Discovery of Potent and Selective Halogen-Substituted Imidazole-Thiosemicarbazides for Inhibition of Toxoplasma gondii Growth In Vitro via Structure-Based Design

Employing a simple synthetic protocol, a series of highly effective halogen-substituted imidazole-thiosemicarbazides with anti-Toxoplasma gondii effects against the RH tachyzoites, much better than sulfadiazine, were obtained (IC50s 10.30—113.45 µg/mL vs. ~2721.45 µg/mL). The most potent of them, 12, 13, and 15, blocked the in vitro proliferation of T. gondii more potently than trimethoprim (IC50 12.13 µg/mL), as well. The results of lipophilicity studies collectively suggest that logP would be a rate-limiting factor for the anti-Toxoplasma activity of this class of compounds.


Introduction
Toxoplasma gondii is a common zoonotic infection of humans, and estimates prove that up to one third of the world's population is chronically infected [1,2]. Although largely asymptomatic, chronic infection can be fatal or lead to serious problems in fetuses and immunocompromised patients [3][4][5][6][7][8][9]. Current administration of the therapeutics, like the combination of pyrimethamine and sulfadiazine, shows high rates of toxicity and side effects, including intolerance or allergic reaction to the sulfa component [10][11][12][13][14][15][16]. Other serious problems, such as the emergence of drug resistance and the incidence of relapses after discontinuation of therapy, in some cases are observed as well [17,18].
In this context, the thiosemicarbazide scaffold has emerged as a promising structure for the lead optimization process. In the search for new drug leads for toxoplasmosis, we are exploring the thiosemicarbazide scaffold as a promising lead structure for developing potent and selective anti-Toxoplasma gondii medicines. Preliminary screening of the imidazole-thiosemicarbazides has revealed several potent inhibitors of tachyzoite growth in vitro with much higher potency when compared to sulfadiazine [19]. Among them, the best anti-toxoplasma response was noted for those with electron-withdrawing nitro and chloro substitution at the N4 phenyl ring (Figure 1). Although new chemotypes were provided, low selectivity in the parasite inhibition over host cells, defined as this goal by further exploiting the N4 phenyl position of the imidazole-thiosemicarbazide core with electron-withdrawing halogen substitution. In fact, based on our initial results, it is reasonable to suppose that the deactivation of the N4 phenyl ring, through the inductive withdrawing effect of halogen atoms, should result in compounds with potent activity against T. gondii tachyzoites growth. From the viewpoint of rational drug design, other factors, such as impact of halogenated compounds on membrane permeability, were of high importance. Indeed, for many years, the effective applied strategy for the hit-to-lead or lead-to-drug optimization process involved the insertion of halogens during the synthesis of final compounds [20][21][22]. This strategy is based on the observation that incorporation of the halogen atoms into a new bioactive chemical entity improves membrane permeability and oral absorption [23]. Further, halogenation enhances the blood-brain barrier permeability, which is a pre-requisite for drugs that need to reach the CNS, like anti-toxoplasma drugs and many others [24]. In this paper, employing the halogenation strategy, a series of highly effective halogenated-substituted imidazole-thiosemicarbazides, with much better anti-Toxoplasma gondii effects against the RH tachyzoites than sulfadiazine, were identified. The most potent of these imidazole-thiosemicarbazides blocked the in vitro proliferation of T. gondii more potently and selectively than pyrimethamine, as well. In further studies we show that the observed trend in their anti-Toxoplasma gondii activity depends significantly on the lipophilicity factor.   [19]. *SR-selectivity ratio; defined as the ratio of the 50% cytotoxic concentration (CC50) to the 50% antiparasitic concentration (IC50).

