leu 255, and weak hydrogen bonds with Val238 and Tyr202. The last residue is not accessible for the colchicine molecule due to the rigidity and less elongated shape of the alkaloid derivative compared to rhodanine derivatives. Only slightly less prominent in the predicted binding is 4. The longer linker between rhodanine and carboxyl group increases the flexibility of the molecule with no additional interactions observed.
Despite the elongated shape of the albendazole molecule, it does not penetrate the binding site as deep as the rhodanine derivatives, which could explain its lower position in the ranking list. Interestingly, the molecule of 16 adopts an opposite orientation compared to other studied rhodanine derivatives. The lowest position in the ranking and observed different pose may be related to the relative flexibility of the diethylamine substituent, which can cause steric hindrance in the deeper regions of the tubulin β-2B binding pocket. With this distinct orientation of the molecule, the carboxylic functional group interacts with Tyr 202, forming a strong O-H. . . O hydrogen bond, which does not tip the balance in favour of this pose. and less elongated shape of the alkaloid derivative compared to rhodanine derivatives. Only slightly less prominent in the predicted binding is 4. The longer linker between rhodanine and carboxyl group increases the flexibility of the molecule with no additional interactions observed. Despite the elongated shape of the albendazole molecule, it does not penetrate the binding site as deep as the rhodanine derivatives, which could explain its lower position in the ranking list. Interestingly, the molecule of 16 adopts an opposite orientation compared to other studied rhodanine derivatives. The lowest position in the ranking and observed different pose may be related to the relative flexibility of the diethylamine substituent, which can cause steric hindrance in the deeper regions of the tubulin β-2B binding pocket. With this distinct orientation of the molecule, the carboxylic functional group interacts with Tyr 202, forming a strong O-H…O hydrogen bond, which does not tip the balance in favour of this pose.

Prediction of Selected Physicochemical Properties
The above-described docking results suggest higher affinities of compounds 2, 4, and 14 towards postulated target protein, tubulin β-2B, compared to albendazole. However, it still does not explain why only 2 exhibits the anthelmintic activity in the biological test. To find the answer to the observed results, we decided to estimate selected physicochemical properties, which can shed some light on the obtained experimental results. We examined predicted measures for solubility (log S) and lipophilicity (log P). The obtained results (Table 3, data for colchicine are shown only for comparison) suggest only moderate solubility with albendazole being the leader amongst the studied compounds.  An interesting relation can be spotted for log P predicted values. The active compound 2 has the lowest value of log P for the studied compounds, suggesting the least lipophilic properties of this molecule. As it was suggested by Kerboeuf and Guégnard [32], anthelmintics can act as substrates or modulators of the P-glycoprotein (P-gp) activity in the nematode. This ABC-transporter protein is responsible for drug resistance, including observed lowering anthelmintic activity in mammals [33]. P-gp recognizes its substrates in the inner region of the cell membrane and transports lipophilic or amphiphilic molecules [34]. Thus, modification of molecular properties (especially the change in lipophilicity level) by inserting different substituents for a series of bioactive compounds can modulate the compound's impact on P-gp activity, which can be called a "log P-response" relationship [35]. For this reason, the stronger anthelmintic activity observed for 2 in comparison to albendazole could be explained by the lower log P value and the related increased bioavailability. This can also suggest that the higher lipophilicity for the remaining rhodanine derivatives (4, 14 and 16) may be the limiting factor forbidding the molecules to reach the target protein in vivo.

Chemical Synthesis
Reagents used for synthesis were purchased from Sigma-Aldrich (Munich, Germany) and Across Organics and were used without further purification.
The melting points were determined on a Boetius apparatus (Jena, Germany) and are uncorrected.
An AmaZon EDT apparatus (Bremen, Germany) equipped with an electrospray ion source and an ion trap analyzer was used to perform the mass spectrometry analysis. The samples were introduced into the apparatus after dissolving them in a mixture of methanol and chloroform (1: 1 v/v) with the addition of 0.1% vol of formic acid. The scanning was carried out in the range of 100-700 m/z, accumulating approximately 300,000 ions in one cycle. Positive or negative ion polarization was used, depending on the properties of the tested sample.
The IR spectra were recorded on a JASCO FT IR-670 Plus instrument in KBr pellets (1 mg sample/400 mg KBr).
General Procedures for the Synthesis of Final Compounds 1-11 The appropriate rhodanine or rhodanine-3-alkanoic acid (5.0 mmol), molecular sieves 4A (5 g), isopropyl alcohol (30 cm 3 ), appropriate aldehyde (5.5 mmol) and triethylamine (25.0 mmol) were placed into a flask. The mixture was heated under a reflux condenser for 4-5 h in argon. Upon completion of the reaction, the hot solution was immediately filtered. The filtrate was cooled and 2 M hydrochloric acid solution was added in order to obtain an acidic solution. Next, the obtained precipitate was filtered using a Büchner funnel and crystallized from acetic anhydride or glacial acetic acid. graphite monochromator, while for 4 were collected using the XtaLAB Synergy-S diffractometer, equipped with the CuKα (1.54184 Å) radiation source and graphite monochromator. Positions of all non-hydrogen atoms were determined by direct methods using the SIR-2014 [36] program. Refinement and further calculations were carried out using the SHELXL-2018 program [37]. All non-hydrogen atoms were refined anisotropically using weighted full-matrix least-squares on F 2 . The hydrogen atoms bonded to carbons were included in the structure at idealized positions and were refined using a riding model. Hydrogen atoms attached to oxygen atoms were found from the difference Fourier map and refined without restraints. For molecular graphics the MERCURY [38] program was used.

