1,3-Substituted Imidazolidine-2,4,5-triones: Synthesis and Inhibition of Cholinergic Enzymes

A series of novel and highly active acetylcholinesterase and butyrylcholinesterase inhibitors derived from substituted benzothiazoles containing an imidazolidine-2,4,5-trione moiety were synthesized and characterized. The molecular structure of 1-(2,6-diisopropyl-phenyl)-3-[(1R)-1-(6-fluoro-1,3-benzothiazol-2-yl)ethyl]-imidazolidine-2,4,5-trione (3g) was determined by single-crystal X-ray diffraction. Both optical isomers are present as two independent molecules in the triclinic crystal system. The lipophilicity of the compounds was determined as the partition coefficient log Kow using the traditional shake-flask method. The in vitro inhibitory activity on acetylcholinesterase from electric eel and butyrylcholinesterase isolated from equine serum was determined. The inhibitory activity on acetylcholinesterase was significantly higher than that of the standard drug rivastigmine. The discussed compounds are also promising inhibitors of butyrylcholinesterase, as some of the prepared compounds inhibit butyrylcholinesterase better than the internal standards rivastigmine and galanthamine. The highest inhibitory activity (IC50 = 1.66 μmol/L) corresponds to the compound 1-(4-isopropylphenyl)-3-[(R)-1-(6-fluorobenzo[d]thiazol-2-yl)ethyl]imidazolidine-2,4,5-trione (3d). For all the studied compounds, the relationships between the lipophilicity and the chemical structure as well as their structure-activity relationships are discussed.

Many low-molecular-weight drugs cross biological membranes through passive transport, which strongly depends on their lipophilicity, therefore the experimental log P (or log K ow ) n-octanol/water partition coefficients were determined. Structure-activity relationships between the chemical structure, physical properties and biological activities of the evaluated compounds are discussed.

Crystallography
The exemplary compound 3g crystallizes in the triclinic crystal system, achiral point group P-1 with two independent molecules (different enantiomers) within the unit cell ( Figure 1). The opposite orientation of the C8 atom seems to be the only remarkable difference between these two molecules. The interatomic distances and angles including torsion and interplanar angles are similar to the typical values found in the literature for similar atom combinations [22].
There is a quite extensive π-π stacking network causing the 3D structure formation with interactions between the heterocyclic rings of 3.372(3)Å and carbonyl groups of 3.121(3)Å, see Figure 2. Although one might expect very tight organization due to these interactions, remarkable cavities were found within the crystal lattice of compound 3g, see Figure 1. No solvent can be accommodated in these cavities, probably because both hydrophilic and hydrophobic parts are present in the molecule of the mentioned compound. Another reason could be free rotability of the bulky diisopropylphenyl group. The π-π stacking interactions of aromatic rings and C=O groups in the molecules are shown in Figure 2. Molecules of compound 3g also form interesting supramolecular architecture, see Figure 3.   (3), C3 S1 C7 88.1(2), S1 C3 C4 111.0(3), C4 N1 C7 106.9(4), N1 C7 S1 117.8(3), N2 C8 C7 108.2(3).

Lipophilicity
Lipophilicity is a property that has a major effect on the absorption, distribution, metabolism, excretion and toxicity properties, as well as pharmacological activity, because drugs cross biological membranes through passive transport, which strongly depends on their lipophilicity. Lipophilicity has been studied and applied as an important drug property for decades [37]. Hydrophobicities (log P/Clog P) of the compounds were calculated using the commercially available program ChemOffice Ultra 11.0. The experimental partition coefficient log K ow (n-octanol/water) was determined using the traditional shake-flask method. The n-octanol/water partition coefficient P (also referred to as K ow ) is a measure of the propensity of a neutral compound to differentially dissolve in these immiscible phases. It is usually referred to as the logarithmic ratio, log P. The partition coefficient serves as a quantitative descriptor of lipophilicity and is one of the key determinants of pharmacokinetic properties. The partition coefficient log K ow (n-octanol/water) indicates potential for crossing the blood-brain barrier for direct inhibition of brain cholinesterases [38]. Experimentally it is done by partitioning the molecule between water and the hydrophobic solvent n-octanol and determining the P value as the ratio of the concentration of the compound in n-octanol and in water. Log K ow can be used as the lipophilicity index converted to in silico log P scale. The results are shown in Table 1. Table 1. Comparison of calculated lipophilicities (log P/Clog P) with determined log K ow values, Hammett's σ parameters of prepared substituted imidazolidine-2,4,5-triones and their AChE and BChE inhibition in comparison with standards rivastigmine (RIV) and galanthamine (GLT). ChE inhibitions are expressed as mean ± SD (n = 3 experiments), and log K ow data of the compounds are expressed as mean ± SD (n = 3 experiments). The results obtained with all the compounds show that the experimentally-determined lipophilicities (log K ow ) of the discussed compounds are in poor accordance with the calculated values of compounds. This fact suggests significant intramolecular interactions within the whole series of the compounds. Compounds 3e (4-Cl) and 4f (4-CN) showed the lowest lipophilicity, while compounds 3a (H) and 3d [4-CH(CH 3 ) 2 ] demonstrated the highest. Benzyl derivative 3k showed lower lipophilicity than phenyl derivative 3a. Therefore it can be assumed that the determined log K ow data specify lipophilicity within this series of the discussed compounds.