Molecular Design and Synthesis
As mentioned in the Introduction, in our previous study, a series of imidazole-thiosemicarbazides was tested to optimize compounds effective against tachyzoite proliferation [19]. We discovered that the variations at the N4 position of the thiosemicarbazide core led to the differentiation of the biological response. For example, compounds with the N4 aliphatic chain had poor activity (IC50 ≥ 125 µg/mL), while compounds with electron donating substitution at N4 aryl position were generally less potent than those with an electron withdrawing group. The best results for the inhibition of tachyzoites proliferation were obtained for the nitro derivatives 1 and 2 ( Figure 1, left), and the difference with the control sulfadiazine was significant (IC50~2721.45 µ g/mL). To better understand the structural and electronic determinants responsible for the observed trend in activity, a computational approach was subsequently performed; this approach led to the Figure 1. Structures of previously reported imidazole-thiosemicarbazides with potent inhibitory activity against the proliferation of Toxoplasma gondii [19]. *SR-selectivity ratio; defined as the ratio of the 50% cytotoxic concentration (CC 50 ) to the 50% antiparasitic concentration (IC 50 ).

Molecular Design and Synthesis
As mentioned in the Introduction, in our previous study, a series of imidazole-thiosemicarbazides was tested to optimize compounds effective against tachyzoite proliferation [19]. We discovered that the variations at the N4 position of the thiosemicarbazide core led to the differentiation of the biological response. For example, compounds with the N4 aliphatic chain had poor activity (IC 50 ≥ 125 µg/mL), while compounds with electron donating substitution at N4 aryl position were generally less potent than those with an electron withdrawing group. The best results for the inhibition of tachyzoites proliferation were obtained for the nitro derivatives 1 and 2 ( Figure 1, left), and the difference with the control sulfadiazine was significant (IC 50~2 721.45 µg/mL). To better understand the structural and electronic determinants responsible for the observed trend in activity, a computational approach was subsequently performed; this approach led to the conclusion that the inductive withdrawing effect of substituents around the N4 phenyl ring, rather than its substitution pattern or geometry of molecules, are the key functionalities required for potent anti-Toxoplasma gondii activity. To test this assumption, a subsequent series of structural analogues of the nitro chemotypes 1 and 2 was designed and tested. Particularly, we investigated a panel of R groups in the context of the halogen substituents. A synthetic route for the preparation of the halogen-substituted series of imidazole-thiosemicarbazides 5-16 is presented in Scheme 1. The compounds were prepared under the routine protocol described elsewhere [25], in the one-step reaction of 4-methyl-imidazole-5-carbohydrazide with appropriate haloaryl isothiocyanate. conclusion that the inductive withdrawing effect of substituents around the N4 phenyl ring, rather than its substitution pattern or geometry of molecules, are the key functionalities required for potent anti-Toxoplasma gondii activity. To test this assumption, a subsequent series of structural analogues of the nitro chemotypes 1 and 2 was designed and tested. Particularly, we investigated a panel of R groups in the context of the halogen substituents. A synthetic route for the preparation of the halogen-substituted series of imidazole-thiosemicarbazides 5-16 is presented in Scheme 1. The compounds were prepared under the routine protocol described elsewhere [25], in the one-step reaction of 4-methyl-imidazole-5-carbohydrazide with appropriate haloaryl isothiocyanate.

Cytotoxicity against L929 Cells
Since T. gondii is an obligate intracellular parasite that requires invasion of mammalian host cells to proliferate, compounds that inhibit parasite growth might be toxic to host cells as well. Therefore, as an initial indicator for host cell toxicity, we firstly evaluated compounds 5-16 for their ability to inhibit the growth of mouse fibroblast (L929) cells line according to the international standards: ISO 10993-5:2009(E). In Table 1, the results are presented as the percent of viable cells ± standard derivation in the concentrations range of 5-16 between 0 to 1000 µ g/mL, together with the CC50 values. To calculate the reduction of viability compared to the untreated blank the equation was used: viability (%) = 100% × sample OD 570 (the mean value of the measured optical density of the treated cells)/blank OD 570 (the mean value of the measured optical density of the untreated cells). CC 50 -represents the concentration of tested compounds that was required for a 50% cell proliferation inhibition in vitro. The effect of the tested compounds on the cell lines was measured using MTT assay according to the international standard: ISO 10993-5:2009(E). CC 50 values were determined based on the plotted curves using GraphPad Prism program (version 8.0.1).
From the cytotoxicity assays against L929s, none of compounds 5-16 showed significant toxic effects on cells after 24 h of incubation. All of them were non-toxic at concentration less than~195 µg/mL which makes them a good candidate for anti-T. gondii activity assay in vitro. From the point of view of rational drug design, two other important conclusions from these studies should also be mentioned. Firstly, the cytotoxic effect of 5-16 generally increases with halogen size from fluorine to iodine (F < Cl <Br < I). In other words, the observed cytotoxic effect increases when electronegativity of the halogen decreases. Secondly, like steric and electronic effects, the halogen distribution pattern has considerable effect on cytotoxicity. Clearly, except the fluoro isomers 5-7, the general scale of toxicity is that para-isomer shows more toxicity than its meta counterpart, which in turn is more toxic than the isomer ortho (para > meta > ortho). Likewise, within a series of the fluoro isomers 5-7, the ortho-isomer is the most toxic, followed by the meta-isomer, whereas the para-isomer is the least toxic.