Anthelmintic Activity Assay
Nematodes of the genus Rhabditis sp. were used for the anthelmintic activity experiments. The nematodes were cultured following the procedure developed by A. Bogucka-Kocka and P. Kołodziej at the Department of Biology and Genetics of the Medical University of Lublin [39]. Test derivatives were suspended in 60% glycerol and sonicated using an ultrasonic homogenizer (Hielscher Ultrasound Technology, Teltow, Germany). The initial homogeneous emulsion was prepared at a concentration of 100 µg/µL. The experiment was performed on 24-well plates. The test compounds were added to nematode cultures at the following concentrations: 0.2 µg/µL; 1.1 µg/µL; 3.3 µg/µL; 5.5 µg/µL; 11.1 µg/µL. Additionally, separate cultures were carried out in 0.6% NaCl, 60% glycerol as the negative control and also the positive control with the addition of the substance with antiparasitic activity, namely albendazole. Each experiment was repeated 3 times. After a 24-h exposure, the culture of nematodes of the genus Rhabditis sp. was observed in terms of development, deformity, damage and motility using a stereoscopic and light microscope (Olympus, Tokyo, Japan). To determine the viability of nematodes, they were stained with methylene blue. Live and dead specimens were counted in Bürker (Brand, Wertheim, Grossostheim, Germany) counting chambers.
Statistical analysis was performed using the GraphPad Prism 5.02 software (GraphPad Software, San Diego, CA, USA) using the ANOVA variance analysis and the Tukey test (statistically significant differences were adopted with the coefficient value of p < 0.05 *, p < 0.01 **, p < 0.001 ***) [39-42].

Protein Preparation
A tubulin structure in complex with colchicine (PDB ID 5NM5 [27]) was selected and prepared using Maestro (Version 12.5.139) [43]. The dimer consisting of chains A (tubulin α-1B) and B (tubulin β-2B) in complex with colchicine and GTP were separated to mimic the closest to the natural binding environment for the docked ligands. The hydrogen atoms were added in the idealized positions. The missing side chains, located a long distance from the colchicine's binding side, were added using the conformers' library. Colchicine and water molecules were removed for the next step of the experiment.
Additional note: the 5NM5 structure is a Bos Taurus tubulin, which shares 88.99% sequence identity with tubulin β-2B from Caenorhabditis elegans-the organism closely related to the one used in the biological activity tests. The sequence alignment was obtained with Clustal Omega [44,45].
Most of the non-identical residues are a long distance from the colchicines binding site. Only four residues are in the proximity of the binding site, however, the character of the mutation does not indicate strong differences in the plausible interactions.

Ligands Preparation
The initial 3D geometries of ligands were taken from the crystal structure data. Colchicine randomized 3D structure for the re-docking experiment, as well as the albendazole molecule, wereobtained from SMILES with OpenBabel (version 2.4.1) [46] and their geometries were initially minimized with the MM+ force field in HyperChem Professional 8.0 [47] integrated molecular modelling software.

Molecular Docking Procedure
Docking procedures were conducted in GOLD [48]. The binding site was centred at Leu255 (CD1 atom) and included all residues within the sphere of 10 Å radii, fully covering the colchicine's residing space. During the docking process, ligands were allowed flexibility to find the most probable conformation (genetic algorithm (GA) implemented in GOLD [48]). Only the chain linking the rhodanine and aromatic moiety was kept rigid due to its nature (conjugation) and to guarantee the fully extended conformation, without alternative isomers consideration. Selected active residues constituting the original colchicines binding site (Cys241, Leu242, Leu248, Asn258, Ile318, Lys352) were treated as flexible during the docking procedure. The selection was adjusted stepwise based on multiple docking procedures and careful inspection of whether the alternative site chains conformations are generated during ligand binding. The empirical ChemScore scoring function [49] was applied to assess the binding affinity.

Log P and Log S Prediction
The log P and log S values for the investigated compounds were predicted using the ALOGPS 2.1 Java applet of the Virtual Computational Chemistry Laboratory (http: //www.vcclab.org; calculations have been performed on 3 February 2022) [50]. All obtained results are shown in Table 3.

Conclusions
New cinnamaldehyde derivatives of rhodanine were synthesized to search for new compounds with anthelmintic activity against Rhabditis sp. In the series of 16 compounds, only one shows very high activity, namely 5-(4 -methoxycinnamylidene)-rhodanine-3etanoic acid (2). Anthelmintic activity of compound 2 is higher than that of the commercial drug-albendazole ( Figure 5). We set the hypothesis that the mechanism of action of 2 is through the inhibition of microtubule polymerisation, just as it is postulated for albendazole. The molecular modelling studies have shown good binding of selected compounds to the colchicine highly hydrophobic pocket, which is a known albendazole putative binding site. The lack of anthelmintic activity of other investigated compounds can be connected with their higher lipophilicity in comparison to active compound 2 and albendazole (Table 3). It is worth noting that the lipophilicity of albendazole is higher than that of compound 2, which may explain its lower anthelmintic activity. The mechanism of action of 2 will be verified in the next step of our research. Thus, compound 2 is a good drug candidate with anthelmintic activity. Data Availability Statement: All data generated or analyzed during this study are included in this published article. Moreover, the datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.