Inhibition of Cholinergic Enzymes
All the prepared carbamate-like compounds were tested for their inhibition of AChE and BChE. The activities of the compounds were compared with the internal standards rivastigmine (RIV, Exelon ® ) and galanthamine (GLT, Reminyl ® ), see Figure 4. These standards were chosen by reason of the different structures of both drugs. While rivastigmine is a classical acylating pseudo-reversible carbamate cholinesterase inhibitor that inhibits both acetylcholinesterase and butyrylcholinesterase, galanthamine is a non-acylating competitive reversible cholinesterase inhibitor and also an allosteric ligand at nicotinic acetylcholine receptors. The choice of these reference drugs with different mechanisms of action can provide relevant results. The results are summarized in Table 1 and expressed as 50% inhibitory concentration (IC 50 [μmol/L]), or the concentration of inhibitor required for 50% inhibition of the mentioned enzymes.
Based on these facts, it can be concluded that AChE is inhibited by para-and meta-substutited phenyl rings, while BChE is preferentially inhibited by a para-substutited phenyl ring. These observations describing steric/positional aspects of substitution on benzene are in agreement with recently published results by Chiou et al. indicating that that AChE prefers para-and meta-substitution to ortho-substitution, whereas BChE prefers para-substitution to ortho-and meta-substitution. These results imply that steric differences in the active sites of both enzymes can be found [40].
It is noteworthy that compounds with high inhibitory activity possess a branched substituent. For example, in case of the AChE inhibitors these are compounds 3d The dependence of AChE inhibition (log L/IC 50 [mol/L]) on log P is illustrated in Figure 5A. The set of 11 tested compounds can be divided into para-substituted and/or unsubstituted, where a bilinear dependence can be found, and meta-substituted, where activity sharply increases with a slight lipophilicity increase. The dependence of BChE inhibition (log L/IC 50 [mol/L]) on log P is illustrated in Figure 5B, and from these relationships it is evident that lipophilicity is only a secondary parameter, although it seems that BChE inhibition activity decreases with lipophilicity increase within meta-substituted series. However, inhibition of cholinergic enzymes is also significantly dependent on the electronic properties of R phenyl substituents expressed as Hammett's σ parameters. Figure 6A shows an evident general bilinear trend: with the increase of electron-withdrawing effect of individual substituents to the value σ = 0.23 (3e, 4-Cl), which is the optimum, AChE-inhibiting activity increases to the value IC 50 = 13.8 μmol/L, and with the following increase of electron-withdrawing effect the activity decreases. Also other compounds 3g [2,6-CH(CH 3

General
All reagents and solvents were purchased from commercial sources (Sigma-Aldrich, Merck, Acros Organics, Lach-Ner CZ). Commercial grade reagents were used without further purification. The reactions were monitored and the purity of the products was checked by thin-layer chromatography plates coated with 0.2 mm silica gel 60 F 254 (Merck, Darmstadt, Germany). TLC plates were visualized by UV irradiation (254 nm). All the melting points were determined on a Melting Point B-545 apparatus (Buchi, Germany) and are uncorrected. Infrared spectra (ZnSe ATR experiments) were recorded on a FT-IR spectrometer (Perkin Elmer, USA) in the range of 600-4000 cm −1 . The NMR spectra were measured in DMSO-d 6 solutions at ambient temperature on a Bruker Avance III 400 MHz spectrometer (Karlsruhe, Bruker, Germany, 400 MHz for 1 H, 100 MHz for 13 C and 376.5 MHz for 19 F). Proton chemical shifts in DMSO-d 6 are related to the middle of the solvent multiplet (δ = 2.50). 13 C-NMR spectra were measured using APT pulse sequences. Carbon chemical shifts are referenced to the middle of the solvent multiplet (δ = 39.5 in DMSO-d 6 ). 19 F-NMR spectra were measured using waltz-16 proton decoupling and were standardised against fluorobenzene as the secondary external standard (δ = −113.1 against CFCl 3 as the primary standard [41], td 64k zero filled to 128k). Chemical shifts for AB systems were calculated as weighed average of the line positions weighted by the peak intensities.