In Vitro Activity and Cytotoxicity
Drug resistance in Toxoplasma gondii is a challenge not only for treatment failure and recurrent infections but also for the correct constructions of basic conclusions from the in vitro bioassays. The correct interpretation of the bioassay results, i.e., proof of effectiveness and proof of safety for the samples tested, depends, partly, on the drug susceptibility of a strain used in the experiments. Therefore, we started our biological studies with a prior determination of the resistance of the reference RH strain of Toxoplasma (ATCC ® PRA 310™) to sulfadiazine and trimethoprim; two routinely used positive controls used in the in vitro bioassays. As presented in Figure 2, this highly virulent RH strain is sulfadiazine resistant with an IC 50 value higher than 2500 µg/mL. Trimethoprim was able to inhibit the tachyzoites production with a low concentration. However its cytotoxicity concentration (CC 50 > 32.00 µg/mL) was unfavorably close to the inhibitory effect (IC 50 = 12.13 µg/mL). Finally, only the combination of sulfadiazine and trimethoprim in the ratio 5:1 produced a non-toxic (CC 50 = 2454.14 µg/mL) synergistic effect against Toxoplasma gondii growth in vitro, with an IC 50 of 37.15 µg/mL.
Having determined the resistance of the reference RH strain to sulfadiazine and trimethoprim, we further investigated the efficacy of 5-16 in blocking T. gondii proliferation. Additionally, T. gondii proliferation was tested using a [ 3 H]-uracil incorporation assay, which is specific for the labelling of the nucleic acids of the parasite. Sulfadiazine, trimethoprim and their combination in ratio 5:1, respectively, were used as positive controls, while DMSO (dimethyl sulfoxide) at a concentration of 0.1% was used as the negative control (data not shown). Results of the screen are summarized in Figure 3, where compounds are ranked by relative potency.
In these cellular assays, several prominent trends were observed. As shown in Figure 3 and Table 2, among the tested set of compounds 5-16, all of them were more inhibitory to the growth of T. gondii than sulfadiazine. In addition, three of these, meta-bromo 12 (IC 50 = 15.64 µg/mL), para-bromo 13 (IC 50 = 14.57 µg/mL), and meta-iodo 15 (IC 50 = 10.30 µg/mL) were of exceptionally high potency with growth IC 50 values comparable to, or even better than, those of trimethoprim (IC 50 = 12.13 µg/mL). Compounds with ortho-fluoro 5, meta-fluoro 6, para-fluoro 7, and para-chloro 10 groups were the weakest inhibitors with IC 50 at~69 µg/mL or even higher. Having determined the resistance of the reference RH strain to sulfadiazine and trimethoprim, we further investigated the efficacy of 5-16 in blocking T. gondii proliferation. Additionally, T. gondii proliferation was tested using a [ 3 H]-uracil incorporation assay, which is specific for the labelling of the nucleic acids of the parasite. Sulfadiazine, trimethoprim and their combination in ratio 5:1, respectively, were used as positive controls, while DMSO (dimethyl sulfoxide) at a concentration of 0.1% was used as the negative control (data not shown). Results of the screen are summarized in Figure 3, where compounds are ranked by relative potency. Assessment of the results in terms of structural features leads to the following conclusions: (i) the inhibitory effect against T. gondii proliferation increases with halogen size from fluorine to iodine; compounds with bulky iodo and bromo groups were definitely more active than those with smaller chloro and fluoro substituents; (ii) the halogen distribution pattern has a considerable effect on the trends in the bioactivity. Clearly, except the bromo-substituted thiosemicarbazides 11-13, the general scale of the inhibitory potency is that the meta-isomer shows more activity than its ortho counterpart which, in turn, is more effective than the isomer para (meta > ortho > para). In turn, within series 11-13 the meta-isomer is almost as effective as its para analogue (IC 50~1 5 µg/mL), whereas the ortho-isomer is two times weaker in inhibitory action than its counterparts. In these cellular assays, several prominent trends were observed. As shown in Figure 3 and Table 2, among the tested set of compounds 5-16, all of them were more inhibitory to the growth of T. gondii than sulfadiazine. In addition, three of these, meta-bromo 12 (IC50 = 15.64 µ g/mL), para-bromo 13 (IC50 = 14.57 µ g/mL), and meta-iodo 15 (IC50 = 10.30 µ g/mL) were of exceptionally high potency with growth IC50 values comparable to, or even better than, those of trimethoprim (IC50 = 12.13 µ g/mL). Compounds with ortho-fluoro 5, meta-fluoro 6, para-fluoro 7, and para-chloro 10 groups were the weakest inhibitors with IC50 at ~69 µ g/mL or even higher.
Assessment of the results in terms of structural features leads to the following conclusions: (i) the inhibitory effect against T. gondii proliferation increases with halogen size from fluorine to iodine; compounds with bulky iodo and bromo groups were definitely more active than those with smaller chloro and fluoro substituents; (ii) the halogen distribution pattern has a considerable effect on the trends in the bioactivity. Clearly, except the bromo-substituted thiosemicarbazides 11-13, the general scale of the inhibitory potency is that the meta-isomer shows more activity than its ortho counterpart which, in turn, is more effective than the isomer para (meta > ortho > para). In turn, within series 11-13 the meta-isomer is almost as effective as its para analogue (IC50~15 µ g/mL), whereas the ortho-isomer is two times weaker in inhibitory action than its counterparts. It is a well-established principle that compounds exhibiting large therapeutic windows between the inhibition of T. gondii proliferation and human cell growth are expected to be more effective and safer during in vivo treatment. In practice, the selectivity ratio, defined as the ratio of the 50% cytotoxic concentration (CC 50 ) to the 50% antiparasitic concentration (IC 50 ), is as widely used as the parameter to express a compound's in vitro efficacy. Comparison of the test results in Table 2 shows our best inhibitor meta-I 15 to be more selective and more potent in inhibiting T. gondii cell growth in vitro than trimethoprim. This particular comparison favoring the meta-I 15 over trimethoprim in potency measure is not the only example favoring the halogen-substituted thiosemicarbazides listed in Table 2. For example, metaand para-bromo derivatives 12 and 13 show higher selectivity ratios and only slightly lower inhibitory potency than control trimethoprim. Again, although the weakest compound in our study (7) is not particularly impressive in its inhibition of T. gondii proliferation, it still displays a more favorable selectivity ratio than those for trimethoprim.  Table 2 shows our best inhibitor meta-I 15 to be more selective and more potent in inhibiting T. gondii cell growth in vitro than trimethoprim. This particular comparison favoring the meta-I 15 over trimethoprim in potency measure is not the only example favoring the halogen-substituted thiosemicarbazides listed in Table 2. For example, meta-and para-bromo derivatives 12 and 13 show higher selectivity ratios and only slightly lower inhibitory potency than control trimethoprim. Again, although the weakest compound in our study (7) is not particularly impressive in its inhibition of T. gondii proliferation, it still displays a more favorable selectivity ratio than those for trimethoprim. our best inhibitor meta-I 15 to be more selective and more potent in inhibiting T. gondii cell growth in vitro than trimethoprim. This particular comparison favoring the meta-I 15 over trimethoprim in potency measure is not the only example favoring the halogen-substituted thiosemicarbazides listed in Table 2. For example, meta-and para-bromo derivatives 12 and 13 show higher selectivity ratios and only slightly lower inhibitory potency than control trimethoprim. Again, although the weakest compound in our study (7) is not particularly impressive in its inhibition of T. gondii proliferation, it still displays a more favorable selectivity ratio than those for trimethoprim. vitro than trimethoprim. This particular comparison favoring the meta-I 15 over trimethoprim in potency measure is not the only example favoring the halogen-substituted thiosemicarbazides listed in Table 2. For example, meta-and para-bromo derivatives 12 and 13 show higher selectivity ratios and only slightly lower inhibitory potency than control trimethoprim. Again, although the weakest compound in our study (7) is not particularly impressive in its inhibition of T. gondii proliferation, it still displays a more favorable selectivity ratio than those for trimethoprim. vitro than trimethoprim. This particular comparison favoring the meta-I 15 over trimethoprim in potency measure is not the only example favoring the halogen-substituted thiosemicarbazides listed in Table 2. For example, meta-and para-bromo derivatives 12 and 13 show higher selectivity ratios and only slightly lower inhibitory potency than control trimethoprim. Again, although the weakest compound in our study (7) is not particularly impressive in its inhibition of T. gondii proliferation, it still displays a more favorable selectivity ratio than those for trimethoprim. potency measure is not the only example favoring the halogen-substituted thiosemicarbazides listed in Table 2. For example, meta-and para-bromo derivatives 12 and 13 show higher selectivity ratios and only slightly lower inhibitory potency than control trimethoprim. Again, although the weakest compound in our study (7) is not particularly impressive in its inhibition of T. gondii proliferation, it still displays a more favorable selectivity ratio than those for trimethoprim.  Table 2. For example, meta-and para-bromo derivatives 12 and 13 show higher selectivity ratios and only slightly lower inhibitory potency than control trimethoprim. Again, although the weakest compound in our study (7) is not particularly impressive in its inhibition of T. gondii proliferation, it still displays a more favorable selectivity ratio than those for trimethoprim.