Synthesis
General Procedure for the Synthesis of Compounds 3a-k The appropriate 1-[(1R)-1-(6-fluoro-1,3-benzothiazol-2-yl)ethyl]-3-substituted phenyl urea 2a-k (5 mmol) was added to a cooled solution (0-5 °C) of oxalyl chloride (7 mmol) in dichloromethane (20 mL) and mixture was stirred at temperature 0-5 °C for 1 h. The mixture was left to warm up to the ambient temperature and stirred for another 2 h. The reaction mixture was filtered and concentrated under reduced pressure. The addition of n-hexane caused precipitation of products 3a-k. The precipitated solid was collected by filtration, dried in vacuum to give the title compounds in 80-85% yields as light yellow crystalline solids (unless stated otherwise).

Determination of Crystallography
The X-Ray data for colourless crystals of compound 3g were obtained at 150 K using Oxford Cryostream low-temperature device on a Nonius KappaCCD diffractometer with MoK α radiation (λ = 0.71073 Å), a graphite monochromator and the φ and χ scan mode. Data reductions were performed with DENZO-SMN [42]. The absorption was corrected by integration methods [43]. Structures were solved by direct methods (Sir92) [44] and refined by full matrix least-square based on F 2 (SHELXL97) [45]. Hydrogen atoms were mostly localized on a difference Fourier map, however to ensure uniformity of the treatment of the crystal, all hydrogen atoms were recalculated into idealized positions (riding model) and assigned temperature factors H iso (H) = 1.2 U eq (pivot atom) or of 1.5 U eq for the methyl moiety with C-H = 0.96, 0.98 and 0.93 Å for methyl, methine and hydrogen atoms in the aromatic rings, respectively.
Selected crystallographic data for compound 3g: Crystallographic data for structural analysis have been deposited with the Cambridge Crystallographic Data Centre (deposition number CCDC 837618). Copies of this information may be obtained free of charge from the Director, CCDC, 12 Union Road, Cambridge CB2 1EY, UK (fax: +44-1223-336033; e-mail: deposit@ccdc.cam.ac.uk or http://www.ccdc.cam.ac.uk).

Determination of Partition Coefficient K ow
Before the partition coefficient is determined, the two solvents are mutually saturated at the temperature of the experiment. To do this, it is practical to shake two large stock bottles, one containing n-octanol and a sufficient quantity of water, and the other containing water and a sufficient quantity of n-octanol, for 24 hours on a mechanical shaker and then to let them stand long enough to allow the phases to separate [46].
The procedure of determination was the following: n-Octanol (2 mL) was placed in a test tube. Then an octanol solution of the chosen inhibitor (15 μL, 0.01 M) was added. Mixture was intensively shaken for 15 min. This mixture (1 mL) was placed into the cell, and its absorbance at the absorption maximum wavelength was measured. The reference solution was n-octanol. The value of absorbance corresponding to 100% of the chosen inhibitor in n-octanol was obtained. An n-octanol solution of the chosen inhibitor (0.01 M, 15 μL) was added into the mixture of n-octanol and water (1:1, total volume 4 mL). The mixture was intensively shaken for 15 min and then centrifuged (3,000 rpm, 10 min). One mL of the n-octanol layer was put into the cell, and its absorbance at the wavelength of absorption maximum was measured. The comparative solution was n-octanol again. The percentage content of chosen inhibitor in the octanol layer (%) was obtained. The n-octanol/water partition coefficient is defined as P ow = c 1 /c 2 , where c 1 and c 2 are molar concentrations of tested compounds in n-octanol and water. For each compound, at least three determinations were performed. The log K ow values of the individual compounds are shown in Table 1.

Lipophilicity Calculations
Log P, i.e., the logarithm of the partition coefficient for n-octanol/water, was calculated using the program CS ChemOffice Ultra ver. 11.0 (CambridgeSoft, Cambridge, MA, USA). Clog P values (the logarithm of n-octanol/water partition coefficient based on established chemical interactions) were generated by means of the same software. The results are shown in Table 1.