11
and only slightly lower inhibitory potency than control trimethoprim. Again, although the weakest compound in our study (7) is not particularly impressive in its inhibition of T. gondii proliferation, it still displays a more favorable selectivity ratio than those for trimethoprim.

Physicochemical Characterization of 5-16
Membrane permeability for a drug intended for an intracellular target is a key metric at the early stage of the anti-Toxoplasma drug development processes [26]. Indeed, a molecule with good in vitro activity but poor membrane permeability will have low or nonexistent efficacy in vivo. Therefore, a detailed understanding of the preferential partitioning of a drug candidate into the membrane is vitally important from a rational drug design standpoint. When lipophilicity and its relationship with passive drug transfer through physiological barriers has been extensively investigated, numerous significant correlations between logP and drug passive permeation have been established. In fact, absorption by passive diffusion permeation is generally considered optimal for compounds having a moderate logP, with a range between 1.5 to 2.7 for blood-brain barrier penetration and a range between 0 to 3 for optimal gastrointestinal absorption [27,28]. Compounds with a lower logP are more polar and have poorer lipid bilayer permeability while compounds with a higher logP are more lipophilic and thus have better membrane permeability. At the same time, however, serious liabilities for more lipophilic compounds can also be incurred, including poor water solubility, increased toxicity, and faster metabolic clearance [29].
In the light of all the facts mentioned above, for the purpose of the study, the lipophilicity of 5-16 was measured experimentally by the HPLC technique and its correlation with antiparasitic activity was probed. The logP parameters for 5-16 are shown in Table 2. The comparison of the logP and IC 50 shows that the general trend in anti-Toxoplasma gondii activity of 5-16 is dependent upon lipophilicity and a significant incremental increase in the partition coefficient resulting from an increase in activity is observed, with the Spearman's rank correlation coefficient 0.81 and p value 0.001 ( Figure 4). In fact, except for 10, the weakest inhibitors in our studies (5)(6)(7) are the least lipophilic, while the most lipophilic compound (15) proved to be the best inhibitor of T. gondii growth in vitro. Thus, the results collectively suggest that lipophilicity would be a rate-limiting factor for the anti-Toxoplasma activity of the halogen-substituted imidazole-thiosemicarbazides. From the point of view of rational drug design, other important conclusions from these studies should also be mentioned: (i) fluoro-substituted thiosemicarbazides (5-7) are the least lipophilic followed by the ortho isomers, 8, 14, 11; (ii) lipophilicity of the meta and para isomers increases with halogen size from chlorine to iodine; (iii) para isomers tend to have lower lipophilicity than their meta counterparts (para-chloro < meta-chloro < para-bromo < meta-bromo < para-iodo < meta-iodo).
Molecules 2019, 24, x FOR PEER REVIEW 9 of 15 membrane is vitally important from a rational drug design standpoint. When lipophilicity and its relationship with passive drug transfer through physiological barriers has been extensively investigated, numerous significant correlations between logP and drug passive permeation have been established. In fact, absorption by passive diffusion permeation is generally considered optimal for compounds having a moderate logP, with a range between 1.5 to 2.7 for blood-brain barrier penetration and a range between 0 to 3 for optimal gastrointestinal absorption [27,28]. Compounds with a lower logP are more polar and have poorer lipid bilayer permeability while compounds with a higher logP are more lipophilic and thus have better membrane permeability. At the same time, however, serious liabilities for more lipophilic compounds can also be incurred, including poor water solubility, increased toxicity, and faster metabolic clearance [29].
In the light of all the facts mentioned above, for the purpose of the study, the lipophilicity of 5-16 was measured experimentally by the HPLC technique and its correlation with antiparasitic activity was probed. The logP parameters for 5-16 are shown in Table 2. The comparison of the logP and IC50 shows that the general trend in anti-Toxoplasma gondii activity of 5-16 is dependent upon lipophilicity and a significant incremental increase in the partition coefficient resulting from an increase in activity is observed, with the Spearman's rank correlation coefficient 0.81 and p value 0.001 ( Figure 4). In fact, except for 10, the weakest inhibitors in our studies (5-7) are the least lipophilic, while the most lipophilic compound (15) proved to be the best inhibitor of T. gondii growth in vitro. Thus, the results collectively suggest that lipophilicity would be a rate-limiting factor for the anti-Toxoplasma activity of the halogen-substituted imidazole-thiosemicarbazides. From the point of view of rational drug design, other important conclusions from these studies should also be mentioned: (i) fluoro-substituted thiosemicarbazides (5-7) are the least lipophilic followed by the ortho isomers, 8, 14, 11; (ii) lipophilicity of the meta and para isomers increases with halogen size from chlorine to iodine; (iii) para isomers tend to have lower lipophilicity than their meta counterparts (para-chloro < meta-chloro < para-bromo < meta-bromo < para-iodo < meta-iodo).

Chemistry
All commercial reactants and solvents were purchased from either Sigma-Aldrich (Saint Louis,

Chemistry
All commercial reactants and solvents were purchased from either Sigma-Aldrich (Saint Louis, MS, USA) or Alfa Aesar (Karlsruhe, Germany) with the highest purity and used without further purification. The melting points were determined on a Fischer-Johns block and are uncorrected. Analytical thin layer chromatography was performed with Merck (Darmstadt, Germany) 60F 254 silica gel plates and visualized by UV irradiation (254 nm). Elemental analyses were determined by a AMZ-CHX elemental analyzer (PG, Gdańsk, Poland). Analyses indicated by the symbols of the elements were within ±0.4% of the theoretical values. The 1 H-NMR were recorded on a Bruker Avance (300 MHz) spectrometer. For representative model compounds (8,13, and 16) 13 C-NMR spectra were also recorded on a Bruker Avance spectrometer. MS were recorded on a Bruker microTOF-Q II mass spectrometer (BioSpin, GmbH, Rheinstetten, Germany) using APCI method. The physicochemical characterization of compounds 5-7, 9, 10 were presented in [19,30,31].

Procedure for Synthesis of the Imidazole-Thiosemicarbazides 5-16
A solution of 4-methylimidazole-5-carbohydrazide (0.01 mol) and an equimolar amount of haloaryl isothiocyanate (0.01 mol) in 25 mL of anhydrous ethanol was heated under reflux for 10-30 min. After cooling, the solid formed was filtered off, dried, and crystallized from ethanol.